^ 2: C/> ^ yjt

UTiON NoiiniiisNi NViNOSHilws S3iavyan libraries Smithsonian institution Noiiniii!

c/i -j. ■-» (/) ~ !s> ~

^aan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws saiavaan librar

UTION NOlifUliSNI NVINOSHillNS SBIdVaan LIBRARIES SMITHSONIAN INSTITUTION NOIlfUli

X
CO

i wm

^aan libraries Smithsonian institution NoiifuiiSNi NviNOSHims^^sa i a va a

(/) — to

1_LI B RAR
z

UTION NoiiniiisNi_NvmosHiUMS^S3 1 ava a n libraries smithsonian^institution Noiiniii
o m V 2 TO TO i

to ' X CO fL to

BRARIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillNS S3iavaail LIBRAR

UTION NOliniliSNI NVINOSHilWS S3iavaan libraries SMITHSONIAN INSTITUTION NOlinili

^ cr ^i-7

O

_l X -I

/aan libraries Smithsonian institution NoiiniiiSNi NViNOSHims saiavaan librari

UTION NOliniliSNI NViNOSHims S3iavaaii libraries Smithsonian institution Noiiniii

(/)

z <0 2 to

1 libraries SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHillMS S3iavaa

2 ^ Z ^ 5 ^

to ^ . c/5 ^

CO

n_ LI B RAR
z

NOSHiiiMS S3iavdan libraries Smithsonian institution N0iiniiiSNi_NviN0SHiiiNS saiav

z: (/) — (/^ z \ ^ ^

» < — \K.v^ -A/'''. r“ \L .

■:tN

Z -J

ITHSONIAN INSTITUTION NOliOiliSNI NVINOSHlIl^S S3iaVHan LIBRARIES SMITHSONIAN INSTIT
“ z t- z z; —

CD

CO f: c/> \ £ CO — CO

NOSHims S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshihns S3iav

CO Z -,•• CO Z V' ^ ^ ^

1
CO

o

2 Vo'*^iy^^y t '"■ ?

> '■W^' 5 '" > ^ >

Z U) ’• 'Z. C/)‘Z CO z

THSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3iavaail LIBRARIES SMITHSONIAN INSTIT

<D — CO ~ CO 2

O

z

NOSHiiws S3 lava an libraries Smithsonian institution NoiiniiiSNi NViNOSHims S3iav

z I— z Z 2 ''' ^

CD
:d

V. >

~ V

m xjwiix ^ m

CO _ CO

THSONIAN INSTITUTION NOIifUliSNI NVINOSHimS S3iavaai1 LIBRARIES SMITHSONIAN INSTIT

CO z (o z , ^2 z

< s: 2 <

NOSHilws S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshii/js S3ia\i

THSONIAN INSTITUTION NOIifUliSNI NVINOSHlIl^S S3iavaan LIBRARIES SMITHSONIAN INSTIT

z

CO

O
z

z CO*Z CO*^ Z CO z

THS0N1AN_1NSTITUTI0N NOlinillSNI NVINOSHIIIAIS S3iavaan LIBRARIES SMITHSONIAN _ INSTIT

CD = CO Z CO

CO

VOLUME 30

Palaeontology

1987

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION

LONDON

Dates of Publication of Parts of Volume 30

Part I, pp. 1-206, pis. 1-23
Part 2, pp. 207-428, pis. 24-53
Part 3, pp. 429 -646, pis. 54 72
Part 4, pp. 647-868, pis. 73-91

6 March 1987
29 May 1987
21 August 1987
18 December 1987

THIS VOLUME EDITED BY M. J. BENTON, P. R. CROWTHER, D. EDWARDS, L. B. HALSTEAD, T. J. PALMER,

C. R. C. PAUL, AND P. A. SELDEN

Dates of Publication of Special Papers in Palaeontology

Special Paper No. 37, 4 December 1987
Special Paper No. 38, 11 December 1987

© The Palaeontological Association, 1987

Printed in Great Britain
at the University Printing Home, Oxford
by David Stanford
Printer to the University

CONTENTS

Part Page

Adams-Tresman, S. M. The Callovian (Middle Jurassic) marine crocodile Metriorhynchus

from central England I 179

Adams-Tresman, S. M. The Callovian (Middle Jurassic) teleosaurid marine crocodiles from

central England I 195

Acer, D. V. Why the rhynchonellid brachiopods survived and the spiriferids did not: a

suggestion 4 853

Benton, M. J. See Wright, A. D. and Benton, M. J.

Berg-Madsen, V. A new cyclocystoid from the Lower Ordovician of Oland, Sweden I 105

Blows, W. T. The armoured dinosaur Polacantinis foxi from the Lower Cretaceous of the

Isle of Wight 3 557

Brinkman, D. B. and Sues, H.-D. A staurikosaurid dinosaur from the Upper Triassic

Ischigualasto Eormation of Argentina and the relationships of the Staurikosauridae 3 493

Bromley, R. G. See D’Alessandro, A, and Bromley, R. G.

Bull, E. E. Upper Llandovery dendroid graptolites from the Pentland Hills, Scotland 1 117

Chalifa, Y. See Raab, M. and Chalifa, Y.

Chatterjee, S. See Grande, L. and Chatterjee, S.

Clack, J. A. Two new specimens of Anthracosaiinis (Amphibia: Anthracosauria) from the

Northumberland Coal Measures I 1 5

D’Alessandro, A. and Bromley, R. G. Meniscate trace fossils and the Muensteria Taenidiion

problem 4 743

Donovan, S. K. See Jefferies, R. P. S., Lewis, M. and Donovan, S. K.

Doyle, P. The Cretaceous Dimitobelidae (Belemnitida) of the Antarctic Peninsula region I 147

Doyle, P. Early Cretaceous belemnites from southern Mozambique 2 311

Drygant, D. M. See Selden, P. A. and Drygant, D. M.

Gao Zhifeng and Thomas, B. A. A re-evaluation of the plants Tingia and Tingiostachya from

the Permian of Taiyuan, China 4 815

Grande, L. and Chatterjee, S. New Cretaceous fish fossils from Seymour Island, Antarctic

Peninsula 4 829

Hancock, J. M. See Kennedy, W. J., Wright, C. W. and Hancock, J. M.

Hickey, D. R. Skeletal structure, development, and elemental composition of the Ordovician

trepostome bryozoan Peronopora 4 691

Hofmann, H. J. See Narbonne, G. M. and Hofmann, H, J. 4 647

Jago, j. B. Idamean (Late Cambrian) trilobites from the Denison Range, south-west Tasmania 2 207

Jefferies, R. P. S., Lewis, M. and Donovan, S. K. Piotocysfites Menevemis — a stem-group

chordate (Cornuta) from the Middle Cambrian of South Wales 3 429

Jefferson, T. H. The preservation of conifer wood: examples from the Lower Cretaceous of

Antarctica 2 233

Kat, P. W. Biogeography and evolution of African freshwater molluscs: implications of a

Miocene assemblage from Rusinga Island, Kenya 4 733

Kennedy, W. J. Ammonites from the type Santonian and adjacent parts of northern Aquitaine,

western Prance 4 765

Kennedy, W. J., Wright, C. W. and Hancock, J. M. Basal Turonian ammonites from west

Texas 1 27

Lewis, M. See Jefferies, R. P. S., Lewis, M. and Donovan, S. K.

Li Jun. Ordovician acritarchs from the Meitan Pormation of Guizhou Province, south-west

China 3 613

Long, J. A. An unusual osteolepiform fish from the Late Devonian of Victoria, Australia 4 839

McNamara, K. H. Taxonomy, evolution, and functional morphology of southern Australian

tertiary hemiasterid echinoids 2 319

Mapes, R. H. and Sneck, D. A. The oldest ammonoid ‘colour’ patterns: description,

comparison with Nautilus, and implications 2 299

iv CONTENTS

Martill, D. M. a taphonomic and diagenetic case study of a partially articulated ichthyosaur 3 543

Mateer, N. J. a new report of a theropod dinosaur from South Africa 1 141

Mawson, R. Early Devonian conodont faunas from Buchan and Bindi, Victoria, Australia 2 251

Mitchell, C. E. Evolution and phylogenetic classihcation of the Diplograptacea 2 353

Mutterlose, J., Pinckney, G. and Rawson, P. F. The belemnite Acroteuthis in the Hibolites

Beds (Hauterivian Barremian) of north-west Europe 3 635

Narbonne, G. M. and Hofmann, H. J. Ediacaran biotas of the Wernecke Mountains, Yukon,

Canada 4 647

Owen, A. W. The Scandinavian Middle Ordovician trinucleid trilobites 1 75

Palmer, D. See Siveter, D. J., Vannier, J. M. C. and Palmer, D.

Pinckney, G. See Mutterlose, J., Pinckney, G. and Rawson, P. F.

Quayle, W. j. English Eocene Crustacea (lobsters and stomatopod) 3 581

Raab, M. and Chalifa, Y. A new enchodontid fish genus from the upper Cenomanian of

Jerusalem, Israel 4 717

Raup, D. M. Mass extinction; a commentary 1 1

Rawson, P. F. See Mutterlose, J., Pinckney, G. and Rawson, P. F.

Scrutton, C. T. a review of favositid affinities 3 485

Selden, P. a. and Drygant, D. M. A new Silurian xiphosuran from Podolia, Ukraine, USSR 3 537

Siveter, D. J., Vannier, J. M. C. and Palmer, D. Silurian myodocopid ostracodes: their

depositional environments and the origin of their shell microstructures 4 783

Skelton, P. W. and Wright, V. P. A Caribbean rudist bivalve in Oman: island-hopping

across the Pacific in the late Cretaceous 3 505

Sneck, D. a. See Mapes, R. H. and Sneck, D. A.

Sues, H.-D. See Brinkman, D. B. and Sues, H.-D.

Taylor. M. A. A reinterpretation of ichthyosaur swimming and buoyancy 3 531

Thomas, B. A. See Gao Zhifeng and Thomas, B. A.

Tunnicliff, S. P. Caradocian bivalve molluscs from Wales 4 677

Vannier, J. M. C. See Siveter, D. J., Vannier, J. M. C. and Palmer, D.

Wright, A. D. and Benton, M. J. Trace fossils from Rhaetic shore-face deposits of

Staffordshire 2 407

Wright, C. W. See Kennedy, W. J., Wright, C. W. and Hancock, J. M.

Wright, V. P. See Skelton, P. W. and Wright, V. P.

f

¥

'if

Published by

The Palaeontological Association • London
Price £23 00

THE PALAEONTOLOGICAL ASSOCIATION

The Association was founded in 1957 to promote research in palaeontology and its allied sciences.

COUNCIL 1986-1987

President. Dr. L. R. M. Cocks, Department of Palaeontology, British Museum (Natural History), Cromwell Road,

London SW7 5BD

Vice-Presidents'. Dr. M. G. Bassett, Department of Geology, National Museum of Wales, Cardiff CFl 3NP
Dr. D. E. G. Briggs, Department of Geology, University of Bristol, Bristol BS8 IRJ
Treasurer. Dr. M. Romano, Department of Geology, University of Sheffield, Sheffield SI 3JD
Membership Treasurer. Dr. A. T. Thomas, Department of Geological Sciences, University of Aston. Birmingham B4 7ET
Institutional Membership Treasurer. Dr. A. W. Owen, Department of Geology, The University, Dundee DDl 4HN
Secretary: Dr. P. W. Skelton, Department of Earth Sciences, The Open University, Milton Keynes MK7 6AA
Circular Reporter: Dr. D. J. Siveter, Department of Geology, University of Hull, Hull HU6 7RX
Marketing Manager: Dr. V. P. Wright, Department of Geology, University of Bristol, Bristol BS8 IRJ
Public Relations Officer: Dr. M. J. Benton, Department of Geology, The Queen’s University of Belfast, Belfast BT5 6FB

Editors

Dr. P. R. Crowther, City of Bristol Museum and Art Gallery, Bristol BS8 IRL
Dr. D. Edwards, Department of Plant Science, University College, Cardiff CFl IXL
Dr. L. B. Halstead, Department of Geology, University of Reading, Reading RG6 2AB
Dr. T. J. Palmer. Department of Geology, University College of Wales. Aberystwyth SY23 2AX
Dr. C. R. C. Paul. Department of Geology, University of Liverpool, Liverpool L69 3BX
Dr. P. A. Selden, Department of Extra-Mural Studies, University of Manchester, Manchester M13 9PL

Other Members

Dr. H. A. Armstrong, Newcastle upon Tyne Professor B. M. Funnell, Norwich

Dr. M. E. Collinson, London Dr. P. D. Taylor, London

Overseas Representatives

Australia: Professor B. D. Webby, Department of Geology, The University, Sydney, N.S.W., 2006
Canada: Dr. B. S. Norford, Institute of Sedimentarv and Petroleum Geology, 3303-33rd Street NW., Calgary, Alberta
Japan : Dr. I. Hayami. University Museum, University of Tokyo, Hongo 7-3-1, Bunkyo-Ku, Tokyo
New Zealand: Dr. G. R. Stevens, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt
U.S.A.: Dr, R. J. Cuffey, Department of Geology, Pennsylvania State University, Pennsylvania 16802
Professor A. J. Rowell, Department of Geology, University of Kansas, Lawrence, Kansas 66045
Professor N. M. Savage, Department of Geology, University of Oregon, Eugene, Oregon 97403
South America: Dr. O. A. Reig, Departamento de Ecologia, Universidad Simon Bolivar, Caracas 108. Venezuela

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1987 are;

Institutional membership . . . £50 00 (U.S. $79)

Ordinary membership . £21 00 (U.S. $38)

Student membership . £1 1 -50 (U.S. $20)

Retired membership . £10-50 (U.S. $19)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr. A. W. Owen,
Department of Geology, The Liniversity, Dundee DDl 4HN. Student members are persons receiving full-time instruction at
educational institutions recognized by the Council. On first applying for membership, an application form should be obtained
from the Membership Treasurer, Dr. A. T. Thomas, Department of Geological Sciences, University of Aston, Aston Triangle,
Birmingham B4 7ET. Subscriptions cover one calendar year and are due each January; they should be sent to the Membership
Treasurer. All members who join for 1987 will receive Palaeontology. Volume 30, Parts 1 4. Back numbers still in print
may be ordered from Marston Book Services, P.O. Box 87, Oxford 0X4 ILB, England.

Cover: Pedicle valve of the brachiopod Strophonella euglypha (Dalman, 1828) from the Wenlock Limestone of Dudley,
West Midlands; x 2. Photography by Harry Taylor of the British Museum (Natural History) Photographic Studio. One of
the specimens illustrated in the Atlas of Invertebrate Macrofossils published by the Association.

MASS EXTINCTION: A COMMENTARY

by DAVID M. RAUP

(Twenty-ninth annual address, delivered 19 March 1986)

Abstract. Four neocatastrophist claims about mass extinction are currently being debated; they are that: 1,
the late Cretaceous mass extinction was caused by large body impact; 2, as many as five other major extinctions
were caused by impact; 3, the timing of extinction events since the Permian is uniformly periodic; and 4, the
ages of impact craters on Earth are also periodic and in phase with the extinctions. Although strongly
interconnected the four claims are independent in the sense that none depends on the others. Evidence for a
link between impact and extinction is strong but still needs more confirmation through bed-by-bed and
laboratory studies.

An important area for future research is the question of whether extinction is a continuous process, with
the rate increasing at times of mass extinctions, or whether it is episodic at all scales. If the latter is shown to
be generally true, then species are at risk of extinction only rarely during their existence and catastrophism, in
the sense of isolated events of extreme stress, is indicated. This line of reasoning can only be considered an
hypothesis for testing.

In a larger context, palaeontologists may benefit from a research strategy that looks to known Solar System
and Galactic phenomena for predictions about environmental effects on earth. The recent success in the
recognition of Milankovitch Cycles in the late Pleistocene record is an example of the potential of this research
area.

During the past five years we have seen an enormous increase in debate about the causes of
mass extinction. This has spilled over into disciplines far removed from palaeontology, including
astrophysics, and the popular press has been covering the story avidly— perhaps too avidly. Further-
more, the extinction at the end of the Cretaceous has made a strange contribution to contemporary
politics. The modelling of atmospheric effects of the putative meteorite impact has led directly to
the concept of Nuclear Winter, at least in its American version.

The literature of the past five years has contained four strong claims about extinction:

1. That the terminal Cretaceous extinctions were caused by a collision between Earth and a
comet or asteroid.

2. That as many as five other extinction events were similarly caused by large body impact.

3. That the principal extinction events of the Mesozoic and Cenozoic are uniformly periodic,
with a spacing of 26-30 million years;

4. That the ages of the major impact craters on Earth are also uniformly periodic and in phase
with the extinctions.

These four claims are independent of each other in the sense that any one could be wrong without
affecting the others. The Cretaceous extinction could be impact-produced but not the others, or
extinctions could be periodic but not driven by impacts. On the other hand, some people have
chosen to put the four together in an integrated theory of mass extinction calling for a Solar System
or Galactic ‘clock’ which causes periodic showers of comets producing both extinction and cratering
at regular intervals.

To many palaeontologists the developments I have just described represent a most disturbing
reversion to the worst kind of catastrophism: wild explanations calling on mysterious forces that
nobody has ever really seen; hypothesis piled upon hypothesis with vanishingly little evidence for
each one. Many palaeontologists are annoyed to see people from other disciplines charging into

(Palaeontology, Vol. 30, Part 1, 1987, pp. 1-13.|

© The Palaeontological Association

2

PALAEONTOLOGY, VOLUME 30

palaeontology with expensive instruments but no experience with rocks and fossils. To other
palaeontologists, however, the new developments are plausible and supported by field and labora-
tory data. To these optimists, palaeontology is making new contributions to our knowledge of
Earth history and to astronomy, to say nothing of increased understanding of extinction as it relates
to the evolution of life.

In this essay, I will review briefly the present state of the four neo-catastrophist claims and then
look a bit deeper into the general phenomenon of extinction in the fossil record. Finally, I will
suggest a somewhat different approach to research into the possible extraterrestrial influences on
past life.

EXTINCTION BY METEORITE IMPACT

The most important element in the current debate is the proposal of a very large impact at the
Cretaceous-Tertiary boundary. This was based initially on the high concentrations of iridium in
K-T boundary sections in Italy, Denmark, and New Zealand (Alvarez et al. 1980). The 1980
Alvarez paper made a powerful case by any normal standards. The reported anomalies were
substantial and because concentrations of iridium in some meteorites are known to be high and
because iridium is nearly absent in the Earth’s crust, the Alvarez paper convinced many people
immediately. Also, the finding was completely plausible because it is well known that meteorite
impacts (comets plus asteroids) have been relatively common throughout the Phanerozoic. It is
estimated that there have been about twelve collisions with bodies of 10 km or greater diameter
during Phanerozoic time and up to 3600 collisions with 1 km objects (Shoemaker 1984).

If the iridium story had been an isolated research result, concerned only with meteorite impact, I
doubt that anyone would have bothered to check the data or interpretation further. However,
because the Alvarez group went on to urge that the impact caused the late Maastrichtian extinctions,
the original interpretation of impact was hotly debated and the scientific community insisted on
more evidence.

An irony is that the whole idea is not new. I refer in particular to a paper published in Nature in
1973 by the Nobel chemist, Harold C. Urey. Using the dates of Cenozoic tektites and comparing
them statistically with the ages of extinctions, Urey argued that large body impacts were responsible
for even relatively minor extinction events. He ended the paper with a prediction that tektites and
other evidence of impact would ultimately be found at the top of the Cretaceous. Urey’s statistical
case was fairly strong. The strange thing is that the paper was virtually ignored.

Since 1980, iridium anomalies have been confirmed at scores of K-T boundary sections but more
importantly, other kinds of evidence have been found. Data from osmium isotopes (Luck and
Turekian 1983), probable microtektites (Smit and Klaver 1981; Montanari et al. 1983), and shocked
quartz (Bohor et al. 1984) have provided what might be called ‘overkill’ in establishing the fact of
an impact. This is not to say that there is complete unanimity. There are still dissenters, most
notably C. B. Officer and C. L. Drake (1983, 1985) who have argued for a volcanic origin of the
iridium and shocked quartz. All in all, however, I think that the case for extraterrestrial impact is
very strong.

The causal link between the Cretaceous impact and mass extinction is much less secure. All we
really have is a match or near-match in timing between two different events and there is much
legitimate argument about how good this time match is. The problem of deciding on cause and
effect comes down to arguments of probability. We have two relatively rare events in Earth history;
a mass extinction and a very large impact. What are the chances that these two could have occurred
at about the same time by pure chance? Fortunately, the coincidence problem can be tackled in a
reasonably rigorous fashion. We can pose the following question: if there were twelve impacts of
10 km bodies during the Phanerozoic, as has been estimated, what is the probability that one of
these should coincide with the Cretaceous extinctions, within the uncertainties of the geologic dating
of the extinctions?

RAUP: MASS EXTINCTION

3

It turns out that if the extinctions are dated only to the nearest six million years, the average
length of a post-Palaeozoic stage, the coincidence should be expected to occur by chance alone
10 % of the time, and the causal argument is not compelling. On the other hand, if the extinctions
can be limited to the final two million years of the Cretaceous, the probability of chance co-
occurrence is only 3 %, that is within the range normally considered to be statistically significant.
Attaching a confidence level to the cause-and-effect hypothesis is thus possible but depends on the
accuracy of the estimates of Phanerozoic impact rates and on our assessment of the dating of
the mass extinction. The exercise emphasizes, among other things, the importance of improving
stratigraphic control and of deciding whether the extinctions were sudden or gradual.

The best way to decide whether impacts cause extinctions is to see whether other extinctions are
associated in time with evidence of large body impact. Five other cases have been alleged. Iridium
anomalies have been found at the Eocene-Oligocene boundary (Alvarez et al. 1982; Ganapathy
1982), the Middle-Upper Jurassic boundary at the top of the Callovian (Brochwicz-Lewinski et al.
1984), the Permo-Triassic boundary (Sun et al. 1984), the Frasnian-Famennian boundary in the
Devonian (Playford et al. 1984), and at the base of the Cambrian (Hsu 1986). The last of these may
not, of course, be an extinction event. The Eocene and Jurassic cases are not really major mass
extinctions, but they certainly qualify as points of important faunal turnover and are used as series
boundaries. Other extraterrestrial signatures have also been found in these Eocene and Jurassic
cases. There is the possibility, of course, that the iridium and other impact evidence is ubiquitous in
the geologic record and that it is found only at extinctions because this is where people have been
looking. This problem is less serious than it once was, however, because so much more systematic
searching has been done throughout the column.

All of the cases listed have real problems. In the Jurassic and Devonian cases the iridium is found
only in stromatolites, making it possible that we are dealing with a purely biological enrichment.
The Permo-Triassic case is based on Chinese analyses and work on portions of the same samples in
other laboratories has not revealed the iridium.

In summary the link between impact and extinction is likely but not confirmed beyond reasonable
doubt.

The proposition that extinction events are uniformly periodic (Raup and Sepkoski 1984, 1986) is
alive and well although somewhat nervous. Periodicity is based on statistical inference from noisy
and uncertain data so there are inevitable doubts and questions.

Text-fig. 1 shows the familial and generic extinction records from the late Permian onward. These
are based on Sepkoski’s compendium of stratigraphic ranges of marine families (Sepkoski 1982)
and on his new compilation for genera (Raup and Sapkoski 1986). In the family data, eight events
stand significantly above background: late Permian (Guadalupian or Dzulfian), late Triassic (Norian
or Rhaetian), Pliensbachian, Tithonian, Cenomanian, Maastrichtian, Upper Eocene, and Middle
Miocene. The error bars show the uncertainty in the extinction metric and are used to say whether
the peaks are statistically above background. In the periodicity hypothesis, these eight events mark
eight of the ten possible 26-ma cycles— with the missing events being in the Middle Jurassic and
Lower Cretaceous. Interestingly, the Callovian iridium anomaly mentioned above is about where it
should be for one of these gaps.

The generic data show the same peaks but the extinction events are generally more pronounced—
even though the error bars are longer. The Upper Eocene and Middle Miocene events are much
more convincing. In addition, the generic data show a small peak in the Aptian and this may fill
the other gap in the 26-ma periodicity.

Text-fig. 2 summarizes the periodicity interpretation and shows the eight events in their proposed
positions with worst case error bars for the stratigraphic and radiometic dating. The straight line is
a 26-ma periodicity in best fit position. The youngest four events are the most accurately known.
Only time will tell whether periodicity is confirmed by other kinds of data. To this end, Sepkoski is
expanding the generic data set and is adding somewhat better stratigraphic resolution.

The analysis of the timing of these events is complex. Some people have claimed that the statistical
analysis is flawed whereas others have accepted it. I think this is almost inevitable in the statistics

4

PALAEONTOLOGY, VOLUME 30

250 200 150 100

Geologic time (ma)

TEXT-FIG. 1. Per cent extinction for marine families (a) and genera (b) for the Permian to
Recent interval. Extinction events which are identified by letters stand significantly above
the local background. (From Raup and Sepkoski 1986.)

RAUP: MASS EXTINCTION

5

GEOLOGIC TIME (10® YRS.)

TEXT-FIG. 2. The ages of the eight principal extinction events in the Permian-Recent interval
plotted against position in a perfect 26-ma periodicity (with two missing events). The straight line
represents the 26-ma periodicity in best-fit position. The horizontal error bars are ‘worst case’
uncertainties in stratigraphic and radiometric dating. (Modified from Sepkoski and Raup 1986.)

of time series. One surprising reaction has come from some astronomers: they say that the periodicity
is too perfect to accommodate some of the astronomical explanations that have been proposed.

Many palaeontologists and biostratigraphers have argued that the data on ranges of families and
genera are inadequate to justify this sort of statistical analysis. People familiar with the taxonomy
of a group appreciate all the guesswork and uncertainty and are appalled to see their taxa used as
data for global, statistical analysis. Obviously, I think the taxonomic and stratigraphic data are
worth using. Any data base of this sort can tolerate quite a bit of uncertainty as long as the
uncertainty is not systematic. If it could be shown that monographic problems could produce a
26-ma cycle where none actually exists, we would have a real problem. But as long as the uncertain-
ties are randomly distributed, their effect should only be to blur or cloud any signal or pattern that
the data may have. Uncertain data can destroy a regular pattern but they cannot create one. This is
analogous to any situation where one works with very large samples of uncertain data, and the
problem actually pervades many fields of science. Consider the somewhat parallel situation in
clinical medicine wherein medical histories of large numbers of patients are used to infer the causes
of disease. For any given patient the information may contain large uncertainties and errors, yet
the analysis of the whole data set often shows useful and important patterns. Closer to home, most

6

PALAEONTOLOGY, VOLUME 30

normal biostratigraphic problems contain uncertain and conflicting data on fossil occurrences or
time ranges, yet the whole system, operating in consensus fashion, is remarkably effective in
defining Earth history. So, although the extinction data are sometimes very shaky, I suggest that
they can and should be used to search for broad patterns and trends in the history of life.

The periodicity question is closely related to the parallel question about the ages of impact craters
(Rampino and Stothers 1984; Alvarez and Muller 1984) and here, too, the range of opinion is
broad. Suffice it to say that the cratering data are not as robust as the extinction data.

In 1983 two papers were published claiming periodicities in reversals of the Earth’s magnetic field
compatible with the extinction and cratering records (Negi and Tiwari 1983; Mazaud et al. 1983).
My own contribution (Raup 1985a) came to the same conclusion but was challenged on statistical
grounds by Lutz (1985). Subsequently, however, two papers have appeared supporting the reversal
periodicity (Pal and Creer 1986; Stothers 1986). This is important because there is some tempting
evidence that large body impacts can cause magnetic reversals.

To summarize, the several claims about extinction are surrounded by uncertainty and controversy
but I submit that they represent legitimate and sound science. Competent workers can disagree on
some or all of them but there is no basis to discard the claims as unscientific. They are important
hypotheses in the process of testing.

EXTINCTION: CONTINUOUS OR EPISODIC

There are many deficiencies in our knowledge of the extinction record but one of the most striking
is our inability to say for sure whether extinction is best characterized as a continuous or episodic
process. Are all species at risk of extinction all the time, with that risk fluctuating up and down
through time, or does the risk of extinction alternate between high and low in a series of jumps? It
has been traditional to use the continuous model and by this view, mass extinctions are simply

TEXT-FIG. 3. Stratigraphic ranges of fifty brachiopod species near the K-T boundary at Nye Klov in Denmark.
(Prom Raup 1986; redrawn from Surlyk and Johansen 1984.)

RAUP: MASS EXTINCTION

7

intervals where the rates increase. On the other hand, the possibility of impact-caused extinction is
more compatible with the episodic alternative.

Text-fig. 3 shows the record of fifty brachiopod species across the K-T boundary in one of the
better Danish sections, based on work by Surlyk and Johansen (1984). The plot covers the brachio-
pods from about 5 m below the boundary to about 10 m above. The points at which collections
were made are shown on the left. A group of species appears to go extinct precisely at the K-T
boundary and several others drop out immediately below the boundary. The first question is
whether the ones that go out below the boundary actually lived to the end of the Cretaceous but
are missing in the boundary sample itself. This is important because it makes the difference between
progressive extinction over a metre of section and sudden extinction at the last instant of the
Cretaceous.

One could claim that the extinction is actually sharper than it looks because the rock record fails
to record the actual last occurrence of these species. This backward smearing of an extinction event
is known to all palaeontologists and has recently been called the Signor Lipps Effect (Raup 1986).
It is dilficult to deal with because it depends so much on negative evidence. The Signor-Lipps Effect
is a problem at all taxonomic and stratigraphic scales. At the family level, for example, the late
Permian extinctions are spread over most of the final 10 million years of the Permian. The sedimen-
tary record is poor and incomplete, however, and it is certainly possible that the true extinctions
are clustered at a single point near the end of the Permian. It is difficult to say where the truth lies.

This plot also illustrates another common problem, called the Lazarus Effect by David Jablonski
(in Elessa and Jablonski 1983). Notice that six species in the middle of the plot disappear at or near
the K-T boundary, only to reappear several metres up into the Danian. Obviously, these species
did not go extinct, assuming that their taxonomy is correct, but something happened to prevent
their preservation in this section. These species are called Lazarus taxa. They raise the possibility
that the apparently sharp episode of extinction at the K-T boundary could be an artifact of the same
environmental or taphonomic change that eliminated the six Lazarus taxa. By this interpretation the
extinctions may actually have taken place over a considerable time in the early Danian. The Lazarus
Effect is also found at other scales. The absence of several important groups from the Early Triassic
is an example. They are known to have survived the Permian extinctions but are not found again
until later in the Triassic.

The problem comes down to whether we trust the fossil record as a complete record or not. As
such, it is reminiscent of some of the arguments about gaps in the record that have surrounded the
punctuated equilibrium question. There is some hope of controlling the problem. Jablonski (1986)
has suggested that the record of Lazarus taxa can be used to evaluate the Signor-Lipps effect in a
given case. That is, if there are several Lazarus taxa, as is the case here, we should be especially
cautious about taking the record of last occurrences literally. The whole problem should be tackled
with as much rigour as possible.

Let us consider the problem of continuous versus episodic extinction in a different way. Text-fig.
4a shows a cohort survivorship curve for species of planktonic foraminifera. The data come from a
recent paper by Hoffman and Kitchell (1984). The horizontal scale is millions of years before present
starting at the base of the Oligocene on the left. We start with the list of foraminiferid species
present in the record at a point 30 ma ago and call this list a ‘cohort’. This is plotted as 100 %. We
then monitor the decay of this cohort by extinction of its constituent species through the remainder
of the Cenozoic. The per cent of species remaining is plotted at approximately 5-ma intervals. By
the end of the Cenozoic, the cohort is almost gone. Origination of new species during this time is
ignored merely because we are interested only in extinction.

The fit of a straight line to the seven points is excellent and suggests a continuous process of
extinction, with the species always being at risk. The straight line on the semi-logarithmic plot is
analogous to radioactive decay, with a ‘decay constant’— the slope of the line— indicating a species
half-life of 5-3 ma. This is an excellent example of Van Valen’s Law of constant extinction (Van
Valen 1973; Raup 1975). This argues strongly for extinction as a continuous process, but when we
add more data and plot them differently, an entirely different picture emerges.

PALAEONTOLOGY, VOLUME 30

38 30 20 10 0

10^ YEARS BEFORE PRESENT

A

10® YEARS BEFORE PRESENT

B

TEXT-FIG. 4. A, survivorship pattern for a polycohort of planktonic foraminiferal species from 30 ma bp
(Early Oligocene) to the present, sampled at intervals of approximately 5 ma. The fitted straight line implies
continuous, background extinction following Van Valen’s Law. (From Raup 1986; data from Hoffman and
Kitchell 1984.) b, survivorship patterns for a set of polycohorts of planktonic foraminiferal species from 30
ma BP to the present. Sampling points are shown on the inside of the lower, horizontal axis. The data points
for survivorship have been connected by straight-line segments. This method of plotting emphasizes the
episodic pattern of extinction in contrast to the method used in a. (From Raup 1986; redrawn from Hoffman

and Kitchell 1984.)

Text-fig. 4b is also from Hoffman and Kitchell data but differs in three ways. First, more sampling
points are used and these are indicated inside the horizontal axis. Secondly, more cohorts are
followed, each differing in starting time. And thirdly, the points in the decay of each cohort are
connected by straight line segments. Here, we see a sort of stair-step pattern. The horizontal parts
(the treads) are times of no extinction and the steep portions (the risers) are times of simultaneous
extinction of several species. There is an especially pronounced extinction event at about 12 ma
BP where the risers are steep and high, the Middle Miocene event discussed above. The overall
pattern suggests an episodic alternation between virtually no extinction and substantial extinction.
By this interpretation the foraminifera experienced long periods of safety punctuated by short
intervals of the proverbial panic.

The same exercise can be carried out at a larger scale. Text-fig. 5 shows a nest of cohort curves
for about 2300 fossil marine families for the whole Phanerozoic (data from Sepkoski 1982). The
biggest mass extinctions show as cliffs or scarps in this diagram. Especially striking are those at the
ends of the Permian and Cretaceous but the graph shows many smaller events. It was this diagram
that prompted Sepkoski and me to test the Mesozoic-Cenozoic record for periodicity of extinction.
The Palaeozoic pattern is either less episodic or blurred by poor taxonomic and stratigraphic
resolution.

The question of episodic versus continuous extinction is fundamental to questions of the causes
of extinction. Are we dealing with progressive deterioration of environments, as is implied by most

1 00

RAUP: MASS EXTINCTION

9

10‘ YEARS BEFORE PRESENT

TEXT-FIG. 5. Survivorship patterns for Phanerozoic polycohorts of marine families plotted in the manner of text-fig. 4b. Major extinction events appear
as ‘cliffs’ and the intervals of negligible extinction appear as horizontal segments. (From Raup 1986; data from Sepkoski 1982.)

10

PALAEONTOLOGY, VOLUME 30

hypotheses of extinction based on sea-level or climatic change, or are we dealing with sudden,
isolated events of environmental stress? The answer to this question does not necessarily involve
extraterrestrial forces because the Earth is quite capable of producing sudden environmental
shock by itself, but it is clear that extraterrestrial hypotheses are favoured by the episodic view of
extinction.

I have no great confidence in any simple generalization on this question. In our present state of
knowledge, it is too much in the eye of the beholder, but the problem represents a most important
area for future research.

EVOLUTION IN A COSMIC ENVIRONMENT

Palaeontologists have usually adopted a defensive stance regarding proposals that cosmic factors
may have influenced the history of life. When a proposal is made, the reaction is usually to look for
flaws. The proposal is, in effect, ‘guilty until proven innocent’. This is probably good practice in
any field of science faced with radically new hypotheses, but it may mean that we miss some
important opportunities for progress.

Let me propose, at least as a thought experiment, that we turn the situation around. Let us look
at our cosmic environment to see what is going on that might be important to life on Earth and
might leave a signature in the fossil record. With this approach, it may be possible to make specific
predictions which can be tested in the fossil record.

The Solar System, and the Milky Way Galaxy of which it is a part, are dynamic and ever-
changing. The Earth’s orbit around the Sun changes over geologic time. The Sun’s heat output
probably changes substantially, with most estimates calling for as much as a 30 % increase since the
early Pre-Cambrian. The Sun and other stars are constantly moving through space. We encounter
interstellar clouds of gas and dust and we pass through the dense plane of the Galaxy every 30-33
ma. Other stars pass close to us on a random basis and these close encounters may well upset the
orbits of Solar System comets. Nearby stars may explode producing supernovae close enough to
Earth to produce environmental effects— and so on. There is a lot going on out there and some of
this undoubtedly impinges on life on Earth. It would be surprising if it did not.

The time scales of these phenomena are often more appropriate to geologic data than to astro-
nomical observation. The Phanerozoic has seen two complete rotations of the Galaxy, or two
galactic years. At a smaller scale, the orbital changes in the Earth-Moon-Sun system produces
cycles of between 20 000 and 400 000 years. And these shorter cycles have been spectacularly well
recognized in recent work on Milankovitch Cycles in relation to ice ages over the past 700 000 years
(Hays et al. 1976; Imbrie and Imbrie 1980).

Astronomers are surprisingly uncertain about most of the processes that operate on geologic
time scales. An astrophysicist told me recently that if extinctions can be related to our position in
the Galaxy, palaeontology will have provided the best available information on critical aspects of
the form and dynamics of the Galaxy. In other words astronomy, that most sophisticated of all the
so-called hard sciences, will welcome our help.

To be a bit more specific, we probably know most about large body impacts — whether or not
their link to extinction can be confirmed. We have data from about 100 impact craters on Earth
(Grieve 1982) as well as from those on the Moon, planets, and their satellites. Also, many present-
day asteroids with Earth-crossing orbits have been sighted and their orbits plotted. Our knowledge
of the comet population is not as good but it is growing.

From a palaeontological viewpoint, there are two main questions. What are the probable effects
of an impact of a given size, and how many impacts have there been in the Phanerozoic? On the
first question, we must rely on numerical simulations and on extrapolations from laboratory
experiments. A great deal of this work has been done since the 1980 Alvarez paper and there is a
reasonable consensus that impact by an object 10 km in diameter would have devastating, global
effects— either through choking the atmosphere with dust or by nitrogen oxide pollution or by other

RAUP: MASS EXTINCTION

II

means. A 10 km object falling in the ocean, which has an average depth of only 5 km, would be
like throwing a brick into a mud puddle. A 1 km object would also have large effects but there is
no consensus on their magnitude. Most of the attention has been on the 10 km size because this
is the size estimated from the iridium concentrations at the K-T boundary. The question of size is
important because impacts by 1 km objects have been several times more common in the Phanero-
zoic than those by 10 km objects (Shoemaker 1984). Are we talking just about the five or six really
major mass extinctions or about extinction events at the level of the stratigraphic series, stage, or
zone? This is not yet clear.

In this context, a recent analysis of two Jurassic extinctions (Pliensbachian and Tithonian) by
Hallam (1986) is of interest. Hallam presents strong evidence that these events were regional rather
than global and argues that this indicates that causes other than comet or asteroid impact were
operating. An alternative explanation is that these events resulted from the purely regional effects
of relatively small impacting objects. Lacking positive evidence for impacts at these times, we
cannot, of course, claim that they were caused by small impacts. My only point is that impacts
need not produce global environmental effects and that small impacting objects are known to be
more frequent than large ones. Although the Pliensbachian and Tithonian events were apparently
only regional extinctions, they were sufficient to produce statistically significant peaks in global
compilations of taxonomic data (Raup and Sepkoski 1986).

I have emphasized large body impacts and extinction but this is only part of the story. Our cosmic
environment has many other possibilities for influencing life. The success of palaeontology with
Milankovitch Cycles is an example. These cycles of orbital change have been known since Ptolemy,
but their application to the geologic past was pure speculation until the work on Pleistocene climate
and palaeoecology by Imbrie, Shackleton, and their co-workers. No significant extinctions are
involved here but the results are nevertheless palaeontologically important.

Another example not involving extinction is the use of growth banding in fossils to deduce day
length and the history of the Earth-Moon system (Wells 1963; Scrutton 1965). Although contro-
versial in some aspects, this is a clear case of palaeontology being able to test predictions of
astronomy in order to say something interesting about the history of the Solar System.

To summarize, there is every reason to think that processes and events in space have had
influences on the ecology and evolution of life and that a more positive approach should yield
results.

CONCLUSION

I have covered a wide range of extinction related topics. I hope that I have demonstrated that the
extinction mania of the past five years stems from valid research questions and that palaeontologists
can play a constructive and positive role. Most needed are detailed field studies to describe and
define the extinction sequences as precisely as possible. These studies should be co-ordinated with
cognate research in geochemistry, geophysics, and astronomy.

Acknowledgements. Some of the research described here was supported by Grant NAG-237 of the National
Aeronautics and Space Administration (USA). I thank A. Hallam and an anonymous reviewer for helpful
criticism of the manuscript. Figures from Raup and Sepkoski 1986 and Raup 1986 are Copyright 1986 by the
AAAS and are reproduced with permission.

REFERENCES

ALVAREZ, L. w., ALVAREZ, w., ASARO, E. and MICHEL, H. V. 1980. Extraterrestrial cause for the Cretaceous-
Tertiary Extinction. Science (IVas/i.), 208, 1095-1 108.

ALVAREZ, w., ALVAREZ, L. w., ASARO, F. and MICHEL, H. V. 1982. Iridium anomaly approximately synchronous
with terminal Eocene extinctions. Ibid. 216, 886-888.

12 PALAEONTOLOGY, VOLUME 30

ALVAREZ, w. and MULLER, R. A. 1984. Evidence from crater ages for periodic impacts on the earth. Nature,
308, 718-720.

BOHOR, B. F., FOORD, E. E., MODRESKi, p. J. and TRIPLEHORN, D. 1984. Mineralogic evidence for an impact event
at the Cretaceous -Tertiary boundary. Science {Wash.), 224, 867-869.

BROCHWicz-LEwiNSKi, w., GASiEWicz, A., suFFCZYNSKi, s., szATKOWSKi, K. and ZBiK, M. 1984. Lacunes et
condensations a la limite Jurassique moyen-superieur dans le Sud de la Pologne: manifestation d’un phen-
omene mondial? C.R. Acad. Sci. Paris, 299, 1359-1362.

FLESSA, K. w. and JABLONSKi, D. 1983. Extinction is here to stay. Paleobiology, 9, 315-321.

GANAPATHY, R. 1982. Evidence for a major meteorite impact on the Earth 34 million years ago: implications
for Eocene extinctions. Science (Wash.), 216, 885-886.

GRIEVE, R. A. F. 1982. The record of impact on Earth; implications for a major Cretaceous/Tertiary impact
event. In silver, l. t. and schultz, p. h. (eds.). Geological Implications of Impacts of Large Asteroids and
Comets on the Earth. Geol. Soc. Amer., Spec. Paper, 190, 25-38.

HALLAM, A. 1986. The Pliensbachian and Tithonian extinction events. Nature, 319, 765-768.

HAYES, J. D., imbrie, j. and SHACKLETON, N. J. 1976. Variations in the earth’s orbit: pacemaker of the ice ages.
Science ( Wash.), 194, 1121-1132.

HOFFMAN, A. and KiTCHELL, J. A. 1984. Evolution in the pelagic planktic system: a paleobiologic test of models
of multispecies evolution. Paleobiology, 10, 9-33.

HSU, K. J. 1986. Environmental changes in times of biotic crisis. In raup, d. m. and jablonski, d. (eds.).
Patterns and Processes in the History of Life, pp. 297-312. Springer-Verlag, Berlin.

imbrie, j. and imbrie, j. z. 1980. Modelling the climatic response to orbital variations. Science {Wash.), 207,
943-953.

JABLONSKI, D. 1986. Causes and consequences of mass extinctions; a comparative approach. In elliott, d.
(ed.). Dynamics of Extinction, pp. 183-229. John Wiley & Sons, New York.

LUCK, J. M. and turekian, k. k. 1983. Osmium- 187/Osmium- 186 in manganese nodules and the Cretaceous-
Tertiary boundary. Ibid. 222, 613-615.

lutz, t. m. 1985. The magnetic reversal record is not periodic. Nature, 317, 404 407.

MAZAUD, A., LAJ, c., de SEZE, L. and VEROSUB, K. L. 1983. 15-Myr periodicity in the frequency of geomagnetic
reversals since 100 Myr. Nature, 304, 328-330.

MONTANARI, A., HAY, R. L., ALVAREZ, W., ASARO, F., MICHEL, H. V., ALVAREZ, L. W. and SMIT, J. 1983. SpheroidS
at the Cretaceous-Tertiary boundary are altered impact droplets of basaltic composition. Geology, 11, 668-
671.

NEGi, J. G. and TiWARi, R. K. 1983. Matching long term periodicities of geomagnetic reversals and galactic
motions of the solar system. Geophys. Res. Lett. 10, 713-716.

OFFICER, c. B. and DRAKE, c. L. 1983. The Cretaceous-Tertiary transition. Science {Wash.), 219, 1383-1390.

1985. Terminal Cretaceous environmental events. Ibid. 227, 1 161-1167.

PAL, p. c. and CREER, K. M. 1986. Geomagnetic reversal spurts and episodes of extraterrestrial catastrophism.
Nature, 320, 148-150.

PLAYFORD, p. E., MCLAREN, D. J., ORTH, c. J., GILMORE, J. s. and GOODFELLOW, w. D. 1984. Iridium anomaly in
the Upper Devonian of the Canning Basin, Western Australia. Ibid. 226, 437-439.

RAMPINO, M. R. and STOTHERS, R. B. 1984. Terrestrial mass extinctions, cometary impacts and the Sun’s motion
perpendicular to the galactic plane. Nature, 308, 709-712.

RAUP, D. M. 1975. Taxonomic survivorship curves and Van Valen’s Law. Paleobiology, 1, 82-96.

1985a. Magnetic reversals and mass extinctions. Nature, 314, 341-343.

19857). Rise and fall of periodicity. Ibid. 317, 304 305.

1986. Biological extinction in Earth history. Science {Wash.), 231, 1528-1533.

and SEPKOSKi, j. j., jr. 1984. Periodicity of extinctions in the geologic past. Proc. Nat. Acad. Sci. {USA),

81,801-805.

1986. Periodic extinction of families and genera. Science {Wash.), 231, 833-836.

SCRUTTON, c. T. 1965. Periodicity in Devonian coral growth. Palaeontology, 7, 552-558.

SEPKOSKI, J. J., jr. 1982. A compendium of fossil marine families. Contr. Milwaukee Pub. Mus. 51, 1-125.

and RAUP, D. M. 1986. Periodicity in marine extinction events. In elliott, d. (ed.). Dynamics of Extinction,

pp. 3-36. John Wiley & Sons, New York.

SHOEMAKER, E. M. 1984. Large body impacts through geologic time. In Holland, h. d. and trendall, a. f.
(eds.). Patterns of Change in Earth Evolution, pp. 15-40. Springer-Verlag, Berlin.

SMIT, J. and KLAVER, G. 1981. Sanidine spherules at the Cretaceous-Tertiary boundary indicate a large impact
event. Nature, 292, 47-49.

RAUP; MASS EXTINCTION

13

STOTHERS, R. B. 1986. Periodicity of the Earth’s magnetic reversals. Nature, 322, 444-446.

SUN, Y.-Y., CHAI, Z., MA, S., MAO, X., XU, D., ZHANG, Q., YANG, Z., SHENG, J., CHEN, C., RUl, L., LIANG, X., ZHAO, J.
and HE, j. 1984. The discovery of iridium anomaly in the Permian-Triassic boundary clay in Changxing,
Zhejiang, China and its significance. Acad. Sinica Contr. to 27th Int. Geol. Congr., Moscow, 1984, pp. 235-
245.

SURLYK, F. and johansen, m. b. 1984. End-Cretaceous brachiopod extinctions in the Chalk of Denmark.
Science ( Wash.), 223, 1174-1177.

UREY, H. c. 1973. Cometary collisions and geological periods. Nature, 242, 32-33.

VAN VALEN, L. 1973. A ncw evolutionary law. Evol. Theory, 1, 1-30.

WELLS, J. w. 1963. Coral growth and geochronometry. Nature, 197, 948-950.

Typescript received 19 March 1986
Revised typescript received 20 May 1986

DAVID M. RAUP

Department of Geophysical Sciences
University of Chicago
Chicago, Illinois 60637, USA

-m ^ ’ V "TB

'1

: .Hf •■.

— ii ,,

, 4:

■ *• •• ‘i '

■‘f>

• :-• jffc.ji^', "ff.i

'« I ‘

f'-

► , ■ V

<,

1 ■ ■ - 1 '■ ' 4J

IMmk

■■ t

'-■■-•• ■ ■/' i

TWO NEW SPECIMENS OF ANTHRACOSAU RUS
(AMPHIBIA: ANTHRACOSAURI A) FROM THE
NORTHUMBERLAND COAL MEASURES

by j. A. CLACK

Abstract. Two specimens from the Low Main Seam at Newsham, Northumberland are attributed to the
genus Anthracosaurus. The first is a skull table (Hancock Museum specimen G 13.78) previously attributed to
‘Eogyrinus' (= PhoUderpeton, Clack 1987). This skull table was used in the past to demonstrate the presence
of a ‘kinetic line' between the embolomere skull table and cheek. The holotype of A. nisselli lacks such a
kinetic line, but the apparent paradox can be resolved if the ‘kinetic line’ was not mobile in the later embolomere
families Eogyrinidae and Anthracosauridae, but acted as a butt joint to resist vertical forces.

The second specimen (Hancock Museum specimen G 24.39) is an isolated right jugal. It shows an unusually
excavated orbital margin, which leaves the bone very narrow below it. This unique character is confirmed as
present in the holotype of A. nisselli. Incorporation of this jugal into new reconstructions shows the skull to
be higher and narrower than formerly appreciated, and without any conspicuous ‘lateral flare’ in the suspen-
sorial region.

Characters of the skull table including dermal ornament are used to construct a cladogram of the embolo-
meres, including the isolated skull table pertaining to Pteroplax.

Anthracosaurus was a large carnivorous anthracosaur related to the embolomerous Eogyrini-
dae, found in the British Coal Measures of Westphalian A and B (Upper Carboniferous) age. It is
now known from several specimens. The holotype of A. nisselli Huxley (1863) consists of an almost
complete skull lacking the lower jaw, from the Airdrie or Mushet’s Blackband Ironstone. After
Huxley’s brief description, that of D. M. S. Watson in 1929 gave a fuller account of the specimen
in his review of Scottish Coal Measures Amphibia. The specimen was thoroughly prepared and
redescribed by Panchen (1977). A second partial skull consisting of the snout and well-preserved
anterior palate derives from the Top Busty seam of the Durham Coalfield, Lower Coal Measures
(communis zone, Westphalian A) (Panchen 1977), and a partial lower jaw was discovered in the
Hancock Museum, probably from the Low Main Seam at Newsham (Westphalian B). This was
described by Panchen in 1981. The skull is thus known in some detail.

The postcranial skeleton of Anthracosaurus remains a mystery. Until recently some vertebrae and
an interclavicle were attributed to the genus (Panchen 1977), but these must now be discounted.
The dorsal vertebrae ((BGS)GS 56580, GS 56581) from Airdrie are undiagnostic and may equally
well pertain to PhoUderpeton (= "Eogyrinus', Clack 1987) which also occurs at this locality. The
attribution of embolomerous vertebrae made by Panchen was based on the presence of occluded
notochordal pits in the centra. However, this character has been discovered in both PhoUderpeton
and Archeria (Clack 1987), and cannot be used as a taxonomically significant character because of
its wide distribution and sporadic occurrence. The interclavicle (RSM 1971.11.4) which probably
derives from Airdrie is clearly that of the eogyrinid PhoUderpeton, whose interclavicle is now
known from the holotype of P. scutigerum (Clack 1987). Thus there is no certain evidence that
Anthracosaurus was either fully embolomerous or long-bodied.

The genus is characterized by a number of features unique among anthracosaurs (autapomorphies)
which distinguish it from the contemporary eogyrinids and result in its being placed in a separate
family. These features are associated with its evident carnivorous adaptations and include consoli-
dation of the palate with elimination of the mid-line suture, extension of the pterygoids laterally to

[Palaeontology, Vol. 30, Part 1, 1987, pp. 15-26.|

© The Palaeontological Association

16

PALAEONTOLOGY, VOLUME 30

cover the palatines, massive dentition with reduced tooth count especially on the premaxilla, and a
lateral shelf on the surangular suggesting elaboration of the adductor musculature. Also unique is
the single large Meckelian fenestra, presumed to be derived from confluence of the two fenestrae
found in the eogyrinids. Two other characters are relevant to the present discussion. The holotype
specimen appears to lack the skull table-cheek kinetism seen in other embolomeres, having the
supratemporal and squamosal firmly sutured, and in dorsal view the suspensorium of the holotype
appears laterally flared.

The skull table now referred to Anthracosaurus (specimen (HM)G 13.78 in the Hancock Museum)
was first mentioned by Hancock and Atthey (1868) and later the underside was figured by Atthey
(1877) as Pteroplax coniuta. In the same paper, he also described a smaller skull table and included
it in the same species. The underside of (HM)G 13.78 was described in detail but the upper side
was mentioned only briefly. In 1926 Watson separated this skull table from the smaller one described
by the two earlier authors, and assigned it (as DMSW 35) to Eogyrinus attheyi, though noting
that the determination was ‘not quite certain’. He described the dorsal surface as ‘so much eroded
that the sculpture is completely destroyed’ and, like Atthey, described in detail and figured the
underside.

The specimen is notable for being one of two skull tables, the other being the holotype of
" Palaeogyrinus' (= Palaeoherpeton, Panchen 1970), which showed ‘a shallow groove . . . (on the
underside of the supratemporal) . . . which received the squamosal, that bone not being attached to
any bone of the table’. Thus Watson concluded that the skull table was kinetically articulated with
the cheek, a primitive condition deriving from the osteolepiform kinetism, and characteristic of
embolomeres. Romer (1963) chose the smaller of Hancock and Atthey’s two skull tables as the
lectotype of Pteroplax cornuta but did not distinguish between the genera Pteroplax and Eogyrinus.
In 1964, Panchen began his series of papers redescribing the British and American embolomeres
with a paper on ' Palaeogyrinus' and in it he accepted Watson’s separation of Hancock and Atthey’s
two skull tables, assigning the larger again to Eogyrinus. The smaller skull table constitutes the only
certain cranial material to pertain to Pteroplax, though Boyd (1980) has attributed postcranial
material. The upper surface of (HM)G 13.78 has not been described in any detail, having always
been dismissed as badly preserved or eroded. Conversely, the underside of other embolomere skull
tables have either been unavailable or not considered as having taxonomic value. My recent study
of Pholiderpeton (Clack 1987) has shown several taxonomically significant features of the underside
of the skull table. A description of the underside of the skull table of Pteroplax which the author
prepared was included in the analysis. Specimen (HM)G 13.78 has been re-examined in the light of
these.

The second specimen to be referred to Anthracosaurus is (HM)G 24.39, an isolated jugal which
has not been described before, but which has been catalogued as this genus by the Hancock Museum
on my advice (Boyd and Turner 1980). Like the skull table, it originated from the Low Main Seam
at Newsham, Northumberland and is therefore also Westphalian B in age.

Abbreviations used for specimens; (HM)G, Hancock Museum, Newcastle upon Tyne; RSM,
Royal Scottish Museum, Edinburgh; (BGS)GS, British Geological Survey, Nottingham; (BMNH)R,
British Museum (Natural History).

SYSTEMATIC DESCRIPTION

Skull table

Specimen (HM)G 13.78 consists of an isolated skull table completely free of matrix. The bone is in
good condition and is robust and paler in colour than the majority of amphibian dermal bone
deriving from the Low Main Seam. The skull table has separated from the snout bones along the
prefrontal-postfrontal and frontal-nasal sutures, and from the cheek along the line usually described
as the lateral kinetism in embolomeres, that is along the lateral margins of the intertemporal,
supratemporal, and tabular bones. The dermal ornament is poorly developed by comparison with

CLACK: ANTHRACOSAURUS FROM NORTHUMBERLAND COAL MEASURES

17

TEXT-FIG. 1 (left). Skull table, (HM)G 13.78, in dorsal view. Low Main Seam, Newsham, Northumberland.
Dotted lines indicate extent of frontals and tabular horns in Dinning’s figure (Atthey, 1877). Scale bar:

10 mm.

TEXT-FIG. 2 (right). Skull table, (HM)G 13.78, in ventral view. Low Main Seam, Newsham, Northumberland.

Scale bar: 10 mm.

the skull tables of eogyrinid embolomeres. In particular the comparison with ^ Eogyritnis' ( =
Pholiderpeton, Clack 1987) forces the description of this ornament as ‘eroded’. However, the
ornament of the holotype of A. russe/li is described by Panchen (1977, p. 454) as ‘very little
developed’ and ‘reduced’. Direct comparison of the two specimens shows them to be extremely
similar. The ‘eroded’ appearance of (HM)G 13.78 is entirely consistent with a specimen pertaining
to Anthracosaurus and does not require the surface to be regarded as damaged, as attribution
to Pholiderpeton has demanded. Ornament type varies between embolomere families more
strongly than was previously recognized (Clack 1987). For example, comparison of ornament
on the parietals alone reveals the following: in Pteroplax, the well-defined pits are discrete and
approximately uniform in size, elongated even near the parietal foramen; in both Archeria and
Proterogyrinus the pits are discrete but elongated only towards the bone margins; in eogyrinids,
however, the pits vary in size and coalesce with one another in an irregular manner, though they
are well defined. They are elongated only towards the lateral margins of the parietals. In

TEXT-FIG. 3. Comparison of three embolomere skull tables showing positions at which measurements in
Table 1 were taken. Left, Pholiderpeton attheyi (holotype); centre, (HM)G 13.78; right, Anthracosaurus russelli

(holotype).

Anthracosaurus the pits are poorly developed and of low profile, such that no definable pattern is
evident. This is one of the features used here to support attribution of (HM)G 13.78 to Anthraco-
saurus.

The type of ornament is the most immediately compelling character supporting attribution
of (HM)G 13.78 to Anthracosaurus, but on examination a number of other similarities become
apparent, particularly with respect to skull table proportions.

The frontals of (HM)G 13.78 are parallel-sided as far as they are preserved, a condition seen in
A. russelli, but not in Pholiderpeton, where the lateral margins diverge anteriorly. The anterior
margin of the frontals in (HM)G 13.78 exhibit marked longitudinal grooves consistent with a
surface for the deeply fimbriated suture between nasals and frontals seen in A. russelli.

Atthey (1877, p. 373) described the frontals as ‘worn’ anteriorly. Comparison of his figure with
the skull table in its present state shows that about 7 mm have been lost from the frontals since
that date. However, even allowing for this and adding a further 3 mm for post-mortem damage, a
comparison of the length of the frontals with the parietal foramen-postparietal length and also
with the width of the skull table shows both A. russelli and (HM)G 13.78 to be similar to each
other. They differ from P. attheyi in having relatively short frontals (Table 1 and text-fig. 3).

A. russelli possesses remarkably short postfrontals when compared with P. attheyi, whether the
comparison is made in terms of postparietal parietal foramen length or in terms of maximum skull
table width. (HM)G 13.78 also shows short postfrontals, closer in proportion to those of A. russelli
than to P. attheyi (Table 1 and text-fig. 3). Consequent upon the shortness of the postfrontals, the
prefrontal forms a substantial portion of the dorsal margin of the orbit. In Pholiderpeton, this bone
contributes dorsally only to the anterior corner of the orbit margin.

Clack (1987) used parietal proportions (length (f): width (g)) to distinguish between certain
embolomere skull tables, for example those of eogyrinids, archeriids, and Pteroplax. In these terms,
skull table (HM)G 13.78 again falls closer to the holotype of A. russelli than to Pholiderpeton

CLACK: ANTHRACOSAURUS FROM NORTHUMBERLAND COAL MEASURES

19

(Table 1 and text-fig. 3). The parietals of (HM)G 13.78 are noticeably broader than long, giving a
characteristically short skull table. In 1977 Panchen noted that the long suspensorium of A. nisselli
was a consequence of the length behind the skull table and that the orbit quadrate distance was
not very different from that of ‘ Eogyrinus\ This must be in part a result of the shortness of the
skull table itself in Anthracosaurus. Another consequence of the short skull table in both A. russelli
and (HM)G 13.78 is the relative position of the parietal foramen. In both these specimens, it is
situated only a little behind the posterior margin of the orbits, whereas in Pholiderpeton it lies some
way behind the orbit margin (text-fig. 3).

The lateral margins of (HM)G 13.78 show a rough emargination at the region of the supratem-
poral-intertemporal suture. In the holotype of A. russelli a similar irregularity appears in the skull
table margin and is occupied by a dorsal growth of the postorbital. This is one of the pieces of
evidence which Panchen (1977) used to support the suggestion that the ‘lateral kinetism’ of embolo-
meres was no longer present in Anthracosaurus, but was sealed by an interdigitating suture. The
lateral margins of the intertemporal in (HM)G 13.78 show pits and sculpture on the underside, also
consistent with a conventional suture having existed between this bone and the postorbital in the
intact skull.

The postparietals and posterior parts of the parietals of (HM)G 13.78 show distinct depressions
about the mid-line. Less regular depressions occur in a similar pjace in the skull table of the holotype
of A. russelli. As these features occur over the thickest bone of the skull table it seems unlikely that
they have resulted from post-mortem compression. In Pteroplax and Proterogyrinus (Holmes 1984)
the parietal foramen appears on a mid-line ridge bounded by longitudinal furrows and such irregu-
larities of the dorsal surface of the table have been used as taxonomic characters in these two
genera (Holmes 1984; Clack 1987). It is probable that the depressions seen in the skull table of
Anthracosaurus are also characteristic.

Both tabular horns are now incomplete in (HM)G 13.78, though Atthey (1877) figured the left as
complete and the right as almost complete. The reader is referred to Dinning’s excellent plate in
Atthey’s paper. However, the specimen has been refigured here as Atthey misinterpreted some of
its features (e.g. left tabular buttress interpreted as the quadrate), with consequent inaccuracies in
the figure. Atthey showed the horns to possess a lower blade similar to those of other embolomeres,
bearing faint longitudinal striations as in Pholiderpeton scutigerum, Pteroplax cornuta, and Palaeo-
herpeton decorum. They are, however, less expanded posteriorly than in those genera. The lower
blades in the holotype of A. russelli are both incomplete, but as far as preserved resemble those of
other embolomeres. Panchen (1977) restored them in the oblique ventrolateral orientation which
they possess in the preserved specimen, which is rather different from the almost horizontal orien-
tation restored in Pholiderpeton by the present author. A characteristic of embolomere tabular
horns is that they are actually or incipiently biramous (Holmes 1984; Clack 1987), with an upper
component consisting of a boss or process separated from a lower component by a pit or notch. In
anthracosaurids, eogyrinids, and archeriids, the lower component is blade-like (Clack 1987). In A.
russelli the upper process has a substantial base and has been restored by Panchen (1977) as a
process orientated ventromesially. Specimen (HM)G 13.78 exhibits only a small boss at this point,
which though separated from the lower blade base by a pit, shows no evidence that it was ever the
substantial process seen in A. russelli and for this reason (HM)G 13.78 is referred only to Anthraco-
saurus sp.

The underside of the skull table shows two characters by which it may be distinguished from that
of an eogyrinid. In all the eogyrinids examined the ventral surface of the parietals shows smooth
bone bearing conspicuous ridges radiating away from the region of the parietal foramen to give
this region of the skull table a fluted appearance (Clack 1987). (HM)G 13.78 shows this feature in
only a poorly developed form and the bone surface is roughened. Strong development of the
fluting has been considered by Clack as a character distinguishing eogyrinids and it is found in
Pholiderpeton, Palaeoherpeton, and Neopteroplax.

In eogyrinids the tabular exhibits a pair of mesially directed facets for attachment of the opistho-
tics. These lie at the ends of well-developed buttresses separated by a deep groove. In (HM)G 13.78,

20

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 4. Isolated right jugal, (HM)G 24.39, in external view. Low Main Seam,
Newsham, Northumberland. Scale bar: 10 mm.

although the tabular bears facets in a similar position to those of eogyrinids, they are not readily
distinguishable as two separate facets, nor are they separated by a noticeable groove as they are in
eogyrinids. This could be interpreted as a result of poor preservation, but in Archeria for example,
the form of the facets is consistent throughout a number of skull tables, despite differential preser-
vation (Clack 1987).

Jugal

The isolated jugal (HM)G 24.39 is complete anteriorly but has part of its posterior border missing.
Part of the posterodorsal border has also been damaged, though probably little is missing. It is
completely exposed over the external surface, though some of the internal surface is still embedded
in matrix.

The dermal ornament is of the same ‘eroded’ nature seen in the skull table referred to Anthraco-
saurus above and that seen in the holotype of A. russelli. The jugal is characterized by a deeply
excavated orbital margin, rendering the jugal very narrow below it. As with the skull table, direct
comparison of the isolated example with the holotype skull reveals an identical shape in the two
jugals, though (HM)G 24.39 is from a slightly smaller individual. The unusual shape of the jugal
does not emerge from Panchen’s reconstruction, because of the foreshortening effect on the sloping
cheek of both dorsal and lateral views. However, the characteristic shape can be seen in the
photograph (Panchen 1977, fig. 1).

The jugal (HM)G 24.39 is damaged along its posterior border at the point where the ‘temporal
fenestra’ occurs in A. russelli. In the holotype the bone here is extremely thin, so such damage in
the isolated specimen is not surprising and is consistent with the presence of a fenestra.

The only segment of lateral line canal to be found in A. russelli occurs along the posteroventral
border of the jugal, parallel with the lower margin. It is shallow but clearly defined. What may be
the anterior end of a lateral line canal is present in (HM)G 24.39 in the corresponding position. In
A. russelli the jugal does not form any part of the jaw margin. The holotype of Pholiderpeton
seutigerum shows that this was also the case in this genus, and a broad flat surface for suture with
the maxilla can be seen along the ventral margin of the bone. In (HM)G 24.39 the ventral margin is
likewise broad and flat, suggesting that this jugal formed no part of the jaw margin, even allowing
for the loss of the posterior part of the bone.

CLACK: ANTHRACOSAU RUS FROM NORTHUMBERLAND COAL MEASURES

21

TABLE 1. Comparison of measurements of three embolo-
mere skull tables, (HM)G 13.78, Anthracosaurus rus-
selli (holotype), and Plwliderpelon attheyi (lectotype).

Measurements
in mm

Specimens

G 13.78

A. russelli

P. attheyi

fr

76

96

106

pp-parfor

43

46

48

sktw

97

120

90

pofr

38

40

52

io

42

62

44

f

50

62

72

g

69

88

62

Ratios

fr : pp-parfor

1-76

208

2-21

fr : sktw

0-78

0-80

117

pofr : sktw

0-39

0-33

0-58

pofr : pp-parfor

0-88

104

108

io : pp-parfor

0-97

1-35

0-92

io : sktw

0-43

0-52

0-48

sktw : pp-parfor

0-44

0-38

0-53

f/g%

72

70

116

Key: fr, frontal length; pp parfor, postparietal parietal for-
amen length; pofr, postfrontal length; sktw, maximum skull
table width; io, minimum interorbital width; f, parietal length;
g, parietal width.

DISCUSSION

The characters on which skull table (HM)G 13.78 may be confidently attributed to Anthracosaurus
can be summarized as follows:

1. Ornament poorly developed— ‘eroded’ in appearance.

2. Short frontals with deeply fimbriated suture for nasals; parallel-sided (diverging anteriorly in
eogyrinids).

3. Short postfrontals.

4. Parietals noticeably broader than long.

5. The two latter characters contribute to a short skull table with parietal foramen only just
behind the level of posterior orbit margin; also prefrontal contributing about 50 % to dorsal orbit
margin.

The following characters exclude (HM)G 13.78 from the Eogyrinidae:

1 . Fluting on underside of parietals poorly developed.

2. Tabular facets almost confluent, not conspicuously double nor separated by a deep groove.

Text-fig. 3 shows outline drawings of the three skull tables under consideration, (HM)G 13.78,
the holotype of A. russelli, and the lectotype of P. attheyi, and shows positions at which measurements
were taken. Table 1 sets out the measurements and proportions derived from these. In all the
proportions examined, except that of the relative interorbital width, (HM)G 13.78 is seen to fall
closer to A. russelli than to Pholiderpeton. The differences which exist between (HM)G 13.78 and
Pholiderpeton may reflect individual variation within one species, and the similarities to A. russelli
no more than coincidence. However, a study of eleven skull tables of Archeria crassidisca (Clack

22

PALAEONTOLOGY, VOLUME 30

1987) shows a remarkable consistency within the members of one species of embolomere. The
resemblances between (HM)G 13.78 and Anthracosaurus russelli are therefore inferred to be signifi-
cant. The differences which do exist, for example in relative interorbital width, could be accounted
for by the fact that (HM)G 13.78 derives from a smaller (?younger) individual.

Reference of skull table (HM)G 13.78 to Anthracosaurus has interesting consequences. This is
one of the two skull tables which Watson (1926) used to establish the presence of a ‘lateral kinetism’
in embolomeres. The under-surface of the supratemporal is smooth throughout and concave except
for the extreme lateral edge which is slightly curled under. Watson interpreted this to mean that the
supratemporal was not securely sutured to the squamosal but that the squamosal facet sat beneath
the supratemporal concavity, with the gap filled by connective tissue. He interpreted ' Palaeogyrinus'
(= Palaeoherpeton, Panchen 1970) in the same way, and embolomeres have been regarded as
‘kinetic’ by most authorities ever since. The kinetism has been perceived as a remnant of the
rhipidistian kinetic mechanism.

If (HM)G 13.78 is correctly assigned to Anthracosaurus, a paradox results in that in the holotype
of A. russelli the kinetism has apparently been eliminated by suturing. The paradox may be resolved
by consideration of the form and possible function of the junction in eogyrinids and anthracosaurids.
In my recent study of Pholiderpeton (Clack 1987) it was shown that the lateral supratemporal-
tabular margin in this genus was, contrary to previous assessments, convex and pitted, exactly
matching the shape of the squamosal facet. The condition is seen in the holotype of P. scutigerum,
P. attheyi (specimen (BMNH)R 8426), and Palaeoherpeton. In these specimens at least, there could
have been no cartilaginous padding present between the skull table and cheek. In all these genera it is
evident that the suture between intertemporal and postorbital was a conventional suture, effectively
immobilizing the skull table-cheek junction. The form of the supratemporal-squamosal junction
with its broad pitted contact resembles that seen between lachrymal and maxilla; both occur in
areas of stress, and appear designed to resist and disperse vertical forces. In these families the
‘kinetic line’ appears to have been modified to become a strengthened butt joint. In both A. russelli
and apparently also in Neopteroplax (Romer 1963), the junction was securely sutured such that the
‘kinetic line’ was completely eliminated. The degree to which this consolidation occurred may have
varied with age or between species, thus no pitting is seen in (HM)G 13.78 at the lateral supratempo-
ral margin, but the intertemporal-postorbital suture was sealed. This condition is presumably
derived from that seen in Proterogyrinus (Holmes, 1984), Pteroplax, and Archeria in which pitting
on the supratemporal margin is also absent. In these non-eogyrinids the intertemporal-postorbital
suture is not sealed {Archeria, Holmes, pers. comm.), perhaps allowing some remnant movement.
Panchen (1977, p. 455) suggests that the sealed ‘kinetic line’ in Anthracosaurus russelli was possibly
a factor related to age and of no taxonomic significance, and in this context it is relevant that
specimen (HM)G 13.78 is from a smaller individual than the holotype specimen.

Examination of the jugal (HM)G 24.39 shows it to be notably tall posterior to the orbit. It is
thereby made clear that the posterodorsal border of the jugal of the holotype of A. russelli has been
damaged or pushed under the postorbital and postfrontal bones. It is also apparent that rather
more crushing has occurred along the lachrymal nasal suture than was allowed for by Panchen
(1977). Another striking feature to emerge from the new restoration of the jugal is the deeply
triangular orbits which Anthracosaurus must have possessed. This represents an additional autapo-
morphy, not obviously an adaptation to carnivory, by which to characterize the genus Anthracosau-
rus. Non-circular orbits are found in a number of Palaeozoic tetrapods, for example, Pholiderpeton
(Clack 1987), Carbonoherpeton (Klembara 1986), Eoherpeton (Panchen 1975); loxommatids
(Beaumont 1977), Crassigyrinus (Panchen 1985), and also apparently in one of the ichthyostegids
{Ichthyostega kochi, Save-Soderbergh 1932). No single explanation may be applicable to them all
and there have been several suggestions (position of a salt gland, muscle insertion, muscle-bulging
space) as to the function of the loxommatid ‘keyhole’ (see Beaumont 1977). Most recently, Bjerring
(1987, in press) suggested that the asymmetrical orbits of these forms also housed an electric organ
by which to locate or stun prey.

The completeness of the palate indicates that the width of the skull is essentially undistorted and

CLACK: ANTHRACOSAURUS FROM NORTHUMBERLAND COAL MEASURES

23

TEXT-FIG. 5. New reconstruction of skull of
Anthracosaurus russelli in dorsal view, from
holotype and (HM)G 24.39. Scale bar: 10
mm.

in dorsal view the anterior part of the skull would have been very little different from Panchen’s
restoration. However, it is now evident that the side of the face was rather deeper than he restored
it. In consequence of a deeper orbital region, the suspensorium must be restored as deeper and as
this region is almost undamaged in the holotype, the skull must also be made narrower at this
point. Thus the ‘lateral flare’ must be reduced as a result. I have attempted new reconstructions
incorporating these features (text-figs. 5 and 6). One further difference between the new reconstruc-
tions and those of Panchen (1977) lies in the interpretation of the quadrate region. Now that this
region is more completely known in the genus Pholiderpeton (Clack 1983, 1987), that of Anthracosau-
rus is seen to be essentially similar to that of the former genus.

This study shows that it is possible to characterize isolated embolomere skull tables more precisely
than in the past. Future discoveries may be identifiable at least as far as family or may, like
Pteroplax, be excluded from any currently defined family. Using characters of the skull table
alone, with the remaining anthracosaur families as the outgroup, it is possible to draw up a

24

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 6. New reconstruction of skull of Anthracosaurus russelli in lateral view, from holotype and (HM)G
24.39. Lower jaw redrawn from Panchen (1981). Scale bar: 10 mm.

cladogram of the embolomeres, and to include the little-known genus Pteroplax. The characters
are as follows:

1. Parietal ornament regular, pits discrete and elongated, especially towards the margins.

2. Tabular horn biramous, differentiated into upper and lower components separated by a
notch or pit.

3#. Tabular bearing double facets for reception of opisthotic.

4. Tabular horn with lower component short, distinctively narrow and finger-like.

<5. Parietal foramen on midline ridge bounded by grooves.)

6. Lower component of tabular horn blade-like.

7. Pits on underside of parietal and postparietal for reception of columella cranii and ?opisth-
otic.

8. Skull table and especially parietals markedly elongate.

<5. Parietal foramen on mid-line ridge bounded by grooves.)

9. Parietals form distinctive arrow-shape.

10. Loss of anterior tabular facet.

1 1 . Intertemporal-postorbital suture.

12. Parietal ornament irregular, pits vary in size, coalescing to form a groove and ridge system.

13. Marked fluting on underside of parietal.

14. Parietal ornament reduced to give ‘eroded’ appearance.

15. Parietals broader than long.

This character distribution supports the scheme of relationships proposed by Clack (1987) and may
be added to the cladogram presented there. The additional characters cited by Clack (1987) are
listed below, and the cladogram presented in text-fig. 7.

CLACK; ANTHRACOSAURUS FROM NORTHUMBERLAND COAL MEASURES

25

4, <5> Proterogyrinidae

16. Two large Meckelian fenestrae, or a derivation thereof.

1 7*. Surangular crest usually developed.

18!. Opisthotic with free lateral processes.

[19. Vomers tuskless.]

20*. Processus alaris of jugal contacts pterygoid.

2 1 $. Pleurocentra ossified as rings.

22$. Oblique glenoid of scapulocoracoid.

23$. Humerus with low degree of twist, condyle only slightly helical.

24. Expanded cleithrum.

25. Nasals excluded from naris by premaxillary-lachrymal suture.

26*. Lachrymal excluded from orbit by long prefrontal-jugal suture.

27. Skull elongate in snout region.

28. Numerous marginal teeth, about sixty.

29. Digitiform process of opisthotic.

30. Medially directed flange of exoccipital.

31. Massive dentition, reduced in number; two premaxillary teeth.

32. Single large Meckelian fenestra, produced from confluence of two.

33. Palate reinforced; pterygoids fused in mid-line; pterygoids cover palatines; pterygoids lack-
ing denticles.

34. Enlarged orbit with ventrally excavated jugal.

35. Lateral shelf on surangular.

26

PALAEONTOLOGY, VOLUME 30

Key:

* Implied reversal in Anthracosaurus.

# Implied reversal in Archeria\ also present in Crassigyrinus.

! Character unknown in Archeria.

< > Character developed in parallel within embolomeres?

[ ] Character distribution uncertain.

$ Unknown in Anthracosaurus.

Acknowledgements. My thanks are due to the staff of the Hancock Museum, Newcastle upon Tyne for
permission to borrow specimens in their care. This paper arose out of work for a doctoral thesis for which Dr
A. L. Panchen was my supervisor. I thank him for reading and commenting upon the manuscript, and for
valuable help and discussion in the past, as I do also Drs A. C. and A. R. Milner and Dr T. R. Smithson.

REFERENCES

ATTHEY, T. 1877. On Pteroplax cornuta (Hancock & Atthey). Ann. Mag. nat. Hist. 20, 369-377.

BEAUMONT, E. H. 1977. Cranial morphology of the Loxommatidae (Amphibia: Labyrinthodontia). Phil. Trans.
R. Soc. B280, 29-101.

BJERRiNG, H. c. 1987 (in press). Electrical tetrapods? Proc. Soc. Herpetol. Europ. 3rd OGM Prague.

BOYD, M. J. 1980. The axial skeleton of the Carboniferous amphibian Pteroplax cornutus. Palaeontology, 23,
273-285.

and TURNER, S. 1980. Catalogue of the Carboniferous amphibians in the Hancock Museum, Newcastle

upon Tyne. Trans, nat. Hist. Soc. North. 46, 1-24.

CLACK, j. A. 1983. The stapes of the Coal Measures embolomere Pholiderpeton scutigerum Huxley (Amphibia:
Anthracosauria) and otic evolution in early tetrapods. Zoo. J. Linn. Soc. 79, 121-148.

1987 (in press). Pholiderpeton scutigerum Huxley, an amphibian from the Yorkshire Coal Measures.

Phil. Trans. R. Soc. B.

HANCOCK, A. and ATTHEY, T. 1868. Notes on the remains of some reptiles and fishes from the shales of the
Northumberland coalfield. Ann. Mag. nat. Hist. 1, 266-278, 346-378.

HOLMES, R. 1984. The Carboniferous amphibian Proterogyrinus scheeli Romer, and the early evolution of
tetrapods. Phil. Trans. R. Soc. B 306, 431-527.

HUXLEY, T. H. 1863. Description of Anthracosaurus russelli, a new labyrinthodont from the Lanarkshire
coalfield. Quart. J. Geol. Soc. 19, 56-58.

KLEMBARA, J. 1986. A new embolomerous amphibian (Anthracosauria) from the Upper Carboniferous of
Florence, Nova Scotia. J. Vert. Paleont. 5, 293-302.

PANCHEN, A. L. 1964. The cranial anatomy of two Coal Measure anthracosaurs. Phil. Trans. R. Soc. B 242,
207-281.

1970. Anthracosauria. Handbuch der Paldoherpetologie, Teil 5a (O. Kuhn, ed.), 1-84. Fischer, Stuttgart.

1975. A new genus and species of anthracosaur amphibian from the Lower Carboniferous of Scotland

and the status of Pholidogaster pisciformis Huxley. Phil. Trans. R. Soc. B 269, 58 1 -640.

1977. On Anthracosaurus russelli Huxley (Amphibia: Labyrinthodontia) and the family Anthracosauri-

dae. Phil. Trans. R. Soc. B 279, 447-512.

1981. A jaw ramus of the Coal Measure amphibian Anthracosaurus from Northumberland. Palaeon-
tology, 24, 85-92.

1985. On the amphibian Crassigyrinus scoticus Watson from the Carboniferous of Scotland. Phil. Trans.

Soc. B 309, 505-568.

ROMER, A. s. 1963. The larger embolomerous amphibians of the American Carboniferous. Bidl. Mus. comp.
Zool. Harv. 128,415-454.

SAVE-SODERBERGH, G. 1932. Preliminary note on Devonian stegocephalians from East Greenland. Medd. om
Gron. 94, 1-10.

WATSON, D. M. s. 1926. Croonian Lecture — The evolution and origin of the Amphibia. Phil. Trans. R. Soc. B
214, 189-257.

1929. The Carboniferous Amphibia of Scotland. Palaont. Hung. 1, 219-252.

J. A. CLACK

Typescript received 12 December 1985 University Museum of Zoology

Revised typescript received 8 May 1986 Downing Street, Cambridge CB2 3EJ

BASAL TURONIAN AMMONITES FROM
WEST TEXAS

by W. J. KENNEDY, C. W. WRIGHT and J . M. HANCOCK

Abstract. A rich ammonite fauna is described from the basal Turonian Pseudaspidoceras flexuosum Zone of
west Texas. Taxa present include typically Boreal, Tethyan, and US Western Interior species and widely
occurring forms of the bivalve Mytiloides. This co-occurrence of key species from different faunal realms and
provinces provides a basis for correlation of the base of the Turonian stage. Species present are the ammonites
Quitmaniceras reaseri Powell, 1963, Kamerunoceras calverteme Powell, 1963, Pseudaspidoceras flexuosum
Powell, 1963, Mammites powelli sp. nov., Vascoceras proprium Reyment, 1954, Fagesia catinus (ManizW, 1822),
Neoptychites sp., Wrightoceras munieri {PQrvincpiihvQ, 1907), Thomasites adkinsi {Y^ummcA and Decker, 1954),
Allocrioceras dentonense Moreman, 1942, A. larvatum (Conrad, 1855), Sciponoceras sp., and Worthoceras
cf. vermiculus (Shumard, 1860). The Mytiloides present is referred to M. columbianus (Heinz, 1935), the
M. opalensis of authors {non Bose, 1923).

Elucidation of the faunal succession across the Cenomanian-Turonian boundary and inter-
regional correlation around the level of this boundary are among the more intransigent problems
of mid-Cretaceous biostratigraphy. The reasons for this are: (i) marked provincialism of ammonite
faunas at this time, with Tethyan ammonite faunas dominated by vascoceratids and Boreal ones
characterized by mammitids and euomphaloceratids, and only limited geographical intermingling
of key taxa; (ii) sequences in the classic areas of western Europe are attenuated and interrupted by
minor non-sequences, while many areas of the world are characterized, at this level, by condensed,
atypical facies and reduced diversity faunas of many groups as a result of an important oceanic
anoxic event (Schlanger et al. in press). It is, for instance, obvious that marked faunal turnover at
this stage boundary in the US Western Interior (Kauffman 1970, fig. 6), discussed at some length
by Kauffman et al. (1978, p. XXIII. 13) is a result of both environmental and biogeographic
complexities.

Hancock and Kennedy (1981) reviewed the indirect evidence for the view that the Lower Turonian
of Tethys was actually equivalent to the Upper Cenomanian of Boreal regions, while Hook and
Cobban (1981), Wright and Kennedy (1981), Amard et al. (1983), and Bengtson (1983) recorded
associations of Boreal Upper Cenomanian and Tethyan Lower Turonian ammonites, but only with
the description of the co-occurrence of Metoicoeeras geslinianum (d’Orbigny, 1850) and Vascoceras
cauvini Chudeau, 1909 in Israel (Lewy et al. 1984) has a precise link between these faunas been
documented. Because of these problems the Cenomanian-Turonian boundary has been drawn at
several different levels on the basis of different macrofossil and microfossil groups (see discussion
in Wright and Kennedy 1981; Birkelund et al. 1984; Kennedy, 1984, 1985). A number of workers
have recently proposed that the most acceptable macrofossil indicator for the base of the Turonian
is the appearance in quantity of the inoceramid bivalve Mytiloides Brongniart, 1822. In particular,
the appearance of the species M. opalensis (Bose, 1923) has been taken as a basal Turonian marker
(e.g. Seitz 1952, 1956; Troger 1967, 1978, 1981; Seibertz 1979; Kauffman et al. 1978 and references
therein). It is, therefore, somewhat unfortunate that M. opalensis (Bose, 1923) is a younger Turonian
species. The basal Turonian ‘M. opalensis' of authors is, following the unpublished work of
Dr W. A. Cobban, M. columbianus (Heinz, 1935). It occurs in abundance in west Texas, where it is
associated with a rich ammonite assemblage described below.

(Palaeontology, Vol. 30, Part 1, 1987 pp. 27-74, pis. 1-10.|

© The Palaeontological Association

28

PALAEONTOLOGY, VOLUME 30

Z

C^*

O

C^*

s

QC

o

u.

Z

o

QC

3

<

“3

<

z

<

z

<

o

z

HI

o

“3

o

(fi

Upper part of unit exposed only in
Mexico

Top of formation not seen in the Quitman
Mountains

Limestone concretions with woollgan
Zone ammonites described by Powell
(1963b) from Cannonball Hill, Chihuahua

Bed B with ammonites of the flexuosum
Zone described in this paper

dark fissile shale with scattered
unfossilferous limestones

Bed A

Tower flaggy unit
many desmoceratid ammonites

>

o

z

<

CO

cc

UJ

a.

a.

3

Z

3

LU

-I

<

X

CO

UJ

-I

a

Q

<

Q

ZD

QQ

TEXT-FIG. 1. Location map and generalized succession of the Cenomanian-Turonian in the southern Quitman
Mountains, Hudspeth County, south-west Texas and Sierra de la Cieneguilla, northern Chihuahua, Mexico.
Based on Powell (1963, 1965), Jones and Reaser (1970), and our own observations.

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

29

BASAL TURONIAN FAUNA IN WEST TEXAS

Elements of this fauna were noted by Adkins (1933) and Powell (1963) and were subsequently made
the basis of a Pseudaspidoceras flexuosum Zone by Powell (1965). At present this ammonite-
inoceramid fauna is known from a series of localities in the southern Quitman Mountains, Hudspeth
County, Texas and in the Sierra de la Cieneguilla, Chihuahua, northern Mexico (text-fig. 1). Powell’s
original material came from three localities in the Cenomanian-Turonian Ojinaga Formation (text-
fig. 1): Arroyo Alamos, near Dos Alamos, 3-25 km east-south-east of Cieneguilla, Municipio de
Guadalupe Bravo in northern Chihuahua; Kelsey’s Crossing, 1 500 m south of the crossing on the
Rio Grande on the west side of the Kelsey Ranch, Chihuahua, and scattered outcrops on either
side of Goat Canyon Arroyo, 3-25 km west of the Kelsey ranch house, Hudspeth County, Texas.
The same fauna occurs abundantly on the east side of Calvert Canyon, from its intersection with
the Rio Grande (Rio Bravo) for half a kilometre northwards, 3 km north-west of Love Triangulation
Station, where we have hundreds of ammonites and inoceramids from a single bed of brown-
weathering black concretions in brown-white weathering black shales.

At Calvert Canyon, the source of the bulk of the material described here, the fauna is limited to
less than 50 cm of section (Bed B of Powell’s account, a term we retain here for clarity). We have
seen only ammonites and inoceramid bivalves plus planktonic foraminifera and calcispheres (some
perhaps calcitized radiolarians), the latter in thin section. The ammonites range from complete
adults to juveniles (the latter dominate). This fauna plus the organic-rich black shale associated
suggest a mass-mortality assemblage preserved under anoxic conditions, and the Ojinaga Formation
as a whole records such conditions, with fossils generally lacking. Indeed, the P. flexuosum Zone at
these localities is in stratigraphic isolation, occurring some 300 m above the top of the Buda
Limestone. Elsewhere, however, as in sections in the the Chispa Summit Formation (a lateral
equivalent of the Ojinaga) on the Stone (formerly Speck) Ranch, Grayton Lake Quadrangle,
Hudspeth County, Texas and between Chispa Summit and Needle Peak, Jeff Davis County, Texas,
the zonal fauna occurs in a fossiliferous sequence, and can be shown to occur above a fauna of the
Neocar diocer as juddii Zone, widely regarded as Upper Cenomanian (see Wright and Kennedy 1981;
Cobban 1983). This relationship has now been proven at sections in New Mexico and elsewhere in
the U.S. Western Interior (Cobban 1983, 1984; Hook and Cobban 1981). The importance of the
west Texas flexuosum Zone fauna is that it provides a wealth of ammonites (at all ontogenetic
stages) of typically Tethyan and Boreal types including species known from as far afield as Montana,
California, Venezuela, Nigeria, Tunisia, England, and Czechoslovakia which provide a datum for
inter-regional correlation. The associated inoceramids are uncrushed and in an expanded sequence
and belong to a form having an even wider distribution than the ammonites. The co-occurrence of
these two key groups of biostratigraphic indicators at a critical level in the mid Cretaceous is the
justification for the present account. A systematic description of the assemblage follows, after which
the inter-regional correlation of this basal Turonian fauna is described.

SYSTEMATIC PALAEONTOLOGY

Location of specimens. This is indicated by the following abbreviations:

OUM, University Museum, Oxford;

USNM, US National Museum, Washington DC.

UTA, University of Texas at Austin Collections, housed in Texas Memorial Museum, Austin, Texas.
BMNH, British Museum (Natural History), London.

Dimensions. Dimensions are given in millimetres, in the following order: diameter (D), whorl breadth (Wb),
whorl height (Wh), and breadth of umbilicus (U); c = costal, ic = intercostal measurements. Figures in
parentheses refer to dimensions as a percentage of diameter.

Suture terminology. The suture terminology of Wedekind (1916) as amplified by Kullmann and Wiedmann
(1970) is followed here: I = Internal lobe; U = Umbilical lobe; L = Lateral lobe; E = External lobe.

30

PALAEONTOLOGY, VOLUME 30

Order ammonoidea Zittel, 1884
Superfamily acanthocerataceae de Grossouvre, 1 894
Family acanthoceratidae de Grossouvre, 1894
(name corrected by Hyatt (1900) from Acanthoceratides de Grossouvre, 1894)

Subfamily acanthoceratinae de Groussouvre, 1 894
Genus quitmaniceras Powell, 1963

Type species. Q. reaseri Powell, 1963, p. 313, pi. 32, figs. 5 and 13; text-fig. 3/!,y, by original designation.

Diagnosis. Variable dwarf compressed acanthoceratines retaining or not the full complement of umbilical,
inner and outer ventrolateral and siphonal tubercles. In most the siphonal line is raised, producing either a
row of fine siphonal clavi or an entire keel. Extreme forms in the young range from nearly smooth oxycones
to compressed, parallel-sided individuals with narrow flat venters from the earliest stage seen.

Discussion. Abundant new material of Q. reaseri includes on the one hand individuals that are
hardly distinguishable in middle growth from compressed Thomelites Wright and Kennedy, 1973
of the Upper Cenomanian and on the other hand smooth oxycones that have been questionably
identified as Choffaticeras Hyatt, 1903. The former type indicates the ancestry of the genus. Some
of the keeled forms look forward to Prohauericeras Nowak, 1913, a poorly understood genus from
the Middle Turonian of France recently discussed by Kennedy et al. (1984).

Occurrenee. Lower Turonian of Chihuahua, west Texas, New Mexico, Colorado, and Wyoming.

Quitmaniceras reaseri Powell, 1963
Plate 1, figs. 1-38; text-fig. 2a-c

1923 Pseudotissotia (Choffaticeras) sp.? Reeside, p. 30, pi. 12, figs. 3-6.

1963 Quitmaniceras reaseri Powell, p. 313, pi. 32, figs. 5 and 13; text-fig. 3/;,y.

1963 Quitmaniceras brandi Powell, p. 314, pi. 32, figs. 6, 8, 11-12, 14-16; text-fig. 3i, p, q.

1977 Metoicoceras aff. whitei Hyatt; Chancellor, Reyment and Tait, p. 91, fig. 5.

1979 Quitmaniceras brandiVo\^eW\Coox>QT,p. 124.

1982 Metoicoceras? sp. Chancellor, p. 83, figs. 5 and 6.

Types. The holotype is UTA 36225 (PI. 1, figs. 14 and 15) from Bed B of the Ojinaga Formation at Powell’s
(1963) Dos Alamos locality, Chihuahua, Mexico. There are seventeen paratypes.

Other material. More than 150 specimens in the Oxford University Museum Collections, from Bed B of the
Ojinaga Formation, Calvert Canyon, Hudspeth County, Texas.

EXPLANATION OF PLATE 1

Figs. 1-38. A variation series of Quitmaniceras reaseri Powell, 1963, arranged in order of increasing whorl
inflation and strength of ornament, x 1. 1 and 2, OUM KT27; 3, OUM KT986; 4 and 5, OUM KT636;
6 and 7, OUM KT550; 8 and 9, OUM KT634; 10 and 1 1, OUM KT21; 12 and 13, OUM KT700; 14 and
15, UTA 36225, the holotype; 16 and 17, UTA 36222, the holotype of Q. brandi Powell, 1963 (a synonym);
18 and 19, UTA 30917, a paratype of Q. brandi; 20 and 21, UTA 36223, a paratype of Q. brandi; 22 and 23,
OUM KT31; 24 and 25, OUM KT602; 26, OUM KT965; 27 and 28, OUM KT738; 29 and 30, OUM
KT517; 31 and 32, OUM KT547; 33 and 34, OUM KT33; 35 and 36, OUM KT590; 37 and 38, OUM
KT575.

All specimens are from the Basal Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation. 1-13, 22-38, from Calvert Canyon, Hudspeth County, Texas; 14-21, from Dos Alamos, in
Arroyo Alomos, Chihuahua, Mexico. All figures are x 1 .

PLATE 1

':^y

&

1

x/

KENNEDY, WRIGHT and HANCOCK, Quitmaniceras

32

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 2. Whorl sections of: a, b, c, Quitmaniceras reaseri Powell, 1963, based on OUM KT21, 27, and 33
respectively; d, Pseudospidoceras flexuosum Powell, 1963, OUM KT352; e, Wrightoceras munieri (Pervinquiere,
1907), OUM KT399; f, g, Mammites powelli sp. nov., based on OUM KT524 and 714 respectively; h,
Kamerunoceras calvertense (Powell, 1963), OUM KT545; i, Thomasites adkinsi (Kummel and Decker, 1954);
j, K, M, N, Fagesia catinus (Mantell, 1822), based on OUM KT662, 261, 237, and 246 respectively; L, Neoptych-

ites sp., based on OUM KT975.

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

33

D

Wh

Wh

Wb\ Wh

U

OUM KT738

—

10-7(-)

15-7(-)

0-64

-(-)

OUM KT549

200(100)

-(-)

7-3(37)

—

6-9(35)

OUM KT774

23-5(100)

6-2(26)

10-0(43)

0-62

7-0(30)

OUM KT751

5-5(-)

8-0(-)

0-69

-(-)

OUM KT543

-(-)

13-K-)

18-7(-)

0-70

-(-)

OUM KT550

-(-)

11'5(-)

15-0(-)

0-76

-(-)

OUM KT638

15-0(100)

3-3(22)

5-0(33)

0-66

5-0(33)

OUM KT700

-U)

11'7(-)

18-8(-)

0-62

-(-)

Description. Since the material covers a wide morphological range, we describe it by drawing
attention to the features of the specimens selected for illustration. At one extreme (PI. 1, figs. 27-
28, 31-38) are moderately evolute, high whorled, and compressed individuals with sides flat and
parallel for inner 75 %, then converging to a narrow, flat venter. Strong primary ribs form bullae
on the umbilical margin, run straight and rectiradiate to an inner ventrolateral tubercle at which
they turn forward to the outer ventrolateral. The ribs cross the narrow venter transversely, slightly
depressed on either side of a very feeble siphonal tubercle. Secondary ribs are intercalated irregularly,
arising near the umbilical margin or at the level of the inner ventrolateral tubercles; the former
have inner and outer and the latter outer ventrolateral tubercles.

At the other extreme (PI. 1, figs. 1-1 1 ) are individuals with a more or less oxyconic whorl section
with entire sharp keel and rounded sinuous ribs. These are linked to the coarse ribbed form by the
holotype of Q. reaseri (PI. 1, figs. 14 and 15) via specimens such as that shown in Plate 1, figs. 20
and 21, the holotype of Q. brandi (PI. 1, figs. 16 and 17), and the individuals shown in Plate 1, figs.
27 and 28. The very simple, little-incised sutures are poorly exposed in all our material.

Discussion. Powell (1963, p. 314) distinguished Q. brandi from Q. reaseri on the basis of the former
having ribbed innermost whorls, stronger tuberculation, and less irregular ribbing. Our abundant
material, albeit fragmentary in many cases, suggests a spectrum whose morphological range goes a
little further than that indicated by Powell both in the direction of feebly ornamented oxycones and
robust individuals with regular strong tuberculation, but which does not appear to have significant
gaps. We therefore unite Powell’s two species.

Occurrence. Basal Turonian of Loma el Macho and Dos Alamos, Chihuahua, Mexico; Calvert Canyon, Stone
Ranch, and Chispa Summit, west Texas; New Mexico and Montana.

Subfamily euomphaloceratinae Cooper, 1978
Genus kamerunoceras Reyment, 1954

Type species. Acanthoceras eschii Solger, 1904, p. 124, pi. 4, figs. 1 -4, by original designation.

Discussion. A full revision of Kamerunoceras has recently been given by Kennedy and Wright
(19796).

Occurrence. Uppermost Cenomanian to Middle Turonian of England; Turonian of France, Spain, north
Africa, the Middle East, Nigeria, Cameroon, Madagascar, Mexico, west Texas and the US Western Interior,
Brazil, and Venezuela.

Kamerunoceras calvertense (Powell, 1963)

Plate 3, figs. 15-18; text-fig. 2h

1963 Acanthoceras calvertense Powell, p. 315, pi. 33, figs. 8 and 9; pi. 34, figs. 6 and 9; text-fig. 2e (?
non pi. 33, figs. 8 and 9).

1963 Acanthoceras sp. Powell, p. 316, pi. 33, fig. 5; text-fig. 5d.

Types. The holotype is UTA WSA227 from Babb Ranch in the Southern Quitman Mountains, and there are
three paratypes from Bed B of the Ojinaga Formation at Powell’s (1963) Arroyo Alamos locality. Chihuahua,
Mexico.

34

PALAEONTOLOGY, VOLUME 30

Other Material. OUM KT405, 417, 426, 429, 435, 437, 545, 679, 733, 976, 978-979, all from Bed B of the
Ojinaga Formation, Calvert Canyon, Hudspeth County, Texas.

Description. The present material consists chiefly of fragments of inner whorls with diameters of
40 mm or less, with only one larger fragment, OUM KT405. They have a more or less square
intercostal section, distant primary ribs with strong umbilical and strengthening inner ventrolateral
spines, slightly weaker clavate outer ventrolaterals, and feeble siphonal clavi; secondary ribs carry
inner and outer ventrolateral and siphonal tubercles. Below a diameter of 20 mm the tubercles are
relatively weaker and the costal whorl section consequently more rounded; occasional supernumer-
ary ribs are present on the venter. The single large fragment shows a broad, nearly flat venter with
weakening primary ribs, slightly adorally convex, with feeble traces of inner ventrolateral and
siphonal tubercles.

The suture line is relatively simple, with broad bifid lobes and saddles.

Discussion. The smaller specimens closely resemble acanthoceratine nuclei from the Upper Cenom-
anian. They also compare very well with the inner whorls of Kanabiceras puebloense as figured by
Cobban and Scott (1972) if those distorted specimens are restored to their natural shape. The larger
fragment agrees well with the body-chamber of an English specimen compared with puebloense
(Wright and Kennedy 1981, p. 56, pi. 14, figs. 3 and 1 1). Powell’s holotype and figured paratype
(1963, pi. 34, figs. 6 and 9) of T. calvertense show a phase between our smaller specimens and OUM
KT405. Another paratype, UTA 5525 (Powell 1963, pi. 33, figs. 8 and 9; see PI. 3, figs. 17 and 18)
differs in having fine untuberculate secondaries between primaries, although this may be merely a
matter of preservation. Apart from this specimen, all the material before us seems to belong to a
single species.

Cobban and Scott’s (1972) K. puebloense is based on crushed composite internal moulds, and
this makes comparison with K. calvertense difficult. K. puebloense was characterized by its regular
ribbing, with nine to sixteen ribs across the venter per half whorl and prominent inner ventrolateral
tubercles that show a tendency towards spinosity. This tendency is seen in some of our specimens,
but the different preservation means that in our material the tips of the septate spines are lost
beyond the basal septum, as is common in ancestral Euomphaloceras. The rib densities of the two
groups of specimens are very similar, and given differences in size and preservation might be within
the range of a single variable species, the trivial name calvertense having priority. Dr W. A. Cobban
points out, however, that K. calvertense is significantly older than K. puebloense, is much more
robust, more involute, and more delicately ornamented and lacks the spinose inner whorls.

Occurrence. Lower Turonian of northern Chihuahua and Calvert Canyon and the southern Quitman Moun-
tains, west Texas.

Genus pseudaspidoceras Hyatt, 1 903
(= Ampakabites CoWxgnon, 1965)

Type species. By original designation. Ammonites footeanus Stoliczka, 1864, p. 101, pi. 52, figs. 1 and 2. See
text fig. 4.

Discussion. See discussioh under P.flexuosum Powell, 1963 below.

Pseudaspidoceras flexuosum Powell, 1963

Plate 2, figs. 1-4, 8-13, 16-17; text-figs. 3a-c, 5, 6c, d, 7a-c

1902 Marnmites footeanus Stol. spec.; Petrascheck, p. 144, pi. 9, fig. 1.

1920 Pseudaspidoceras aff. pedroanum White; Bose, p. 209, pi. 13, fig. 1. pi. 15, fig. I.

1957 Pseudaspidoceras paganum Reyment; Barber, p. 1 1 (pars).

1963 Pseudaspidoceras flexuosum Powell, p. 318, pi. 32, figs. 1,9-10; text-fig. 2a-c,f, g.

1965 Kamerunoceras (Ampakabites) auriculatum Collignon, pp. 29, 31; pi. 388, fig. 1662; pi. 389, fig.
1664 (auriculatus on pp. 31-32).

Ampakabites collignoni Cobban and Scott, p. 81, pi. 29, figs. 1-3; text-figs. 39 and 40.

1972

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

35

TEXT-FIG. 3. External sutures of Pseiidaspidoceras flexuosum Powell, 1963. a is based on the lectotype of
Ampakabites auriculatus Collignon, 1964; b on OUM KT41 1; c on OUM KT352.

1978 Pseudaspidoceras flexuosum Powell; Young and Powell, pi. XXV. 2, figs. 8-10.

1978 Pseudaspidoceras aff. footeanum Petrascheck, 1902; Young and Powell, pi. 3, fig. 16.

Types. The holotype is UTA 30842 and there are thirty-nine paratypes from Bed B of the Ojinaga Formation
at Powell’s (1963, p. 310) Kelsey Crossing locality, Hudspeth County, Texas.

Other material. We have the following specimens from Bed B of the Ojinaga Formation in Calvert Canyon,
Hudspeth County, Texas: OUM KT188-204, 206-207, 352, 403, 407, 411, 413, 415, 422-423, 436, 533, 536,
656, 697, 707, 737, 753, 966, 989, 990.

36

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 4. Plaster cast of the lectotype of Pseudaspidoceras footeanum (Stoliczka, 1864), GSI 213. The original
specimen was figured by Stoliczka 1864 as his pi. 52, fig. 1, and is from the Utatur Group north of Odium,

southern India; x 0-45.

Description. Up to 20 mm diameter. The shell is rather evolute, with a moderately deep, fairly wide
umbilicus. The umbilical walls are steep, with evenly rounded shoulders. The flanks are rather
broad and flat, with an arched venter. Very weak, slightly prosiradiate ribs arise at about mid-flank
and give rise to seven weakly clavate ventrolateral tubercles per whorl. Across the venter the ribs
are arched strongly forwards, with a hint of weak looping. The posterior (adapical) rib, when they
are looped, is ornamented with two minute, bubble-like nodes, closely spaced on either side of the

EXPLANATION OF PLATE 2

Figs. 1-4, 8-13, 16-17. Pseudaspidoceras flexuosum Powell, 1963. 1 and 2, UTA 37124; 3 and 4, OUM KT413;

8 and 9, OUM KT206; 10 and 11, OUM KT737; 12 and 13, OUM KT411; 16 and 17, OUM KT352.

Figs. 5-7. Sciponoceras sp. OUM KT652.

Figs. 14 and 15. Mammitinae gen. et. sp. indet. OUM KT329.

All specimens are from the Basal Turonian P. flexuosiim Zone fauna of Bed B of the Ojinaga Formation; 1
and 2, from the Dos Alamos in Arroyo Alamos, Chihuahua, Mexico; the remainder are from Calvert
Canyon, Hudspeth County, Texas. Figs. 1-7, x 2; figs. 8-17, x 1.

PLATE 2

KENNEDY, WRIGHT and HANCOCK, Pseudaspidoceras, Sciponoceras, Mammitinae

38

PALAEONTOLOGY, VOLUME 30

siphonal line. Occasional specimens have stronger, finer ribs, some of which originate from weak
umbilical bullae, as well as lower ventrolateral tubercles intercalated between the main ribs and
looped across the venter, but lacking the upper ventrolateral nodes found on the main ribs.

20-50 mm diameter. A number of specimens (OUM KT207, 411, 737) represent this growth stage.
OUM KT207 (Text-fig. 6c) shows that the main ribs are now all ornamented with weak umbilical
bullae, whilst the inner ventrolateral tubercles are much more prominent and almost conical. OUM
KT737 (PI. 2, figs. 10 and 1 1) shows that the weak flank ribs are slightly prorsiradiate, with a faint
adorally-convex curvature, and that frequently two ribs arise from a single umbilical bulla, with
one of them generally becoming obsolete before reaching the venter. There are six umbilical bullae
per half-whorl at this stage and eight inner ventrolateral tubercles.

OUM KT41 1 (PI. 2, figs. 12 and 13) has the ventral region well preserved and shows prominent
looping of the ribs across the venter. Between the pair of looped ribs the interspace is unusually
deep, giving the impression that constrictions are present, although this probably is not the case. In
this specimen the inner ventrolateral tubercles are becoming distinctly clavate whilst, across the
venter, intercalated ribs show weak bullate swellings on either side of the siphonal line, so that there
are far more outer than inner ventrolateral tubercles, giving the specimen a markedly Euomphalo-
ceras-WkQ appearance.

OUM KT204 at 55 mm diameter shows considerably weaker ornament than OUM KT411, and
there is clearly much variation in this respect. In OUM KT206, ribs are virtually absent and
ornament comprises irregular weak folds and growth striae and the ventrolateral tubercles. Across
the venter, ribs are lacking but the deepened interspaces between looped ribs (immediately in front
of the outer ventrolateral nodes) still remain.

50-300 diameter. At this growth stage (PI. 2, figs. 16 and 17; text-figs. 6c and 7), the whorl section
is rectangular, compressed, the umbilical wall is slightly overhanging with an abrupt, evenly rounded
umbilical shoulder. The flanks are broad and flattened, subparallel, and the venter is slightly convex.
Ornament comprises distinct umbilical bullae, four per quarter whorl, which give rise to very weak
ribs, either singly or in pairs. The ribs are initially distinctly prorsiradiate, but flex back at about
mid-flank to become rectiradiate. There may be up to two very weak intercalatories between main
ribs. Main ribs are ornamented with weak, clavate ventrolateral tubercles which are joined across
the venter by strongly arched, adorally convex ribs which still show a tendency to be looped. The
outer ventrolateral nodes have normally disappeared by the beginning of this stage.

The external suture line (text-fig. 3b, c), consists of a narrow E, deep, asymmetrically trifid L,
narrow E/L and L/U2.

Discussion. The holotype is a medium-sized body-chamber fragment. This and larger specimens
(text-fig. 7) before us suggest considerable variation in adults of this species, sufficient to encompass
the Czech specimen figured by Petrascheck (1902), the types of Kamerunoceras (Ampakabites)
auriculatum Collignon 1965 (text-figs. 5, 6d), and perhaps Ampakabites collignoni Cobban and
Scott, 1972. Cobban and Scott distinguished the single specimen of A. collignoni from K. auriculatum
by its ‘more arched venter, greater number of umbilical and ventrolateral tubercles, and a third
row of tubercles on the flank’. The arching of the venter in collignoni is clearly due to crushing, the
slightly greater rib and therefore tubercle density does not seem to us to be significant in this
variable species and the feeble inner ventrolateral tubercles of collignoni are not markedly stronger
than those visible in the types of auriculatus (text-figs. 5, 6d). The lectotype of Pseudaspidoceras
footeanum (Stoliczka, 1865) (pi. 52, fig. 1) (text-fig. 4) has a squarer whorl section and straighter
ribs which distinguish it from the present form at specific level, but they appear to be congeneric.
Wright and Kennedy (1981, p. 81) pointed out that the outer whorls of the two species were closely
similar but that the inner whorls of flexuosum were very different from the Mammites-VikQ inner
whorls of the lectotype of footeanum. Subsequent examination of a plaster cast of this specimen
makes it very doubtful that the inner whorls figured by Stoliczka (1865, pi. 52, fig. la, b) actually
belong to the specimen in that figure. If they do, the figure is much restored and incorrectly so. The
innermost whorls visible in the plaster cast of the lectotype show distant ribs comparable with those

KENNEDY. WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

39

TEXT-FIG. 5. The paralectotype of Kamenmoceras (Ampakahites) auriculatum Collignon, 1964, the original of
Collignon 1964 pi. 389, fig. 1664, from the Lower Turonian of Ampakabo (Betioky), Madagascar. Collignon

Collection, Universite de Dijon; x 1.

of flexuosum. Moreover, its outer whorls can be seen to have feebly looped primary ribs of the type
characteristic of flexuosum. Mammites (Pseudaspidoceras) dubertreti Basse, 1937 (p. 183, pi. 10, fig.
3fl, b; pi. 11, fig. 2) may also belong to this genus. It too is distinguished by having straight rather
than flexuous ribs at a comparable growth stage.

- jm

1 ■%/ ^ M

'-‘mk

f ^ ■ > m.

40

PALAEONTOLOGY, VOLUME 30

KENNEDY, WRIGHT AND HANCOCK; TURONIAN AMMONITES FROM TEXAS

41

TEXT-FIG. 7. A-c, Pseudaspidoceras flexuosum Powell, 1963. Adult body-chambers from the Basal Turonian
P. flexiiosum Zone of Calvert Canyon, x 0-5. a, c, OUM KT149; b, OUM KT199.

Occurrence. Lower Turonian of west Texas, northern Mexico, ?Colorado, the German Democratic Republic,
Nigeria, and Madagascar.

Subfamily mammitinae Hyatt, 1900
(= Metoicoceratidae Hyatt, 1903; Fallotitinae Wiedmann, 1960)

Discussion. See Wright and Kennedy 1981, p. 62.

TEXT-FIG. 6. A, B, E, F, Pseudcispidocercis pseudonodosoides (Choffat, 1898); a, b, USNM 337482, from the
Neocardioceras juddii Zone at USGS Mesozoic locality D10114 in NE 1/4 section 13, Township 21S, Range
9W, Luna County, New Mexico; E, f, OUM KYI 095, from the upper part of the Paravascoceras caiivini Zone
of Freund and Raab (1969), Ora Formation, roadside section on south side of Mishor Seifim, 18 km N. of
Elat, Negev, Israel, c, d, Pseudaspidoceras flexuosum Powell, 1963; c, silicone squeeze taken from an external
mould, OUM KT207, from the Lower Turonian P. flexuosum Zone of Calvert Canyon; d, the lectotype of
Kamerunoceras (Ampakahites) auriculalum Collignon, 1964, the original of Collignon 1964, pi. 358, fig. 1662,
from the Lower Turonian of Ampakabo (Betioky), Madagascar; Collignon Collection, Universite de Dijon
collections. All figures are x 1.

42

PALAEONTOLOGY, VOLUME 30

Genus mammites Laube and Bruder, 1887
(= SchluetericerasViydLiX, 1903)

Type species. Ammonites nodosoides Schliiter, 1871, p. 19, pi. 8, figs. 1-4, by monotypy.

Discussion. See Wright and Kennedy 1981, p. 75 for a full review of this genus.

Occurrence. Turonian, widespread in Eurasia, West Africa, and North and South America.

Mammites powelli sp. nov.

Plate 3, figs. 1-14; Plate 4, figs. 16 and 17, text-fig. 2f, g

1963 Mammites nodosoides (Schlotheim); Powell, p. 316, pi. 33, figs. 1, 3, 4, 6, 10, 1 1; text-fig. 3m-o,
t, u.

1978 Mammites nodosoides (Schlotheim); Young and Powell, pi. 2, fig. 2.

Types. The holotype is OUM KT404; paratypes are OUM KT205, 408, 412, 421, 425, 431, 434, 438, 524, 527,
534, 541, 678, 703, 714, 724, 968-970. All specimens are from Bed B of the Ojinaga Formation, Calvert
Canyon, Hudspeth County, Texas.

Origin of Name. For J. Dan Powell, who first described the Calvert Canyon fauna.

Description. A small, evolute, coarsely ribbed Mammites with slightly depressed subquadrate to
subtrapezoidal whorl section. There are about twenty ribs to a whorl, of which seven or eight are
primaries. All ribs bear inner and outer ventrolateral clavi, which on the mature body-chamber
coalesce and finally form prominent ear-shaped ventrolateral bullae which almost join across the
venter. The umbilical wall is rounded and slightly undercut.

The rather simple sutures are too poorly exposed for description.

Discussion. M. powelli differs from M. nodosoides (Schliiter, 1871) in being much smaller and in
laeking the exaggerated ventrolateral horn directed upwards and outwards on the last whorl.
Morrowites species, e.g. M. wingi (Morrow, 1935) (p. 467, pi. 51, fig. 2; pi. 52, fig. 2a-c, text-fig. 2)
of whieh M. clwuberti (Collignon, 1967) (p. 41, pi. 22, figs. 1, iff) and M. costatus (Matsumoto and
Kawashita, 1978) (p. 5, pi. 1, figs. 1 and 2; pi. 2, figs. 1 and 2; text-fig. 1-3) are probably synonyms,
are also large, coarsely ornamented species with broad flat venters and outward directed ventrola-
teral horns in maturity. Morrowites nuclei are constricted.

M. depressus Powell, 1963 (see Cobban and Hook 1979) is a large, very evolute Middle Turonian
species with a depressed whorl section much wider than high.

M. dixeyi Reyment (1955, p. 50, pi. 9, fig. 4; pi. 11, fig. 2; text-figs. 20 and 21) of which M.
mutabilis Reyment, 1955 (p. 51, pi. 10, fig. \a, b) is probably a synonym, is mature at about the
same size and is closely related to the present species. It differs in having more outer than inner
ventrolateral tubercles and in not developing the long ventrolateral bullae on the mature body-
chamber.

Occurrence. Apart from the type occurrence, the species occurs at Dos Alamos, Chihuahua, northern Mexico
and Chispa Summit, Jeff Davis County, Texas. There is a specimen in the British Museum from the Turonian
of Novo Redondo, Angola.

EXPLANATION OF PLATE 3

Figs. 1-14. Mammites powelli sp. nov. 1 and 2, UTA 35524; 3 and 4, UTA 35532; 5 and 6, UTA 35539; 7 and
8, UTA 30968; 9 and 10, paratype, OUM KT524; 11 and 12, paratype OUM KT969; 13 and 14, the
holotype, an adult body-chamber, OUM KT404.

Figs. 15-18. Kamerunoceras calvertense (Powell, 1963). 15 and 16, paratype UTA 35526; 17 and 18, paratype
UTA 35525.

All specimens are from the Basal Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation. 1-8, 15-18, from Dos Alamos in Arroyo Alamos, Chihuahua, Mexico; 9-14, from Calvert
Canyon, Hudspeth County, Texas. Figs. 1-8,11-12, x 2; figs. 9-10, 13- 18, xl.

PLATE 3

KENNEDY, WRIGHT and HANCOCK, Mammites, Kamenmoceras

44 PALAEONTOLOGY, VOLUME 30

Mammitinae gen. et sp. indet.

Plate 2, figs. 14 and 15

Material. Two fragments, OUM KT329 and 416, from Bed B of the Ojinaga Formation, Calvert Canyon,
Hudspeth County, Texas.

Description. The material assigned here comprises two very small fragments which show a subquad-
rate whorl section with flattened, subparallel flanks and a slightly convex venter. Ornament com-
prises prominent, rounded, slightly prorsiradiate main ribs separated by one to two intercalatories.
At the ventrolateral shoulders the ribs flex forwards and pass strongly across the venter, where they
are prominently raised, forming a linguoid adorally-convex arch. The ribs are wider than the
interspaces. Some ribs show the faintest hint of umbilical bullae, but there is no sign of any other
tuberculation.

Discussion. The present material shows features unlike those of any described genus of Turonian
ammonite, but is reminiscent of the body-chamber ornament of certain Protacantlioceras species
(Wright and Kennedy 1980). Until adequate material is available we consider the material generi-
cally indeterminate.

Family vascoceratidae H. Douville, 1912

(name corrected and translated by Spath, 1925, from Vascoceratines H. Douville, 1912).
Discussion. See Wright and Kennedy 1981, p. 84.

Subfamily vascoceratinae H. Douville, 1912
Genus vascoceras Choffat, 1898

( = Paravascoceras Furon, 1935; Pachyvascoceras Furon, 1935; Paracanthoceras Furon, 1935; Brog-
giiceras Benavides-Caceres, 1956; Discovascoceras CoWxgnon, 1957; Provascoceras Coop&v, 1979).

Type species. Vascoceras gamai Choffat, 1898, p. 54, pi. 7, figs. 1-4; pi. 8, fig. 1; pi. 10, fig. 2; pi. 21, figs. 1 and
5, by the subsequent designation of Roman 1938, p. 452.

Discussion. Inspection of any large fauna of Vascoceras or of any of the many published descriptions
of such faunas shows that shell shape and ornament vary widely and, if there are enough specimens,
continuously. By picking out and figuring individuals at intervals through the morphological spec-
trum it is easy to give an impression of a wealth of species or subspecies or even genera or subgenera.
The Calvert Canyon material studied here, abundant, uncrushed, and from a single narrow horizon,
makes it perfectly clear that there is present but one variable population of Vascoceras, ranging
from rare smooth cadicones through abundant intermediates to rare ventrally ribbed, moderately
compressed specimens with rectangular whorl section, the whole population being linked together
by a comparable degree of involution. As a result of the wide variation in this and other populations,
there seems no point in separating off segments into the genera noted above.

explanation of plate 4

Figs. 1-15, 18-19. Vascoceras proprium (Ktyrntni, 1954). 1-3, OUM KT992, a juvenile with slender, strongly
ribbed inner whorls; 4 and 5, OUM KT480, a nucleus showing the strong constrictions and associated ribs
at this stage of development; 6 and 7, OUM KT278, showing the succeeding ribbed stage; 8 and 9, OUM
KT613, a relatively finely ribbed juvenile with flattened venter; 10-12, OUM KT272, a juvenile with well-
differentiated primary and secondary ribs; 1 3-15, OUM KT589, showing innermost smooth whorls followed
by constricted stage; 18 and 19, OUM KT366, a body-chamber of a possible microconch.

Figs. 16 and 17. Mammites powelii sp. nov. Paratype, OUM KT417.

All specimens are from the Basal Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas. Figs. 4-7, 13-15, x 2; the remainder are x 1.

PLATE 4

KENNEDY, WRIGHT and HANCOCK, Vascoceras, Mammites

46 PALAEONTOLOGY, VOLUME 30

Occurrence. Upper Cenomanian and Lower Turonian. Generally Tethyan in occurrence, but a few occur in
higher and lower latitudes.

Vascoceras proprhim (Reyment, 1954^)

Plated, figs. 1-15, 18-19; Plates 5-6; text-figs. 8a-c, 9

1920 Vascoceras angermaiini Bose, p. 217, pi. 16, figs. 1 and 3 (non 2 and 4); pi. 17, fig. 1.

1920 Neoptychites aff. cephalotus (Courtiller); Bose, p. 221, pi. 18, figs. 3, 10, 13.

1931 Thomasites sp. Adkins, p. 56, pi. 2, figs. 16 and 17.

19546 Pachyvascoceras propriion Reyment, p. 258, pi. 5, fig. 1; text-fig. 3<f.

19546 Pachyvascoceras proprium plenum Reyment, p. 258, pi. 5, fig. 5; text-figs. 3c and 6.

19546 Pachyvascoceras costatum Reyment, p. 257, pi. 3, fig. 6; pi. 4, fig. 3; pi. 5, fig. 2; text-figs. 3a-6, 5.
19546 Pachyvascoceras globosum Reyment, p. 259, pi. 3, fig. 3; pi. 5, fig. 4; text-figs. 3c and 7.

1955 Paravascoceras costatum Reyment, p. 65, pi. 14, figs. 2 and 4.

1957 Vascoceras globosum globosum (Reyment); Barber, p. 21, pi. 7, fig. 1; pi. 9, fig. 4; pi. 28, figs. 1
and 2.

1957 Vascoceras globosum plenum (Reyment); Barber, p. 23, pi. 7, fig. 2; pi. 9, fig. 2; pi. 28, figs. 3-5.
1957 Vascoceras globosum proprium (Reyment); Barber, p. 25, pi. 7, fig. 3; pi. 28, figs. 6 and 7.

1957 Vascoceras globosum compressum Barber, p. 25, pi. 7, fig. 4; pi. 9, fig. 1; pi. 28, figs. 10 and 1 1.
1957 Vascoceras globosum carter! Barber, p. 25, pi. 8, fig. 2; pi. 28, figs. 8 and 9.

1957 Paravascoceras costatum costatum (Reyment); Barber, p. 35, pi. 14, fig. 1; pi. 32, figs. 1-3.

1957 Paravascoceras costatum tectiforme Barber, p. 37, p. 14, fig. 4; pi. 15, figs. 1-3; pi. 16, fig. 1; pi.
32, figs. 4-7.

1963 Pachyvascoceras compressum Barber; Powell, p. 321, pi. 32, figs. 2-4, 7; pi. 34, figs. 8 and 10;
text-fig. 36-^/, /.

1963 Pachyvascoceras globosum Reyment; Powell, p. 321, pi. 34, figs. 7-1 1; text-fig. 3^.

1977 Paravascoceras carter! (Barber); Chancellor, Reyment and Tait, p. 92, figs. 15-17.

1982 Paravascoceras carter! (Barber); Chancellor, p. 102, figs. 35-37.

71982 Paravascoceras compressum (Powell not Barber); Chancellor, p. 106, figs. 49 and 50.

Holotype. BMNH C47302 from the Lower Turonian of Gulani, Gongola Valley, Bornu Province, Nigeria.

Material. More than sixty specimens in the Oxford University Museum Collections, from Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas.

Dimensions.

D

OUMKT369 54-8(100)

OUMKT378 74-0(100)

OUMKT376 120-0(100)

Wb Wh

35-0(63-7) 25-8(47

80-0(1-08) 32-5(43

87-0(72-5) 57-5(47

Wb: Wh

U

1)

1-36

7-5(13-7)

9)

2-46

16-0(21-6)

9)

1-51

18-3(15-4)

Description. Specimens of all whorl sections are relatively involute; the umbilical diameter ranges
from 13 % to 22 % of total diameter. The umbilical wall is vertical to slightly undercut, with a well-
rounded umbilical shoulder. The whorl section is variable, from slightly compressed, flat, and
parallel sided with evenly rounded to almost flat venter, to depressed and increasingly cadicone (see
PI. 5). After an initially smooth, moderately compressed stage (PI. 4, figs. 13-15), constrictions
appear at a diameter of 7 or 8 mm (PI. 4, figs. 4-12), three or four to a whorl, deepest on the outer

EXPLANATION OF PLATE 5

Figs. 1-8. Vascoceras proprium (Reyment, 1954). A series of juveniles showing the range of ornament and
whorl section. 1-3, OUM KT380; 4-6, OUM KT369; 7 and 8, OUM KT378.

All specimens are from the Basal Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas. All figures are x 1.

PLATE 5

KENNEDY, WRIGHT and HANCOCK, Vascoceras

48

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 8. Whorl sections of Vascoceras proprium (Reyment, 1954). a, OUM KT378; b, OUM KT382;

c, OUM KT992.

part of the whorl and the venter, with a strong rib behind and in front which, rarely, develop weak
to moderate umbilical bullae (PI. 4, fig. 7). Fold-like ribs may appear after the first constrictions
and strengthen to varying degrees. In some specimens they become distinct on the inner part of the
flanks by a diameter of 1 1 or 12 mm, but normally they are strong on the outer and weak on the
inner flank (PI. 4, figs. 10-12). They cross the venter uninterrupted (PI. 5, fig. 6), with a slight

EXPLANATION OF PLATE 6

Figs. 1-6. Vascoceras proprium (Reyment, 1954). 1 and 2, OUM KT387, a juvenile of intermediate inflation;
3 and 4, OUM KT267, a slender, ribbed body-chamber; 5 and 6 OUM KT377, a moderately depressed
specimen with body-chamber.

All specimens are from the Lower Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas. All figures are x 1.

PLATE 6

KENNEDY, WRIGHT and HANCOCK ^ascoceras

50

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 9. Vascoceras proprium (Reyment, 1954) the largest of the specimens known from the Basal Turonian
Pseudaspidoceras ftexuosum Zone fauna of Calvert Canyon, OUM KT358; x 0-8.

forward bend. Ribs, where present, generally decline at maturity (PI. 6, figs. 5 and 6; text-fig. 9),
and the adult body-chamber is virtually smooth for the last part.

Some possibly adult specimens are only 75 mm in diameter (PI. 4, figs. 18 and 19; ?P1. 6, figs. 3
and 4); others reach 180 mm (text-fig. 9), but our material is inadequate to confirm size dimorphism
like that to be documented below in Fagesia.

Discussion. The large number of specimens, from a single horizon and locality, at all stages of
ontogeny, intergrading in whorl section, nature and persistence of ribbing and presence or absence
of umbilical tubercles in early stages are clearly part of a single population. They throw a clear
light on the problems of vascoceratine taxonomy at species and to some extent at generic level. It is
obvious that descriptions of vascoceratine taxa are of little value unless they cover abundant
material and are under tight stratigraphic control. Intermediate specimens in our material match
well with the holotype of Reyment’s (1954^, pi. 5, fig. 1) Vascoveras proprium. With this we would
synonymize a range of associated forms distinguished by Reyment (1954/?, 1955) and Barber (1957)
(see synonymy).

Powell (1963) described a considerable number of specimens from the horizon of the present
material, referring them to Pachyvascoceras compressum Barber, 1957 and P. globosum Reyment,

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS 51

1 954^. These fit in well with our material, and lead to the extreme cadicone variants of our collections
(PI. 5, figs. 7 and 8).

Similar forms are found in several Vascoceras faunas, and by themselves are difficult to dis-
tinguish. V. harttii (Hyatt, 1870 p. 386; 1875, p. 370; White 1887, p. 226, pi. 19, figs. 1 and 2; pi. 20,
fig. 3) is a globose cadicone but is distinctly more evolute and has an umbilicus with steeply sloping
sides instead of one with overhanging umbilical edges as in the present species. The lectotype of V.
angermanni Bose, 1920 (p. 217, pi. 16, figs. 2 and 4 only) designated by Chancellor, Reyment, and
Tait (1977, p. 92) seems to us indistinguishable from V. harttii, whereas specimens described by
Chancellor et al. (1977) and Chancellor (1982) as Paravascoceras carteri (including V. angermanni
B5se 1920, pi. 16, figs. 1 and 3 {non 2 and 4)) seem to be conspecific with our material. The
Neoptychites aff. cephalotus of Bose (1920) also belongs here.

If the material from various localities and horizons is put together it may well be possible to
arrange it to show continuous variation in degree of involution, as with whorl shape and ornament.
The Calvert Canyon population, however, is uniform in involution throughout all its whorl shapes.
Consequently we do not synonymize it with populations that have a significantly different degree
of involution.

The Damergou fauna, described by Chudeau (1909), Furon (1935), and Schneegans (1943),
includes specimens with a similar range of whorl shapes to those of our material but most individuals
are less involute and in the ornamented, compressed individuals the ribbing is much stronger and
more persistent. Schobel (1975) has analysed statistically a large population from this locality and
synonymized under Paravascoceras cauvini Chudeau, 1909 most of Furon’s and Schneegans’s
species and subspecies of this group. This is probably correct but his population (unpublished study
by R. I. Kirby and M. R. Cooper) is strongly dimorphic as regards inflation and also independently
dimorphic as regards involution. Since the material was collected from surface scree it is possible
that it does not represent part of a single contemporaneous population.

Freund and Raab (1969) distinguished five species of Paravascoceras occurring in their P. cauvini
Zone, of which four were placed, probably correctly, in the synonymy of cauvini by Schobel.

Broggiiceras olssoni Benavides-Caceres, 1956 and B. humboldti Benavides-Caceres, 1956 (pp. 469-
470) appear to resemble V. cauvini rather than our material.

Occurrence. Lower Turonian of Texas, northern Mexico, and Nigeria.

Genus fagesia Pervinquiere, 1907
(= Plesiovascoceras Spdiih, 1925)

Type species. Olcostephanus siiperstes Kossmat, 1897, p. 26 (133), pi. 6, (17), fig. 1, by original designation.
Discussion. See Wright and Kennedy 1981, p. 87.

Fagesia catinus (Mantell, 1822)

Plate 7, figs. 1-13; Plate 8, figs. 1-4, 6-9; text-figs. 2j, k, m, n, 10

1822 Ammonites catinus Mantell, p. 198, pi. 22, fig. 10 {non fig. 5, attributed in error; = Ammonites
navicularis Mantell).

1981 Fagesia catinus (Mantell, 1822); Wright and Kennedy, p. 88, pi. 26, fig. 2; text-figs. 31-36 (with
full synonymy).

1982 Fagesia haarmanni Bose; Chancellor, p. 106, figs. 43-48 (with full synonymy).

1982 Fagesia levis Renz, p. 78, pi. 22, fig. 20; pi. 23, figs. 1-3; text-figs. 53 and 59a, c.

1982 Fagesia aff. superstes (Kossmat); Renz, p. 78, pi. 22, fig. 19; pi. 23, fig. 4; text-fig. 59b.

1983 Fagesia cf. thevesten.sis Peron; Renz, p. 79, pi. 22, fig. 15.

1982 Fagesia aff. haarmanni Bose; Renz, p. 79, text-fig. 60.

1982? Fagesia sp. indet. Renz, p. 79, pi. 24, fig. 1 1.

Holotype. By monotypy: BMNH C3379, refigured by Wright and Kennedy 1981, text-fig. 31, from the Middle
Chalk of Lewes, Sussex.

52

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 10. Fagesia catiims (Mantell, 1822), juvenile macroconch, OUM KT218, from the Basal Turonian
Pseudaspidoceras flexiiosiim Zone of Calvert Canyon; x 0-72.

Description. Inner whorls moderately evolute with a subtrapezoidal to semilunate, depressed whorl
section. There are 7-11 prominent umbilical tubercles per whorl, from which generally 1-2, oc-
casionally 3, coarse, prorsiradiate ribs arise, frequently with an intercalatory between nodes. In
some juveniles there is a distinct tendency to form very weak ventrolateral bullae (PI. 7, figs. 8-11).
There are about 20-30 ribs per whorl. Beyond 40 mm diameter the ribbing weakens considerably
and is eventually lost at greatly varying diameters. In maturity the shell becomes very evolute
(umbilicus 50% of the diameter), with the umbilical wall strongly inclined and the whorl section
very depressed coronate. The umbilical tubercles may be retained to maturity or not.

EXPLANATION OF PLATE 7

Fig. 1-13. Fagesia catimis (Mantell, 1822). 1-3, OUM KT977, a pathological juvenile with a lateral tubercle
on the side figured; 4 and 5, OUM KT246, a strongly ribbed fragment of a cadicone juvenile macroconch;
6 and 7, OUM KT250, a feebly ribbed juvenile; 8 and 9, OUM KT223, a coarsely ribbed juvenile with
massive umbilical bullae; 10 and 11, OUM KT248, a very depressed juvenile with incipient ventral bullae;
12 and 13, UTA 20842, the holotype of F. texana Adkins, 1933.

The original of figs. 12 and 13 is from an unspecified horizon in the Lower Turonian of the Van Horn
Mountains, Culberson County, Texas. The remainder are from the Basal Turonian Pseudaspidoceras flexuo-
sum Zone fauna of Bed B of the Ojinaga Formation, Calvert Canyon, Hudspeth County, Texas. All figures
are x 1 .

PLATE 7

KENNEDY, WRIGHT and HANCOCK, Fagesia

54

PALAEONTOLOGY, VOLUME 30

The most noteworthy variation in the material before us is in the whorl section and the ontogenetic
stage at which ribbing finally becomes obsolete. The holotype of F. texana Adkins, 1931, loses all
its ribbing at about 50 mm diameter (PI. 7, figs. 12 and 13) and is closely matched by OUM KT250
(PI. 7, figs. 6 and 7) in which ribbing is also lost at the same diameter. Individuals that lose their
ribbing early are also characterized by being somewhat more involute and having a semilunate to
reniform whorl section. In addition, the largest individual of this type, OUM KT245 (PI. 8, figs. 3
and 4), attains an adult diameter of only 100 mm whereas the evolute individuals with sloping
umbilical walls and strongly depressed, coronate whorl section attain maturity at much larger
diameters (see text-fig. 10 for an incomplete juvenile of this form). This latter form is represented
by the holotype of F. haarmatmi Bose, 1920, ‘F.’ stantoni Reeside, 1923, and Ammonites catinus
Mantell, 1822, and frequently attains diameters in excess of 200 mm. The fact that juveniles of the
two forms are indistinguishable up to 40 mm diameter, whereafter the shells show somewhat
different adult ornament and mature at strikingly different diameters, suggest to us that we are
dealing with dimorphs. They may be distinguished as follows:

Microconchs: Adult at diameters up to 100 mm. Shell rather involute (U = 28-35 % of the
umbilicus) with ribbing lost at or considerably weakened after about 50 mm diameter. Umbilical
tubercles are retained on the body-chamber as rather weak bullae. Whorl section semilunate to
reniform, depressed.

Macroconchs: Adult at diameter of the order of 200 mm, with a very evolute shell (umbilicus
frequently in excess of 50 % of the diameter). The umbilical walls are strongly inclined, so that the
shell becomes cadicone and the umbilical tubercles move to a ventrolateral position. Umbilical
tubercles become exaggerated in maturity and are retained onto the adult body-chamber. Whorl
section strongly depressed, coronate.

Discussion. F. super stes (Kossmat, 1895) (p. 26 (133), pi. 6 (17), fig. 1) the type species of the genus,
differs from F. catinus in having more umbilical tubercles per whorl and invariably two ribs arising
from each tubercle. Intercalated ribs are lacking.

F. tevesthensis (Peron, 1897) (p. 23, pi. 7(1), figs. 2 and 3) is based upon a juvenile which
closely resembles F. superstes, from which it differs only in having steeper umbilical walls, with a
corresponding change in whorl section. The variation in this respect in the Texas material suggests
that F. tevesthensis may prove to be a synonym of Kossmat’s species. Indeed, this was suggested by
Peron (1897, p. 84) who felt that F. superstes and F. tevesthensis could represent extreme variants
of a single species. A. kotoi Yabe 1904 (p. 26, pi. 6, figs. 3 and 4) was included in the synonymy of
F. tevesthensis by Matsumoto (1973).

F. bomba Eck, 1909 (p. 181, text-figs. 1-5), F. rudra (Stoliczka, 1865) (p. 122, pi. 60, fig. 1), F.
involuta Barber 1957 (p. 27, pi. 9, fig. 3; pi. 29, figs. 6 and 7), and F. pervinquieri Bose, 1920 (p. 212,
pi. 14, fig. 3) are all rather involute, globose species with steep umbilical walls which lack umbilical
tubercles from an early growth stage. They show few of the characters of F. superstes and are
possibly better referred to the genus Vascoceras, as suggested by Barber (1957, p. 27).

F. simplex Barber, 1957 (p. 27, pi. 8, fig. 1; pi. 29, figs. 4 and 5) is based upon an immature
specimen showing a rather narrow, steep-sided umbilicus and about ten umbilical tubercles per
whorl, from which arise broad, weak ribs, frequently in pairs and with occasional intercalatories. It

EXPLANATION OF PLATE 8

Figs. 1-4, 6-9. Fagesia catinus (Mantell, 1822). 1 and 2, OUM KT229, a very depressed juvenile; 3 and 4,
OUM KT245, the body-chamber of an adult microconch; 6, OUM KT269, a strongly ribbed juvenile; 7-9,
OUM KT247, an incomplete slender microconch.

Fig. 5. Worthoceras ci. vmm'cM/wi' (Shumard, 1860). OUM KT698.

All specimens are from the Basal Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas. Fig. 5, x 3; fig. 6, x2; figs. 1-4, 7-9, x 1.

PLATE 8

KENNEDY, WRIGHT and HANCOCK, Fagesia, Worthoceras

56 PALAEONTOLOGY, VOLUME 30

differs from the Texas material in being weaker and more coarsely ribbed at comparable diameters.
The differences are slight.

F. pachydiscoides Spath, 1925 (p. 198) (= A. catinus Sharpe {non Mantell), 1855, p. 29, pi. 13,
fig. 1) differs from F. catinus in being still more evolute with gently rounded umbilical shoulders
and a reniform whorl section; the rate of increase in whorl width is even less than in the presumed
macroconchs of the present species but F. pachydiscoides may eventually prove not to be specifically
separable.

F. siskiyouensis Anderson, 1931 (p. 125, pi. 17, figs. 2 and 3) was tentatively referred to the
Campanian genus Anapachydiscus by Matsumoto (1959), whilst F. klamatliensis Anderson, 1958
(p. 248, pi. 28, fig. 3) was considered a synonym of Eupachydiscus haradai (Jimbo, 1894).

F. lenticidaris Freund and Raab, 1969 (p. 36, pi. 6, figs. 3-7; pi. 7, figs. 1-3; pi. 8, figs. 1 and 2;
text-figs. Ih-k, Sa-i, 9a-c) is based upon crushed material and the variation on which Freund and
Raab (1969) based their subspecies elliptica and asymmetrica are largely, if not entirely, due to post-
burial distortion. It is not possible to separate F. lenticidaris asymmetrica (cf. pi. 6, figs. 3-5) from
F. superstes sphaeroidalis (Pervinquiere) (1907, p. 322, pi. 20, figs. 3 and 4) at a comparable growth
stage but, as the holotype of F. lenticidaris is 380 mm diameter, comparison is difficult. In maturity,
F. lenticidaris is much more involute than F. catinus and looses its umbilical tubercles at an early
ontogenetic stage.

The authors do not agree with Chancellor’s (1982) interpretation of the Mexican Fagesia.

Occurrence. Apart from the present occurrences, F. catinus is characteristic of a low level in the Turonian of
southern England, France, Venezuela, northern Mexico, Texas, Montana, and California. It is also present in
the Upper Cenomanian Neocardioceras juddii Zone of SW New Mexico.

Genus neoptychites Kossmat, 1895

Type species. Ammonites telinga Stoliczka, 1865, p. 125, pi. 62, figs. 1 and 2 (= A. cephalotus Courtiller, 1860
p. 248, pi. 2, figs. 1-4), by original designation.

Discussion. See Kennedy and Wright 1979a for the most recent discussion of this genus.

Occurrence. Lower and Middle Turonian, France, Spain, Morocco, Algeria, Tunisia, Syria, Israel, Cameroon,
Nigeria, Madagascar, southern India, Japan, Texas and the US Western Interior, Mexico, Caribbean, Brazil,
Columbia, and Venezuela.

Neoptychites sp.

Plate 9, figs. 5-7, 13-14; text-fig. 2l

Material. OUM KT373, 975, from Bed B of the Ojinaga Formation, Calvert Canyon, Hudspeth County,
Texas.

Dimensions.

OUM KT975
OUM KT373

D

34-2(100)

55-3(100)

Wb

16-2(47)

32-4(59)

Wh

17-8(52)

28-7(52)

Wb: Wh

0- 91

1- 12

U

3-7(11)

-(-)

Description. The shell is very involute, with compressed whorls, the greatest breadth being at the
umbilical shoulder. The whorl section is subtrigonal; slightly convex flanks converge to a narrow.

EXPLANATION OF PLATE 9

Figs. 1-4, 8-12, 15-16. Thomasites adkinsi (Kummel and Decker, 1954). 1 and 2, OUM KT982; 3 and 4,
OUM KT568; 8-9, 12, OUM KT354; 10 and 1 1, OUM KT356; 15 and 16, OUM KT353.

Figs. 5-7, 13-14. Neoptychites sp. 5-7, OUM KT975; 13 and 14, OUM KT313.

All specimens are from the Basal Turonian Pseudaspidoceras flexuosum Zone fauna of Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas. Figs. 1-4, x 2; figs. 5-16, x 1.

PLATE 9

KENNEDY, WRIGEIT and EIANCOCK, Thomasites, Neoptychites

58 PALAEONTOLOGY, VOLUME 30

evenly rounded venter. Faint fold-like ribs are visible on the ventral region at the smallest diameters
visible.

Neither specimen shows the sutures.

Discussion. Pervinquiere (1907) and Kennedy and Wright (1979a) amongst others have commented
on the variability in ornament and proportions of juvenile Neoptychites. The present specimens
probably belong to the cephalotus group but are best left in open nomenclature.

Subfamily pseudotissotiinae Hyatt, 1903
Genus wrightoceras Reyment, 1954
(? = Imlayiceras Leanza, 1967)

Type species. Bauchioceras (Wrightoceras) wallsi Reyment, 1954a, p. 160, pi. 2, fig. 4, pi. 3, fig. 3, by original
designation.

Discussion. See Hirano (1983) for a review of this genus.

1907
1920
non 1954

1956

71967

71967

1975

1979

1982

1982

Wrightoceras tnunieri (Pervinquiere, 1907)

Plate 10, figs. 9-11; text-fig. 2e

Hoplitoides munieri Pervinquiere, p. 217, pi. 10, figs. 1 and 2; text-fig. 83.

Hoplitoides aif. mirahilis Pervinquiere; Bose, p. 225, pi. 19, figs. 1-3.

'Hoplitoides' cf munieri Pervinquiere; Kummel and Decker, p. 317, pi. 33, figs. 1 and 2; text-
figs. la, 10 ( = Herrickiceras costation (Herrick and Johnson, 1900)).

Hoplitoides inca Benavides-Caceres, p. 475, pi. 63, figs. 6-11; text-fig. 54.

Proplacenticeras zeharense Collignon, p. 33, pi. 18, figs. 5 and 6 (77-9).

Imlayiceras washboitrni Leanza, p. 198, pi. 4, figs. 1-4; pi. 6, figs. I, 4-6.

Wrightoceras munieri (PerVmcpiitrey, Wiedmann, p. 144, pi. 2, fig. 2.

Imlayiceras 7 ralphimlayi Etayo-Serna, p. 88, pi. 13, fig. 3; text-fig. 8a.

Wrightoceras cf. munieri (Pervinquiere); Chancellor, p. 1 19, figs. 24, 60-63.

Hoplitoides munieri Pervinquiere; Renz, p. 100, pi. 31, figs. 3-6, 1 1.

Holotype. The original of Pervinquiere 1907, p. 217, pi. 10, fig. 2, from the Lower Turonian of Draa el Miaad,
Tunisia.

Material. OUM KT396-399, 531, 534, 556, 665, 706, 788, from Bed B of the Ojinaga Formation, Calvert
Canyon, Hudspeth County, Texas.

Dimensions

D Wb Wh Wb: Wh U

OUMKT396 85-5(100) 19-7(23 0) 47-4(55-4) 41-6 9-2(10-8)

Description. Very compressed, involute, with narrow sulcate-bicarinate venter.

Discussion. Our material is largely fragmentary, but obviously belongs to Pervinquiere’s species,
the types of which are before us. Imlayiceras washbournei Leanza, 1967 (p. 198, pi. 4, figs. 1-4; pi.
6, figs. 1, 4-6) closely resembles the present species but has feeble umbilical bullae on the inner
whorls, as shown in text-fig. 1 1a-d. It is regarded as a possible synonym.

The Hoplitoides cf. munieri of Kummel and Decker (1954, p. 317, pi. 33, figs. 1 and 2; text-figs.

EXPLANATION OF PLATE 10

Figs. 1-8. Allocrioceras larvatum (Conrad, 1855). 1, OUM KT315; 2, OUM KT314; 3, OUM KT659; 4,
OUM KT736; 5, OUM KT736a; 6 and 7, OUM KT336; 8, OUM KT312.

Figs. 9-11. Wrightoceras munieri (Pervinquiere, 1907). 9 and 10, OUM KT396; 11, OUM KT399.

Figs. 12 and 13. Allocrioceras dentonense Moreman, 1942. OUM KT424.

All specimens are from the Basal Turonian Pseudaspidoceras Jie.xuosum Zone fauna of Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas. Figs. 1-8, 12, 13, x 2; figs. 9-1 1, x 1.

PLATE 10

KENNEDY, WRIGHT and HANCOCK, Allocrioceras, Wrightoceras

60

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1 1 . Imlayiceras washhoumi Leanza, 1967.
The holotype, USNM 132.556, from the Lower
Turonian of Ubate, Colombia; x 1.

KENNEDY, WRIGHT AND HANCOCK; TURONIAN AMMONITES FROM TEXAS 61

la and 10) is referred by Cobban and Hook (1980, p. 22) to Herrickiceras Cobban and Hook, 1980,
a homoeomorphic Middle Turonian eoilopoceratid.

A further possible synonym of the present species is Proplacenticeras zeharense Collignon, 1967
(p. 33, pi. 18, figs. 5-9) from the Turonian of Tarfaya, Morocco, although in the absence of sutures
this is not fully proven, while one of the paratypes (pi. 18, fig. 7) shows a rounding venter.

Hirano (1983) discusses the Nigerian Wrightoceras, none of which closely resemble the present
species.

Occurrence. Lower Turonian of Texas and northern Mexico, Venezuela, Peru, ?Coloinbia, Spain, Tunisia,
and ?Morocco.

Genus thomasites Pervinquiere, 1907

( = Gomheoceras Reyment, 1954u; Koidabiceras Atabekjan, 1966; Ferganites Stankievich and Pojar-

kova, 1969)

Type species. Pachydiscus rollandi Peron, 1 890, p. 25, pi. 17, figs. 1 -3, by original designation.

Discussion. See Wright and Kennedy 1981, p. 98 for a review of the genus and its synonyms.

Thomasites adkinsi (Kummel and Decker, 1954)

Plate 9, figs. I 4, 8-12, 15 16; text-fig. 2i

1954 "Hoplitoides' adkinsi Kummel and Decker, p. 316, pi. 32, fig. 6; pi. 33, fig. 3; text-figs, lb and 8.

1957 Pseudotissotia (Bauchioceras) adkinsi (Kummel and Decker); Barber, p. 47.

1982 Pseudotissotia adkinsi (Kummel and Decker); Chancellor, p. 116, figs. 2i, 66-69 (with synonymy).

Holotype. The original of Kummel and Decker 1954, pi. 32, fig. 6, pi. 33, fig. 3, from the Turonian of Piedra
de Lumbre, Coahuila, Mexico.

Material. Six specimens, OUM KT353-356, 418, 427, from Bed B of the Ojinaga Formation, Calvert Canyon,
Hudspeth County, Texas.

Dimensions

D Wb Wh Wb.Wb U

OUMKT354 61-8(100) 26-3(42-6) 30 1(48-7) 0-87 10-0(16-2)

Description. 0-25 nun diameter. The earliest growth stages are involute (umbilicus 16% of the
diameter), with steep, slightly overhanging umbilical walls and an abruptly rounded umbilical
shoulder. The flattened flanks converge slightly towards the somewhat flattened, trituberculate
venter. On the adoral quarter of the outer whorl there are two prominent umbilical bullae from
which arise weak, slightly prorsiradiate, single ribs. The outer half of the flanks is ornamented with
weak ribs corresponding to the rows of tubercles across the venter.

25-45 mm diameter. This growth stage is represented by several fragments (PI. 9, figs. 1-4, 10-
1 1) which show the whorls to be very compressed, much higher than wide, with flattened flanks. At
this growth stage the venter is weakly fastigiate, with three prominent rows of weakly clavate
tubercles; the siphonal row already forms a weak keel.

45-65 mm diameter. This growth stage is represented by the best preserved specimen, OUM
KT354 (PI. 9, figs. 8-9, 12). The shell is rather involute (umbilicus 16 % of the diameter), compressed,
with an elliptical whorl section. The umbilicus is narrow and deep, with vertical umbilical walls and
an abruptly rounded umbilical shoulder. The broad, slightly convex flanks converge towards the
narrow, truncated venter and maximum width is just above the umbilical shoulder. The venter is
tricarinate-bisulcate although the keels are distinctly nodate. Save for growth striae, the flanks lack
ornament.

60 mm diameter. A single fragment, retaining shell, OUM KT353 (PI. 9, figs. 15 and 16) shows
the late growth stages of this species. The flanks are still flattened and converge towards the venter
which is now evenly rounded. Only the siphonal keel persists at this growth stage although the

62

PALAEONTOLOGY, VOLUME 30

ventrolateral shoulders are again marked by very weak nodes which are all that are left of the
tricarinate middle growth stages. Both the flanks and the venter are crossed by rather prominent
growth striae.

Discussion. Although referred by all previous authors to Pseudotissotia (and its synonyms), a genus
characterized by a tricarinate venter throughout most of its ontogeny, we would refer the present
species to Thomasites for the following reasons: (i) the trituberculate, fastigiate venter is retained to
advanced growth stages; (ii) the tricarinate stage is only developed for a short period during
ontogeny; and (iii) the ventrolateral keels are still weakly nodate on the body-chamber.

T. adkinsi is intermediate between early Thomasites (i.e. 'Gomheoceras') and early Pseudotissotia
(i.e. "Bauchioceras'), but is best included in the fonner. The boundaries, if any, between T. rollandi
(Peron, 1890), T. gongilensis (Woods, 1911) and T. adkinsi can only be decided on the basis of large
populations of well-preserved specimens.

Occurrence. Lower Turonian of west Texas and northern Mexico.

Suborder ancyloceratina Wiedmann, 1966
Superfamily turrilitaceae Gill, 1871
Family ANisocERATiDAE Hyatt, 1900
Genus allocrioceras Spath, 1926

Type species. By original designation: Crioceras ellipticum Woods, 1896, p. 84 {non Mantell), renamed Allocri-
oceras woodsi Spath, 1939 p. 598, = Hamites angustus J. de C. Sowerby 1850, p. 346, pi. 29, fig. 12.

Occurrence. Uppermost Cenomanian to Lower Coniacian. Species are known from western Europe, Nigeria,
Zululand, Japan, the US western interior, Texas, and Mexico.

Allocrioceras dentonense Moreman, 1942
Plate 10, figs. 12 and 13

1942 Allocrioceras dentonense Moreman, p. 209, pi. 34, fig. 4; text-fig. 2h.

1963 Allocrioceras dentonense Moreman; Swensen, p. 78, pi. 1, figs. 7 and 8; pi. 3, fig. 3.

1965 Allocrioceras dentonense Moreman; Clark, p. 32, pi. 1, figs. 7 and 8; pi. 5, fig. 3; text-fig. \2b.

Type. The holotype by monotypy is UTA 19808, from the Britton Formation east of Lewisville, Texas.

Material. OUM KT424, from Bed B of the Ojinaga Formation, Calvert Canyon, Hudspeth County, Texas.

Description. The fragment is part of a loosely coiled, planispiral shell, with a compressed, rectangular
whorl section. The broad, flattened flanks are ornamented with four rather robust, slightly prorsira-
diate ribs in a distance equal to the whorl height. All ribs terminate in small, pointed ventrolateral
tubercles which are joined across the venter by a pair of weakly looped ribs.

Discussion. The present fragment closely matches Moreman’s holotype (see Clark 1965, pi. 5, fig.
3) in whorl section and nature of the ribbing. It is not known, however, whether Moreman’s species
shows the looped ribbing across the venter evident in the fragment before us. However, Moreman
(1942) does note that in his species the ribs are ‘flat’ across the venter; such flattening is perhaps
comparable to incipient looping.

A. annulatum (Shumard, 1860) (p. 595) has an almost circular whorl section and oblique ribbing
across the venter, characters which readily separate it from the present fragment. Common features,
however, are the looped ribs across the venter and the presence of ventrolateral tubercles on every
rib. Cobban and Scott (1972) have included A. pariense (White, 1887) (p. 203, pi. 19, fig. 2) in the
synonymy of A. annulatum.

In A. larvatum (Conrad, 1855) (p. 265) the whorl section is ovate, and only slightly higher than
wide, and only every alternate rib bears ventrolateral tubercles.

Occurrence. Upper Cenomanian of central Texas; basal Turonian of west Texas.

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS 63
Allocrioceras larvatum (Conrad, 1855)

Plate 10, figs. 1-8

1855 Hamites larvatus Conrad, p. 265.

1857 Hamites larvatus Conrad; Conrad in Emory, pi. 21, fig. 8.

1928 Hamites larvatus Conrad; Adkins, p. 209.

1933 Allocrioceras larvatum (Conrad); Adkins, p. 439.

1942 Allocrioceras larvatum (Conrad); Moreman, p. 208, text-fig. 2i.

1963 Allocrioceras larvatum (Conrad); Swensen, p. 80.

1963 Allocrioceras sp. Powell, p. 322, pi. 31, fig. 18.

1965 Allocrioceras larvatum (Conrad); Clark, p. 32.

Type. Holotype, by monotypy, no. 4790 in the collections of the Philadelphia Academy of Sciences, from the
Upper Cenomanian Britton Formation of Dallas County, Texas.

Material. More than 150 specimens in the Oxford University Museum Collections from Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas.

Description. This is a rather common species in the Calvert Canyon fauna, and the ontogenetic
development can be deduced from various fragments.

The initial coiling is in a loose planispire (PI. 10, figs. 3 and 4). Up to a whorl height of 1 mm, the
shell is very weakly ornamented, almost smooth, whereafter there are slightly prorsiradiate ribs,
4-5 in a distance equal to the whorl height. They are very weak across the dorsum, but strengthen
ventrally. At this stage there is no sign of ventrolateral tubercles. Weak ventrolateral tubercles
appear on alternate ribs at a whorl height slightly in excess of 2 mm. At this growth stage the whorl
seetion is eompressed oval.

At larger growth stages (PI. 10, fig. 2) the whorl seetion is elliptical, compressed, and the flanks
show signs of flattening. There are six prorsiradiate ribs in a distance equal to the whorl height
across the flanks, of which three bear small pointed ventrolateral tubercles. There is usually only
one, rarely two non-tubereulate ribs between the tuberculate ribs, and they tend to be somewhat
finer than the tuberculate ones. The ventrolateral tubercles are joined transversely across the venter
by a broad, flattened rib which may show incipient looping.

A still larger growth stage of this species is shown by OUM KT315 (PI. 10, fig. I). This fragment
elearly shows the tendency for the shell to become higher whorled and more eompressed with
growth (see also PI. 10, fig. 8). This in turn results in an increasing rib density, with now seven ribs
in a distance equal to the whorl height. This fragment also shows one instance of two non-
tuberculate ribs between adjacent tuberculate ones.

The most signifieant variation in the material before us is in the increase in whorl height
whieh means that the later growth stages are more densely ribbed, with a more compressed whorl
section than is seen in juveniles. Add to this the absence of ventrolateral tubercles on the earliest
whorls, and the two growth stages could readily be separated into diflerent genera: Hamites and
Allocrioceras.

Discussion. The present species diflers from A. dentonense in being more densely ribbed at compar-
able growth stages and generally with tubereles on every alternate rib. A. annulatum has an almost
eircular whorl section at large growth stages and also has tubercles on every rib.

Occurrence. Upper Cenomian and basal Turonian. West and central Texas and northern Mexico.

Family baculitidae Gill, 1871
Genus sciponoceras Hyatt, 1 894

(= Cv/'/oc/n7t« Meek, 1876 (non Jakowlew, 1875); Cvrtoc/;//£’//o Strand, 1929)

Type species. By original designation: Hamites baculoides Mantell, 1822, p. 123, pi. 23, figs. 6 and 7.
Discussion. See Wright and Kennedy 1981, p. 112 for recent comments on this genus.

64 PALAEONTOLOGY, VOLUME 30

Sciponoceras sp.

Plate 2, figs. 5-7

Material. OUM KT334 and 652, from Bed B of the Ojinaga Formation, Calvert Canyon, Hudspeth County,
Texas.

Description. The shell is slowly expanding with a compressed whorl section (the whorl breadth to
height ratio is 0-62), broadly rounded flanks, and narrowly rounded dorsum and venter. There is a
single concave constriction on one specimen. The sutures are not exposed.

Remarks. The presence of a constriction confirms this as a Sciponoceras rather than Baculites. It is
specifically indeterminate.

Superfamily scaphitaceae Gill, 1871
Family scaphitidae Gill, 1871
Subfamily otoscaphitinae Wright, 1953
Genus worthoceras Adkins, 1928

Type species. Macroscaphites platydorsatus Scott, 1924, p. 18, pi. 5, pi. 6, pi. 9, fig. 6, by original designation.

Diagnosis. Small, strongly dimorphic. Evolute spire followed by moderately to very long straight
shaft and terminal hook; aperture, where known, simple in macroconchs and with long, straight
lappets in microconchs. Shell surface smooth or with very feeble ribs. Sutures simple, quadrilobate,
with bifid saddles and trifid to bifid lobes.

Occurrence. Upper Albian to Upper Turonian of western and central Europe, New Zealand, Texas, and the
US Western Interior.

Worthoceras cf. vcrm/cw/M^ (Shumard, 1860)

Plate 8, fig. 5

compare:

1 860 Scaphites venniculus Shumard, p. 594.

1965 Worthoceras vermiculum (Shumard); Clark, p. 62, pi. 4, figs. 9-11.

1972 Worthoceras vermiculum (Shumard); Cobban and Scott, p. 43 (with synonymy).

Type. Shumard’s original material appears lost. Moreman (1942) established a neotype, UTA 19827; it is from
the Upper Cenomian Britton Formation 0-5 miles east of the Britton-Midlothian highway, 2-7 miles south of
Britton, Texas.

Material. One distorted near complete specimen and two spires on OUM KT698, from Bed B of the Ojinaga
Formation, Calvert Canyon, Hudspeth County, Texas.

Description. These tiny specimens add nothing to previous descriptions, and preservation precludes
confident determination.

Remarks. The trivial name of this form should be vermiculus {Latin, little worm) since it is a noun
in apposition to the genus.

Occurrence. Upper Cenomian and Lower Turonian of central and west Texas, Colorado, Kansas, and Mon-
tana.

DISCUSSION

The fauna of the Pseudaspidoceras flexuosum Zone in west Texas is characterised by abundant
Quitmaniceras reaseri, P. flexuosum, Vascoceras proprium, Fagesia catinus, and A. larvatum; less
common are Mammites powelli, Neoptychites sp., Wrightoceras munieri, T. adkinsi, A. dentonense,
Sciponoceras sp., and Worthoceras cf. vermiculus. Inoceramid bivalves are abundant, with many
Mytiloides (text-fig. 12a-c). The unpublished work of Dr W. A. Cobban (letter of 6 March 1985)

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

65

TEXT-FIG. 12. A-c, MytUoules columbianus (Heinz, 1935), the M. opalensis of authors non Bose and M. pUcatus
(d’Orbigny) that typifies the Basal Turonian. x 1. a, OUM KT488; b, OUM KT510; c, OUM KT470; all are
from the Basal Turonian Pseudaspidoceras fiexuoswn Zone of Calvert Canyon, Texas, d shows the syntypes
of Inoceramus plicatus, no. 5504 in the D’Orbigny Collection, Museum d'Histoire Naturelle, Paris, from ‘Santa
Fe du Bogota, Nouvelle Grenade’, e, a cast of the original of M. opalensis Bose, 1923, p. 184, pi. 13, figs. 2

and 3, from Opal, Zacatecas, Mexico.

shows that this species, the M. opalensis of Seitz and other authors, also referred to M. plicatus
(d’Orbigny, 1842), the surviving syntypes of which are shown as text-fig. 12d, should be called M.
columbianus (Heinz, 1935). True M. opalensis (B5se 1923, pi. 13, figs. 2 and 3 only) (text-fig. 12e) is
a younger Turonian species.

At Calvert Canyon this assemblage is found in stratigraphic isolation, but, as noted in the
introduction to this paper, it occurs elsewhere in west Texas (e.g. Chispa Summit) and New Mexico
(Cobban 1984), immediately above a distinctive assemblage of the Neocardioceras juddii Zone,

66

PALAEONTOLOGY, VOLUME 30

regarded as Upper Cenomanian. The juddii Zone fauna in this region includes N.juddii (Barrois and
Guerne, 1878), P . pseudonodosoides (Choffat, 1898) (text-fig. 6a, b), Kamerunoceras, and Vascoceras
species. We take the base of i\\Q flexuosum Zone as the base of the Turonian stage for reasons
discussed fully elsewhere (Hancock 1984; Kennedy 1984; see also Cobban 1984; Birkelund et al.
1984). Here we discuss the correlation of this boundary with some of the better known boundary
successions in both Boreal and Tethyan Realms.

United States Western Interior

The best documented Cenomanian-Turonian boundary section is Rock Canyon, Pueblo, Colorado
(Cobban and Scott 1972), and the boundary is discussed in detail by Kauffman et al. (1978) who
show the intervention of an anoxic interval around this level. No juddii Zone forms are recorded
from the Pueblo area, but the boundary lies above Bed 77 and below Bed 97, and possibly between
Beds 79 and 86 of the published section (Cobban and Scott 1972, pp. 23-24). However, fossils of
the N. juddii and P. flexuosum Zones do occur in the western interior (Cobban 1983, 1984), while
the presence of an ammonite and inoceramid association typical of the flexuosum Zone as far north
as Montana (Reeside 1923) shows that there is no biogeographic hindrance to recognizing this
boundary throughout the region.

Western Europe

The proposed correlation is shown in text-fig. 13. The wide occurrence of N.juddii is now well
known (Wright and Kennedy 1981; Amedro et al. 1984) and we have new records from Haute
Normandie, Aube, and West Germany. It is unfortunate that around this level there is either
sedimentary condensation or anoxic black shale, and interpretation of many sections is difficult. In
southern England the condensed successions of the Beer district (Wright and Kennedy 1981; Jarvis
and Woodroof 1984) show the juddii Zone overlain by what has been termed a Watinoceras
coloradoense Zone fauna of the Lower Turonian with diverse Watinoceras species, none of which
occurs in Xhc flexuosum Zone in south-west Texas. The ammonite assemblage is nevertheless ac-
companied by diverse inoceramids, including numerous Mytiloides conspecific with the form we
refer to as M. columbianus. The coloradoense Zone is in turn overlain by the Mammites nodosoides
Zone that contains elements in common with the nodosoides Zone in the interior of the United
States, so that it is likely that the coloradoense Zone of western Europe corresponds to parts or all
of the birchbyi dind flexuosum Zones of the USA.

P . flexuosum is known from near Dresden in the German Democratic Republic (the "M. footeanus
Stol. spec.’ of Petrascheck 1902 p. 14, pi. 9, fig. 1), but is, unfortunately, imprecisely dated with
respect to other ammonites in this region.

Iberian Peninsula

Berthou and Lauverjat and their collaborators have described the sequence in the Vascoceras-
dominated successions of the Portugese littoral in a series of papers, summarized by Berthou (1984).
Part of the key to the correlation shown in text-fig. 1 3 is the extension of P. pseudonodosoides up to
Division J of the succession of Choffat (1898); this is an exclusively juddii Zone species in North
America. It is accompanied by the type material of Spat bites subconciliatus (Choffat, 1898) which
thus appears in Portugal, as in England, as low as the juddii Zone. The upper limit of S. subconciliatus
is less certain. In Portugal it does not range above the juddii Zone Level J (Berthou 1984), but in
northern Spain it is recorded by Wiedmann (1978) from the same division as Mytiloides opalensis
(Seitz non Bose, e.g. M. columbianus), the typically flexuosum Zone inoceramid. In western Aquitaine
it probably occurs still higher, the variable Spathites populations of the Mammites nodosoides Zone
yielding individuals close to the types of subconciliatus. In England there is a possible example from
the coloradoense Zone in Devon (Wright and Kennedy 1981). It would seem that S. subconciliatus
is a species with a total range through several zones.

The lowest definite Turonian assemblage in Portugal is found in Level L of Choffat. None of the
species quoted by Berthou has yet been found in Texas and no direct correlation is at present
possible.

Wrightoceras mumeri Pseudohssotia (Bauc'hioceras)

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

67

(N

c3

O

^

2 c

0>
73 X!
c c

J [1, -O

rt ^

C ^ O

§ "S s.

cd CT

C/3

<U

3 6^ '

ON •

' — C

— <u
00 5 o

e ^

T3 nj

(U

c/5

O

<U

T3
C ~ u

« Tf ^

00

U On ^

J- ^ O
^ 3

d! o
^ r *=Jj
— CQ —
^T3 .22
^ C _

03

C3

■3 00 .22
5 os t;
o 00 i2
o — o

« u

c o

S ^ o
c -c y

2 H

o

U- 00

C Os

^ —

• ^ m
c ^ <u
-O

c CJ)

g CO O

§

^ • -n

^ 'S C

03

N

o ,WJ o

«■§!

2 ^

O GO NO
O ON

: >.
: T3

rt
00
o3 2
) t=

c

<D Q
73

a ^
w o>
X J
toj;)

u '?

^ o3

On
c3 —

opinion is that: Fallotiles subconciliatus = Spatbites (Jeanrogericeras) subconciliatus', Gombeoceras gongilense = fhomasites gongilensis:

Paravascoceras costatum = Vascoceras proprium.

68

PALAEONTOLOGY, VOLUME 30

Israel

The sequences in the Negev (Freund and Raab 1969; text-fig. 13) provide a standard for much of
the Middle East and the Sahel. Lewy et al. (1984) have recorded Metoicoceras gesliniamim of
the Upper Cenomanian from the Kanabiceras Zone of Freund and Raab together with early
representatives of V. cauvini, while the cauvini Zone of Freud and Raab yields P. pseudonodosoides
(text-fig. 6e, f) showing it to be equivalent to the juddii Zone. There are no common elements
between the flexuosum Zone faunas discussed here and those of the Israeli pioti Zone, but our
unpublished work in central Tunisia shows that V. durandi (Peron, 1890), a species known from
the pioti Zone, appears above the correlative of the juddii Zone in association with fragmentary P.
cL flexuosum.

Nigeria

The important vascoceratid-dominated faunas of this country provide a key to wider correlation
across the trans-Saharan seaway (see Dufavre et al. 1984). These faunas have been discussed by
Reyment (1954fl, b, 1955) and Barber (1957), whilst Wozny and Kogbe (1983) have reviewed key
sections in the Upper Benue Basin. P. flexuosum occurs in north-east Nigeria (the P. paganum
Reyment of Barber (1957, p. 9) in part belongs to this species), while V. proprium, the commonest
ammonite in {ht flexuosum Zone in west Texas, was first described from Bauchi Province.

Actual correlations are more difficult than might be hoped, in spite of detailed stratigraphy by
Barber (1957) and Wozny and Kogbe (1983). Pseudaspidoceras has not been recorded from an
exact level. V. proprium (as Paravascoceras costatum costatum and P. c. tectiforme by Wozny and
Kogbe and as V. globosum in both papers) occurs in the costatum Zone according to Wozny and
Kogbe but only in the upper half of the Zone according to Barber. This would place the base of the
Turonian at the base or in the middle of the Nigerian costatum Zone.

On the other hand the costatum Zone also contains:

(i) Metengonoceras dumbli (Cragin 1893) which in northern France extends to the upper part of
the gesliniamim Zone and does not range above the Cenomanian in north Texas, the region where
it is best known (Hancock and Kennedy 1981).

(ii) T. gongilensis including subsp. tectiformis and lautus that in Devon occurs below and in the
juddii Zone, i.e. distinctly below the summit of the Cenomanian.

(iii) Nigericeras, a genus often listed as ‘Lower Turonian’, although all the stratigraphically
controlled records are within the Cenomanian, e.g. below the juddii Zone in Devon, from the
Kanabiceras Zone in Israel, and the Upper Cenomanian of Angola (Wright and Kennedy 1981).

It may well turn out that the finer divisions of Nigeria cannot be recognized internationally but,
provisionally, we would place the base of the Turonian as somewhere within the costatum Zone of
Barber, i.e. within Bed 9 of the Pindiga section (Barber 1957, table 3).

Other regions

As yet it is not really possible to make detailed comparisons with other regions famous for their
mid-Cretaceous ammonites.

In Japan there are too few ammonites from critical levels to allow useful discussion (Matsumoto
1977, 1982).

In Soviet central Asia there are almost no ammonite speeies in common with the other regions
discussed here (Pojarkova 1984), further emphasized since Wright and Kennedy (1981) have main-
tained that T. koulabicum (Kler), the most common ammonite in Pojarkova’s Lower Turonian, is
distinct from all the Nigerian forms.

In southern India there are several zones missing in the succession close to the Cenomanian-
Turonian boundary aceording to Ayyasami and Banerji (1984).

The rich Madagascan assemblages suffer from inadequate stratigraphic control (Besairie and
Collignon 1972). We have not yet made a full study of our material from Tunisia.

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS

69

Acknowledgements. We thank Don Reaser of Arlington, Texas, for guiding us in the Quitman Mountains,
Ray Parish of London for driving us there, and Mann and Pat Bramblette for allowing us to wander freely
over their ranch. Our collections were initially sorted by Dr M. R. Cooper (now at the University of Natal),
but we found ourselves in disagreement with his taxonomy. We have benefitted enormously from discussions
with Dr W. A. Cobban of the United States Geological Survey, Denver, Colorado, who kindly allowed us to
quote his unpublished work on Lower Turonian Inoceramus. The financial support of the Natural Environment
Research Council is gratefully acknowledged. This is a contribution to IGCP project 58 on ‘Mid-Cretaceous
Events’.

REFERENCES

ADKINS, w. s. 1928. Handbook of Texas Cretaceous fossils. Bull. Univ. Tex. 2838, 385 pp., 37 pis.

1931. Some Upper Cretaceous ammonites in western Texas. Ibid. 3101, 35-72, pis. 2-5.

1933. The Mesozoic systems in Texas, 239-518. In sellards, e. h., adkins, w. s. and plummer, f. b. The

geology of Texas 1. Stratigraphy. Ibid. 3232, 1007 pp., 1 1 pis. (mis-dated 1932).

AMARD, B., COLLIGNON, M. and ROMAN, J. 1983. Etude stratigraphique et paleontologique du Cretace superieur
et du Paleocene du Tinrhert-W et Tademai't-E (Sahara Algerien). Docum. Lab. geol. Lyon, Hor. Ser. 6, 15-
173, 17 pis. (for 1981).

ANDERSON, F. M. 1931. The genus Fagesia in the Upper Cretaceous of the Pacific Coast. J. Paleont. 5, 121-126,
pis. 15-16.

1958. Upper Cretaceous of the Pacific coast. Mem. geol. Soc. Am. 71, xi + 378 pp., 75 pis.

ATABEKJAN, A. A. 1966. New genus Koulahiceras gen. nov. from the Turonian of the eastern parts of Central
Asia. Isvest. Akad. Nauk Armen. S.S.R. 19, 75-78. [In Russian.]

AYYASAMi, K. and BANERji, R. K. 1984. Cenomanian-Turonian transition in the Cretaceous of southern India.
Bull. geol. Soc. Denmark, 33, 21-30.

BARBER, w. 1957. Lower Turonian Ammonites from north-eastern Nigeria. Bull. geol. Surv. Nigeria, 26, 86
pp., 34 pis.

BARROis, c. and guerne, j. de. 1878. Description de quelques especes nouvelles de la Craie de I’Est du Bassin
de Paris. Annls Soc. geol. N. 5, 42-64, 3 pis.

BASSE, E. 1937. Les Cephalopodes cretaces des massifs cotiers syriens, pt. 1. Notes Mem. Ht.-Comm. Srrie
2, 165-200, pis. 8-11.

benavides-caceres, v. e. 1956. Cretaceous system in northern Peru. Bull. Am. Mus. nut. Hist. 108, 353-494,
pis. 31-66.

BENGTSON, p. 1983. The Cenomanian -Coniacian of the Sergipe Basin, Brazil. Fossils and Strata, 11, 1-78.
BERTHOU, p. Y. 1983. Lcs liiuites et subdivisions des etages Albien a Turonian dans le bassin occidentale
portugais avec un interet special pour la limite Cenomanien-Turonien dans les deux facies: a ammonites
(Vascoceratides) ou a rudistes, 21-24. In Abstracts. Cretaceous Stage Boundaries. University of Copenhagen,

210 pp.

1984. Albian-Turonian stage boundaries and subdivisions in the western Portugese Basin, with special

emphasis on the Cenomanian-Turonian boundary in the ammonite facies and rudist facies. Bull. geol. Soc.
Denmark, 33, 41 - 55.

BESAIRIE, H. and COLLIGNON, M. 1972. Geologie de Madagascar 1. Les terrains sedimentaires. Annls geol.
Madagascar, 35, 552 pp.

BIRKELUND, T., HANCOCK, J. M., HART, M. B., RAWSON, P. F., REMAINE, J., ROBASZYNSKI, F., SCHMID. F. and SURLYK, F.

1984. Cretaceous stage boundaries— proposals. Bull. geol. Soc. Denmark, 33, 3-20.

BOSE, E. 1920. On a new ammonite fauna of the Lower Turonian of Mexico. Univ. Texas Bull. 1856, 173-257,
pis. 12-20 (misdated 1918).

1923. Algunas faunas Cretacicas de Zacatecas, Durango y Guerrero. Boln Inst. geol. Mex. 42, iv-l-1-

219, pis. 1-19.

BRONGNiART, A. 1822. Sur quelques terrains de Craie hors du Bassin de Paris, 80-101. In cuvier, g. and
BRONGNIART, A. Description geologique des environs de Paris, 3rd edn., 428 pp., 1 1 pis. Chez G. Dufour et E.
d’Ocagne, Librairies, Paris.

CHANCELLOR, G. R. 1982. Cenomanian —Turonian ammonites from Coahuila, Mexico. Bull. geol. Instn Univ.
Upsala, NS 9, 77 129.

REYMENT, R. A. and TAiT, E. A. 1977. Notes on Lower Turonian ammonites from Loma el Macho,

Coahuila, Mexico. Ibid. 7, 85-101.

70

PALAEONTOLOGY, VOLUME 30

CHOFFAT, p. 1898. Recueil d’etudes paleontologiques sur la faune cretacique du Portugal. I, especes nouvelles
ou peu connues. Deuxieme serie, Les Ammonees du Bellasien, des couches a Neolobites Vibrayeanus, du
Turonien et du Senonien. Trav. geol. Portugal, (1898), 41-86, pis. (cephalopodes) 3-22.

CHUDEAU, R. 1909. Ammonites du Damergou (Sahara meridional). Bull. Soc. geol. Fr. (4), 9, 67-71, pis. 1-3.
CLARK, D. L. 1965. Heteromorph ammonoids from the Albian and Cenomanian of Texas and adjacent areas.
Mem. geol. Soc. Am. 95, viii + 99 pp., 24 pis.

COBBAN, w. A. 1983. Mid-Cretaceous ammonite zones. Western Interior, United States, 37-38. In Abstracts.
Cretaceous Stage Boundaries. University of Copenhagen, 210 pp.

1984. Mid-Cretaceous ammonite zones. Western Interior, United States. Bull. geol. Soc. Denmark, 33,

71-89.

and HOOK, s. c. 1979. Collignoniceras woollgari woollgari (Mantell) ammonite faunas from Upper Cre-
taceous of Western Interior, United States. Mem. Inst. Min. Teclmol. New Mex. 37, 51 pp., 12 pis.

1980. The Upper Cretaceous (Turonian) ammonite family Coilopoceratidae Hyatt in the Western

Interior of the United States. Prof. pap. US geol. Surv. 1192, 28 pp., 21 pis.

1983a. Mid-Cretaceous (Turonian) ammonite fauna from Fence Lake area, west-central New

Mexico. Mem. Inst. Min. Teclmol. New Mex. 41, 50 pp., 14 pis.

1983^. Pseudaspidoceras pseudonodosoides — common Upper Cretaceous guide fossil in southwest

New Mexico. Rep. Inst. Min. Teclmol. New Mex. 1981-1982, 37-40.

and SCOTT, g. r. 1972. Stratigraphy and ammonite fauna of the Graneros Shale and Greenhorn Limestone

near Pueblo, Colorado. Prof. pap. US geol. Surv. 645, 108 pp., 41 pis.

COLLiGNON, M. 1957. Cephalopodes neocretacees du Tinrhert (Fezzan). Annls Paleont. 43, 113-136, pis. 16-
18.

1965. Atlas des fossiles caractaistiques de Madagascar (Ammonites), XII (Turonien), iv + 82 pp., pis. 376-

413. Service Geologique, Tananarive.

1967. Les cephalopodes cretaces du bassin cotier de Tarfaya. Notes Mem. Serv. Mines Carte geol. Maroc.

175, 7-148, 35 pis. (1966).

CONRAD, T. A. 1855. Descriptions of eighteen new Cretaceous and Tertiary fossils. Proc. phila. Acad. Nat. Sci.
7, 265-268.

COOPER, M. R. 1979. Ammonite evolution and its bearing on the Cenomanian-Turonian boundary problem.
Paldont. Z. 53, 120-128.

COURTiLLER, M. A. 1860. Description de trois nouvelles especes d’ammonites du terrain cretace. Mem. Soc.
Imp. Agric. Sci. Arts Angers, 3, 246-252, pis. 1-3.

CRAGiN, F. w. 1893. A contribution to the invertebrate paleontology of the Texas Cretaceous. Tex. geol. Surv.
4th Annual Report (1892), 139-246, pis. 24-46.

DOUVILLE, H. 1912. Evolution et classification des Pulchelliides. Bull. Soc. geol. Fr. (4) 11 (for 1911), 285-
320.

DUFAURE, p., FOURCADE, E. and MASSA, D. 1984. Realitc des communications marines trans-sahariennes entre
la Tethys et I’Atlantique durant le Cretace superieur. C.r. hebd. Seanc. Acad. Sci. Paris, (2) 298, 665-670.
ECK. o. 1909. Bemerkungen fiber drei neue Ammoniten aus der oberen egyptischen Kreide. Sber. Ges. Natur.
Freunde Berl. 1909, 179-191.

ETAYO-SERNA, F. 1979. Zonation of the Cretaceous of central Columbia by Ammonites. Pubis geol. espec.
Ingeominas, Bogota, 2, 186 pp., 15 pis.

FREUND, R. and RAAB, M. 1969. Lower Turonian ammonites from Israel. Spec. Pap. Palaeont. 4, v + 83 pp.,
10 pis.

FURON, R. 1935. Le Cretace et le Tertiaire du Sahara soudanais (Soudan, Niger, Tshad). Archs. Mus. natn.
Hist. nat. Paris, (6) 13, 1 -96, pis. 1 -7.

GILL. T. 1871. Arrangement of the Families of Mollusks. Smiths. Misc. Coll. 227, xvi+49 pp.

GROSSOUVRE, A. DE. 1894. Recherches sur la craie superieure. 2: Paleontologie— les ammonites de la craie
superieure. Mem. Serv. Carte geol. det. Fr. ii + 264 pp., 39 pis. (misdated 1893).

HANCOCK, j. M. 1984. Some possible boundary-stratotypes for the base of the Cenomanian and Turonian
stages. Bull. geol. soc. Denmark, 33, 123-128.

and KENNEDY, w. j. 1981. Upper Cretaceous ammonite stratigraphy: some current problems. In house,

M. R. and SENIOR, i. r. (eds.). The Ammonoidea. Spec. Vol. Syst. ^45^. 18, 531-553.

HEINZ, R. 1935. Unterkreide-Inoceramen von der Kapverden-Insel Maio. Neues Jb. Miner. Geol. Paldont.
BeilBd. 13 Abt. B, 302-311.

HERRICK, c. L. and JOHNSON, D. w. 1900. Geology of the Alberquerque Sheet. Bull. Univ. New Mex. geol. Ser.
1, 1-67, 32 pis.

KENNEDY, WRIGHT AND HANCOCK: TURONIAN AMMONITES FROM TEXAS 71

HiRANO, H. 1983. Revision of two vascoceratid ammonites from the Upper Cretaceous of Nigeria. Bull. Sci.
Eng. Res. Lab., Waseda Univ. 105, 44-79, 5 pis.

HOOK, s. c. and cobban, w. a. 1981. Late Greenhorn (Mid Cretaceous) discontinuity surfaces, southwest New
Mexico. Circ. Bur. Min. Technol. New Mex. 180, 5-21, pis. 1-3.

1983. Mid-Cretaceous mollusc sequence at Gold Hill, Jeff Davis county, Texas, with comparison

to New Mexico. Ibid. 185, 48-54.

HYATT, A. 1870. Report on the Cretaceous fossils, 385-393. In hartt, c. f. Geology and physical geography of
Brazil. Field, Osgood and Co., Boston.

1894. Phylogeny of an acquired Characteristic. Proc. am. Phil. Soc. 32, 349-647, pis. 1-14.

1900. Cephalopoda, 502-604. In zittel, k. a. von. 1896-1900. Textbook of Palaeontology, transl. east-

man, c. R. Macmillan, London and New York.

1903. Pseudoceratites of the Cretaceous. Monogr. US geol. Surv. 44, 351 pp., 47 pis.

JAKOWLEW, B. 1875. Hemiptera and Homoptera of the Russian Fauna. Bull. Soc. Nat. Moscou, 49, 248-285.
[In Russian.]

JARVIS, I. and woodroof, p. b. 1984. Stratigraphy of the Cenomanian and basal Turonian (Upper Cretaceous)
between Branscombe and Seaton, S.E. Devon, England. Proc. Geol. 95, 193-215.

JIMBO, K. 1894. Beitrage zur Kenntniss der Fauna der Kreideformation von Hokkaido. Palaeont. Abh. nf 2,
147-194, pis. 17-25.

JONES, B. R. and reaser, d. f. 1970. Geology of Southern Quitman Mountains, Hudspeth County, Texas.
Geology of the Southern Quitman Mountains area trans-Pecos Texas, Permian Basin Section of Soc. econ.
Pal. Min. 70-12, 31-54.

KAUFFMAN, E. G. 1970. Population systematics, radiometrics and zonation— a new biostratigraphy. Proc. N.
Amer. Paleont. Conv. September 1969, part F, 612-666.

COBBAN, w. A. and EICHER, D. 1978. Albian through lower Coniacian strata, biostratigraphy and principal

events, western interior United States. Ann. Mus. Hist. nat. Nice, 4 (for 1976), XXIII. 1-52, 17 pis.

KENNEDY, w. J. 1984. Ammonite faunas and the ‘standard zones’ of the Cenomanian to Maastrichtian stages
in their type areas, with some proposals for the definition of the stage boundaries by ammonites. Bull. geol.
Soc. Denmark, 33, 147-161.

1985. Integrated macrobiostratigraphy of the Albian to basal Santonian, 91-108. In reyment, r. a. and

BENGTSON, p. (Compilers). Mid-Cretaceous Events; report on results obtained 1974-1983 by IGCP Project
58. Pubis. Palaeont. Inst. Univ. Uppsala, Spec. Vol. 5, 132 pp.

AMEDRO, F., BADiLLET, G., HANCOCK, J. M. and WRIGHT, c. w. 1984. Notes On late Cenomanian and

Turonian ammonites from western Prance. Cret. Res. 5, 29-45.

and WRIGHT, c. w. 1979u. Vascoceratid ammonites from the type Turonian. Palaeontology, 22, 665-683,

pis. 82-86.

\919b. On Kamerunoceras Reyment, 1954 (Cretaceous: Ammonoidea). J. Paleont. 53, 1 165- 1 178, 4 pis.

and HANCOCK, j. m. 1983. Ammonite zonation and correlation of the uppermost Cenomanian and

Turonian of southern England and the type areas of Sarthe and Touraine in France. Mem. Mus. natn. Hist,
nat. Paris, Ser. C, 49, 175-181.

KOSSMAT, F. 1895-1898. Untersuchungen iiber die Siidindische Kreideformation. Beitr. Paldont. Geol. Ost.-
Ung. 9(1895), 97-203 ( 1-107), pis. 15-25 (1-11); 11 (1897), 1-46(108-153), pis. 1-8 (12-19); II (1898), 89-
152 (154-217), pis. 14-19 (20-25).

KUMMEL, B. and DECKER, J. M. 1954. Lower Turonian ammonites from Texas and Mexico. J. Paleont. 28, 310-
319, pis. 30-33.

LAUBE, G. c. and bruder, g. 1887. Ammoniten der bohmischen Kreide Palaeontographica, 33, 217-239, pis.
23-29.

LEANZA, A. F. 1967. Algunos ammonites nuevos o poco conocidos del Turoniano de Colombia y Venezuela.
Acta. Geol. Lilloana, 9, 189-213, 7 pis.

LEWY, z., KENNEDY, w. J. and CHANCELLOR, G. R. 1984. Co-occurrence of Metoicoceras geslinianum (d’Orbigny)
and Vascoceras cauvini Chudeau (Cretaceous Ammonoidea) in the Southern Negev (Israel) and its strati-
graphic implications. Newsl. Stratigr. 13, 67-76.

MANTELL, G. A. 1822. The fossils of the South Downs; or illustrations of the geology of Sussex, xvi + 327 pp.,
42 pis. Lupton Relfe, London.

MATSUMOTO, T. 1959. Upper Cretaceous ammonites of California. Part 1. Mem. Fac. Sci. Kyushu Univ., Ser.
D. (Geol.), 8, 91-171, pis. 30-45.

1973. Vascoceratid ammonites from the Turonian of Hokkaido. Proc. palaeont. Soc. Japan, ns 89,

27-41.

72 PALAEONTOLOGY, VOLUME 30

MATSUMOTO, T. 1977. Zonal correlation of the Upper Cretaceous of Japan. Spec. Pap. palaeont. Soc. Japan,
21, 63-74.

1982. Upper Cretaceous ammonites from the Monobe area, Shikoku. Spec. Pap. palaeont. Soc. Japan,

25, 31-52, pis. 1-7.

and KAWASHiTA, Y. 1978. In matsumoto, t., kawashita, y., fujishima, y. and miyauchi, t. 1978.

Manimites and allied ammonites from the Cretaceous of Hokkaido and Saghalien. Mem. Fac. Sci. Kyushu
Univ., Ser. D. (Geol.), 24, 1-24, pis. 1-6.

MEEK, F. B. 1876. A report on the invertebrate Cretaceous and Tertiary fossils of the upper Missouri county.

In FIAYDEN, F. V. Rep. US geol. geogr. Surv. Territ. 9, lxiv + 629 pp., 45 pis.

MOREMAN, w. L. 1942. Paleontology of the Eagle Ford group of north and central Texas. J. Paleont. 16, 192-
220, pis. 31-34.

MORROW, A. L. 1 935. Cephalopods from the Upper Cretaceous of Kansas. Ibid. 9, 463-473, pis. 49-53.

NOWAK, J. 1913. Untersuchungen fiber die Cephalopoden der oberen Kreide in Polen. Ill Teil. Bull. int. Acad.
Cracovie (Acad. pol. Sci.), (1913), 335-415, pis. 40-45.

ORBiGNY, A. d’. 1 850. Prodrome de Paleontologie stratigraphique universelle des animaux mollusques el rayonnes,
2, 428 pp. Masson, Paris.

PERON, A. 1890-1893. Description des mollusques fossiles des terrains Cretaces de la region sud des Hauts-
Plateaux de la Tunisie recueillis en 1885 et 1886 par M. Phillippe Thomas. Explor. sci. Tunisie, xii + 405 pp.,
35 pis. xii + 1-103 (1890); 105-327 ( 1891); 328-405 ( 1893). Masson, Paris.

1896-1897. Les ammonites du Cretace superieur de I’Algerie. Mem. Soc. geol. Fr. Paleont. 17, 88 pp., 18

pis. 6, 1-24, pis. 14-19 (1-6) (1896); 7, 25-88, pis. 7-18 (1897).

PERViNQUiERE, L. 1907. Etudes de paleontologie tunisienne. 1. Cephalopodes des terrains secondaires. Carte
geol. Tunisie, v + 438pp., 27 pis. De Rudeval, Paris.

PETRASCHECK, w. 1902. Die Ammoniten der sachsischen Kreideformation. Beitr. Paldont. geol. Ost-Ung. 14,

131- 162, pis. 7-12.

POJARKOVA, z. N. 1984. The Cenomanian and Turonian in northeastern Central Asia. Cret. Res. 5, 1-14.
POWELL, J. D. 1963. Cenomanian-Turonian (Cretaceous) ammonites from Trans-Pecos Texas and north-
eastern Chihuahua, Mexico. J. Paleont. 37, 309-322, pis. 31-34.

1965. Late Cretaceous platform-basin facies, northern Mexico and adjacent Texas. Bull. Am. Petrol.

Geol. 49, 511-525.

REESiDE, J. B. 1923. A new fauna from the Colorado group of Southern Montana. Prof. Pap. US geol. Surv.

132- 8,25-33, pis. 11-21.

RENZ, o. 1982. The Cretaceous ammonites of Venezuela, 132 pp., 40 pis. Maraven, Basel.

REYMENT, R. A. 1954a. New Turonian (Cretaceous) ammonite genera from Nigeria. Colon, geol. Surv. Min.
Resour. Div. 4, 149-164, 4 pis.

1954^. Some new Upper Cretaceous ammonites from Nigeria. Ibid. 248-270, 5 pis.

1955. The Cretaceous Ammonoidea of southern Nigeria and the Southern Cameroons. Bull. geol. Surv.

Nigeria, 25, 112 pp., 25 pis.

ROMAN F. 1938. Les ammonites jurassiques et cretacees. Essai de genera, 554 pp., 53 pis. Masson, Paris.
SCHLANGER, s. o., ARTHUR, M. A., JENKYNS, H. c. and SCHOLLE, p. A. (in press) The Cenomanian-Turonian
anoxic event. 1. Stratigraphy and distribution of organic carbon-rich beds and the marine ‘^C excursion. Jl
geol. Soc. Lond.

SCHLUTER c. 1871-1876. Cephalopoden der oberen deutschen Kreide. Palaeontographica, 21, 1-24, pis. 1-8
(1871); 21, 25-120, pis. 9-35 (1872); 24, 1-144 (121-264)-f x, pis. 36-55 (1876).

SCHNEEGANS, D. 1943. Invertebres du Cretace superieur du Damergou (Territoire du Niger). In Etudes stratigra-
phiques et paleontologiques sur le Bassin du Niger. Bull. Div. Mines Afr. occid. fr. 7, 87- 1 50, 8 pis.

SCHOBEL, J. 1975. Ammoniten der Familie Vascoceratidae aus dem unteren unterturon des Damergou-gebeites,
Republique du Niger. Spec. Publ. Palaeont. Inst. Univ. Uppsala, 3, 136 pp., 6 pis.

SCOTT, G. 1924. Some gerontic ammonites of the Duck Creek Formation. Texas Christian Univ. Quart. 1(1),
31 pp., 9 pis.

SEiBERTZ, E. 1979. Biostratigraphie im Turon des SE Miinsterlands und Anpassung an die internationale
Gliederung aufgrund von Vergleichen mit anderen Oberkreide-Gebieten. Newsl. Stratigr. 8, 111-123.

SEITZ, o. 1952. Die Oberkreide-gliederung in Deutschland nach ihrer anpassung an das internationale schema.
Z. dt. geol. Ges. 104, 148-151, 1 pi.

1956. liber Ontogenie, Variabilitat und Biostratigraphie einiger Inoceramen. Paldont. Z. 30, 3-6,

1 pi.

SHARPE, D. 1853-57. Description of the fossil remains of Mollusca found in the Chalk of England. I, Cephalo-

KENNE DY, WRIGHT AND HANCOCKiTURONlAN AMMONITES FROM TEXAS

73

poda. Monogr. palaeontogr. Soc. 68 pp., 27 pis. 1-26, pis. 1-10, 1853; 27-36, pis. 11-16, 1855; 37-8,
pis. 17-27, 1857.

SHUMARD, B. F. 1860. Descriptions of new Cretaceous fossils from Texas. Trans. Acad. Sci. St. Louis, 1, 590-
610.

SOLGER, F. 1904. Die Fossilien der Mungokreide in Kamerun und ihre geologische Bedeutung, mit besonderer
Beriicksichtigung der Ammoniten. In esch, e., solger, f., oppenheim, p. and jaekel, o. Beitrdge zur Geologic
von Kamerun, 2, 85-242, pis. 3-5. E. Schweizerbart’sche Verlagsbuchhandlung (E. Nagele). Stuttgart.
SORNAY, J. 1981. Inocerames (Bivalvia) du Turonien inferieur de Colombie (Amerique du Sud). Annls. Paleont.
{Invert.), 67, 135-148.

sowERBY, j. DE c. 1850. Description of the shells of the Chalk formation, 346-359, pis. 27-29. In dixon, f. The
Geology and Fossils of the Tertiary and Cretaceous Formations of Sussex. 1st edn., xxxii+423 pp., 43 pis.
Richard and John Edward Taylor, London.

SPATH, L. F. 1925. On Tipper Albian Ammonoidea from Portuguese East Africa, with an appendix on Upper
Cretaceous ammonites from Maputoland. Ann. Transv. Mus. 11, 179-200, pis. 28-37.

1926. On new ammonites from the English Chalk. Geol. Mag. 63, 77-83, table.

1939. A monograph of the Ammonoidea of the Gault. Part 13. Monogr. palaeontogr. Soc. 541-608, pis.

59-64.

STANKiEViCH, E. s. and POJARKOVA, z. N. 1969. Vascoccratids from the Turonian of southern Kirgisia and the
Tadzhiksian depression, 86-1 1 1, pis. 1-10. In Kontinental diyye obrazovaniya vostoshnykh rayonov Sredney
Azii i Kasakstana (lithologiya: hiostratografiya). Akad. Nauk SSSR, Inst. Geol. Geokhronol. Dokember.
Leningrad. [In Russian.]

STOLICZKA, F. 1863-1866. The fossil cephalopoda of the Cretaceous rocks of southern India: Ammonitidae,
with revision of the Nautilidae. Mem. geol. Surv. India, Palaeont. indica, 3, 41-56, pis. 26-31 (1863); (2-5),
57-106, pis. 32-54(1864); (6-9), 107-lM, pis. 55-80 ( 1865); (10-13), 155-216, pis. 81-94(1866).

STRAND, E. 1929. Zoological and palaeontological nomenclatorial notes. Latv. Univ. Rak. 20, 3-29.

SWENSEN, A. J. 1963. Anisoceratidae and Hamitidae (Ammonoidea) from the Cretaceous of Texas and Utah.

Geol. Stud. Brigham Young Univ. 9, 53-82, 5 pis. (misdated 1962).

TROGER, K. A. 1967. Zur Palaontologie, Biostratigraphie und faziellen Ausbildung der unteren Oberkreide
(Cenoman bis Turon) Tl. 1: Palaontologie und Biostratigraphie der Inoceramen des Cenomans und Turons
Mitteleuropas. Abh. staatl. Mus. Miner. Geol. 12, 13-207, 14 pis.

1978. Probleme der Palaontologie, Biostratigraphie und Palaobiogeographie oberkretazischer Faunen

(Cenoman-Turon) Westeuropas und der Russischen Tafel. Z. geol. Wiss. Berlin, 6, 557-570.

1981. Zu problemen der Biostratigraphie der Inoceramen und der Untergliederung des Cenomans und

Turons in Mittel-und Osteuropa. Newsl. Stratigr. 9, 139-156.

WHITE, c. A. 1887. Contribution to the palaeontology of Brazil; comprising descriptions of Cretaceous invert-
ebrate fossils mainly of the Provinces of Sergipe, Pernambuco, Para and Bahia. Arch. Mus. natl. Rio de
Janeiro, 7, 1-273, 28 pis.

wiEDMANN, J. 1960. Le Cretace superieur de I’Espagne et du Portugal et ses cephalopodes. C. R. Congres des
Societes Savantes-Dijon, 1959: Collogue sur le Cretace superieur franqais, 709-764, 8 pis.

1966. Stammesgeschichte und System den posttriadischen Ammonoideen; Ein iiberblick. Neues Jb. Geol.

Paldont. Abh. 125, 49-79, pis. 1-2; 127, 13-81, pis. 3-6.

1975. Subdivisiones y precisiones biostratigraficas en el Cretacico superior de las Cadenas Celtibericas.

Actes 1" symposium Cretacico Cordillera Iberica cuenca 1974, 135-153, 3 pis.

1980. Mid Cretaceous Events Iberian Field Conference 77. Guide II. Partie Itineraire geologique a travers

le Cretace Moyen des Chaines vascogotiques et Celtiberiques (Espagne du Nord). Cuad. Geol. Iberica, 5,
127-214, 12 pis. (misdated 1979).

WOODS, H. 1896. The Mollusca of the Chalk Rock: Part I. Q. Jl geol. Soc. Loud. 52, 68-98, pis. 2-4.

1911. The palaeontology of the Upper Cretaceous deposits of northern Nigeria, 273-286, pis. 19-24. In

FALCONER, J. D. The Geology of northern Nigeria. Macmillan, London.
wozNY, E. and kogbe, c. a. 1983. Further evidence of marine Cenomanian, Lower Turonian and Maastrichtian
in the upper Benue Basin of Nigeria (West Africa). Cret. Res. 4, 95-99.

WRIGHT, c. w. 1953. Notes on Cretaceous ammonites. 1. Scaphitidae. Ann. Mag. nat. Hist., (12) 6, 473-476.

and KENNEDY, w. J. 1973. Paleontologie systematique. In juignet, p., Kennedy, w. j. and wright, c. w.

1973. La limite Cenomanien-Turonien dans la region du Mans (Sarthe): stratigraphie et paleontologie.
Annls Paleont. (Invert.), 59, 207-242, 3 pis.

1980. Origin, evolution and systematics of the dwarf acanthoceratid Protacanthoceras Spath, 1923

(Cretaceous Ammonoidea). Bull. Br. Mus. nat. Hist., (Geol.), 34, 65-107.

74

PALAEONTOLOGY, VOLUME 30

WRIGHT, c. w. and Kennedy, w. j. 1981. The Ammonoidea of the Plenus Marls and the Middle Chalk.
Monogr. palaeontogr. Soc. 148 pp., 32 pis.

and HANCOCK, J. m. 1984. Stratigraphic Introduction. In weight, c. w. and Kennedy, w. j. A

Monograph of the Ammonoidea of the Lower Chalk. Ibid. Part 1, pp. 1-126, pis. 1-40.

YABE, H. 1904. Cretaceous Cephalopoda from the Hokkaido. Part II. J. Coll. Sci. imp. Univ. Tokyo, 20 (2), 1-
46, pis. 1-6.

YOUNG, K. and POWELL, J. D. 1978. Late Albian-Turonian correlations in Texas and Mexico. Ann. Mus. Hist,
not. Nice, 4 (for 1976), XXV, 1-36, 9 pis.

ziTTEL, K. A. 1884. Handbuch der Paleontologie . . . Abt. 1, 2, (Lief 3), Cephalopoda, 329-522. R. Oldenbourg,
Munich and Leipzig.

W. J. KENNEDY
C. W. WRIGHT

Geological Collections
University Museum
Parks Road, Oxford 0X1 3PW

Typescript received 2 August 1985
Revised typescript received 1 December 1985

J. M. HANCOCK

Department of Geology
Royal School of Mines
Imperial College of Science and Technology
Prince Consort Road
London SW7 2BP

THE SCANDINAVIAN MIDDLE ORDOVICIAN
TRINUCLEID TRILOBITES

by ALAN W. OWEN

Abstract. Thirteen species of trinucleid are described from the lower Llanvirn to lower Caradoc platform
successions of Norway and Sweden. Of these, five are established taxa, four are described under open
nomenclature, and four are new: Bergamia Johanssoui sp. nov., Botrioides impostor sp. nov., B. simplex sp.
nov., and B. margo sp. nov. The genus Botrioides is stabilized by choosing a neotype for the type species
'Trinucleus' coscinorinus Angelin and placing it in the synonymy of B. hronnii (Boeck). Two species groups are
recognized within Botrioides centred on B. bronnii and B. foveolatus (Angelin). Trinucleids were largely
restricted to the western, deepest water, parts of the platform. Although the oldest species, Bergamia johanssoni
sp. nov., represents an early Llanvirn immigration of an Anglo-Welsh genus, the Baltic stocks were endemic
until the mid-Llandeilo when Botrioides spread into the Gondwanan province and possibly ReedoUthus
extended into Scoto-Appalachian faunas. Middle and late Caradoc immigrations of Broeggerolithus and
Tretaspis into Scandinavia were from the Anglo-Welsh basin and North America respectively.

Norwegian trinucleids have long held an important place in studies of this stratigraphically
important group especially since the work of Stormer in 1930. The Swedish species, however, have
been largely neglected and the review of the family by Hughes et al. (1975) brought to light some
fundamental problems involving the taxonomy and biogeographical affinity of the Scandinavian
Trinucleidae. The present work follows studies of the Norwegian late Caradoc and Ashgill trinuc-
leids (Owen 1980a, h, 1983) and involved an examination of all the available Llanvirn to lower
Caradoc material from Scandinavia. It forms part of a revision of the whole trilobite fauna and
stratigraphy of this part of the succession in Norway. The illustrated trinucleid specimens are housed
in the Paleontologisk Museum, Oslo (PMO), Riksmuseum, Stockholm (RM), Paleontologiska
Institutionen, Uppsala (UM), Sveriges Geologiska Undersokning (SGU), British Museum (Natural
History) (BM), and the departments of Geology at the Universities of Lund (LO) and Copenhagen
(MGUH).

SETTING

The Ordovician rocks of Scandinavia essentially belong to two distinct tectonic settings: the
thick siliciclastic and volcanic sequences of the allochthonous Caledonides and the much thinner,
carbonate-dominated autochthonous platform successions (see Bruton et al. 1985 for summary).

All but the lowest nappes of the Scandinavian Caledonides are far travelled and bear little or no
sedimentological or provincial relations to the platform rocks and faunas. Diverse Arenig-Llanvirn
faunas in the Trondheim Region in western Norway (text-fig. 1) show marked North American
affinities and are interpreted as representing environments around oceanic islands far removed from
their present position (Bruton and Harper 1985 and in press). Only in the upper Ordovician
sequences of this part of Norway are there faunas similar to those of the Baltic platform. With the
exception of the lowermost allochthon in Jamtland, Sweden, the only middle Ordovician trinucleids
from the Caledonide belt are of uncertain provenance and comprise a cephalon of ReedoUthus in
a glacial erratic and a specimen of Stapeleyellal forosi (Stormer 1932; see also Hughes et al. 1975,
pp. 559-560) found in a roofing slate. As the origins and ages of these specimens are unknown,
neither can be used in models of faunal provincialism or migration.

(Palaeontology, Vol. 30, Part I, 1987, pp. 75-103, pis. 11-I4.|

© The Palaeontological Association

76

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1. Map of Baltoscandia showing the Ordovician outliers and their position within
the confacies belts defined by Jaanusson (1976). Districts of the Oslo Region, following
Stormer (1953) are as follows: SL = Skien-Langesund, ES = Eiker-Sandsvaer, M =
Modum, R = Ringerike, El = Hadeland, Mj = Mjosa (Toten, Hamar-Nes, Ringsaker),

OA= Oslo-Asker.

The platform successions are preserved in a series of outliers which Jaanusson (1976) interpreted
as remnants of extensive areas of largely uniform, persistent, lithofacies and biofacies termed
confacies belts (text-fig. 1) and showing an overall westward deepening. The successions of the Oslo
Region do not fit readily into this scheme being tectonically more complex, thicker, and much
more lithologically variable. Moreover, there is a general eastward deepening in the region. The
stratigraphy of the Oslo Region has been undergoing revision in recent years with the long estab-
lished but confused ‘etasje’ system being replaced by a modern lithostratigraphical terminology
(e.g. Owen 1978, 1979). A full revision by the author and Norwegian based workers covering all
the districts of the region is at an advanced state of preparation. The Arenig to Caradoc succession
of the Oslo-Asker district is summarized in text-fig. 2 along with those of Scania, Jamtland, and
Vastergotland in Sweden.

Before the Caradoc the faunas of the Baltic platform were distinct not only from those of
the North American and Celtic provinces but also from those of Gondwanaland, including the
Anglo-Welsh area (Cocks and Fortey 1982; Dean 1985). This latter separation may have been
the result of an oceanic barrier whose suture is now represented by the Tornquist zone (text-fig.
1). This was advocated by Cocks and Fortey (1982, p. 467), but in a review of the Tornquist
zone, Pegrum (1984) considered it to have had a much longer history and to have acted
primarily as a major transform lineament during the Caledonian Orogeny. The Baltic province
( = Asaphid Province of Whittington) shared a few deep-water and rare pelagic trilobites

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

77

TEXT-FIG. 2. Correlation between the Arenig-Caradoc successions of Oslo-Asker, Scania, Jamtland, and
Vastergotland. Swedish units based on Bergstrom (1982), Jaanusson (1982u, h), and Jaanusson and Karis
(1982). Many of the terms applied to the Oslo-Asker succession are new or recently introduced and replace an
existing terminology as follows: Bestum Formation = ‘Orthoceras Limestone’ sensu Into, Fines Formation
= Upper Didymograptus Shale and Ogygiocaris Shale, Vollen Formation = Ampyx Limestone, Arnestad
Formation = Lower Chasmops Shale, Rodelokken Formation = Lower Chasmops Limestone, Nakholmen
Formation = Upper Chasmops Shale, Solvang Formation = Upper Chasmops Limestone.

with other provinces but it was not until the Caradoc that significant mixing of shallower benthos
took place.

Some sixteen species of trinucleid, distributed amongst six or seven genera are known from the
Llanvirn to Caradoc rocks of the platform and lowermost allochthon in Scandinavia (text-fig. 3).
Llanvirn and Llandeilo species were largely restricted to the westernmost (i.e. deeper) parts of the
area: Scania, the Oslo Region, and the lowermost allochthon in Jamtland. The short-lived appear-
ance of trinucleids in the lower Llandeilo Gullhdgen Formation in Vastergotland in the Central
Confacies Belt is related to a brief eastward transgression of the western facies (Jaanusson 1982^,
p. 168). By the late Caradoc, however, trinucleids were more widespread, if rare, in the Central
Confacies Belt.

78

PALAEONTOLOGY, VOLUME 30

EVOLUTION AND AFFINITIES

The oldest Scandinavian trinucleid is Bergamia johcmssoni sp. nov. from the uppermost part of the
Komstad Limestone and basal Upper Didymograptus Shale (low Llanvirn) in Scania. This consti-
tutes the first undoubted record of a genus otherwise restricted to the Arenig to lower Llandeilo of
the Anglo-Welsh area. Cocks and Fortey (1982, p. 470) noted that Bergamia was a component of a
fairly deep-water biofacies and thus its extension to Baltica at a time of transgression (A. Nilssen,
Copenhagen, pers. comm.) is consistent with the model proposed by Fortey (1984).

Bergamia may have given rise to Botrioides possibly by paedomorphosis, a heterochronic process
which is well documented in many trilobite groups (McNamara 1983) including trinucleids (Owen
1980r/). The oldest known species of Botrioides is B. simplex sp. nov. from the low-mid Llanvirn of
the Oslo Region and like most other species, its small size and simple fringe shape and pitting
suggest a paedomorphic origin. The lateral eye tubercles of Botrioides may have been derived from
the eye ridges present in many juvenile trinucleines and even in the adults of Bergamia johanssoni.

Llanvirn

Llandeilo

Caradoc

Botrioides broeggeri (St0rmer)

- 0

7

Botrioides bronnii (Boeck)

— i o,s

Botrioides impostor sp. nov.

0

Botrioides simpiex sp. nov.

; f

O.J

Botrioides sp. A

- 0

Botrioides sp. B

- o

Botrioides foveoiatus (Angelin)

0

Botrioides effiorescens (Hadding)

•

S,J

9

Botrioides margo sp. nov.

- V

Bergamia Johanssoni sp. nov.

- S

Trinucleid gen. et sp. indet.

-. 0

Reedoiithus sp.

?

V

Reedoiithus carinatus (Angelin)

o,?v

Broeggeroiithus discors (Angelin)

0

Broeggeroiithus aff. discors (Ang.)

J.SI -

Tretaspis ceriodes (Angelin)

o,v -

TEXT-FIG. 3. The range and suggested phylogeny of trinucleid trilobites in the middle Ordovician of Norway
and Sweden. Widely spaced dots indicate very tentative derivation of one species from another; closely
spaced dots show more certain relationships. O = Oslo Region, S = Scania, V = Vastergotland, SI = Siljan,
J = Jamtland. Species of Broeggerolithus and Tretaspis described by Owen (I980u, 1983); the remainder are

treated herein.

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

79

Nine species of Botrioides are known from horizons in Scandinavia ranging up to the mid Llandeilo
(text-fig. 3). Towards the end of its time range the genus crossed the Tornquist divide and extended
into south-east Ireland and probably Cornwall and north-eastern Newfoundland. Most of the
Scandinavian species along with those from outside the area have a narrow fringe with no more
than two E arcs and two I arcs. B. foveolatus (Angelin), B. ejflorescens (Hadding), and B. margo sp.
nov., however, have up to five I arcs and up to four E arcs.

A trinucleid cephalon from the lower Elnes Formation (low-mid Llanvirn) in the Hadeland
district of the Oslo Region is here considered under open nomenclature but may have important
phylogenetic implications. Hughes et al. (1975, pp. 562-563) considered it close to the common
ancestor of Tretaspis and Botrioides and ascribed it to the latter genus. However, it differs in several
respects from that genus and also shares characters with Reedolithus which first appeared in the
middle Llandeilo of Scandinavia and Canada. There is, however, a time gap between the Hadeland
specimen and the first known appearance of Reedolithus and as is discussed below, there are doubts
as to the age of R. quehecensis Staiible, the oldest Scoto-Appalachian species. Thus, the derivation
of Reedolithus from the Hadeland form or even its origin within the Baltic province is only tentatively
suggested here.

The appearance of Broeggerolithus in Scandinavia in the middle Caradoc represents the immi-
gration of an essentially Anglo-Welsh genus late in its history. The Scandinavian material was
described in an earlier study (Owen 1983) and although Broeggerolithus occurs in Jamtland, Siljan,
and several districts of the Olso Region, it is only abundant in the deepest water facies of the latter
region (Harper et al. 1985, pp. 298 299). As with Bergamia in the early Llanvirn, the appearance of
Broeggerolithus in Scandinavia was probably associated with marine transgression but in addition,
major interchanges of faunal elements between provinces were also taking place during the late
Caradoc as the provincial barriers disappeared.

During the latest Caradoc, Tretaspis appeared in Scandinavia but had a long history in the North
American province extending back to the early Caradoc and possibly the Llandeilo. The earliest
species known of Tretaspis is T. canadensis Staiible which occurs with R. quehecensis Staiible in
clasts in a melange in the Citadel Formation in Quebec. As is discussed below under R. carinatus
(Angelin), this part of the formation contains graptolites of the Nemagraptus gracilis Zone but the
age of the melange clasts is not known with certainty. Tretaspis occurs widely in the upper Caradoc
and Ashgill of both Norway and Sweden and shares several species in common with the British
Isles. In the deepest water facies in the Ashgill of Sweden the trinucleid Nankinolithus ('Tretaspis
granulatus') is associated with Tretaspis such as in the fauna of the Ulunda Mudstone in Vastergot-
land described by Bergstrom (1973). This fauna also includes rare cyclopygids and probably occu-
pied a broadly similar niche to that of the Opsimasaphus-Nankinolithus Association of Price (1981 )
in the Ashgill of the Llyn Peninsula in North Wales. Cryptolithus was a late immigrant from North
America to the Oslo Region occurring in the upper Rawtheyan of Oslo in inner shelf regressive
mudstones. This represents both migration consequent on the narrowing of the lapetus Ocean and
a change to a shallower water habitat. Poorly preserved specimens from the upper Ordovician of
the Trondheim area may also belong in Cryptolithus (Stormer, 1932, pi. 28, figs. 2 and 3).

SYSTEMATIC PALAEONTOLOGY

The terminology used herein is that advocated by Hughes et al. (1975) and unless otherwise stated,
pit counts refer to half-fringe values.

Family trinucleidae Hawle and Corda, 1847
Subfamily trinucleinae Hawle and Corda, 1847
Genus botrioides Stetson, 1927

Type species. Trinucleus coscinorinus, Angelin, 1854, p. 65, pi. 34, fig. 4, ?from the Lower Dicellograptus Shale
of Scania, south-west Sweden (= B. hronnii (Boeck, 1838)); by original designation of Stetson (1927, p. 97).

80

PALAEONTOLOGY, VOLUME 30

Emended diagnosis. Fringe narrow, declined. Up to four I arcs and four E arcs present. Pits on
upper lamella in deep radial sulci; on lower lamella absent or restricted to E arcs. Genal prolongation
absent. Lateral eye tubercles present.

Discussion. Stetson (1927) established Botrioides to encompass a group of narrow-fringed Scandina-
vian trinucleines typified by Trinucieus coscinorinus Angelin, 1854. Although Stormer (1930, p. 13)
considered that Botrioides could not be distinguished from Trinucleus Murchison, 1839, Hughes et
al. (1975, p. 561) resurrected Stetson’s genus in their revision of the Trinucleidae. However, as
Angelin’s original material of T. coscinorinus is lost, Hughes et al. preferred to refer to allied species
as "BotrioidesT . This is clearly unsatisfactory as most of the species they included under this name
form a close plexus which conforms in most of its diagnostic characters both to features of Angelin’s
illustration of T. coscinorinus and to Stetson’s concept of Botrioides. In attempting to stabilize the
genus, however, several taxonomic problems have to be resolved.

The type horizon and locality of T. coscinorinus was given by Angelin as being "Regio C.’ at
Fagelsang in Scania. This was accepted by Hadding (1913, p. 75) as being the ‘Orthoceras Limestone’
(= Komstad Limestone of modern usage) but Funquist ( 1919, pp. 35, 39) thought it more probable
that the species was not from Fagelsang and was from a higher unit, the Lower Dicellograptus
Shale. This view was adopted by subsequent workers. The only known trinucleid from the Komstad
Limestone is the recently discovered material described below as Bergamia johanssoni sp. nov. which
differs considerably from the accepted concept of T. coscinorinus in particular and Botrioides in
general. Several workers described material as T. coscinorinus from the Lower Dicellograptus
Shale in Scania including Funquist (1919) who included some Norwegian specimens amongst his
illustrations. Stormer (1930, p. 19) considered T. coscinorinus a ']\xxnox synonym of T. hronnii (Boeck,
1 838): a view which was accepted by most subsequent workers in Sweden. As a result, stratigraphical
terms used in Scania such as ‘Coscinorinus limestone’ were changed to (for example) ‘Bronni
limestone’ (e.g. Hadding 1958, p. 217). The present study shows that the specimens which Stormer
(1930) described as T. bronnii from Norway are not conspecific with the lectotype which he sub-
sequently chose (on good grounds) for Boeck’s species (1940). However, the lectotype morphology
is the same as that of most of the cephala from Scania although preservation of the Swedish material
is poor. Thus in order to stabilize Botrioides, a neotype for ‘T.’ coscinorinus is here chosen from the
Killerod Formation (formerly included in the Lower Dicellograptus Shale) at Killerod Quarry in
Scania. The specimen is numbered LO 5717T and is illustrated on Plate II, fig. 5. This specimen
falls well within the range of variation seen in B. hronnii as diagnosed herein and thus B. coscinorinus
becomes its junior subjective synonym. The assignment to Botrioides of the other species described
and discussed below now becomes unequivocal in terms of the reservations held by Hughes et al.
(1975). All the named species which they provisionally assigned to Botrioides are here confirmed as
belonging to that genus. Of the indeterminate materal which Hughes et al. listed, only the forms
described originally by Dean (1971) and Sadler (1974) from the Llandeilo of northeastern New-
foundland and Cornwall respectively probably belong to Botrioides. The fragmentary material from
the upper Arenig of northeastern Newfoundland described by Dean (1974; see also Neuman 1976
for age) is too incomplete for adequate determination but the presence of arc E3 suggests it belongs
elsewhere. The form which Hughes et al. (1975, pi. 4, figs. 48-51) listed as Botrioidesl sp. from the
Llanvirn of Hadeland, Norway is described below as trinucleid gen. et sp. indet.

Two species groups are here recognized within Botrioides (Table 1). One, centred on B. bronnii
(Boeck) has a very narrow fringe comprising up to two I arcs and two E arcs. The radial sulci in
which the fringe pits are dispersed are broad in proportion to their length and number between 12
and 174 per half-fringe in the material available. Pygidia of the B. hronnii species group have only
three to four axial rings. The second group, centred on B. foveolatus (Angelin), has up to five I arcs
and up to four E arcs arranged in long, narrow sulci which are commonly more numerous (16^-
224 in the available specimens). Pygidia of the B. foveolatus group are more segmented, with seven
to nine axial rings present. This latter group comprises B. foveolatus, B. ejflorescens (Hadding), and
B. inargo sp. nov. and may ultimately prove worthy of separate generic status.

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

TABLE 1. Summary of fringe pit arcs and pygidial axis segmentation in the named Scandinavian
species of Botrioides. C = complete arc, A = arc present anteriorly, L = arc present laterally,
P = arc present posteriorly, al = arc present anterolaterally, x = arc absent. Number of sulci
refer to half-fringe values on the upper lamella. Only B. hroeggeri shows deep sulcation on the
lower lamella where it is restricted to the E arcs.

El

E2

E3

E.

In

I2

I3

I4

sulci

axial

rings

B. bronnii species group

B. bronnii (Boeck)

C

c

X

X

c

X -al

X

X

X

12|-16

4

B. impostor sp. nov.

C

X

X

X

c

al-C

X

X

X

131-171

3

B. broeggeri (Stormer)

c

A-C

X

X

c

X

X

X

X

12-15

—

B. simplex sp. nov.

c

X

X

X

c

X-P

X

X

X

14-161

3-4

B. foveolatus species group

B. foveolatus (Angelin)

c

X

X

X

c

c

c

c

X -L

17 221

^7

B. efflorescens (Hadding)

c

A

X

X

c

c

A-fL

X

161-211

7 9

B. margo sp. nov.

c

c

A?-rL

A? + L

c

c

c

-

—

22

--

Botrioides bronnii Species Group

Botrioides bronnii (Boeck, 1838)

Plate II, figs. 1 17

1838 Trilobites Bronnii Boeck, p. 144.

1854 Trinucleus coscinorinus Angelin, p. 65, pi. 34, fig. 4.

1854 Trinucleus hucculentus Angelin, p. 84, pi. 41, fig. 1 .

?1857 Trinucleus Bronnii., Kjerulf, p. 94.

?1887 Trinucleus hucculentus, Ang., Brogger, p. 17.

non 1913 Trinucleus coscino(r)rhinus [wV] ANG; Hadding, pp. 74-75, pi. 7, figs. 18, 20 ( = B. efflorescens
(Hadding)), 19 ( = B. simplex sp. nov.).

1919 Trinucleus coscinorrhinus [5/c] ANG; Funquist, pp. 34-35, pi. I, figs. 7-9, 11-22, non 10, lOu
( = B. impostor sp. nov.).

1927 Botrioides bucculentes [i'/c] (Angelin); Stetson, pi. 1, fig. 1 1 .

non 1927 Botrioides coscinorrhinus [5/c] (Angelin); Stetson, pi. 1, fig. 12 ( = B. efflorescens (Hadding)).

1930 Trinucleus hucculentus Ang.; Stormer, pp. 21-24, pi. 2, figs. 8-15; text-figs. 7-11, 16<?.

1930 Trinucleus bronni (Sars and Boeck); Stormer, text-fig, 6; non pp. 19-21, pi. 2, figs. 1-7; text-
figs. 5, 16c, 43 ( = B. impostor sp. nov.).
non 1934 Trinucleus bronni', Stormer, p. 331 ( = 5. impostor sp. nov.)

1940 Trinucleus bronni (Sars and Boeck MS); Stormer, p. 147, pi. I, fig. 18.

1940 Trinucleus bronni (Sars and Boeck MS); Grorud, p. 160, text-figs. 1, 2c, /.

1952 Trinucleus bronni Sars and Boeck; Nilsson (pars), pp. 684, 691-692 (some specimens B. efflor-
escens (Hadding)).

non 1953 Trinucleus bronni, Stormer, pp. 61, 83.

1958 Tr. bronni, Hadding, pp. 217-218.

1975 Botrioidesl bronnii (Sars and Boeck in Boeck); Hughes el al., p. 561 (pars), non pi. 4, fig. 44
( = B. impostor sp. nov.).

1975 Botrioides"! bucculentus (Angelin); Hughes et al., p. 562.

1982 Botryoides coscinorhinus [^/c] (Angelin); Bergstrom (pars), p. 192 (some specimens B. efflor-
escens Hadding).

non 1984 Botrioides! hucculentus (Angelin); Wandas, pp. 234-235, pl.l2G, j; pi. 13f (= B. simplex sp.
nov.)

References to Swedish material cited, but not illustrated, by earlier workers were given by Funquist (1919,
p. 35). Note that Angelin (1854) used the spelling coscinorinus in the text but coscinorhinus in the plate descrip-
tion of his work; hence the confusion of subsequent workers.

82

PALAEONTOLOGY, VOLUME 30

Lectotype. Selected by St0rmer (1940, p. 147); the internal mould of a slightly crushed cephalon (PMO 61752)
labelled 'Wraatz’s Lokke’, Oslo, horizon not known. Note that Stormer earlier (1930, p. 19) indicated that he
had selected a lectotype from Boeck’s ‘cotypes’ but he gave no further details at that time.

Material, localities, and horizons. In addition to four cephala/cranidia from the type locality, elements of the
cephalon and one articulated thorax and pygidium are known from the upper part of the Elnes Formation
(the ‘Ogygiocaris Shales’) elsewhere in Oslo- Asker. Specimens illustrated by Grorud (1940) from the lowest
part of the overlying Vollen Formation (‘Ampyx Limestone’) at Tortberg, Oslo also belong here, as does a
cranidium from 1 m above the conglomerate at the base of this unit at Kulerud, Ringerike. The species occurs
more commonly in the Lower Dicellograptus Shale (including the Killerod Formation) in Scania, south-east
Sweden.

Emended diagnosis. Pseudofrontal lobe subspherical, overhanging narrow fringe which comprises
complete arcs. Ei 2 and I„. Arc Ii may be present anterolaterally. Pits set in short sulci. Lateral eye
tubercle very subdued, situated opposite L2 or posterior part of S2. External surface of pseudofron-
tal lobe in larger specimens bears concentric ridges superimposed on coarse reticulation. Median
tubercle on thoracic axial rings. Pygidium transverse bearing up to four axial rings, four pleural
ribs, and a broad border.

Description. Cephalon (excluding genal spines) semicircular in outline. Glabella weakly swollen posteriorly,
very strongly so in front of S2. Occipital ring ridge-like, transversely directed except distally where it is
deflected forwards slightly. Occipital furrow shallow but distinct mesially, deepening into occipital pit distally.
Occiput short (sag., exsag.), very weakly swollen mesially, narrower (tr.) than occipital ring, and clearly defined
from axial furrow. SI deep, directed abaxially rearwards at about 75 % to the sagittal line. L2 triangular in
outline, broadening abaxially, and poorly differentiated from axial furrow. S2 deep except distally where

EXPLANATION OF PLATE 1 1

Figs. 1-17. Botrioides bronnii (Boeck). 1, Boeck ‘cotype’, PMO 113.224, dorsal view of latex east of small
cranidium, horizon unknown, Wraatz’s Lokke, Oslo, x 5. 2, Boeck ‘cotype’, PMO 113.225, dorsal view of
internal mould of cephalon and partial thorax, same locality as 1, x4- 3, LO 2948?, latex cast showing
dorsal view of cranidium and ventral views of two lower lamellae, probably Elnes Formation, Vestre Aker
Church, Oslo, x4; also figured by Funquist (1919, pi. 1, fig. 18). 4, lectotype, PMO 61752, dorsal view of
internal mould of cephalon, same locality as 1, x 3; also figured by Stormer (1940, pi. 1, fig. 18). 5, LO
5717?, dorsal view of latex cast of cranidium, Killerod Formation, Killerod Quarry, Scania, x 5; here
selected as neotype for "Trinucleus' coscinorinus Angelin. 6, LO 5718?, dorsal view of internal mould of
cranidium, same locality and horizon as 5, x 4. 7, LO 5719?, dorsal view of latex cast of cephalon, Killerod
Formation, level k of Nilsson (1952), same locality as 5, x 3. 8, LO 5720?, dorsal view of latex cast of
pygidium, same horizon and locality as 7, x 5. 9, LO 5721?, dorsal view of internal mould of cranidium,
same horizon and locality as 5, x 4. 10, PMO H482, dorsal view of pygidium and posterior thorax, upper
Elnes Formation, Engervik, Asker, x4-5; also figured by Stormer (1930, pi. 2, figs. 14 and 15). 11 and 12,
PMO H481, dorsal and lateral views of partially exfoliated cephalon, same horizon and locality as 10, x 3;
also figured by Stormer (1930, pi. 2, fig. 13). 13, PMO H480, dorsal view of partially exfoliated cephalon,
same horizon and locality as 10, x 3; also figured by Stormer (1930, pi. 2, figs. 8-10 as neotype of "T.'
bucculentus Angelin). 14, LO 2946?, dorsal view of pygidium. Lower Dicellograptus Shale, Tommarp, Scania,
x4; also illustrated by Funquist (1919, pi. 1, fig. 16). 15, LO 2947?, dorsal view of partially exfoliated
artieulated individual, same horizon and locality as 14, x 4; specimen also figured by Funquist (1919, pi. 1,
fig. 17). 16, LO 2950?, dorsal view of latex cast of cephalon and anterior thorax, same horizon and locality
as 14, X 3; also figured by Funquist (1919, pi. 1, fig. 21). 17, LO 2937?, dorsal view of internal mould of
crushed cephalon, same horizon and locality as 14, x 2; also figured by Funquist (1919, pi. 1, fig. 7).

Figs. 18-20. B. impostor sp. nov. 18, holotype, PMO H0566, dorsal view of latex cast of cranidium and first
thoracic segment, upper Elnes Formation between Hakavik and Bjerkasholmen, Asker, x 2-5; also figured
by Stormer (1930, pi. 2, fig. 3) and Flughes et al. (1975, pi. 4, fig. 44). 19, paratype, PMO H401, ventral
view of lower lamella, same horizon as 18, Gomnaes, Ringerike, x4-5; also figured by Stormer (1930, pi. 2,
fig. 5). 20, paratype, PMO H0574, dorsal view of latex cast of articulated individual, same horizon as 18,
near Vollen, Asker, x 3; also figured by Stormer (1930, pi. 2, fig. 1).

PLATE 11

2Kp5t^W

j[j

|E

fl|HP/ -4

OWEN, Botrioides

84

PALAEONTOLOGY, VOLUME 30

it shallows abruptly, oval in outline, and directed abaxially forwards at about 60° to the sagittal line. Pseudo-
frontal lobe subspherical, overhanging the fringe anteriorly, and bearing a median node a short distance behind
its mid-point. In small specimens the pseudofrontal lobe is not as swollen and thus less well differentiated from
the posterior part of the glabella. Axial furrow almost parallel to sagittal line, broad and shallow, bearing
deep anterior fossula. Genal lobe quadrant-shaped, strongly convex (tr., exsag.) bearing a very subdued eye
tubercle opposite L2 or the posterior part of S2, well away from the axial furrow. Small specimens also show
a very weakly developed eye ridge directed towards the mid-part of the pseudofrontal lobe. Posterior border
transversely directed; narrow (exsag.) proximally, broadening considerably distally. Posterior border furrow
broad (exsag.) and shallow, transversely directed and bearing a deep posterior fossula distally.

Fringe narrow, bearing 12|-16 sulci (n = 4) which contain arcs Ei„2 and I„ over the whole fringe. A short
Ii arc is also developed anterolaterally in some specimens. Individual pits are difficult to discern on the upper
lamella, especially the presence of two E arcs, but are clear on the lower lamella and in section (e.g. Stormer
1930, text-figs. 10 and 1 1). Posterior margin of upper lamella of fringe transversely directed or deflected very
gently abaxially forwards. Lower lamella bears a distinct girder, a broad anterior band, and a genal spine of
unknown length. Hypostoma not known.

External surface of glabella and genal lobe bears a fine reticulation. Larger specimens also show concentric
ridges superimposed on the reticulation of the pseudofrontal lobe and the posterior part of the glabella is
smooth. Internal mould smooth.

Thorax tapering very slightly rearwards. Each axial ring strongly convex (tr.) and bears a distinct, sagitally
elongate median tubercle on its anterior two thirds. Posterior edge of ring transversely directed; anterior edge
arched gently forwards mesially and curving abaxially forwards at about 60° to the sagittal line distally.
Details of articulating half-ring not known. Axial furrow broad and shallow anteriorly, narrowing a little
posteriorly. Thoracic segments transversely directed, each parallel-sided except over its distal 20 % where the
anterior edge is deflected downwards slightly and abaxially rearwards through about 20°. The posterior edge
is deflected a little more gently rearwards. Anterior band very narrow (exsag.) proximally, broadening a little
over its outer 35 %. Pleural furrow very narrow proximally, broadening gently over its proximal half and
more strongly so distally. Broad posterior band ridge-like, directed abaxially rearwards from just behind the
anteromesial corner of the pleura to the posterolateral corner. The posterior face of this band bears a shallow
but distinct furrow in front of the very narrow posterior border. External surface of posterior part of
axial ring and the highest parts of the pleural bands bear a subdued, dense, granulation. Remaining areas
smooth.

Pygidium smooth, transverse in outline having a sagittal length in dorsal view (including border) equal to
about a third of the maximum width. Axis gently convex (tr.), tapering rearwards at about 30° and comprising
an anterior articulating half-ring and four rings, the anterior one of which bears a subdued median tubercle.
Ring furrows progressively less well defined posteriorly along the axis, bearing distinct apodemes distally. Up
to six more pairs of apodemes are also seen on the subdued but distinct extension of the axis on to the border.
Axial furrows narrow and shallow. Up to four pleural ribs present, dying out abaxially but only the anterior
two are well developed. The first of these is directed abaxially rearwards at about 80° to the sagittal line, the
second at about 60°. Anterolateral parts of pleural fields directed abaxially rearwards at 60° and declined
steeply forwards. Posterior border broad, steeply declined but occupying up to 25 % of the sagittal length of
the pygidium in dorsal view.

Discussion. The lectotype of B. bronnii selected by Stormer (1940) differs in several respects from
the morphology of the specimens which he ascribed to the species in his major work on the
Scandinavian trinucleids (1930) and which are described below as B. impostor sp. nov. Boeck’s
original description (1838; see also Stormer 1940, p. 147) was not accompanied by an illustration
and could be applied equally well to either species bearing in mind the closeness of the two E arcs.
None the less, there is no reason to doubt that the lectotype was chosen from Boeck’s original
material and is representative thereof. An unfortunate, but unavoidable, consequence of Stormer’s
concept of B. bronnii is that the strata in Norway which he and subsequent workers termed the
"bronni beds’ are not characterized by an abundance of that species but by B. impostor sp. nov.

B. bronnii is here considered to be the senior synonym of ‘T.’ bucculentus (Angelin) which was
based on Norwegian material. The neotype selection for T. coscinorinus Angelin made in the
discussion of Botrioides (above) also places this species in the synonymy of B. bronnii. Angelin’s
illustration of T. coscinorinus does not show the marked anterior overhang of the glabella typical
of B. bronnii, but his illustration may have been based on a crushed specimen in which this feature

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

85

was not preserved. Such material is fairly common amongst samples from Scania and includes
some of the specimens described by Funquist (1919). However, associated material and other
morphological features show that they belong in B. bromiii. The neotype of ‘T.’ coscinorimis comes
from material which Nilsson (1952) ascribed to ‘T.’ hronnii. Examination of this material shows
that B. bronnii is present in his samples from horizons j, k, and 1 in the Killerod Formation at
Killerod Quarry. The former two levels also contain B. ejfiorescens (Hadding) which is described
below.

Botrioides impostor sp. nov.

Plate 11, figs. 18-20

71857 Trinucletis Bronnii', Kjerulf, p. 94.

71887 Trinucleus hucculentus Ang.; Brogger, p. 17.

1930 Trinucleus bronni (Sars and Boeck); Stormer, pp. 19-21, pi. 2, figs. 1-7; text-figs. 5, 16c, 43;
non 6 ( = 5. bronnii (Boeck)).

1934 Trinucleus bronni', Stormer, p. 331.

1953 Trinucleus bronni', Stormer, pp. 61, 83.

1975 Botrioidesl bronnii (Sars and Boeck in Boeck); Hughes et al., p. 561 (pars)', pi. 4, fig. 44.

Holotype. The external mould of a cranidium and first thoracic segment (PMO H0566) from the upper Elnes
Formation (the ‘Ogygiocaris Shale’) between Hakavik and Bjerkasholmen, Asker.

Paratypes. The external mould of an articulated individual (lacking most of the fringe) (PMO H0574) from
the uppermost Elnes Formation near Vollen, Asker and a lower lamella (PMO H401) from the upper Elnes
Formation at Gomnaes, Ringerike.

Material, localities, and horizons. Disarticulated exoskeletal elements of this species are the most common
fossils in the upper Elnes Formation especially in the uppermost, sandy, part of the formation throughout
Oslo-Asker. The species occurs more rarely in the upper Elnes Formation of Ringerike. Fragmentary material
in the Paleontologisk Museum, Oslo, from the Elnes Formation near Fure, Modum and the Kirkerud Group
near Skiaker, Hadeland may also belong in this species.

Derivation of name. Impostor — a charlatan, referring to the incorrect belief of earlier workers that specimens
of this species belonged in B. hronnii.

Diagnosis. Pseudofrontal lobe oval, only slightly overhanging narrow fringe which comprises com-
plete arcs El, I„, and in some specimens Ii. Pits set in short sulci except posteriorly. Very subdued
lateral eye tubercle opposite L2. External surface of glabella reticulate. Median tubercle only on
anterior thoracic axial ring. Transverse pygidium bearing three axial rings, two pleural ribs, and a
broad border.

Description. Cephalon (excluding genal spines) approximately semicircular in outline. Glabella increasing in
convexity (tr.) forwards towards the pseudofrontal lobe which has little independent convexity. Occipital ring
and furrow, occiput, SI and S2 similar to those of B. bronnii. Distal part of F2 joined to the pseudofrontal
lobe as a very weakly developed composite lobe. Pseudofrontal lobe elongately oval and overhangs the fringe
only very slightly. Axial furrow broad and shallow bearing deep fossula anteriorly. Quadrant-shaped genal
lobe moderately convex (tr., exsag.) bearing subdued eye tubercle at its highest point, opposite L2.

Narrow fringe gently declined with a broad border. Arcs E] and I,i invariably complete, containing 13|-
17^ (X = 15^, n = 1 1) pits in the half fringe. Arc T may be complete or restricted to the anterolateral and
lateral parts of the fringe. The holotype shows only three pits which may belong in this arc on the posterior
part of the fringe. A few adventitious pits are included in the posterior few radii of the most specimens. Pits
situated in deep sulci anteriorly on the upper lamella but this sulcation breaks down abaxially although radial
alignment is maintained. Posterior margin of upper lamella of fringe directed very slightly rearwards. Lower
lamella bears a genal spine directed parallel to the sagittal line and equal in length to at least two and a half
times that of the sagittal cephalic length.

External surface of genal lobes, the pseudofrontal lobe, and the mesial part of the glabella behind S2 finely
reticulate.

Hypostoma not known. Thorax essentially similar to that of B. bronnii except that the first axial ring is
arched concave forward, there are no median tubercles and there is no surface granulation.

86

PALAEONTOLOGY, VOLUME 30

Pygidium transverse, sagittal length approximately one third of maximum width. Weakly convex axis tapers
evenly rearwards at about 40° and is extended weakly on to the very broad border. Four axial rings present,
only the anterior three of which are well developed. Axial furrow narrow and shallow. Two pleural ribs
present, the first directed at about 80° to the sagittal line, the second at about 65°. Surface of pygidium
smooth.

Discussion. As is noted in the discussion of B. bronnii, B. impostor sp. nov. is based on the material
which Stormer (1930) described as ‘T. bronni' and which characterizes the so-called ‘bronni beds’ in
Oslo-Asker. B. impostor differs from B. bronnii primarily in its much less swollen pseudofrontal
lobe which lacks concentric ridges even in large specimens, in the absence of arc E2, the common
presence of an extensive Ii arc, in the shape of the first axial ring of the thorax, the absence of
median tubercles and granulation on the thoracic segments, and in having fewer axial rings (3, cf.
4) and pleural ribs (2, cf. 4) on the pygidium.

Botrioides broeggeri (Stormer, 1930)

Plate 12, figs. 1-6

1930 Trimicleus hihernicus REED var. broggeri Stormer, pp. 24-27, pi. 3, figs. 1-14; text-figs. 12,
13, \6d.

1953 Trinucleus hihernicus broggeri, Stormer, pp. 62, 84.

1953 Trinucleus cf. hihernicus broggeri', Stormer, pp. 72?, 73.

1953 Trinucleus alT. hibernicus', Stormer, p. 80 (pars).

1975 5.? hibernicus broeggeri (Stormer); Hughes et al., p. 562.

Holotype. A cephalon (PMO H553) from the Vollen Formation (‘Ampyx Limestone’) at Gullerud near
Norderhov, Ringerike.

Material, localities, and horizons. Elements of the cephalon occur abundantly in the Vollen Formation at
Gullerud and are especially common in the conglomerate at the base of the formation. The thorax is not
known and pygidia are rare. Cephalic elements also occur in the lowest part of the formation in Oslo-Asker
and in the uppermost parts of the Fines Formation in Eiker-Sandsvsr and Skien-Langesund.

Emended diagnosis. Pseudofrontal lobe subcircular outline, weakly differentiated from rest of glabella, not
overhanging fringe. Arcs Ei and complete, E2 commonly incomplete posteriorly situated in sulci with Ei
on the lower lamella; ridges developed between these sulci. No other pit arcs developed. Marginal band of
fringe very long mesially. Lateral eye tubercle prominent, situated opposite L2. External surface of genae and
glabella coarsely reticulate.

Discussion. Stermer’s detailed description of this species (1930) need not be repeated here other
than to note the presence of a distinct composite lateral glabellar lobe and to describe the fringe
pitting in more modern terms. The upper lamella shows 12-15 radial sulci (X = 14-^-, n = 9)
comprising arcs Ei and over the whole fringe and E2 anteriorly at least, but this second E arc is
commonly difficult to discern. On the lower lamella, however, arcs Ei and E2 are clearly seen to
share oval sulci which become smaller abaxially and are separated by low but distinct ridges. E2 is
commonly incomplete posteriorly and comprises only 4^ pits (half-fringe) in one specimen. No
other arcs are developed and the ‘zone of complication’ is restricted to a single adventitious pit seen
in a few specimens.

Stormer considered this form to be a subspecies of B. hibernicus (Reed, 1895) from the Tramore
Limestone (Llandeilo-lowest Caradoc, see Carlisle 1979). Specimens of Reed’s species illustrated
by Hughes et al. (1975, pi. 4, figs. 45-47) and preliminary analysis of material in the Murphy
Collection at the National Museum of Ireland show that the Irish species has a more spherical
pseudofrontal lobe which overhangs the inner part of the fringe. The sulci on the upper lamella of
the fringe are long and E2 is distinet from Ei, these arcs do not share sulci anteriorly on the lower
lamella and the lateral eye tubercle is much less prominent than in B. broeggeri which is thus given
full specific status.

B. broeggeri differs from B. bronnii and B. impostor sp. nov. in the shape of the glabella, the

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

87

distinct composite lateral lobe, the development of S3 on the side of the pseudofrontal lobe, the
prominent lateral eye tubercle, and the marked abaxial tapering of the marginal band of the fringe.
It is closest to B. bronnii in fringe composition but differs in consistently lacking 1 1, in the E pit
sulci on the lower lamella, and in commonly having arc E2 incomplete posteriorly.

Stormer (1953, p. 80) recorded material as T. aff. hibernicus from the upper part of the Elnes
Eormation to the basal parts of the Fossum Formation between 46-6 and 63 0 m in the section at
Flata, Eiker-Sandsvaer. Examination of Stormer’s material and collections made recently indicates
that samples between 46-6 and 510 m belong in B. broeggeri, but specimens from 63 0 m differ
significantly from this species and are described below as B. sp. B.

Botrioides simplex sp. nov.

Plate 12, figs. 7-11

1913 Trinucleus coscino{r)rhinus [iic] ANG.; Hadding (pars), pp. 74-75 (‘Orthoceras Limestone’
material only), pi. 7, figs. 19, non 18, 20 (= B. efflorescens (Hadding)).

71963 Trinucleus sp.; Skjeseth, p. 63.

1982 Botrioidesi bucculentus,'^dLndks,,p. 138.

1984 Botrioides'! bucculentus (Angelin); Wandas, pi. 12g, j; pi. 13f.

Holotype. A cranidium (PMO 83021) from the basal part of the Elnes Formation (7-75 m above base) at
Fumes, Mjosa.

Paratypes. Two articulated specimens (PMO 104064), possibly representing successive instars of the same
animal, from 8 0 m above the base of the Elnes Formation at Vikersundbakken, Modum.

Material, localities, and horizons. Other complete and disarticulated specimens occur between 7-9 and 12-5 m
above the base of the Elnes Formation at Vikersundbakken (see Wandas 1984) and a second cranidium is
known from the type locality. Rare cephala and cranidia are also known from the upper part of the Iso
Fimestone (= ‘Orthoceras Limestone’) and possibly lower Anderso Shale in the Lower Allochthon in
Jamtland.

Derivation of name. Simplex — simple, referring to the morphology of the fringe of this species.

Diagnosis. Pseudofrontal lobe subspherical, overhanging inner part of narrow fringe which com-
prises short sulci containing arcs Ei and I„. Ei pits in lower lamella decrease in size abaxially.
Ridges present between E pits on lower lamella. Composite lateral glabellar lobe distinct. Lateral
eye tubercle well developed, located opposite S3.

Description. Sagittal length of cranidium equal to about 40 % of posterior width. Glabella similar to that of
B. bronnii except that the pseudofrontal lobe overhangs the inner part of the fringe a little less, S2 is less
extensive abaxially and thus there is a well-developed composite lateral glabellar lobe. There is also a shallow
S3 discernible on the sides of the pseudofrontal lobe slightly behind level of median tubercle. Genal lobe
quadrant-shaped, bearing distinct lateral eye tubercle opposite S3.

Fringe narrow, comprising pits of arcs Ei and set in 14-16^ (n = 3) sulci in a half-fringe. One specimen
also shows three pits in the position of L posteriorly. On lower lamella Ej pits enlarged anteromesially where
they are separated by distinct ridges when viewed in ventral view; Ej size and distinctiveness of ridges decreases
abaxially. Marginal band of lower lamella very broad mesially, tapering abaxially. Genal spine curving
rearwards in a very broad arc from being directed abaxially proximally to gently adaxially distally, extending
well beyond the level of the tip of the pygidium.

Internal mould of glabella smooth whilst gena bears a subdued but distinct reticulation. Details of external
surface not known other than the reticulation on the occiput and posteromesial part of the genal lobe seen in
the holotype.

Hypostoma not known. Thorax similar to that of B. bronnii except that there are no median tubercles on
the axial rings. Transverse pygidium too poorly known other than to note the broad border and the presence
of three or four rings on the axis and two pleural ribs.

Discussion. Although based on only a few rather poorly preserved specimens, B. simplex sp. nov.
clearly differs significantly from the other narrow fringed species of Botrioides. The shape of the

PALAEONTOLOGY, VOLUME 30

glabella is closest to that of B. bronnii but the pseudofrontal lobe is less spherical and overhangs
the fringe less, S2 is less extensive abaxially, and thus there is a well-developed composite lateral
glabellar lobe. The very simple fringe, comprising just Ei and I,, is closest to that of those specimens
of B. impostor which lack Ii but the marked abaxial decrease in size of Ei and the development of
ridges on the lamella between these pits distinguish B. simplex. The ridges are only otherwise seen
in B. broeggeri. The forwardly placed lateral eye tubercle distinguishes B. simplex from all these
other species.

Botrioides sp. A
Plate 12, figs. 13 and 14

Material, locality, and horizon. A cephalon and a pygidium from the middle part of the Elnes Formation
(uppermost ‘4ao(2’ of earlier usage) road section at the ski jump at Slemmestad, Asker.

Description. Pseudofrontal lobe subspherical but not significantly wider than posterior part of glabella; only
slightly overhanging fringe. Arcs Ei, T, and complete, and comprise sixteen radii, E developed laterally at
least. Pits arranged in sulci except laterally where Ei becomes discrete. Lateral eye tubercle prominent, situated
far back on the genal lobe, opposite SI. Transverse pygidium with four axial rings, two pleural ribs, and a
narrow border.

Discussion. The broad sulci containing only two complete I arcs and the segmentation of the pygidial
axis suggest that this material belongs in the B. bronnii species group although the number of sulci
is near the upper end of the range seen in the group and an albeit incomplete I2 arc is present. The
shape of the glabella and the large lateral eye tubercle suggest an affinity to B. broeggeri. Although
that species has a very prominent lateral eye tubercle, that of B. sp. A is more posteriorly placed,
the genal lobe is smooth and in addition to the Ei development, the fringe comprises more I arcs
(3, cf. 1), there are no E2 pits and the number of radii and E and I arcs is closest to the condition in
B. impostor although this species lacks I2 pits and all the arcs or just become discrete posteriorly.

EXPLANATION OF PLATE 12

Figs. 1-6. Botrioides broeggeri (Stormer). 1, PMO H563, ventral view of lower lamella external to girder,
Vollen Formation, Gullerud, Ringerike, x4; also figured by Stormer (1930, pi. 3, fig. 10). 2, PMO 103.978,
latex cast showing ventral view of lower lamella and dorsal view of cephalon, uppermost Elnes Formation
50-5 m above base of section, Flata, Eiker-Sandsvaer, x 5. 3 and 4, PMO 66633, dorsal view of latex cast of
cranidium and ventral view of latex cast of lower lamella, same locality as 2, 46-6 m above base of section,
X 6, X 3. 5, holotype, PMO F1553, dorsal view of partially exfoliated cephalon, same horizon and locality
as 1, x4; also figured by Stormer (1930, pi. 3, fig. 1). 6, PMO 81903, dorsal view of partially exfoliated
cephalon, same horizon and locality as 1, x 6.

Figs. 7-11. B. simplex sp. nov. 7, LO 2541?, dorsal view of internal mould of cranidium, upper Iso Limestone,
Anderson, Jamtland, x4; also figured by Hadding (1913, pi. 7, fig. 19). 8, holotype, PMO 83021, dorsal
view of partially exfoliated cranidium, 7-75 m above base of Elnes Formation Fumes, Mjosa, x 4; also
figured by Wandas (1984, pi. 12j). 9, PMO 104.055, dorsal view of latex cast of two articulated individuals,
7-9 m above base of Elnes Formation, Vikersundbakken, Modum, x 3; also figured by Wandas (1984, pi.
12g). 10, SGU Type 5064, dorsal view of internal mould of cranidium, probably lower Anderso Shale,
Verkon, Jamtland, x3. 11, Paratype, PMO 104.064, dorsal view of internal mould of two articulated

specimens, possibly successive instars of the same animal which died during ecdysis, 8 0 m above base of
Elnes Formation, same locality as 9, x 3; also figured by Wandas (1984, pi. 13f).

Fig. 12. B. sp. B, PMO 66650, dorsal view of latex cast of cephalon and anterior thorax, basal part of Possum
Formation, 63 0 m above base of section, Flata, Eiker-Sandsvaer, x 3-5.

Figs. 13 and 14. B. sp. A. 13, PMO 81265, dorsal view of latex cast of pygidium, middle Elnes Formation,
road section at ski jump, Slemmestad, Asker, x9-5. 14, PMO 82168, dorsal view of partially exfoliated
cephalon, same horizon and locality as 13, x 6.

Figs. 15 and 16. B. foveolatus (Angelin), neotype, RM Ar2310, dorsal and lateral views of cranidium, upper
Elnes Formation, probably Oslo-Asker, x 6, x 7-5; also figured at Stormer (1930, pi. 1, figs. 4 and 5).

PLATE 12

OWEN, Botrioides

90

PALAEONTOLOGY, VOLUME 30
Botrioides sp. B
Plate 12, fig. 12

1953 Trinucleus aff. hibernicus', Stormer, p. 80 (pars).

Material, locality, and horizon. A fringe fragment and a cephalon with five thoracic segments attached from
the basal part of the Possum Formation, 63 m above the base of the section at Plata, Eiker-Sandsvaer.

Discussion. As noted in the discussion of B. broeggeri, Botrioides occurs at various levels in the
section at Plata and whilst the specimens up to 51 m above the base of the section belong in
Stormer’s species, those from the 63 m level differ in several respects. The lateral eye tubercle is
much less prominent, the reticulation on the gena is much finer, arcs Ii and I2 are present, and the
sulcation on the upper lamella breaks down posteriorly. It is not known whether E2 is developed.
The shape of the glabella and genal lobe is very close to that of B. hibernicus but the subdued eye
tubercle, fine reticulation, and fringe pitting are closest to the morphology of B. impostor sp. nov.

Botrioides foveolatus Species Group

Botrioides foveolatus (Angelin, 1854)

Plate 12, figs. 15 and 16; Plate 13, figs. 1-4

1854 Trinucleus foveolatus Angelin, p. 84, pi. 41, fig. 2.

1857 Trinucleus foveolatus Ang.; Kjerulf, p. 94.

1927 Botrioides foveolatus (Angelin); Stetson, pi. 1, fig. 10.

1930 Trinucleus foveolatus ANG.; Stormer, pp. 16-18, pi. 1, figs. 4-13; text-figs. 2a-c (non d = B.
ejflorescens (Hdidding)), 3, 16a.

1930 Trinucleus foveolatus ANG. var. intermedius Stormer, pp. 18-19, pi. 1, figs. 1-3; text-figs. 4,
166, 39.

1934 Trinucleus foveolatus', Stormer, p. 331.

1975 5.? foveolatus (Angelin); Hughes et al., p. 562.

1975 5.? foveolatus intermedius (Stormer); Hughes et al., p. 562.

non 19826 Botryoides [sic] foveolatus', Jaanusson, p. 168 (= B. margo sp. nov.).

1985 Botrioides foveolatus Angelin; Owen, table 1, text-fig. 56.

Neotype. Selected by Stormer (1930, p. 16), a cranidium (RM Ar2310) from the upper part of the Pines
Formation, probably Oslo-Asker, Norway.

Material, localities, and horizons. Cephalic elements and pygidia occur in the middle part of the Pines Forma-
tion (‘4a«2’ and ‘4a«3’ of earlier usage) in Oslo-Asker. Cranidia are also known from this formation in
Ringerike and the Mjosa districts.

Emended diagnosis. Pear-shaped glabella evenly increasing in convexity to the mid-part of pseudo-
frontal lobe, overhanging fringe only very slightly. Arcs Ei, Ii_3, I„ situated in long sulci over
whole fringe. U present laterally in some specimens. Fringe narrows only slightly laterally. Small
but distinct lateral eye tubercle just in front of level of S2. External surface of glabella and genae
finely reticulate, fine concentric ridges also on mesial part of glabella. Pygidium with narrow border;
about seven axial rings and five pleural ribs.

Description. Pear-shaped glabella increases evenly in width and convexity to the mid-part of the pseudofrontal
lobe. Occipital ring and furrow arched very gently rearwards. Occiput weakly developed. SI deep proximally,
shallowing abaxially and directed rearwards at about 60° to the sagittal line. LI short (tr.), extended as part
of a very poorly swollen composite lobe directed abaxially forwards at about 30° to the sagittal line. S2 pit-
like. S3 shallow depression at side of pseudofrontal lobe just behind the mid-line of the latter. Very weak
median tubercle located at about same level as S3 and at the anterior end of a weak carina in many specimens.
Genal lobe broader posteriorly than maximum exsagittal length; inner part rising gently from axial furrow,
outer parts steeply declined. Lateral eye tubercle small but distinct, located just in front of level of S2.

Fringe steeply declined, narrowing very slightly laterally, comprising long sulci containing arcs Ei, Ii_3,
and I„. These arcs are complete, each containing \1-T1\ pits in the half-fringe (X = 19|, n = 9). Arc I4 is

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

91

developed laterally in two of the five specimens where this can be determined confidently. Posterior margin of
fringe directed very slightly abaxially forwards.

External surface of mesial part of glabella and inner part of genal lobe finely reticulate. This reticulation
becomes more subdued away from the median and lateral eye tubercles and is associated with weak concentric
ridges on the glabella. The surface sculpture is extremely subdued in the holotype and only specimen ascribed
to B. foveolatus intermedins by Stormer (1930) and is no more than a fine pitting on the genal lobe and hence
the eye tubercle appears more prominent. This sculptural difference is the only feature which distinguishes
Stormer’s specimen from other specimens of B. foveolatus and is here considered insufficient to warrant formal
taxonomic status.

Hypostoma and thorax not known. Pygidium triangular in outline, sagittal length (including fairly narrow
border) equal to about 35 % of maximum width. Axis tapers evenly rearwards at 25° and terminates on the
inner part of the border. Seven axial rings present along with an anterior articulating half-ring and a short
terminal piece which bears faint impressions of further apodemes. Axial furrow weakly raised. Five pleural
ribs developed, the anterior three extending to edge of border.

Botrioides ejfiorescens (Hadding, 1913)

Plate 13, figs. 5-13

1913 Trinucleus coscino(r)rhinus ANG; Hadding, pp. 74-75, pi. 7, figs. 18, 20, non 19 ( = B. simplex
sp. nov.).

1913 Trinucleus ejfiorescens Hadding, p. 75, pi. 7, fig. 21a-c.

1927 Trinucleus ejfiorescens (Hadding) [x/c]. Stetson, pi. 1, fig. 4.

1927 coxc/norr/;/nM5 [xic] (Angelin); Stetson, pi. l,fig. 12.

1930 Trinucleus foveolatus ANG., Stormer, p. 18 (pars Jiimtland material only), text-fig. 2d.

1952 Trinucleus bronni SARS and BOECK; Nilsson (pars), pp. 684, 691-692 (layers, f, j (pars), and
k (pars)).

1982 Botryoides [x/c] bronni (Sars); Karis, p. 58.

1982 Botryoides ejflozescens [x/c] (Hadding); Karis, p. 58.

1982 Botryoides [x/c] coscinorhinus (Angelin); Bergstrom, p. 190 (pars — material recorded by Nilsson
1952).

Holotype. By monotypy, an external mould of a cranidium (LO 2543T) from the Anderso Shale at Anderson,
Jamtland.

Material, localities, and horizons. Cephalic elements and rare pygidia are known from the type unit and locality
and from the Killerod Formation in Killerod Quarry, Scania.

Emended diagnosis. Pear-shaped glabella with moderately swollen pseudofrontal lobe overhanging
fringe very slightly. Arcs Ei, Ii, and complete, E2 and I2 present anteriorly and, in the case of I2
laterally, I3 present anterolaterally. Fringe narrows markedly posteriorly. Small but distinct lateral
eye tubercle opposite antereomesial edge of L2. External surface of gena smooth. Pygidium with
narrow border, seven to nine rings and up to three, weak, pleural ribs.

Discussion. B. ejfiorescens was placed in the synonymy of B. foveolatus by Stormer (1930, p. 3) and
the two are clearly closer to each other than to other species of Botrioides. There is an overall
similarity in the shape of the glabella and genae, the number of radii in the fringe (16^-21^ in B.
ejfiorescens (n = 3)), and the presence of arcs I2 and I3. Features which distinguish the cephalon of
B. ejfiorescens from B. foveolatus include the more rounded anterior part of the glabella, the slightly
more posteriorly situated lateral eye tubercle, the presence of E2 pits (although these are small and
difficult to discern on the dorsal surface), and the absence of I2 and I3 pits posteriorly thus producing
a markedly tapered fringe. The narrow pygidial border of B. ejfiorescens also allies it to B. foveolatus,
but the sparser and more subdued ribbing of pleural area serves to distinguish the Swedish species.

Botrioides margo sp. nov.

Plate 13, figs. 23-25; Plate 14, figs. 1 and 2

19826 Botryoides [sic] foveolatus, iavLnm^on,p. 168.

92

PALAEONTOLOGY, VOLUME 30

Holotype. A cranidium (RM Ar52033) from the lowest 10 cm of the Gullhogen Formation, at Flallekisbrottet,
Kinnekulle, Vastergotland, Sweden.

Paratypes. Six cephala/cranidia/lower lamellae from the type formation at the type locality and at Gullhogen
Quarry, Vastergotland (all specimens UM).

Derivation of name. Margo— hum, referring to the presence of such a feature on the outer part of the fringe of
this species.

Diagnosis. Fringe broad, declined with a distinct brim. Ei_4 developed anterolaterally but only Ei
extending to posterior margin, at least three I arcs present anterolaterally to posteriorly. Pits on
upper lamella in very narrow sulci. Sulci on lower lamella restricted to very shallow depressions
linking E pits in ventral view. Lateral eye tubercle distinct.

Description. Glabella increasing evenly in convexity forwards to the swollen pseudofrontal lobe which over-
hangs most of the mesial fringe. LI small, marked anteriorly by short (tr.), transversely directed SI. Adaxial
swollen part of L2 almost square in outline; outer part confluent with pseudofrontal lobe as a very narrow
(tr.) composite lobe outside the pit-like S2. Pseudofrontal lobe subcircular in outline, bearing a prominent
median node a short distance behind its longitudinal mid-point. Axial furrow broad (tr.) and shallow, directed
abaxially forwards at about 25° to the sagittal line to opposite S2 in front of which it curves gently adaxially
towards the anterior fossula. Genal lobe quadrant-shaped except anteromesially where it is gently indented,
gently inclined from axial furrow, more steeply so from posterior border furrow, steeply declined towards
fringe. Lateral eye tubercle situated at highest point of genal lobe, opposite and slightly in front of S2. One
internal mould (PI. 14, fig. 1) shows a weak eye ridge directed adaxially forwards from the eye tubercle. This
specimen also shows five subparallel ridges directed rearwards from the abaxial side of the tubercle towards

EXPLANATION OF PLATE 13

Figs. 1-4. Botrioides foveolatns (Angelin). 1 and 2, PMO H391, dorsal and frontal views of cranidium, upper
Fines Formation, promontary south of Engervik, Asker, x 2-5; also figured by Stormer (1930, pi. 1, figs. 1-
3) as holotype oVTrinucleus' foveoiatus intermedins. 3, PMO H0538, dorsal view of latex cast of cranidium,
same horizon as 1, Huk, Bygdoy, Oslo, x 4: also figured by Stormer (1930, pi. 1, fig. 8). 4, PMO H0538,
dorsal view of pygidium, on same block as 3, x 4-5.

Figs. 5-13. B. ejfiorescens (Hadding). 5, LO 5723t, oblique ventral view of lower lamella, Anderso Shale,
Anderson, Jamtland. 6 and 7, holotype, LO 2543 F, dorsal and oblique frontal views of latex cast of cephalon
lacking upper lamella of fringe frontally, same horizon and locality as 5, x 4; also figured by Hadding
(1913, pi. 7, fig. 21). 8, LO 2540r, dorsal view of partially exfoliated cranidium, same horizon and locality
as 5, X 5; also figured by Hadding (1913, pi. 7, fig. 18). 9, LO 2542/, dorsal view of internal mould of
pygidium, same horizon and locality as 5, x 5; also figured by Hadding (1913, pi. 7, fig. 20). 10, SGU Type
5065, oblique posterolateral view of partially exfoliated cranidium, same horizon and locality as 5, x 4. 11,
SGU Type 5066, dorsal view of partially exfoliated cranidium, same horizon and locality as 5, x 2. 12, LO
5722/, dorsal view of pygidium, Killerod Formation, level k of Nilsson (1952), Killerod Quarry, Scania,
X 5. 13, SGU Type 5067, anterolateral view of latex cast of cranidium, same horizon and locality as 5, x 5.

Figs. 14-22. Bergamia johanssoni sp. nov. 14, RM Ar53002, anterolateral view of latex cast of etched cranidium
showing reticulation on glabella, uppermost part of Komstad Limestone, Fagelsang, Scania, x 5. 15, RM
Ar53003, dorsal view of pygidium, same horizon and locality as 14, x 7. 16, RM Ar53004, oblique

posterolateral view of juvenile cranidium, same horizon and locality as 14, x 10. 17, RM Ar53005,

approximately sagittal section through cephalon, same horizon and locality as 14, x 5-5. 18, RM Ar53006,
oblique anterolateral view of latex cast of cranidium, same horizon and locality as 14, x 4. 19, RM Ar53007,
lateral view of latex cast of cranidium, same horizon and locality as 14, x4-5. 20 and 22, holotype, RM
Ar53001, dorsal and oblique frontal views of unwhitened cranidium, same horizon and locality as 14, x 3.
21, MGUH 16872, anterolateral view of gena and fringe, 30 cm above base of Upper Didymograptus Shale,
same locality as 14, x 5.

Figs. 23-25. Botrioides margo sp. nov. 23, holotype, RM Ar52033, dorsal view of cranidium, lowest 10 cm of
Gullhogen Formation, Hiillekisbrottet, Kinnekulle, Vastergotland, x 4. 24, UM Vg983/1, ventral view of
lower lamella, lower part of Gullhogen Formation, Gullhogen, Vastergotland, x 12. 25, UM Vg981, ventral
view of part of lower lamella external to girder, same horizon and locality as 24, x 9.

PLATE 13

OWEN, Botrioides, Bergamia

94

PALAEONTOLOGY, VOLUME 30

the posterior border furrow. The holotype shows a single, weak ridge extending from the eye tubercle to the
posterior border furrow on the left genal lobe which is partially exfoliated. External surface of glabella bears
a hne, weak reticulation on which is superimposed a series of subdued concentric ridges on the pseudofrontal
lobe. External surface of genal lobe smooth or weakly reticulate. Internal mould of glabella and genal lobe
smooth or finely pitted.

Inner part of fringe steeply declined, outer parts flattening out to form a distinct brim and bounded by a
broad marginal band. On upper lamella, pits arranged in very narrow radial sulci (twenty-two sulci in the half
fringe of two specimens). The width of these sulci increases slightly distally and in this broader part up to four
pit arcs can be discerned anterolaterally, otherwise individual pit arcs cannot be determined. Fragments of
lower lamellae show that at least three discrete I arcs are present laterally and up to four E arcs, arranged in
very shallow such in ventral view, are developed anterolaterally although this number of E arcs clearly
decreases both adaxially and abaxially. Only one E arc is present over the posterior few radii. There is no
zone of complication or genal prolongation. The girder is prominent and the lower lamella is extended as a
spine of unknown length, approximately parallel to the sagittal line.

Hypostoma, thorax, and pygidium not known.

Discussion. The narrowness of the sulci, the extensive E arc development causing a distinct brim on
uncompressed cephala, and the anterior overhang of the glabella serve to distinguish B. margo sp.
nov. from B. foveolatus and B. efflorescens. The presence of more than two E arcs in a trinucleine is
otherwise recorded in Stapeleyella Whittard, 1955 and Decordinaspis Harper and Romano, 1967
(see Hughes et al. 1975, pp. 559, 566). Dean (1974, pp. 5-6) also noted a short E3 arc in specimens
which he termed Berganiial sp. but which Hughes et al. (1975) reassigned to Botrioides. Both
Stapeleyella and Decordinaspis have four E arcs but the pit development of both E and I arcs is
much more complex than the simple radial sulci of B. margo. Dean’s material from the upper
Arenig of north-east Newfoundland is too incomplete for adequate generic assignment although
unlike species of Botrioides, the sulcation breaks down for the inner I arcs posteriorly.

Genus bergamia Whittard, 1955

Type species. Bergamia rhodesi Whittard, 1955, pp. 32-35, pi. 3, figs. 8-13, from the uppermost part of the
Mytton Flags (D. hirundo Zone), Bergam Quarry, Shelve, England; by original designation of Whittard (1955,
p. 31).

Bergamia johanssoni sp. nov.

Plate 13, figs. 14-22

Holotype. A cranidium (part and counterpart) (Rm Ar53001) from the uppermost part of the Komstad
Limestone at Fagelsang, Scania, south-west Sweden.

Paratypes. Seven cranidia/cephala and one pygidium from the type horizon and locality. One cephalon from
a thin limestone bed in the basal part of the Upper Didymograptus Shale, 30 cm above the Komstad Limestone
at the type locality.

Derivation of name. After Mr J. V. Johansson of Skollersta who collected most of the material.

Diagnosis. Swollen pseudofrontal lobe partially overhanging fringe. Internal mould of gena smooth.
Short eye ridges present. Arcs Ei, Ei, Ii, and I„ complete, I2 and I3 extensive but incomplete
anteriorly.

Description. Cranidium sub-semicircular in outline. Occipital ring not known. Occipital furrow broad (sag.,
exsag.) and shallow; transversely directed. SI deep, oval in outline, directed abaxially rearwards at about 45°
to sagittal line. S2 directed abaxially forward at about 60°, in which direction it deepens before shallowing
abruptly such that a narrow (tr.) composite glabellar lobe is developed. LI expands (exsag.) abaxially and is
connected to the pseudofrontal lobe by the composite lobe. S3 developed as shallow indentation in the
pseudofrontal lobe at the axial furrow. In transverse profile the preoccipital part of the glabella rises evenly to
the highest part of the swollen pseudofrontal lobe where the glabellar node is developed, approximately level
with S3. In front of this the glabellar profile is markedly convex forwards and the glabella overhangs the

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

95

cephalic fringe slightly. In dorsal view the pseudofrontal lobe is almost circular in outline with a maximum
width only slightly less than the sagittal length. In a juvenile specimen (PI. 13, fig. 16) the pseudofrontal lobe
is more oval in outline. Axial furrow broad (tr.) and shallow posteriorly, directed forwards at about 20° in
which direction it narrows, deepens, and bears a deep anterior fossula a short distance in from the fringe.
Quadrant-shaped genal lobe bears a subdued eye ridge near its inner edges. This ridge is directed abaxially
rearwards from the dorsal furrow at S3 at about 80°. On internal moulds the glabella and genal lobes of large
specimens are smooth. The preservation of these specimens in pure limestone is such that the calcareous
exoskeleton remains with the counterpart to the internal mould. Careful etching of a specimen with 10 % HCl
revealed a fine reticulation on the external surface but this was largely seen as colour mottling and because of
the close compositional similarity between skeleton and matrix further etching failed to produce a true external
mould. A latex cast of the resulting mould shows a faint reticulation on the glabella (PI. 13, fig. 14). The
juvenile cranidium shows reticulation even on the internal mould (PI. 13, fig. 16). Posterior border furrow
shallow, broadening (exsag.) abaxially where it bears a deep fossula. Ridge-like posterior border transversely
directed over most of its length, deflected very gently rearwards distally.

Cephalic fringe broadens abaxially; steeply and fairly evenly declined mesially. Less steeply declined laterally
where the outer parts flatten out as a distinct brim. The juvenile specimen shows this brim much less well
pronounced. No complete fringes are available but the pit distribution is constant between the various
incomplete fringes. Arcs Ei, E2, L, and I„ complete with E2 and L becoming progressively more distinct
abaxially from Ej and I„ respectively. Approximately fifteen pits are present in the half fringe of I„. I2 missing
only one pit mesially in the holotype, I3 beginning at Re. Both arcs continuous posteriorly. Pits arranged in
sulci at about 90° to the inner edge of the fringe except posteriorly on the distinct, short, genal prolongation.
The sulci here are arranged at an acute angle to the first set and are thus at a high angle to the posterior
border rather than parallel to it. Lower lamella bears a robust genal spine of unknown length set at a slight
angle to the sagittal line. Remaining details of lower lamella not known.

Hypostoma and thorax not known. Pygidium transverse, sagittal length (including border) equal to about
35 % of the maximum width. Convex (tr.) axis occupies 20 % of the anterior width of the pygidium, tapering
rearwards at 35°. Three distinct and two more subdued rings present in front of the border. Axis continues on
to posterior border where it is very gently swollen and bears a further six weak rings. Pleural field bears two
gentle ribs extending and dying out rearwards from the first and second axial rings at angles to the sagittal
line of 80° and 60° respectively. Articulating facet short (tr.) and directed rearwards at 60° to the sagittal line.
Inner edge of border marked by a narrow ridge anterolaterally; this ridge dying out towards the axis. Border
steeply declined, becoming broader adaxially and changing from concave outwards to convex outwards
towards the axis; thus the mesial part is swollen. Internal mould of pygidium smooth.

Discussion. As noted above in the discussion of Botrioides, Angelin’s original material of the type
species of that genus, Trinucleus coscinorinus, was thought to be from the Komstad Limestone at
Fagelsang. The present material differs from the specimen illustrated by Angelin primarily in having
a much broader fringe and a well-developed genal prolongation. The latter feature, along with the
absence of lateral eye tubercles, distinguishes the Komstad Limestone specimens from species
ascribed to Botrioides. Thus on both morphological and pragmatic grounds the present material is
not considered conspecific with ‘T.’ coscinorinus. The neotype selected above for Angelin’s species
is from a higher horizon in Scania but is more in keeping with Angelin’s illustration and the accepted
concept of Botrioides.

The presence of pit arc Ei and the short simple pygidium exclude the Komstad Limestone
material from Trinucleus Murchison, 1839 and indicate that it belongs in Bergamia. Hughes et al.
(1975, p. 556) also gave the absence of eye ridges as a diagnostic feature of Bergamia although they
did figure a specimen of B. prima (Elies, 1940) on which they drew attention to the presence of
short, faint ridges (1975, pi. 2, fig. 25). B. johanssoni sp. nov. has these ridges more pronounced,
resembling the condition in juvenile specimens of Trinucleus (e.g. Hughes et al. 1975, p. 556) and
even in adults of T. bicallis Rushton & Hughes, 1981.

Unlike B. johanssoni, most other described species of Bergamia have a narrow fringe with, in
addition to Ei and E2, no more than two I arcs, L and I„; the latter being incomplete frontally or
laterally. E2 is also incomplete in some species. B. wliittardi Hughes, 1971 has an anterior and
anterolateral pit development similar to that of B. johanssoni, but E2 is restricted to the anterior
part of the fringe and a fifth I arc (I4) is developed laterally. Hughes’s species, from the lower

96

PALAEONTOLOGY, VOLUME 30

Llandeilo of central Wales, also has more numerous pit radii, a larger genal prolongation with a
more complex pit distribution, and a more elongate pygidium.

B. johanssoni shows some resemblance to ‘T.’ sedgwicki Salter, a species from the middle Arenig
of south Wales which Whittard (1955) and subsequent workers tentatively placed in Bergamia. Drs
R. A. Fortey and R. M. Owens, however, propose to include Salter’s species along with a new
middle Arenig species in a new genus in their forthcoming revision of the Arenig of south Wales.
Dr Owens informs me (pers. comm. 1985) that the new genus is distinguished from Bergamia
(including B. johanssoni) in having inter-radii developed in the I pit series, no deep sulci, and a
narrower pygidial axis and border.

The overall cephalic morphology and fringe pit distribution of B. johanssoni is similar in many
respects to some of the younger Ordovician trinucleines, especially the late Caradoc Tretaspis
ceriodes (Angelin) (see Owen 1980u) which was also an immigrant to Scandinavia, at a time
of probable regression, and which also occurs in limestone units. The two forms are, however,
homeomorphs and are not directly related. The absence of lateral eye tubercles, the strong brim on
the fringe, the absence of fringe lists, and the pitting on the genal prolongation all serve to distinguish
B. johanssoni from the much later form.

Subfamily reedolithinae Hughes, Ingham, and Addison, 1975
Genus reedolithus Bancroft, 1929

Type species. Trimicleiis siibradiatiis Reed, 1903, pp. 12 14, pi. 2, hgs. 1-6; from the Balclatchie Group (lower
Caradoc) near Girvan, south-west Scotland; by original designation of Bancroft (1929, p. 77).

Reedolithus carinatus (Angelin, 1854)

Plate 14, figs. 4-10, 13-16; text-fig. 4
1854 Trinucleus carinatus Angelin, p. 65, pi. 34, fig. 3, 3a.

EXPLANATION OF PLATE 14

Figs. 1 and 2. Botrioides margo sp. nov. 1, UM Vg982, dorsal view of partially exfoliated cephalon, lower part
of Gullhogen Formation, Gullhogen, Vastergotland, x4-5. 2, UM Vg983/2, oblique posterolateral view of
internal mould of cranidium, same horizon and locality as 1, x 8-5.

Fig. 3. Trinucleid gen et sp. indet. PMO 87252, anterolateral view of cephalon, about 16 m above base of
Fines Formation, Hovodden, Hadeland, x7; also figured by Hughes et al. (1975, pi. 4, figs. 48-51) and
Wandas (1984, pi. 12h, k., l).

Figs. 4-10, 13-16. Reedolithus carinatus (Angelin). 4, neotype, PMO H460, dorsal view of partially exfoliated
cranidium, Vollen Formation, Gomnres, Ringerike, x4-5; also figured by Stormer (1930, pi. 4, figs. 10 and
11). 5, PMO H441, dorsal view of cranidium, same horizon as 4, x 5-5; also figured by Stormer (1930, pi.
4, figs. 5-7). 6, PMO 69176, dorsal view of latex cast of cephalon, lower Fossum Formation, borehole at
Saltboden, Frierfjord, Skien-Langesund, x4. 7, PMO 81684, dorsal view of internal mould of cranidium,
Arnestad Formation, Bjornsviken, Gyssestad, Asker, x6-5. 8, PMO 81682, oblique anterolateral view of
internal mould of cranidium, same horizon and locality as 7, x 6-5. 9, PMO 81772, lateral view of cephalon
showing lower lamella of fringe, same horizon and locality as 7, x 3. 10, RM Ar49461, posterolateral view
of latex cast of cranidium, labelled Skagen Limestone, Vastergotland, x 5. 13, PMO 81688, dorsal view of
cephalon showing lower lamella of fringe, same horizon and locality as 7, x 5. 14, PMO H449, dorsal view
of lower lamella of fringe and impression of thorax and pygidium, same horizon and locality as 4, x 5-5.
15, PMO 69549, dorsal view of deformed cranidium, 1 m below top of Arnestad Formation, Lindoya, Oslo,
X 2-5. 16, PMO 31264, lateral view of latex cast of cephalon, erratic block, Skattvold, Trondelag, x 4-5.

Figs. 11 and 12. Reedolithus sp. 11, UM Vg986, anterior view of fringe, Gullhogen Formation, near Viske-
Kleva crossroads, Mosseberg, Vastergotland, x3-5. 12, UM Vg987, dorsal view of internal mould of

cranidium, same horizon and locality as 1 1, x4.

Figs. 17 and 18. Reedolithus quehecensis Staiible. BM It7360, dorsal and oblique posterolateral view of partially
exfoliated cranidium. Citadel Formation, Quebec City, Canada, x4.

PLATE 14

MM

wi

OWEN, Botrioides, Reedolitlius, trinucleid indet.

98

PALAEONTOLOGY, VOLUME 30

1930 Reedolithus carinatus (ANG.); Stormer, pp. 30-39, pi. 4, figs. 1-13; pi. 5, figs. 1-18; text-figs.
17-20.

1934 Reedolithus carinatus; Stormer, p. 331.

1940 Reedolithus carinatus (Ang.): Grorud, pp. 159-160, text-fig. 2a.

71948 Reedolithus carinatus (Ang.); Thorslund in Waern et al., pp. 344, 347.

1953 Reedolithus carinatus, Stormer, pp. 62, 73, 81, 83, 84.

?1982fi Reedolithus carinatus; J'danusson, p. 168.

Neotype. A cranidium (PMO H460) from the Vollen Formation ('Ampyx Limestone’) at Gomnaes, Ringerike;
selected by Stormer (1930, p. 31), but see discussion below.

Material, localities, and horizons. Cephala, cranidia, and lower lamellae occur throughout the Vollen Forma-
tion and Arnestad Formation ('Lower Chasmops Shale’) of Oslo-Asker and in the former unit in Ringerike
where incomplete thoraxes and pygidia are known also. The species is known also from broadly comparable
levels in Skien-Langesund and Eiker-Sandsvaer. A cephalon from a glacial erratic of unknown provenance at
Skattvold, Trondelag, is included here as is a cranidium from Vastergotland which is labelled ‘Skagen
Limestone’ but may be from a lower horizon (see discussion below).

Description. The morphology and ontogeny of this upper Llandeilo-lower Caradoc species were described in
detail by Stormer (1930) on the basis of Vollen Formation specimens. The species is also recorded here from
the overlying Arnestad Formation but no redescription is necessary other than a reassessment of the fringe
pitting in more modern terms. Pits progressively decreasing in size from arcs Ei to most discrete although
pits in El and L share shallow sulci in a few specimens. A first internal pseudogirder is developed laterally
(e.g. Stormer 1930, pi. 5, fig. 18). Most specimens show a general radial alignment of pits anteriorly and
anterolaterally but this breaks down posteriorly. As text-fig. Aa shows, arcs Ei, L_2, and invariably
developed mesially where up to two more I arcs may also be present. By the anterior end of the axial furrow

n-28
5 -

postero-

laterally

3

4 ' 5 ' 6 ' 7 '

15 -
10 -
5 -

n- 1 8

in front of
axial furrow

' 3 ' 4 ' 5

^ 6 ' 7 ^

20 1
15 -
10 -
5 -

n-32

anteriorly

n-

1 1

1

9 ' 10

1 1 ’

12

13

Pits along
posterior margin

e

5 -I

c

n-12

Number of I arcs Pits in arc Ip

TEXT-FIG. 4. a-c, histograms showing the range of variation in selected characters of the fringe of Reedolithus
carinatus (Angelin) (note a, c refer to half-fringe data), d, e, sketches showing the extensive complex zone of
pit arcs between L and I3, based on PMO 69176 (see PI. 14, fig. 6) and PMO 81772 (PI. 14, fig. 9) respectively.

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

99

most specimens have arcs Ei, Ii_4, developed. The insertion of incomplete I arcs on the anterior and
anterolateral parts of the fringe conforms to the ‘normal’ trinucleid pattern outlined by Hughes et al. (1975,
p. 544), i.e. immediately outside arc Thus the arc numbering given above corresponds to that applied to
homologous structures in other trinucleids. Posterolaterally, however, the course of H commonly becomes
difficult to dehne unequivocally and up to three short arcs are developed between E and I3 (text-fig. Ad, e).
This area may be viewed as an extended zone of complication but in contrast to other trinucleids, in
R. carinatus it includes distinct arcs and very few adventitious pits. Thus the posterolateral fringe in R. carinatus
contains up to seven I arcs (text-fig. Aa) and where Ei can also be counted it is seen to contain a few more pits
than Nine to thirteen pits are present along the posterior margin of the fringe, excluding the posterior
fossula (text-fig. Ab).

Discussion. Although the neotype of R. carinatus was selected, without discussion, by Stormer from
Norway, Angelin’s original material was from loose blocks in Vastergdtland, Sweden. Angelin’s
illustration (1854, pi. 34, fig. 3, 3a) is of a cranidium with a swollen, carinate, glabella, an occipital
spine base, genal nodes close to the posterior border, a distinct genal prolongation, and a broad
fringe with large, discrete pits. Thus the concept of R. carinatus based on Norwegian material
does not differ from that exemplified by Angelin’s missing specimen. The only specimen from
Vastergotland seen by the writer which can be placed unequivocally in Angelin’s species is a
cranidium (PI. 14, fig. 10) collected by Jarvik in 1938. The horizon given on the specimen label is
the Skagen Limestone which in its present definition is low-mid Caradoc (Jaanusson 1982«, fig. 4)
but its looser usage in the past may have included lower horizons. Specimens of Reedolithus collected
by Jaanusson from the Gullhogen Formation (Llandeilo) in Vastergotland are too fragmentary for
detailed determination and are illustrated here under open nomenclature (PI. 14, figs. 1 1 and 12),
although of the named species, they are closest to R. carinatus.

R. carinatus differs from the type species R. subradiatus (Reed, 1903; see also Hughes et al. 1975,
pi. 5, figs. 64-69) primarily in its more transverse cephalic outline, posteriorly placed genal nodes,
and larger L-In pit size. In R. subradiatus pits in arcs Ei and L are considerable larger than those
of the remaining I arcs especially laterally, and are separated from them by a list on the upper
lamella and a well-developed first internal pseudogirder on the lateral and posterior parts of the
lower lamella. Only I2 laterally approaches the size of the two arcs outside it. Examination of
Reed’s syntypes in the British Museum (Natural History) shows that there are five to seven I arcs
in front of the axial furrow (cf. 3-5 in R. carinatus). This increases to eight laterally before the
extensive zone of complication in which the regular pit arrangement between arcs I3 and the
innermost three I arcs breaks down completely. R. carinatus and R. subradiatus have an overlapping
stratigraphical range: Llandeilo-lowest Caradoc for the former and lower Caradoc (see Tripp 1980,
table 1) for the latter.

The only other species undoubtedly ascribed to Reedolithus is R. cjuebecensis Staiible, 1952 (PI.
14, figs. 17 and 18) from the Citadel Formation (= Quebec City Formation, see Globensky and
Riva 1982, pp. C39-49) of Quebec, Canada. This species is from limestone blocks in melange
horizons in the lower part of the formation which also contains graptolites of the Neniagraptus
gracilis Zone (Riva 1972; Globensky and Riva 1982). The age of the blocks is not clear but as
earlier assessments of their shelly fauna suggested levels above the N. gracilis Zone (e.g. Osborne
1956; Globensky and Riva 1982) they are unlikely to be significantly older than that zone. Its
distinction from R. carinatus was discussed fully by Staiible (1953, pp. 1 14-1 17). Suffice it to note
here that examination of Staiible’s illustrations and a topotype sample in the British Museum
(Natural History) shows that the fringe of R. quebecensis has a distinct brim, pits in arcs Ei and L
are enlarged compared to those in the other I arcs (although the contrast is not as marked as in R.
subradiatus), and share shallow sulci. Four or five I ares are present anteriorly (n = 7) and in front
of the axial furrow (n = 9), five to seven laterally (n = 7). A zone of complication between arcs I2
and the innermost three I arcs is developed posterolaterally and although pitting in this area is
eommonly very irregular, a few specimens (e.g. Staiible 1953, fig. 3) show the development of short
ares. Its fringe characters are thus intermediate between those of R. carinatus and R. subradiatus
but on balanee they are eloser to the former.

100

PALAEONTOLOGY, VOLUME 30

Trinucleid gen. et sp. indet.

Plate 14, fig. 3

1975 Botrioidesl sp.; Hughes et al., pp. 563-564, figs. 48-51.

1984 Botrioides sp.; Wandas, pp. 234-235, pi. 12h, k, l.

Material, locality, and horizon. A cephalon (PMO 87252) from about 16 m above the base of the Elnes
Formation (Kirkerud Group) at Hovodden, Hadeland.

Description. Maximum width of pear-shaped glabella equal to about 75 % of sagittal length. Preoccipital
part of glabella increases markedly and evenly in convexity forwards to the strongly swollen pseudofrontal
lobe which overhangs the fringe and axial furrow. Weakly swollen ala directed abaxially forward at 30°
situated opposite weakly inflated occiput. SI and S2 deep. Composite lateral lobe swelling and broadening
forwards. S3 is a very gently depressed smooth area a short distance in front of S2. Genal lobe quadrant-shaped;
bearing a large lateral eye tubercle opposite S2, well away from the glabella. External surface of glabella
bearing a coarse reticulation mesially. This becomes finer and dies out abaxially. Genal lobe bears a similar
reticulation around the eye tubercle and this reticulafion also dies out towards the outer parts of the lobe
except for a broad (exsag.) strip which extends to the axial furrow opposite, and for a short distance in front
of S3.

Fringe narrow mesially, broadening abaxially, steeply declined with a narrow brim marginally. Anterior
arch and genal prolongation distinct. Approximately 19 pits in and about 22 in Ei. Mesially Ei and 2-3 I
arcs are represented in sulci although there is some irregularity both in the radial disposition of the sulci and
in one sulcus bifurcating distally. This latter feature may reflect a repaired injury (see Owen 1985 for other
examples). Beyond the axial furrow, individual pits become clearly distinguishable and additional arcs appear
such that by R12 there are 5 1 arcs and 2 E arcs with sulci restricted to a very shallow depression containing
El and E2. Weak lists are present between I2-I3 and I3-I4 here and these persist posteriorly where a sixth I
arc appears along with an I1-I2 list. A few adventitious pits are also present posteriorly and about 9 pits are
present along the posterior margin of the fringe. Marginal band broad. Long genal spine parallel to sagittal
line.

Discussion. Although the glabella and genal lobes of this specimen suggest an affinity to trinucleines
such as Botrioides the fringe shows few characters in common with the genus, points of difference
being the anterior arch, the irregularity of the sulci anteriorly, the absence of major sulci and
breakdown of regular radii laterally and posteriorly, the probable restriction of E2 to the posterior
part of the fringe, and the development of an appreciable genal prolongation.

Hughes et al. (1975) considered the Hadeland specimen to be close to the common ancestor of
Botrioides and Tretaspis. They noted that the anterior part of the fringe has sulci incorporating pits
of each arc, as in Botrioides, and the outermost arc is probably Ei whereas posteriorly the sulcation
is restricted to Ei and a short E2 arc, as in many species of Tretaspis. Equally, however, the lateral
and posterior fringe morphology (with the exception of the probable E2 pits) and the overall shape
of the glabella and genal lobes are close to the morphology of R. carinatus. The possibility that the
Hadeland cephalon is a reedolithine may be even more likely in view of the palaeogeographical
difficulties of the earliest Tretaspis species being found in Scoto-Appalachian faunas in the early
Caradoc and not appearing in Baltica until late in the Caradoc (Hughes et al. 1975, p. 585). In
contrast, Reedolithus appears virtually simultaneously in both Scoto-Appalachian and Baltic faunas
during the Elandeilo.

Acknowledgements. I am very grateful to Drs D. L. Bruton and J. K. Ingham for their very helpful criticisms
of an earlier draft of this paper and to Dr R. M. Owens and Messrs R. K. Melville, D. Rudkin, J. Johansson,
and A. Nilsson for their comments and assistance with various aspects of the study. I also thank the following
for access to specimens in their care: D. L. Bruton (PMO), V. Jaanusson (RM), K. Lindholm and P. Ahiberg
(LO), S. Laufeld and Y. Grahn (SGU), A. Nilsson (MGUH), S. Stuenes (UM), and S. F. Morris (BM(NH)).
The work was undertaken during the tenure of a NERC Research Grant which is gratefully acknowledged
and the manuscript was expertly typed by Jenny Orr.

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

101

REFERENCES

ANGELIN, N. p. 1854. PalaeolUologica Scandinavica 1: Crustacea fonnationis transitionis. Ease. 2, 21 -92.
BANCROFT, B. B. 1929. Some new species of Cryptolithus {s.I.) from the Upper Ordovician. Meat. Proc. Maaclir.
Ht. phil. Soc. 73, 67-98.

BERGSTROM, J. 1973. Palaeoecologic aspects of an Ordovician Tretaspis fauna. Acta. Pal. Polon. 23, 180-206.

1982. Scania. In bruton, d. l. and williams, s. h. (eds.). Field excursion guide. IV Int. Symp. Ordovician

System. Paleont. Contrib. Univ. Olso, 279, 184-197.

BOECK, c. 1838. Uebersicht der bischer in Norvegen gefunden Formen der Trilobiten Familie. In keilhau,
B. M. (ed.). Gaea Norvegica 1. Christiania.

BR0GGER, w. c. 1887. Geologisk kart over oerne ved Kristiania. Nyt. Mag.f. Natwvid. 31, 1-36.

BRUTON, D. L. and HARPER, D. A. T. 1985. Early Ordovician (Arenig-Llanvirn) faunas from oceanic islands in
the Appalachian-Caledonide Orogen, 359-368. In gee, d. g. and sturt, b. a. (eds). The Caledonide Orogen—^
Scandinavia and Related Areas. John Wiley & Sons.

(in press). Arenig-Llandovery stratigraphy across the Scandinavian Caledonides. In Harris, a. l.

and FETTES, D. j. (eds.). The Caledonide/Appalachian Orogen. Spec. Piihl. geol. Soc. Load.

LiNDSTROM, M. and OWEN, A. w. 1985. The Ordovician of Scandinavia, 273-282. In gee, d. g. and

STURT, B. A. (eds.). The Caledonide Orogen—Scandinavia and Related Areas. John Wiley & Sons.

CARLISLE, H. 1979. Ordovician stratigraphy of the Tranmore area. County Waterford, with a review of
Ordovician correlation for south-east Ireland. In Harris, a. l., Holland, c. h. and leake, b. e. (eds.). The
Caledonides of the British Isles— reviewed. Spec. Puhl. geol. Soc. Load. 5, 545-554.
cocks, l. r. m. and fortey, r. a. 1982. Faunal evidence for oceanic separations in the Palaeozoic of Britain.
J. geol. Soc. Load. 139, 465-478.

DEAN, w. T. 1971. Ordovician trilobites from the Central volcanic Mobile Belt at New World Island, north-
eastern Newfoundland. Bull. geol. Surv. Can. 210, 1-37, pis. 1-7.

1974. Lower Ordovician trilobites from the Summerford Group at Virgin Arms, New World Island,

northeastern Newfoundland. Ibid. 240, 1 -43.

1985. Relationships of Cambrian-Ordovician Faunas in the Caledonide-Appalachian Region, with

Particular Reference to Trilobites, 17-47. In gayer, r. a. (ed.). The Tectonic Evolution of the Caledonide-
Appalachian Orogen. Vieweg & Sohn, Braunschweig/Wiesbaden.
elles, g. 1940. The stratigraphy and faunal succession of the Builth-Llandrindod inlier, Radnorshire. Q. Jl
geol. Soc. Load. 95 (for 1939), 383-445.

FORTEY, R. A. 1984. Global earlier Ordovician transgressions and regressions and their biological implications.

In BRUTON, D. L. (ed.). Aspects of the Ordovician System. Palaeont. Contrib. Univ. Oslo, 295, 37-50.
FUNQUiST, H. p. A. 1919. Asaphus regionens omfattning: sydostra Skane og pa Bornholm. Kgl. Fisiograf.
Sdllsk. Handl.31{\),\-55.

GLOBENSKY, Y. and RiVA, J. 1982. Field Trip C. Ordovician Stratigraphy and Paleontology of St Lawrence
Lowlands and the Frontal Appalachians near Quebec City. In globensky, Y. (ed.). North American
Paleontological Convention Field Trip Guidebook. Montreal 1982, C1-C47.

GRORUD, H. F. 1940. Et profil gjennem Ogygiaskifer og Ampyxkalk pa Torberg, Frogner ved Oslo. Norsk geol.
Tidsskr. 20, 157-160.

HADDiNG, A. 1913. Undre Dicellograptussilferen i Skane jamte Nigra Darmed Ekvivalenta Bildningar. Funds.
Univ. Arsskr. N.F. A vd. 2, 9 ( 1 5), 1 -90.

1958. The Pre Quaternary Sedimentary Rocks of Sweden. VII. Ibid. 54 (5), 1 -262.

HARPER, D. A. T., OWEN, A. w. and WILLIAMS, s. H. 1985. The Middle Ordovician of the Oslo Region, Norway.
34. The Type Nakholmen Formation (upper Caradoc), Oslo, and its faunal significance. Norsk geol. Tidsskr.
34 (for 1984), 293-312.

HARPER, J. c. and ROMANO, M. 1967. Decordinaspis—a. new Caradoc trinucleid trilobite from the Ordovician of
Ireland. Proc. R. Ir. Acad. 65B, 305-308.

HAWLE, I. and CORDA, A. J. c. 1847. Prodrom einer Monographic der bdhinischen Trilobiten, 176 pp. Prague.
HUGHES, c. p., INGHAM, J. K. and ADDISON, R. 1975. The morphology, classification and evolution of the
Trmucleidae (Trilobita). Phil. Trans. R. Soc. Fond. B272, 537-607.

JAANUSSON, v. 1976. Faunal dynamics in the Middle Ordovician (Viruan) of Balto-Scandia, 301-326. In
BASSETT, M. G. (ed.). The Ordovician System: proceedings of a Palaeontological Association symposium,
September 1974. University of Wales Press and National Museum of Wales, Cardiff.

102

PALAEONTOLOGY, VOLUME 30

JAANUSSON, V. 1982a. Introduction to the Ordovician of Sweden. In bruton, d. l. and williams, s. h. (eds.).
Field excursion guide. IV Int. Symp. Ordovician System. Paleont. Contrib. Univ. Oslo, 279, 1-9.

1982^. Ordovician in Vastergotland. In bruton, d. l. and williams, s. h. (eds.). Field excursion guide.

IV Int. Symp. Ordovician System. Ibid. 164-183.

and KARis, L. 1982. Introduction to the Ordovician in Jamtland. In bruton, d. l. and williams, s. h.

(eds.). Field excursion guide. IV Int. Symp. Ordovician System. Paleont. Contrib. Univ. Oslo, 279,43-46.
KARIS, L. 1982. The sequence in the lower Alloehthon in Jamtland. In bruton, d. l. and williams, s. h. (eds.).

Field excursion guide. IV Int. Symp. Ordovician System. Ibid. 47-54.

KJERULF, T. 1857. iiber die Geologie des siidlichen Norvegens. Nyt. Mag. f. Naturvid. 9, 193-333.

MCNAMARA, K. J. 1983. Progenesis in trilobites. In briggs, d. e. g. and lane, p. d. (eds.). Trilobites and other
early arthropods: papers in honour of Professor H. B. Whittington, FRS. Spec. Pap. Palaeont. 30, 59-68.
MURCHISON, R. I. 1839. The Silurian System, founded on geological researches in the counties of Salop, Hereford,
Radnor, Montgomery, Caermarthen, Brecon, Pembroke, Monmouth, Gloucester, Worcester and Stafford; with
descriptions of the coalfields and overlying formations, xxxii + 768 pp. London.

NEUMAN, R. B. 1976. Early Ordovician (late Arenig) brachiopods from Virgin Arm, New World Island,
Newfoundland. Bull. geol. Surv. Can. 261, 11-61.

NILSSON, R. 1952. Till kannedomen om ordovicium i sydostra Skane. Geol. Foren. Stockh. Fork. 73 (for 1951),
682-694.

OSBORNE, F. F. 1956. Geology near Quebec City. Naturaliste Can. 83, 157-213.

OWEN, A. w. 1978. The Ordovician and Silurian Stratigraphy of central Hadeland, south Norway. Norges
geol. Unders. 338, 1-23.

1979. The upper Ordovician succession at Norderhov and on Frognoya in Ringerike, Norway. Norsk

geol. Tidsskr. 58 (for 1978), 245-258.

1980a. A new species of Crvptolithus (Trilobita) from the late Ordovician of Norway. J. Paleont. 54,

144-148.

19806. The trilobite Tretaspis from the upper Ordovician of the Oslo region, Norway. Palaeontology, 23,

715-747.

1983. The trinucleid trilobite Broeggerolithus in the Ordovician of Scandinavia. Geol. Foren Stock. Fork.

105, 49-53.

1985. Trilobite abnormalities. Trans R. Soc. Edinb.: Earth Sci. 76, 255-272.

PEGRUM, R. M. 1984. The extension of the Tornquist Zone in the Norwegian North Sea. Norsk geol. Tidsskr.
64, 39-68.

PRICE, D. 1981. Ashgill trilobite faunas from the Llyn Peninsula, North Wales, U.K. Geol. J. 16, 201-216.

REED, F. R. c. 1903. The Lower Palaeozoic trilobites of the Girvan District, Ayrshire. Palaeontogr. Soc.
[Monogr.], 1, 1-48.

RIVA, J. 1972. Geology of the environs of Quebec City. XXXIV Int. Geol. Congr. Excursion
B-19.

RUSHTON, A. w. A. and HUGHES, c. p. 1981. The Ordovician trilobite fauna of the Great Paxton Borehole,
Cambridgeshire. Geol. Mag. 118, 623-646.

SADLER, p. M. 1974. Trilobites from the Gorran Quartzites, Ordovician of South Cornwall. Palaeontology, 17,
71-93.

SKJESETH, s. 1963. Contributions to the geology of the Mjosa District and the classical sparagmite area in
Southern Norway. Norges geol. Unders. 220, 1-126.

STAUBLE, A. 1952. Les Cryptolithides de Quebec. Naturaliste Can. 79, 285-319.

1953. Two new species of the family Cryptolithidae. Ibid. 80, 83-1 19, 201-220.

STETSON, H. c. 1927. The distribution and relationships of the Trinucleidae. Bull. Mus. comp. Zool. Harv. 68,
87-104.

ST0RMER, L. 1930. Scandinavian Trinucleidae with special reference to Norwegian species and varieties. Skr.

Norsk Vidensk-Akad. Mat. Naturv. Kl. 1930 (4), 1-111.

1932. Trinucleidae from the Trondheim area. Ibid. 1932 (4), 169-175.

1934. Cambro-Silurian zones of the Olso Region, with a brief correlation between British and Norwegian

sections. Proc. Geol. ^4^5. 45, 329-337.

1940. Early descriptions of Norwegian trilobites. Norsk geol. Tidsskr. 20, 1 13-151.

1953. The Middle Ordovician of the Oslo Region, Norway, 1. Introduction to Stratigraphy. Ibid. 31,

37-141.

TRIPP, R. p. 1980. Trilobites from the Ordovician Balclatchie and lower Ardwell groups of the Girvan district,
Scotland. Trans. R. Soc. Edinb.: Earth Sci. 71, 123-145.

OWEN: MIDDLE ORDOVICIAN TRINUCLEIDS

103

w^RN, B., THORSLUND, p. and HENNiNGSMOEN, G. 1948. Deep boring through Ordovician and Silurian Strata
at Kinnekulle, Vestergotland. Bull. Geol. Inst. Univ. Uppsala, 32, 337-475.
wandAs, b. t. g. 1984. The Middle Ordovician of the Oslo Region, Norway. 33. Trilobites from the lowermost
part of the Ogygiocaris Series. Norsk geol. Tidsskr. 63 (for 1983), 21 1-267.

WHiTTARD, w. F. 1955. The Ordovician trilobites of the Shelve Inlier, West Shropshire. Palaeontogr. Soc.
[Monogr.], 1, 1-40.

Typescript received 6 October 1985
Revised typescript received 18 January 1986

ALAN W. OWEN

Department of Geology
The University
Dundee DDl 4HN

A NEW CYCLOCYSTOID FROM THE LOWER
ORDOVICIAN OF OLAND, SWEDEN

by VIVIANNE BERG-MADSEN

Abstract. Mo)wcy chides oekiudicus gen. et sp. nov. is described from upper Arenig age beds on Oland,
Sweden (Volkhov Stage, Langevoja Substage). The genus is monotypic and is the earliest known cyclocystoid
echinoderm. Palaeogeographical origins of the cyclocystoids remain uncertain, but Monocycloides is regarded
as a primitive sister group of the North American Ordovician Cyclocystoides.

Forty years have elapsed since the first Swedish cyclocystoids were described from the Silurian of
Gotland (Regnell 1945). As the oldest cyclocystoids then known were from the Kirkfieldian (Middle
Ordovician) of North America and the Ashgill (Upper Ordovician) of England and Scotland, the
migration route seemed evident. The cyclocystoids apparently reached Sweden during the Silurian
and Belgium-Germany in the Lower and Middle Devonian (Regnell 1948; Sieverts-Doreck 1951).

Smith and Paul recently (1982) revised the class Cyclocystoidea thoroughly, now to include eight
genera and twenty-five species. The early Middle Cambrian species from Australia {^Cyclocystoides'
primotica Henderson and Shergold, 1971) was removed from the class, and the range in age
thus re-established to Middle Ordovician-Middle Devonian. The stratigraphical and geographical
distribution was discussed and the cyclocystoids regarded as probably having evolved in North
America during the Lower Ordovician. This did not alter the assumed migration pattern.

Now the presence of fragments of a new cyclocystoid from the high Lower Ordovician of Sweden
extends the range considerably. The fragments represent a new genus with important features
distinct from all known genera but most closely related to Cyclocystoides. Modern palaeogeographi-
cal research questions an early Ordovician migration from Baltoscandia to North America and the
new genus Monocycloides is regarded as a primitive group of cyclocystoids which originated from
an ancestor in common with Cyclocystoides. There is no evidence of a close genetic relationship
with the genera from the Silurian of Gotland and thus the later migration theory still seems valid.

GEOLOGICAL SETTING

Oland is situated in the Baltic Sea, south-east of mainland Sweden (text-fig. la). Halludden is an
approximately 2 km long cliff section in the north of the island (text-fig. lfi).The name is generally
used for the whole 5 km long beach section north of Byxelkrok. The Lower Ordovician strata here
are about 20 m in thickness. The Halludden section comprises the upper ‘Limbata’, Langevoja,
Hunderum, and Valaste Substages (text-fig. 2) with a total thickness of 6-5 m (Grahn 1982; Jaanus-
son and Mutvei 1982).

The material described here was obtained from a level 10-20 cm below the oolitic bed whose
base forms the boundary between the Langevojan and Hunderumian Substages. The boundary is
accessible only at low-water level. For convenience the level is referred to as the E-level because of
its content of echinoderm fragments.

Lithology, methods, and material. Lithologically the Halludden sequence is identical to the similar
interval in the Boda Hamn core where the fauna and microlithology were described by Bohlin and
Jaanusson (in Bohlin 1955). Boda Hamn lies just outside the margin of the map, south-east of
Halludden (text-fig. \h).

(Palaeontology, Vol. 30, Part 1, 1987, pp. 105-116, pis. 15-16.|

© The Palaeontological Association

106

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1. a, map showing the geographical location of Oland in southern Baltoscandia. b, the northern-
most part of Oland and the bedrock geology at Halludden. Limestones of (A) Latorp and Volkhov Stages,
(B) Kunda Stage. The remaining part (white) is Middle Ordovician limestones. Modified from Jaanusson

and Mutvei ( 1982).

The Langevoja limestone is a grey, predominantly sparitic calcarenite interbedded with marl
layers and with a moderate to high glauconite content. The single beds are 2-5 cm thick. A regression
starting in Latorp times (text-fig. 2) reached its maximum on Oland during the early Valastean.
The water depth during late Langevojan time is estimated to have been no more than 30 m,
shallowing during the Hunderumian (Grahn 1982).

A series of limestone samples (45) covering the complete section, each between 50 and 100 g, was
dissolved in 10 % acetic acid. The sampling levels correspond to those of Grahn (1982) who kindly
provided most of the material.

In order to retain as much calcite as possible in partly glauconitized echinoderm fragments,
the dissolution was interrupted twice a week, the residue removed, and cleaned. Once dried the
phosphatized fossils become very fragile, and therefore the decanted sediment was used directly for
manual extraction of fossils without sieving. Mechanical sieving was carried out during the first
investigation of the Halludden fauna (H. Mutvei, Swedish Museum of Natural History, Stockholm,
pers. comm. 1983). This may explain why hardly any microfossils were obtained.

Microfossils do not occur in all samples. Large numbers are found in the samples just below and
above the ‘Limbata’-Langevoja and Langevoja Hunderum boundaries. The Hunderum-Valaste
boundary is practically unfossiliferous, chitinozoans and macrofossils excepted. In many cases
microfossils are found at levels where macrofossils are absent.

THE FAUNA IN GENERAL

The Halludden cliff section was sampled for fossils both qualitatively and quantitatively in 1948-
1950 by H. Mutvei. Some of the results have been published by Jaanusson (1957), Skevington (1963,
1965n, h), and Grahn (1980, 1982) who also gave a short review of additionally described or
unpublished fauna in general.

The macrofossils are mainly benthic and gradually become more diverse towards the end of
Langevojan time, with a maximum diversity reached during the late Hunderumian. Both soft-
bottom and hard-bottom forms occur in the shelly fauna. Predominantly fossils from the Kunda
Stage have been described (text-fig. 2).

BERG-MADSEN: LOWER ORDOVICIAN CYCLOCYSTOID FROM SWEDEN

107

Stages

Substages

Formational units
on Oland

Graptolite Zones

z

GIGAS

cc

Aluoja

>

z

<

-1

KUNDA

OBTUSICAUDA

Didymograptus

”bifidus

-J

Valaste

RANICEPS

* “ “

Hunderum

EXPANSUS

- “

Langevoja

LEPIDURUS

Didymograptus

hirundo

VOLKHOV

"Limbata”

LIMBATA

o

z

UJ

cc

<

Didymograptus

extensus

LATORP

Billingen

BILLINGEN

Hunneberg

HUNNEBERG

(Tetragraptus

approximatus)

TEXT-FIG. 2. Diagram showing the stages and substages in Arenig and Lower Llanvirn of Sweden correlated
with the British graptolite zones. The star indicates the position of the E-level. Modified from Grahn (1980,

1982).

Apart from chitinozoans (Grahn 1980, 1982) no fossils have been obtained previously from the
E-level (H. Mutvei, pers. comm. 1985). Predominant among the fossils now found are conodonts,
ostracodes, gastropods/monoplacophorans, and echinoderm fragments.

Anita Lofgren (Department of Historical Geology and Palaeontology, Lund) kindly identified
the conodonts, all of which are known from the Microzarkodina fiabellum parva Zone (Lindstrom
1984; Lofgren 1985). As the conodonts from the Halludden section have not previously been
described the taxa are recorded here; M. f. parva, Prionodus (Baltoniodus) prevariabilis norrlandicus,
Paroistodus originalis, Drepanoistodusl cf. venustus, D. basiovalis, Protopanderodus cf. rectus, Senii-
acontiodus corniformis, Drepanodus arcuatiis, Scalpellodus latus, and Cornuodus longibasis.

The ostracodes are preserved as internal moulds which are worn and difficult to determine.
Primitia sp. and Protallinnella sp. were identified by Ake Bruun (Geological Survey of Sweden,
Uppsala).

Coiled internal moulds are either gastropods or monoplacophorans or both; closer assignment is
impossible due to the lack of muscle scars.

Echinoderms. Echinoderm macrofossils (Cheirocrinus, Sphaeronites) have been recorded from
the ‘Limbata’ Substage and the upper part of the Hunderum Substage from other localities on Oland

108

PALAEONTOLOGY, VOLUME 30

(Regnell 1945, 1948). The number of echinoderm microfragments increases from the topmost
‘Limbata’ to the early Hunderum, thereafter decreasing drastically to be totally absent in Valastean
samples. Most identifiable fragments can be assigned to inadunate crinoids but some belong to
ophiuroids (A. B. Smith, British Museum (Natural History), London, pers. comm. 1985).

In the E-level about 50 % of the acid resistant residue consists of partly phosphatized, very fragile
echinoderm fragments. Roughly 75 % of the fragments are poorly preserved and only identifiable
as echinoderms due to their rather coarse stereom structure. Among the identifiable remaining
25 %, fragments of brachials and stem ossicles are dominant.

At least three main types of stem ossicles can be distinguished. 1, High, cylindrical forms without
ornament on the latera, narrow to wide lumen and smooth, undifferentiated zygum. 2, High,
pentalobate with variable lumen width. Areola subdivided in five discrete spoon-shaped fields with
prominent culmina connecting the perilumen with the narrow crenularium midway along each side.
Occasionally the culmina extend into minute spines. 3, High, circular with pentastellate narrow
lumen, wide areola, and narrow peripheral crenularium. The ossicles are mostly found isolated but
two or more together do occur. Single ossicles not referable to these types occur, for example
quadrangular forms.

CYCLOCYSTOIDS

In all, thirty-five marginal ossicles from cyclocystoids have been found in the E-level. Additionally,
two single marginal ossicles have been found in two lower levels, 108-1 17 cm and 159-162 cm below
the Langevoja-Hunderum boundary. Unfortunately, the total of thirty-seven marginal ossicles
diminished to twenty when picking and mounting for SEM, with the remaining left more or less
fragmentary. Although additional fragments in the sample residue must belong to cyclocystoids,
no attempt has been made to identify these. The presence of several echinoderm groups within the
same sample prevents identification with reasonable certainty. Furthermore, the small size of the
non-marginal plates of the cyclocystoids probably prevents any identification. All thirty-seven
marginal ossicles are of almost equal size, about 500 /rm long and 250 /.im broad. Variations are
due to etching. Two thin, awl-shaped ossicles without cupules represent newly initiated marginal
ossicles. Etching has obliterated the outer structure and ornament in all specimens but has revealed
the three-dimensional stereom structure.

In spite of the sample size (100 g before etching) and the size of the marginal ossicles the number
make the presence of only one specimen likely. Due to the rarity of cyclocystoids recorded in
general, and the distinct difference between any known genus and species and this material, it seems
well founded to erect a new monotypic genus.

The type material is deposited at the Palaeozoology Section of the Swedish Museum of Natural
History (Naturhistoriska Riksmuseet), Stockholm.

SYSTEMATIC PALAEONTOLOGY

Class CYCLOCYSTOIDEA Miller and Gurley, 1895
Family cyclocystoididae S. A. Miller, 1882
Genus Monocycloides gen. nov.

Type species. Monocycloides oelandicus sp. nov.

Derivation of generic name. Greek mono = alone, single.

Diagnosis. Cyclocystoid with > 35 marginal ossicles which are longer than broad, each with only
one cupule without tubercle and prominent radial facets. Strongly convex dorsally. Crest broader
than long, flat to saddle-shaped. Test and disc unknown.

Occurrence. Arenig (Lower Ordovician), Sweden. As currently known the genus is monotypic.

BERG-MADSEN; LOWER ORDOVICIAN CYCLOCYSTOID FROM SWEDEN

109

TEXT- FIG. 3. Reconstruction of a mar-
ginal ossicle of Monocycloides oelan-
dicus seen from the ventral {left) and
the dorsal side (in life position) to the
right. Pits, pustules, and articulation
ridges may have been slightly different
in life from how they are recon-
structed here.

500 pm

Description. The marginal ossicles are twice as long as broad, the width almost equal distally and proximally.
In radial cross-section they are rectangular to cuneiform, in lateral view characteristically submarine-shaped
(PI. 16, figs. 4 and 5) due to the convex dorsal surface and the high, almost flat-topped crest. The crest is
broader than long and forms 35-40 % of the length of the marginal ossicle. The lateral margins are slightly
raised, in lateral view convex to flattened, but the central part is always depressed, the distal and proximal
edges gently curved. The cupule zone is slightly oblique to the crest, forming 35-40 % of the total length of
the ossicle. The single cupule is squarish as seen directly from above but in fact is spoon-shaped as the straight
cupule walls taper proximally below the overhanging crest. The distal edge is sharp, there is no tubercle or
trace of any such structure. The circumferential canal is very narrow. The extended radial facets are horizontal
to slightly oblique to the crest; they are separated by a shallow groove (PI. 16, fig. 6). In most marginal ossicles
the radial facets form 20-30 % of the entire length of the marginal ossicle but some of the best preserved
ossicles suggests a tripartition (33 x 33 x 33 %) of equal size between cupule zone, crest, and radial facet. The
gap between adjacent crests is less than 30 % of the breadth of the crest. The lateral surfaces have two series
of articulation ridges, each with only a single ridge.

Remarks. The distinct shape of the marginal ossicles and the presence of only one cupule without
tubercle readily distinguish Monocycloides from any other cyclocystoid genus.

Monocycloides oelandicus sp. nov.

Plate 15, figs. 1-5; text-fig. 3

Type material. Holotype RM Ec27420, paratypes RM Ec27421 -27426.

Occurrence. Upper Arenig (Lower Ordovician), Langevoja Substage, Halludden, northern Oland, Sweden.
Derivation of species name. Latin oelandicus = Oland.

Diagnosis. As for genus.

Description. In all, thirty-five marginal ossicles of almost equal size were studied and form the basis for this
description. Average length and breadth based on the fifteen best preserved specimens; Length of marginal
ossicle (ventral) 480 pm, breadth of marginal ossicle (proximal) 190 /im, (distal) 230 pm, length of crest
200 ^m, breadth of crest 220 pm, height (top of crest perpendicular to dorsal surface) 2\0 pm. Measure-
ments according to Smith and Paul (1982).

Based on the dimensions of the wedge-shaped marginal ossicle a hypothetical size and shape of the test has
been calculated: diameter 3-8 mm, circumference 12 mm, and with approximately fifty marginal ossicles. The
disc diameter should form more than 75 % of the test. Calculations of this kind should be accepted with some
reservations. The test might have been ovoid. However, an imprint of a segment of a circle (PI. 16, fig. 8) has
been found in the acid resistant residue of the matrix. The size matches an imprint of the ventral surface, the
ridges representing the gaps between crests. Unfortunately the sample disintegrated when removing it from
the SEM.

The marginal ossicles may have been in contact along most of their entire length but the state of preservation
prevents a reliable judgement. The dorsal surface is occasionally strongly convex (PI. 16, figs. 2, 4, 5) and

110

PALAEONTOLOGY, VOLUME 30

continues into the radial facets without interjacent crescentic facets. This part is the thinnest and is most often
incomplete (PI. 15, hgs. 1 and 3; PI 16, fig. 6). The outer surface is absent due to the etching; pustules and
lateral striae obliterated.

Articulation ridges are seen in only a few ossicles, usually in the less well-preserved specimens, where the
etching has thrown the ridges into relief (PI. 16, fig. 4; text-fig. 3). The distal ridge follows the slope of the
cupule wall upwards to the circumferential canal and from there bends gently downwards ending at a point in
the middle of the crest length. The proximal ridge lies above the distal ridge, distally overlapping this. It is
horizontal to gently arched and extends to the proximal end of the crest in line with the radial fecet.

Remarks. The size of the marginal ossicles (and the estimated test) is small and the possibility that
they belong to a juvenile specimen cannot be discounted. The presence of only one cupule seems to
support this as the number of cupules seems to increase by growth in many other genera. Also,
progressive growth is suggested by the presence of thin ossicles without cupules, and some of the
largest marginal ossicles show evidence of growth in the cupule zone. Here the cupule walls become
increasingly broader towards the lateral faces, flattening dorsally (PI. 16, fig. 7). On the other hand,
the additional two ossicles found at lower levels within the Langevoja Substage which contradict
this, are exactly the same average size as the thirty-five from the E-level and also have only one
cupule. It is hardly believable that only juvenile specimens should be represented in all three levels.

Although the methods for extraction of fossils used here is primarily aimed at the microfauna,
larger fossils also occur in the samples. If larger marginal ossicles representing an adult stage existed
they would probably have been found. Furthermore, the sequence at Halludden and other localities
on Oland, for example the Boda Hamn exposure and core, have been searched specifically for
echinoderms. A large cyclocystoid would not have gone unnoticed. Finally, the size of the fauna in
general in the Langevoja (and most other) samples is worth noting. All fossils are small: ostracodes,
gastropods/monoplacophorans, inarticulate brachiopods, and the inadunate crinoids as estimated
from the size of the fragments. For some reason the fauna seems starved, perhaps due to unfavour-
able environmental conditions. This may also explain the relative paucity of macrofossils in the
Langevoja Substage (Grahn 1982), trilobites excepted. The supposition that M. oelandicus is adult
in spite of its small size therefore seems well founded.

Microstructure. A characteristic of most lower Ordovician echinoderms from Halludden is the
rather coarse labyrinthic stereom (PI. 15, figs. 1-5). The dorsal side of the marginal ossicles in M.
oelandicus are especially coarse (PI. 16, figs. 1-3) whereas the remaining part of the ossicle has a
finer structure. It seems that the stereom gradually becomes more dense towards the centre of the
ossicle. This can be seen in heavily etched ossicles (PI. 16, fig. 4). The radial duct does not appear as
an open channel, nor can any connection between the cupule zone and the radial facets be seen, for
example in the form of especially thin stereom structure. If connected with the respiratory system
the coarse stereom may indicate either an unfavourable environment (low oxygen content) or lack
of the tube-foot/ampulla system used by most larger echinoderms. Both cases are in fact possible
as the coarse stereom is found in almost all echinoderm fragments and Monocycloides is small
enough to make do with diffusion alone (Paul 1977).

Preservation. The three-dimensional stereom is preserved during etching due to the presence of
an acid-resistant component. This component is francolite (carbonate fluorapatite, Caio(P04)e

EXPLANATION OF PLATE 15

Figs. 1-5. Monocycloides oelandicus, holotype RM Ec27420. 1, marginal ossicle in slightly oblique ventral
view showing the cupule zone, the saddle-shaped crest, and the incomplete radial facets, x 200. 2, cupule
zone in distal view showing the narrow circumferential channel and the lateral slopes, x 200. 3, oblique
lateral-proximal view showing the radial duct and the incomplete radial facets, x 200. 4, slightly tilted
lateral view showing the raised lateral slopes and the depressed centre of the crest. The flat dorsal surface is
due to etching, x 150. 5, labyrinthic stereom of the lower left of the crest in fig. 1, x400.

PLATE 15

BERG-MADSEN, Monocycloides

112

PALAEONTOLOGY, VOLUME 30

X (CO3F) X (F,0H)2; XRD analysis by Ulf Sturesson, Palaeontological Institute, Uppsala). In
some cases the francolite appears to be a 2-4 yum thick rind of cement in the pore space, in analogy
with the Middle Cambrian echinoderms reported from Bornholm (Berg-Madsen 1986). Then, the
acid-resistant structure is the negative of the original calcite stereom. Most often, however, a
replacement/recrystallization of the original calcite matter seems to have taken place (PI. 16, figs.
1-3) and the acid-resistant structure then is a true replica of the original skeleton. Phosphatization
is believed to be an early diagenetic process gradually affecting the skeletons. Most of the unaltered
calcite (the cores of the trabeculae) was lost during the dissolution in acetic acid.

It is interesting to note that the oolites at the Langevoja-FIunderum boundary are completely
phosphatized to francolite (U. Sturesson, pers. comm. 1985). Phosphatization of the matrix also
occurs (PI. 16, fig. 8) although only very weakly as the samples are extremely fragile.

Phosphatization and glauconitization of microfossils have been described from the overlying
Middle Ordovician strata at Halludden by Eisenack (1978). He assumed that the original calcite
dissolved completely leaving voids where the francolite was introduced. With regard to glauconitiza-
tion he noted its occurrence especially in the echinoderm fragments. Glauconitized echinoderm
fragments occur in the E-level but are much more common in the ‘Limbata’ Substage and the
overlying Hunderum and lowermost Valaste Substages. The marginal ossicles of M. oelandicus
show no trace of glauconitization but tiny glauconite grains are often squeezed into the cupule and
the distal opening of the radial duct.

DISCUSSION

Palaeogeography. Since Regnell (1945, 1948) proposed his view of migration routes it has been
accepted generally that the cyclocystoids originated in North America, probably during early
Ordovician time (Sieverts-Doreck 1951; Kesling 1966). The known geographical and stratigraphical
distribution is shown by Smith and Paul (1982, fig. 13) together with proposed details of their
phylogeny (fig. 14). The occurrence of the cyclocystoid M. oelandicus in the Arenig of Baltoscandia
demands a critical assessment of this view in the light of recent research in palaeobiogeography. A
short review of the palaeogeography between the opening and closing of the lapetus Ocean is
compiled from McKerrow and Ziegler (1972), Jaanusson (1973a, b). Smith et al. (1973), McKerrow
(1979), Ziegler et al. (1979), Cocks and Eortey (1982), Lindstrom (1984), and Schallreuter and
Siveter (1985). Based on this the origin and possible migration will be discussed.

In early Cambrian times it appears that continental rifting began and the Gondwana, Laurentian,
and Baltic plates drifted apart. The Gondwana continent stretched roughly 50° north and south
of the equator during the Cambrian; it moved gradually polewards during Ordovician to early
Devonian times.

Laurentia (including western Newfoundland, Greenland, north-west Britain, and Ireland) was
already in late Cambrian time established across the equator, reaching from 20° S. to 20° N. Apart

EXPLANATION OF PLATE 16

Figs. 1-8. Monocycloides oelandicus, paratypes. 1, RM Ec27421, incomplete dorsal surface (distal part) of
typical cuneiform marginal ossicle, x 140. 2, RM Ec27422, rectangular and strongly convex dorsal surface,
X 135. 3, detail of fig. 2, showing the coarse labyrinthic stereom structure, x 475. 4, RM Ec27423, lateral
view of heavily etched marginal ossicle with convex dorsal surface, flattened crest and the articulation ridges
revealed by the etching, x 160. 5, RM Ec27424, lateral view of etched marginal ossicle with a more dense
stereom structure showing towards the centre of the ossicle, x 1 50. 6, RM Ec27425, ventral view of marginal
ossicle with prominent radial facets, the break indicating the groove leading to the radial duct, x 150. 7,
RM Ec27426, the cupule zone of a marginal ossicle with narrow cupule walls and lateral growth (arrows)
from the walls outward, x 1 10. 8, imprint of the ventral surface of M. oelandicus in porous, acid resistant
matrix, x 40. Owing to the porosity the sample collapsed when it was removed from the SEM.

PLATE 16

BERG-MADSEN, Monocvcloides

114

PALAEONTOLOGY, VOLUME 30

from a counterclockwise rotation during the Upper Cambrian through the Ordovician the position
was maintained until late in the Silurian.

Regarding the Baltic plate, opinions differ considerably not only with regard to size but also to
its degree of movement according to the authors cited above. Some authors include Baltoscandia,
northern Poland, and the Russian Platform eastwards to Ural, others also the southern part of the
British Isles, a large part of Central Europe, and eastern Newfoundland. However, more recent
opinion suggests that the latter areas were attached to the Gondwana continent (Cocks and Fortey
1982). A third opinion advocates a separate microcontinent consisting of the southern part of the
British Isles and eastern Newfoundland. From a position at cool, temperate latitudes in Middle
Cambrian times Baltica either moved gradually towards the equator and Laurentia or moved
polewards during the late Cambrian and early Ordovician and back towards low latitudes from
Caradoc times onwards.

The continents were separated from each other by oceans. The lapetus Ocean between Laurentia,
Baltica, and Gondwana closed towards the end of Silurian times. Tornquist’s sea is believed to have
separated Baltica from Gondwana from the early Ordovician (or earlier?) until it closed in the
Ashgill (Cocks and Fortey 1982). From then on the Rheic Ocean started opening. The maximum
distance between the continents seems to have been in the early Ordovician, Baltica lying at least
3000 km, and perhaps as much as 8000 km from Laurentia.

Origin and diversification. Regnell (1945, 1948, 1960a, b) reviewed the non-crinoid echinoderms, in
particular from Baltoscandia. Both during the Ordovician and the Silurian about 98 % of the
echinoderm fauna was apparently developed from an eastern stock. Among the exceptions were
the cyclocystoids Apycnodiscus, Polytryphocy chides, and Sievertsia (Smith and Paul 1982) which
were thought to have migrated from North America to Baltoscandia.

Migration from Baltoscandia to North America is known to have taken place at least from
Llanvirn times onwards (Schallreuter and Siveter 1985). There is, however, no evidence of migration
either way in Arenig or earlier times which would be necessary to explain two out of three models
of cyclocystoid evolution: a, Cyclocystoides existed back in the lower Ordovician and gave rise to
Monocycloides as a specialized side branch, or b, Monocycloides was more widely distributed in the
Lower Ordovician and gave rise to Cyclocystoides and hence the other cyclocystoid genera. Al-
though not impossible both models imply benthic animals from a shallow shelf sea environment
migrating across a wide (minimum 3000 km) ocean.

If large scale migration is excluded a third model is left: c, Monocycloides and Cyclocystoides
stem from a common ancestor found in either of the continents Laurentia, Baltica, and Gondwana.
Considering the still more advanced echinoderm groups recorded from the Cambrian the ancestral
group may be as old as, for example, the Middle Cambrian. By this time the lapetus Ocean was
still fairly narrow and migration accordingly would have been more easy than in early Ordovician
times. The cyclocystoids have much in common with edrioasteroids, although more with the
Isorophida than Stromatocystitoidea (Smith and Paul 1982). The early Middle Cambrian cyclocy-
stoid from Australia (Henderson and Shergold 1971) later excluded from the class is now redescribed
as a new genus {Edriodiscus) of stromatocystitids (Jell et al. 1985).

Monocycloides shows similarities with several other cyclocystoid genera but is most closely related
to Cyclocystoides. The articulation probably along the entire length of the marginal ossicle, the
length and shape of the cupule zone, the smooth-floored cupule and absence of tubercle, the crest
broader than long, and the prominent radial facets, are especially comparable with the Middle
Ordovician C. latus and the Upper Ordovician C. halli. However, both these species have two
cupules; in C. latus the cupules are squarish and not tapered proximally and in C. halli strongly
tapered but ovoid. Both have convex crests but C. latus has prominent radial facets which are
absent in C. halli. Also the shape in radial cross-section is different from that of Monocycloides,
which has a flattened crest and more strongly convex dorsal surface. Although the outer details of
this surface are not known, pits rather than a smooth or granular surface are suggested by the
coarse stereom structure.

BERG-MADSEN; LOWER ORDOVICIAN CYCLOCYSTOID FROM SWEDEN

115

There are fewer features in common with other genera, for example Apycnodiscus, Polytryphocy-
cloides, and an undescribed species of Sievertsia from the Silurian of Gotland (Smith and Paul 1982;
Christina Franzen-Bengtson, Swedish Museum of Natural History, Stockholm, pers. comm. 1985).
Whereas the dorsal surface tends to become flat or even concave, several cupules with tubercles are
always present although they seem to disappear secondarily in the above mentioned species of
Sievertsia. Even if flattened crests and radial facets occur the radial cross-section of the Silurian
cyclocystoids differs more from Monocycloides than Cyclocystoides does from Monocycloides.

The presence of only one cupule in all marginal ossicles is the most significant difference between
Monocycloides and Cyclocystoides. One cupule seems to represent a primitive feature, and if Mono-
cycloides was the direct ancestor of Cyclocystoides at least a minor part of the marginal ossicles
would have been expected to have two cupules. Until more (and older) genera or species have been
found Monocycloides is regarded as a primitive ‘sister’ group of cyclocystoids, originating from an
ancestor in common with Cyclocystoides and later genera. The ancestral form may have lived in
either of the three continents, with the widening lapetus Ocean separating the group, or an early
migration may have taken place. There is no reason to believe that Monocycloides is the ancestral
form of the genera found in the Silurian of Gotland; most possibly it became extinct before the end
of Middle Ordovician.

SUMMARY

Marginal ossicles of a cyclocystoid occur in the uppermost of the Langevojan Substage (Arenig),
Lower Ordovician, at Halludden, northern Oland, Sweden. These represent the hitherto oldest
known cyclocystoid. The thirty-five marginal ossicles are believed to belong to a single adult
specimen estimated to have had at least fifty marginal ossicles. The cyclocystoid is also the smallest
species ever recorded with an estimated diameter of 3-8 mm. The similarities with the genus Cyclo-
cystoides are many but the distinct differences including only one cupule, the shape of the crest,
the radial facets, and the articulation ridges warrant the erection of a new Monocycloides.

This is not regarded as the direct ancestor of Cyclocystoides nor of any other cyclocystoid genera
but is thought to represent a more primitive genus living almost parallel in time with the more
advanced Cyclocystoides, becoming extinct before a migration of cyclocystoids from North America
to Europe started.

There is no evidence of migration from Europe (Baltica) to North America (Laurentia) during
the early Ordovician and the common ancestor is believed to have lived in either of the three
continents, Laurentia, Baltica, and Gondwana, lying separated by the lapetus Ocean. An early
migration may also have taken place from Baltica or Gondwana to Laurentia. This does not affect
the migration theory of Regnell (1945, 1948).

Acknowledgements. I wish to thank Drs A. Bruun, C. Franzen-Bengtson, Y. Grahn, A. Lofgren, H. Mutvei,
and U. Sturesson for their assistance. In particular I am grateful to Dr E. T. Alexandersson, Department of
Quaternary Geology, Uppsala, for many stimulating discussions and advice and to Dr A. B. Smith for his
kind help and suggestions to the manuscript. Mrs M. Lindell and Mr T. Westberg gave technical assistance
and Dr M. G. Bassett improved the English. This is a contribution to Project Tornquist (IGCP Accession
no. 86).

REEERENCES

BERG-MADSEN, V. 1986. Middle Cambrian cystoid {sensii lato) stem columnals from Bornholm, Denmark.
Lethaia, 19, 63-80.

BOHLiN, B. 1955. The Lower Ordovician limestones between the Ceratopyge Shale and the Platyums Limestone
of Boda Hamn. With a description of the microlithology of the limestone by V. Jaanusson. Bull. Geol. Instn.
Univ. Upsala, 35, 111-173.

COCKS, L. R. M. and fortey, r. a. 1982. Eaunal evidence of oceanic separations in the Palaeozoic of Britain.
J. Geol. Soc. London, 139, 465-478.

116

PALAEONTOLOGY, VOLUME 30

EiSENACK, A. 1978. Phosphatischc und glaukonitische Mikrofossilien aus dem Vaginatenkalk von Halludden,
Oland. N. Jh. Geol. Paldont. Mh. 1, 1-12.

GRAHN, Y. 1980. Early Ordovician Chitinozoa from Oland. Sver. Geol. Unders. Ser. C, 775, 1-41.

1982. Chitinozoophoran palaeoecology in the Ordovician of Oland. Ibid. 792, 1-17.

HENDERSON, R. A. and SHERGOLD, j. H. 1971 . Cyclocystoides from early Middle Cambrian rocks of northwestern
Queensland, Australia. Pcdaeontology, 14, 704-710.

JAANUSSON, V. 1957. Unterordovizische Illaeniden aus Skandinavien. Bull. Geol. Instn. Univ. Uppsala, 37,
79-165.

1973fl. Aspects of carbonate sedimentation in the Ordovician of Baltoscandia. Lethaia, 6, 1 1 -34.

1973/). Ordovician articulate brachiopods, 19-25. In hallam, a. (ed.). Atlas of Palaeohiogeography.

Elsevier, Amsterdam.

and MUTVEi, h. 1982. Ordovician of Oland: Guide to Excursion 3. IV International Symposium on

the Ordovician System, Oslo 1982, 1-23. Sect, of Palaeozoology, Swedish Museum of Natural History,
Stockholm.

JELL, p. A., BURRETT, c. F. and BANKS, M. R. 1985. Cambrian and Ordovician echinoderms from eastern
Australia. Alcheringa, 9, 183-208.

KESLING, R. V. 1966. Cyclocystoids. In moore, r. C. (ed.). Treatise on Invertebrate Paleontology. Part U.
Echinodermata 3(1), U 188-2 10. Geological Society of America and University of Kansas Press, Boulder,
Colorado and Lawrence, Kansas.

LiNDSTROM, M. 1984. Baltoscandic conodont life environments in the Ordovician: Sedimentologic and paleo-
geographic evidence. Geol. Soc. Am. Spec. Pap. 196, 33-42.

LOFGREN, A. 1985. Early Ordovician conodont biozonation at Einngrundet, south Bothnian Bay, Sweden.
Bull. Geol. Inst. Univ. Uppsala, ns, 10, 1 15-128.

MCKERROW, w. s. 1979. Ordovician and Silurian changes in sea level. /. Geol. Soc. London, 136, 1 17-145.

and ZIEGLER, A. M. 1972. Palaeozoic Oceans. Nature, 240, 92-94.

PAUL, c. R. c. 1977. Evolution of primitive echinoderms, 123-158. In hallam, a. (ed.). Patterns of evolution as
illustrated by the fossil record. Elsevier, Amsterdam.

REGNELL, G. 1945. Non-crinoid Pelmatozoa from the Palaeozoic of Sweden. Meddn. Lunds geol. -miner. Instn.
108, 1-255.

1948. An outline of the succession and migration of non-crinoid pelmatozoan faunas in the Lower

Palaeozoic of Scandinavia. Ark. Kemi. Miner. Geol. A 26 (13), 1-55.

1960a. The Lower Palaeozoic echinoderm faunas of the British Isles and Balto-Scandia. Palaeontology,

2,161-179.

1960/). ‘Intermediate’ forms in Early Palaeozoic echinoderms. Report Intern. Geol. Congress XXI Session,

Norden, 1960, Pt. XXII, 71-80. Copenhagen.

SCHALLREUTER, R. E. L. and siVETER, D. J. 1985. Ostracodes across the lapetus Ocean. Palaeontology, 28, 577-
598.

SIEVERTS-DORECK, H. 1951. Uber Cyclocystoides Salter & Billings und eine neue Art aus dem belgischen und
rheinischen Devon. Senckenbergiana, 2>1, 9-30.

SKEViNGTON, D. 1963. Graptolites from the Ontikan limestones (Ordovician) of Oland, Sweden. I. Dendroidea,
Tuboidea, Camaroidea, and Stolonoidea. Bull. Geol. Instn. Univ. Upsala, 42, 1-62.

1965a. Graptolites from the Ontikan limestones (Ordovician) of Oland, Sweden. II. Graptoloidea and

Graptovermida. Ibid. 43, 1-74.

1965/). Chitinous hydroids from the Ontikan limestones (Ordovician) of Oland, Sweden. Geol. Fdren.

Stockholm Fdrhandl. 87, 152-161.

SMITH, A. B. and PAUL, c. R. c. 1982. Revision of the class Cyclocystoidea (Echinodermata). Phil. Trans. R.
Soc. Fond. B 296, 577-679.

SMITH, A. G., BRiDEN, J. c. and DREWRY, G. E. 1973. Phanerozoic World Maps. Spec. Pap. Palaeontology, 12,
1-42.

ZIEGLER, A. M., SCOTESE, C. R., MCKERROW, W. S., JOHNSON, M. E. and BAMBACH, R. K. 1979. PaleOZOic paleo-
geography. Ann. Rev. Earth Planet. Sci. 7, 473-502.

V. BERG-MADSEN
Geological Institute
Stockholm University

Typescript received 2 January 1986 S_106 91 Stockholm

Revised typescript received 26 March 1986 Sweden

UPPER LLANDOVERY DENDROID GRAPTOLITES
FROM THE PENTLAND HILLS, SCOTLAND

by ELIZABETH E. BULL

Abstract. Well-preserved dendroids and graptoloids of Upper Llandovery, Monoclimacis creimlata Zone age
are reported from the North Esk Inlier in the Pentland Hills, near Edinburgh, Scotland. The effects of
palaeogeography and tectonics at the time of this zone are discussed, and a section on dendroid terminology
is included. One new species Dictyonema peiitlaiulica sp. nov. is described and ten other species (six of which
have not been recorded from Britain before) are discussed. Emended diagnoses, based on material from the
Pentland Hills, for the genus Thallograptus and most species are presented.

The Pentland Hills near Edinburgh consist mainly of coarse clastic sediments and volcanics of
early Devonian age. At certain points denudation has revealed the underlying Silurian rocks which
form part of a chain of inliers of early Llandovery to late Wenlock age in the Midland Valley of
Scotland (text-fig. 1). The North Esk Inlier, near Carlops, consists of a regressive sequence ranging
from fully marine turbidites to terrestrial alluvial-fan deposits (Tipper 1976), and is the source of
all the specimens discussed here (text-fig. 2). These beds are in places highly fossiliferous; extensive
faunal lists were given by the UK Geological Survey (Peach and Horne 1899, pp. 589-606; later
updated by Mykura and Smith 1962, p. 14). Major collections were made by the Survey and by
enthusiastic amateurs (Haswell 1865; Henderson 1874, 1880; Henderson and Brown 1869); Laurie
(1892, 1898, 1899) collected and described many specimens of eurypterid from a locality in the
Gutterford Burn that has also yielded many graptolites. Graptolites were first described from the
Pentland Hills by Lapworth (1874), from the Habbies Howe Inlier which is now known to be a
further outcrop of the Reservoir Formation (Robertson 1985).

TEXT-FIG. 1. Location map showing the Silurian inliers discussed here, in
the Midland Valley of Scotland.

IPalaeontology, Vol. 30, Part 1, 1987, pp. 117-140, pis. 17-20.|

© The Palaeontological Association

PALAEONTOLOGY, VOLUME 30

Interest in the area was rekindled by Lament (1943, 1947, 1952, 1954, 1978), who illustrated
many of the fossils (unfortunately with, at best, sparse descriptions) and was the first to suggest a
late Llandovery age for the marine sequence, previously considered to be Wenlock. The stratigraphy
was redefined by Tipper (1976) who described four formations in the North Esk Group (text-fig.
2). The Reservoir Formation is only sparsely fossiliferous, yielding rare graptolites, brachiopods,
and trilobites throughout. Exposures along the Gutterford Burn, however, contain a diverse fauna,
including eurypterids, starfish, and echinoids (Spencer 1914-1940; Brower 1975). The sediments
and faunas of the overlying Deerhope Formation and highly fossiliferous Wether Law Linn Forma-
tion have been revised by Robertson (1985). The latter formation contains a shallow marine
fauna which becomes progressively restricted towards the boundary with the succeeding Henshaw
Formation (considered to be the Llandovery-Wenlock boundary). The Henshaw Formation is
thought to be fluviatile-terrestrial in origin (Tipper 1976). Robertson (1985) included an additional
division, the Cock Rig Formation, between the Deerhope Formation and Wether Law Linn Forma-
tion.

Many of the fossil groups from the North Esk Inlier are poorly described, with some exceptions:
the trilobites were comprehensively treated by Clarkson et al. (1977) and Clarkson and Howells
( 1981 ), the stylonuroid eurypterids were studied by Waterston (1979), and the bivalves by Robertson
(1985). Davidson’s (1868) work on the brachiopods is greatly in need of revision. I describe here
the graptolite fauna of the Reservoir Formation and sparse graptolite remains from the higher
marine beds (Deerhope, Cock Rig, and Wether Law Linn formations), all of Monoclimacis crenulata
Zone age.

The crenulata Zone assemblages given by two previous authors appear on text-fig. 3, which shows
that in Britain most species, and especially Monograptus spiralis, have long time ranges— except for
Monoclimacis crenulata itself, which is restricted to its zone. On Bornholm, however, Bjerreskov
(1971) recorded ''Monograptus aflf. crenulatus sensu Elies and Wood, 191 T, only from the Mono-
climacis griestoniensis Zone and Monograptus spiralis only from the spiralis Zone (equivalent

\

HENSHAW

OR8

DEERHOPE

WETHER
LAW LtNN

fossi I
local ities

RESERVOIR

TEXT-FIG. 2. Outcrop of the North Esk Inlier, Pentland Hills. The main fossiliferous
localities are in the Reservoir Formation, just south of the dyke in the Gutterford Burn
(arrowed), and in stream sections of the Wether Law Linn Formation.

BULL; LLANDOVERY DENDROID GRAPTOLITES

119

Qraptolite Zones
(BJERRESKOV)

(RICKARDS)

TEXT-FIG. 3. Time ranges of certain key late Llandovery graptoloids
as given by Bjerreskov (1971 ) and Rickards (1976) for the east Baltic
and Britain respectively.

in time to the lower crenulata Zone). Obviously some clarification of this broad zonal scheme is
required, a process I hope to assist by recording the dendroids of the Pentland Hills.

Tectonic setting. Leggett et al. (1979) recognized the Southern Uplands of Scotland as an early
Palaeozoic accretionary prism; the evolutionary history of the Midland Valley has been further
described by Bluck (1983, 1 984) (see text-fig. 4). The North Esk Group, a mainly regressive sequence,
was deposited during late Llandovery times in an interarc basin (Bluck 1983). This partial isolation
produced a relatively low faunal diversity, and possibly some unique faunal groups. The basin was
probably a linear feature stretching from Northern Ireland in the south-west to the Baltic region in
the north-east.

BIOSTRATIGRAPHY

The majority of the specimens described below were found in the Gutterford Burn section of the
Reservoir Formation (text-fig. 2). This formation was interpreted by Tipper (1976) as originating
from contourite or turbidite flow; further work has, in general terms, supported the turbidite model
(Robertson 1985). Considered in the context of the whole sedimentary sequence the formation
represents a prodelta sequence with a constant input of clastic material from a low grade sediment
source, forming thin-bedded turbidites by currents lacking the strength to create significant sole

TEXT-FIG. 4. The tectonic picture in the late Llan-
dovery (simplified after Bluck 1983, p. 130), show-
ing the formation of an accretionary prism, which
is now the Southern Uplands. The arc and fore-
arc basin have not survived. The Midland Valley
inliers were formed in an elongated interarc basin
which occupied what is now the Midland Valley.

120

PALAEONTOLOGY, VOLUME 30

marks. The presence of Ischadites sp. (a calcareous alga) suggests formation in the photic zone at
fairly shallow depth, although below storm-wave base.

The graptolites are found in two lithologies. First, in mudstones associated with either articulated
eurypterids (Waterston 1979) or echinoids (Kier 1973), with their spines virtually undisturbed, and
starfish (Spencer 1914-1940). Secondly, in winnowed bioclastic deposits (the ‘Gutterford Limestone’
Beds) which have evidently been transported, though not far nor for very long since well-preserved
trilobites and other articulated fossils also occur in these beds.

The outcrop of the Reservoir Formation reaches a maximum thickness of 1500 m, all of which
appears to be of cremdata Zone age (sub-stage C of the Telychian Stage: Cocks et al. 1970) as
sparse graptolite remains characteristic of this zone occur throughout, together with other diagnostic
forms such as the brachiopod Eoplectodonta penkillensis (Reed, 1917). The overlying Deerhope and
Wether Law Linn formations also fall within this zone, giving it a total thickness of about 2000 m.

In other parts of the British Isles the cremdata Zone is represented by only a sparse graptolite
fauna and much thinner sequences than that of the North Esk Inlier are found in the Carmichael
(Rolfe 1960) and Girvan (Cocks and Toghill 1973) inliers of the Midland Valley, the Cross Fell
Inlier (Burgess et al. 1970), the Howgill Fells (Rickards 1970), Northern Ireland (Rickards 1973),
and recently in Kirkcudbrightshire (Kemp 1985; Kemp and White 1985).

The zone is much more clearly developed in Europe (Boucek 1953, 1957; Bjerreskov 1971), and
attempts have been made to subdivide it. Bjerreskov noted the equivalence of the cremdata Zone to
her spiralis Zone, and introduced a Cyrtograptus lapworthi Zone between it and the lowermost
Wenlock C. centrifugus Zone. I consider her lapworthi Zone to equate with part of the cremdata
Zone (text-fig. 3). Clarkson and Howells (1981) noted the similarity between the trilobites from the
North Esk Inlier and those from the East Baltic; this has been confirmed by Howells (1982) and
Ramskold (1984).

Graptolite diversity in Scotland is generally lower in the west. Some forms, such as Monograptus
spiralis, appear to show different time ranges across the region, relative to Monoclimacis cremdata,
suggesting that one or more barriers to open access may have existed. At the south-west end of
the Midland Valley conditions seem to have become more restricted with time. Cocks and
Toghill (1973, p. 242) observed quite a thin cremdata Zone succession in the Girvan area, and hence
inferred a short duration for the zone. This is unlikely, but the thin sequence does suggest that the
consistently low graptolite faunal diversity in Britain is real and not caused by an increase in
sedimentation rate.

More details about the cremdata Zone assemblage are emerging but, crucially, M. cremdata itself
remains in need of revision. In the Pentland Hills, the dendroid species dealt with here outnumber
the graptoloids, although Monograptus priodon is by far the most common graptolite. The current
faunal list from the cremdata Zone of the North Esk Inlier is:

Dictyonema pentlandica sp. nov.

Thallograptus cf. arborescens Boucek, 1957
T. inaequalis Boucek, 1957
Coremagraptus kalfusi Boucek, 1957
C. imperfectus Kraft, 1982
C . plexus (VocidL, 1894)

Palaeodictyota pergracilis (Hall and Whitfield, 1872)

Retiolites geinitzianus geinitzianus { Barrande, 1 850)

Monograptus priodon (Bronn, 1835)

M. spiralis (Geinitz, 1842)

Monoclimacis cremdata (sensu Elies and Wood, 191 1)

R. g. geinitzianus occurs only rarely throughout the succession, although it is more common in the
lower beds. Monograptus priodon is common, usually as fragments preserved either flattened or
occasionally as 3D casts and moulds. When the hooked part of the theca is missing, it can be

BULL: LLANDOVERY DENDROID GRAPTOLITES

121

difficult to distinguish M. priodou from Monodimacis', this may have caused confusion in the past
as to the abundance of M. crenulata.

Monograptus spiralis is fairly common in the Reservoir Formation, and its coiled rhabdosomes
are generally well preserved; the long thecal spines (Bulman 1932; Rickards et al. 1977, p. 73, fig.
31a), so rarely preserved, are present on some specimens. The zone fossil Monodimads crenulata is
not common but the specimens available preserve its diagnostic features, e.g. sicula, apertural
eversion, and genicular spines.

MORPHOLOGICAL TERMINOLOGY

The definitions of morphological terms given by Bulman (1970) are mostly followed here, but some terms
applied to dendroids require further clarification; Ruedemann (1947, pp. 23-30) gave a historical review of
terminology, later updated by Boucek (1957, pp. 17-25).

Branches. Two types of branching of the stipe are observed in dendroids; dichotomous branching (bifurcation)
and lateral branching. In dichotomous branching the main stipe splits to form either two new main stipes that
are usually continuous and go on to bifurcate again (e.g. Dictyonema), or two thinner branches that usually
terminate rapidly (e.g. Thallograptus). Lateral branching was defined by Bulman (1970) as a ‘division of the
stipe in which branches diverge at an angle to parent stipe, which continues its original direction of growth’.
Stipes formed in this way can be either the same thickness and rigidity as the main stipe, or thinner side
branches; both types are observed in Coremagraptiis. Ruedemann (1947) and Kraft (1982, 1984o) used the
term ‘monopodial branching’ to describe how lateral branches form from only one side of the main stipe, but
the term is redundant and its use should be abandoned. Genera of the family Acanthograptidae show both
dichotomous and lateral branching.

Due to the particular type of preservation of material described in this paper, some of the internal structures
of the dendroids can be seen, particularly in the acanthograptids Coreniagraptus, Thallograptus, and Palaeo-
dictyota, where the stipes consist of very many elongated tube-like thecae, up to 1 mm long (Bulman 1955, p.
22). When these form a lateral branch as a bundle of tubes, they are known as twigs (Boucek 1957, p. 87). The
nature of these twigs is discussed below (p. 126).

Connection of branches. There are two ways in which the stipes connect or join: by dissepiments and by
anastomosis. A dissepiment was defined by Bulman (1970, p. 22) as a ‘strand of cortical periderm serving to
connect adjacent branches in dendroid rhabdosome (especially in Dictyonema)'-, his definition is followed here.
The various types of dissepiment were discussed by Boucek (1957, p. 51). Bulman (1970, p. 23) defined
anastomosis as ‘temporary fusion, as of adjacent branches to form an ovoid mesh’. The term is used here
only to refer to the approach and subsequent joining of two stipes or branches. Any later separation of the
stipes is considered a distinct branching episode. When two adjacent stipes come together and anastomose,
they then form a single stipe and have not been observed to maintain the separate character of the original
stipes; where separate thecal bundles are described, I consider that these represent the superposition of one
stipe upon another and not anastomosis.

Meshworks of stipes. Regular meshworks of stipes are formed by anastomosis or by formation of dissepiments.
Variation in size and shape of the holes in the meshwork of most dendroids is taxonomically significant, so a
consistent terminology is required. Boucek (1957, p. 2) suggested ‘meshes’ and ‘fenestrulae’ for the larger and
smaller openings in the net respectively. The single term adopted here is fenestellae.

SYSTEMATIC PALAEONTOLOGY

Localities. All specimens were found in the North Esk Inlier, Pentland Hills, Scotland and, unless otherwise
stated, they originate from the ‘Gutterford Limestone’ bed locality in the Gutterford Burn (arrowed on text-
fig. 2).

Repositories of specimens. RSM GY, Department of Geology, Royal Museum of Scotland, Chambers Street,
Edinburgh; EDNCM, Grant Institute of Geology, King’s Buildings, West Mains Road, Edinburgh; GSE,
British Geological Survey, Murchison House, West Mains Road, Edinburgh.

122

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 5. Dictyonema pentlandica sp. nov. EDNCM 20237, conical proximal end; the rhabdosome
is preserved gently folded and reaches a length of 200 mm; superimposed on this specimen are two
fragments of Coremagraptus imperfectus, x 0-75.

Class GRAPTOLITHINA Bronii, 1849
Order dendroidea Nicholson, 1872
Family dendrograptidae Roemer in Freeh, 1897
Genus dictyonema Hall, 1851
Dictyonema pentlandica sp. nov.

Plate 17; text-figs. 5-8

Diagnosis. Rhabdosome broadly conical, up to 400 mm axial length. Terminates proximally with a
short stem, 5-10 mm long, either attached to a solid object or accompanied by fine root fibres.
Some secondary thickening proximally. Stipes slender, 0-22 mm wide, 1 7-22 in 10 mm. Dissepiments
regularly arranged and very thin, 0 06 mm, 1 1-13 in 10 mm. Autothecae not known; bithecae simple
tubes extending alternately to either side of the stipe with aperture pointing distally, forming distinct
swellings of the stipe, closely spaced, 28-30 in 10 mm.

Type material. Holotype RSM GY. 1985.30.1. Paratypes RSM GY. 1985.29.4 and EDNCM 20237.

Other material. RSM GY.1980.51.5, 1985.29.1, 1985.29.2, 1985.30.2, 1897.32 and thirty-four unregistered
specimens in ‘Drawer 61’ of RSM; and three unregistered EDNCM specimens.

Description. The rhabdosome grew as a broad cone but, due to the extremely large size of this species, it is not
possible to tell how wide the cone became. It is commonly preserved with the walls forming broad undulating
folds, and this may reflect the original form of the cone (text-figs. 5 and 6). The base of the cone is sometimes
secondarily thickened, up to double the thickness of the normal stipe, and has a short stem (text-fig. 7a). No

explanation of plate 17

Figs. 1-5. Dictyonema pentlandica sp. nov. Reservoir Formation, Gutterford Burn (see text-fig. 2); crenulata
Zone, upper Llandovery. I, RSM GY. 1985.29.1, numerous fragments superimposed, associated with
Coremagraptus imperfectus Kraft, 1982, xO-5. 2, RSM GY. 1985.29.2, several rhabdosomes showing basal
attachment and rejuvenation of meshwork, x 0-5. 3, RSM GY. 1985.30. Ip, holotype, part of mesh showing
an area of rejuvenation, x 3. 4, RSM GY. 1985.30. 1, part of mesh where thecal structure is preserved; note
that a pair of dissepiments is associated with each bifurcation, x 10. 5, RSM GY. 1985.29.4, showing
secondary thickening of basal region and root fibres, x 1.

PLATE 17

BULL, Dictyonema, Coremagraptus

124

PALAEONTOLOGY, VOLUME 30

thecae were observed on the stem because of secondary thickening. This species is often found attached to
shells (text-fig. 7b) such as Leplaena (which lived partially buried in the sediment) or less well-anchored forms,
when the stem is accompanied by many small root fibres.

Autothecae are not seen, but the short, tubular bithecae give the rhabdosome a distinctive appearance,
being preserved as regular swellings of the stipe (text-fig. 8a) which appear to be Type T bithecae of Bulman
(1933), i.e. simple tubes emerging to the side of the main stipe with apertures pointing distally (see Chapman
and Rickards 1982, p. 220).

Stipes are straight and evenly spaced, bifurcating regularly every 6 mm; this interval does not become
appreciably larger distally. Bifurcation is concentrated on certain stipes, while others are continuous, straight,
and unbroken.

TEXT-HG. 6. Dictyonema pentlamUca sp. nov. Reconstruction showing
the broad conical form; clusters lived attached by a short stem with
root fibres to a solid object.

TEXT-FIG. 7. Dictyonema pentlandica sp. nov. Modes of attachment, a, RSM GY. 1985.29.4,
showing secondary thickening of stem and base region, and root fibres for securing stem
within sediment. B, RSM GY. 1985.30. 1, attached to brachiopod shell securely anchored in the
sediment (boxed area is shown on text-fig. 8a). Scale bars 10 mm.

BULL: LLANDOVERY DENDROID GRAPTOLITES

125

TEXT-FIG. 8. Dictyonema pentlandica sp. nov. RSM GY. 1 985.30.1. A, knobbly appearance of stipe is caused
by presence of large bithecae, x 10. b, dissepiments associated with each bifurcation, x 10.

Dissepiments are of constant length, perpendicular to the stipes, and often form continuous chains across
the rhabdosome. Each point of bifurcation is directly associated with two dissepiments— one on either side of
the zone of bifurcation (text-fig. 8b). The fenestellae formed by dissepiments and stipes are of fairly constant
size (0-6 X 0-4 mm), and generally rectangular. When dissepiments grow from the swelling of a bitheca the
fenestellae may become ovoid. Zones of mesh rejuvenation occur, similar to those described by Bulman (1950)
for other species of Dictyonema, and are probably related to regrowth after damage (PI. 17, hg. 3).

Discussion. This species shows considerable resemblance to both D. delkatulum Lapworth, 1881
(see Boucek 1957 and Kraft 1984/i) and D. elegans Bulman, 1928. It is distinguished by: the very
distinctive swollen or knotted appearance of the stipe, due to the thecal morphology; the large, very
elongated nature of the rhabdosome; and the thinness of the stipes, which are slightly wider spaced
(17-22 in 10 mm, rather than 22-23 or 20-24 respectively).

The general outline of the stipe, caused by the thecal type, closely resembles D. geniculatum
Bulman, 1928, but is easily distinguished from it by the smaller size and spacing of the stipes. Kraft
(1984o) noted the close similarity of D. delkatulum and D. elegans, particularly in poorly preserved
specimens; he stressed that ‘their stratigraphic distribution must be taken as the fundamental
criterion’, D. delkatulum being limited to lower Llandovery and D. elegans limited to Wenlock.
Care should be taken in limiting dendroid faunas to particular stratigraphic horizons; it is uncertain
at present whether these three species are stratigraphically useful, as D. pentlandica appears to be
intermediate both structurally and temporally.

Family acanthograptidae Bulman, 1938

Included genera. I follow the classification of Boucek (1957; after Bulman 1938) and distinguish the genera
Acanthograptus Spencer, 1878, Thallograptus Ruedemann, 1925, Coremagraptus Bulman, 1927, and Palaeo-
dictyota Whitfield, 1902. Holland et al. (1967) referred only Acanthograptus, Coremagraptus, and Palaeodictyota
to this family and assigned Thallograptus to the family Inocaulidae. The evidence presented below shows that
Thallograptus is much more closely related to the Acanthograptidae.

Discussion. Boucek (1957, p. 86) grouped Acanthograptus and Thallograptus as forms with ‘Rhabdo-
some composed of individual branches or completely ramified; branches mostly free, not uniting
with each other’. Similarly, he grouped Coremagraptus and Palaeodictyota as forms with ‘Rhabdo-
some infundibuliform, branches more or less flexuose and regularly united by anastomosis’. As I

126

PALAEONTOLOGY, VOLUME 30

consider Acanthograptus and Palaeodictyota to be the two morphological extremes of the family,
no subfamilies {sensu Boucek) are defined here; since Acanthograptus is not found in the North Esk
Inlier, it is only considered for the sake of completeness.

Acanthograptus is composed of relatively few tube-like thecae that regularly and continuously
branch off from the main stipe in bundles of two autothecae and two bithecae to form very many
small twigs, of approximately the same length, along the main stipes and usually ventrally or sub-
ventrally. It is possible that all thecae terminate by turning away from the stipe and forming a twig,
and that none terminate as pores on the surface of the branch. Boucek’s diagrams (1957, e.g. fig.
37/) show some anastomosis of main stipes, but this is probably due to one branch being superim-
posed on another; indeed the lack of anastomosis seems characteristic of this genus.

ThaUograptus is an intermediate form. Like Acanthograptus, a major distinguishing feature is the
lack of anastomosis, but ThaUograptus has much less regularity in the arrangement of its twig-like
side branches. The main stipes maintain a constant thickness. They bifurcate and also form lateral
branches, but the latter gradually become thinner, consisting of fewer and fewer thecae until only
one or two remain (text-fig. 1 Ic); this gives the rhabdosome a very spiny appearance (text-fig. 9b).
In some species the thinning begins abruptly, giving the branch a swollen appearance (text-fig. 10b)
at the base of the terminating twig.

Coremagraptiis also has many side branches that thin as the constituent thecae successively reach
their full length, although in this genus most thecae terminate together abruptly without one or two
thecae becoming distally isolated. Corernagraptus shows quite frequent but not always very regular
anastomosis. Sometimes two stipes are connected by a number of lateral branches from one joining
with another. Side branches are increasingly less abundant in ThaUograptus and Corernagraptus,
and not all thecae terminate as discrete twigs (as in Acanthograptus). In both genera thecal apertures
appear also as pores on the surface of the stipe (text-fig. 13).

Palaeodictyota is characterized by complete anastomosis, and represents the other morphological
end member of the family. All thecae terminate within the stipe or as pores on its surface; none
form twigs or discrete branches. The main stipes are fairly continuous and rarely bifurcate; the
meshwork is formed by lateral branches joining other stipes, creating regular fenestellae with a
distinctive and constant size and shape.

Genus thallograptus Ruedemann, 1925

Type species. T. succulentus (Ruedemann, 1904). Lower Ordovician of Deep Kill, New York and Point Levis,
Quebec, Canada.

Etnended diagnosis. Rhabdosome shrub-like, stipes bifurcating and forming lateral branches irregu-
larly; branches discrete, no anastomosis; thecae tubular, adnate throughout their length, opening
as pores on the surface; branches taper to a single theca distally.

Discussion. Ruedemann (1947) assigned ThaUograptus to the family Inocaulidae, together with the
genera Inocaulis, Medusaegraptus, and Diplospirograptus; I follow Boucek (1957, p. 88) in abandon-
ing this classification. Neither the diagnosis by Bulman (1938) nor Ruedemann (1947, p. 230)
adequately summarized all of the features of the genus. Ruedemann’s reference to ‘hair-like fila-
ments’ (see Boucek 1957, p. 97) appears either to be a misidentification or to refer to a different genus.
Boucek (1957, pp. 97-98) discussed the lack of branch connection by cross-bars or anastomosis, and
noted the intermediate nature of this genus between Acanthograptus and Corernagraptus without
formally emending the diagnosis.

ThaUograptus is recorded from both the Ordovician and Silurian. Boucek (1957) considered it to
peak in numbers during the Ludlow, but it is now known to be common also in the Llandovery
(Kraft 1982; this paper). Numerous species have been described (Boucek 1957; Ruedemann 1947;
Kraft 1982, 1984u), and these are distinguished by thickness of stipe, length of branches, whether
branches appear inflated before termination, angle of stipe, bifurcation, and branching.

BULL: LLANDOVERY DENDROID GRAPTOLITES 127

TEXT-FIG. 9. Thallograptus arborescens Boucek. RSM GY. 1985.28.2/?. A, rhabdosome with well spaced, taper-
ing branches, x 2. b, enlargement of boxed area on a, showing preservation of thecal tubes and branches

tapering to a single theca, x 25.

Thallograptus arborescens Boucek, 1957

Plate 18, figs. 1-4; text-figs. 9 and 10

1957 Thallograptus arborescens Boucek, pp. 106-107, pi. 23, fig. 4.

Emended diagnosis. After Boucek (1957). Rhabdosome up to 30 mm wide, partially flabellate and
shrub-like in form. Main branches approximately 1 mm thick, bifurcating regularly at 30-70°,
thinning gradually to 0-5 mm, and nowhere perceptibly inflated. Lateral branching frequent, ter-
minal branchlets 0-4-0-5 mm long, thinning to single theca, and separated by about 2 mm. Branches
entirely free without anastomosis or dissepiments.

Material. RSM GY. 1985.28.2- 1985.28.4 and nine unregistered specimens all located in ‘Drawer 62’ of the
RSM collections; GSE 14161 and two unregistered specimens; and EDNCM 1984.87.3-1984.87.5 and one
unregistered specimen.

Description. RSM GY. 1985.28.2 is undoubtedly T. arborescens and has terminating branchlets tapering from
a maximum thickness of 0-2 mm to a single theca thickness (text-fig. 9b) of 0 04 mm, but exceptionally
elongated up to 3 mm long. Branchlets curve out at progressively increasing angles from the stipe, which is
not straight but changes course slightly at each branching episode. The mean angle of stipe bifurcation is 70°.

Other specimens examined, such as RSM GY. 1985.28.3 and 1985.28.4 (text-fig. 10), GSE 14161 (PI. 18, fig.
4), and EDNCM 1984.87.3-1984.87.5 (PI. 18, figs. 1-3), are distinguished by a slight thickening at the base of
some of the branchlets. All other dimensions are very similar to T. arborescens s.s.. including main stipe

128

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 10. Thallograptus arborescens Boucek. a, RSM GY. 1985.28.3/), showing some
complete, well-preserved branches, sometimes superimposed, b. RSM GY. 1985.28.4,
specimen displaying branching pattern and apparent swelling of branch before end. Scale

bar 10 mm.

thickness of 0-6 0-8 mm, stipe divergence angle of 45-60°, and branchlets 1-2 mm apart and greater than 1-5
mm long; overall rhabdosome size up to 40 x 40 mm.

Discussion. Boucek noted only four specimens of this species (the holotype and three fragments),
gave a diagnosis (emended here), and provided a clear figure (1957, pi. 23, fig. 4) from whieh it has
been possible to identify the specimens figured here. All the present specimens are similar enough
to be considered one species. Boucek noted that the branching pattern and lack of inflation of the
branches made T. arborescens quite unmistakably distinct. Even allowing for a very slight inflation,
this species is unlike any other.

Some of the specimens figured here bear previous identification labels. GSE 14161 carries a label
by O. M. B. Bulman stating ' Koremagraptus sp B’; his identification is undated (possibly 1957) and
incorrect. Hardie identified RSM GY. 1985.28.4 as Callograptus cf. salteri Hall, while Lamont
identified RSM GY. 1985.28.2 as Calyptograptus digitatus Lapworth, 1881. However, T. {= Calyp-
tograptus) digitatus, as currently recognized, has a much smaller rhabdosome and thinner branches,
despite the resemblance of RSM GY. 1985.28.2 to one of Lapworth’s figures.

EXPLANATION OF PLATE 18

Figs. 1-4. Thallograptus arborescens Boucek, 1957. 1 and 4, Reservoir Formation, Gutterford Burn (see text-
fig. 2); 2 and 3, Unit D of Tipper (1976), Wether Law Linn Formation; all crenulata Zone, upper Llandovery.
1, EDNCM 1984.87.3, branches show traces of thecal structure; fine lateral branches taper to a single theca,
x25. 2, EDNCM 1984.87.3, part of rhabdosome with regular branching, x3. 3, EDNCM 1984.87.4,
attached to a pebble and growing out in two directions, x3. 4, GSE 14161, sparsely branched specimen,
X 3.

Figs. 5 and 6. T. inaequalis Boucek, 1957. Reservoir Formation, Gutterford Burn (see text-fig. 2); crenulata
Zone, upper Llandovery. 5, RSM GY. 1985.28. 10, compressed rhabdosome, x 1-5. 6, RSM GY. 1985.28.9,
distal portion of rhabdosome, x 2.

PLATE 18

BULL, Thaltograptus

130

PALAEONTOLOGY, VOLUME 30

ThaHograptiis inaequalis Boucek, 1957

Plate 18, figs. 5 and 6; text-figs. 1 1 and 12

1957 ThaHograptiis inaequalis Boucek, p. 107, text-fig. 47/?; pi. 21, figs. 4 and 5.

Emended diagnosis. After Boucek 1957. Rhabdosome of medium to large size, composed of separate
flabellately shrub-like stocks attached by individual stems to a central stem. Each stock composed
of unequally ramifying main branches and thinner lateral branches, forming a dense network. Main
branch width 0-7-0-8 mm. Lateral and terminal branches decrease in thickness from 0-5 to 0-2 mm.
Branches rarely irregularly inflated. Nine to ten branches in 10 mm.

Material. RSM GY. 1985.28.5, 1985.28.7, 1985.28.9, and 1985.28.10, and GY. 1897.32 from the Hardie Collec-
tion, all located in ‘Drawer 62’ of the RSM.

Description. The main central stem is sometimes attached to a shell (text-fig. 11b); secondary stems attach
separate stocks to the central stem. Individual stocks number up to six per specimen and are preserved
compressed above one another. Branches taper to a single thecal thickness (text-fig. 1 Ic) and are not percep-
tibly inflated.

Discussion. Boucek (1957) suggested that specimens may differ due to the irregular growth of the
branches. As each stock arises from an individual stem, they must have been vulnerable to displace-
ment on deposition (text-fig. 1 1a), so producing a misleadingly irregular appearance.

Genus coremagraptus Bulman, 1927
Type species. C. onniensis Bulman, 1927. Upper Llandovery, Shropshire, England.

Description. The stipe consists of many generations of tightly packed, elongate thecae twisted into a rope-like
form. Each theca is about 0 04 mm wide and average side branch width (preserved compressed) is 0-3 mm.

TEXT-FIG. 11. ThaHograptiis inaequalis Boucek. RSM GY. 1985.28.7. a, at least six stocks attached to single
stem, which is attached to a shell and associated with some root fibres (boxed area shown in c), x 1-5. b,
attachment area, xl-5. c, thecal structure, outline of thecal thickness, and thecal distribution are clear;

branches taper to single theca, x 1 5.

BULL: LLANDOVERY DENDROID GRAPTOLITES

131

TEXT-FIG. 12. Thallograptus inaequalis Boutek. RSM
GY. 1985.28.5, single isolated stock. Branches appear to
anastomose but were merely compressed and superim-
posed at deposition. Scale bar 10 mm.

Coremagraptus can develop quite large rhabdosomes, at least 130 x70 mm. Anastomosis is common but
irregular, and usually only the lateral branches are involved. Sometimes two stipes are united by unidirectional
lateral branches from one stipe joining another. Numerous species of Coremagraptus have been described
(Bulman 1927; Boucek 1957; Kraft 1982), the distinguishing features being thickness of stipe, degree of
anastomosis, thickness, length, and abundance of lateral branches, and overall size.

Discussion. The nature of the thecae has been discussed at length (e.g. Bulman 1955, p. V27; Boucek
1957, pp. 87, 1 14). Thecae are not normally observed in unsectioned material, but specimens figured
here are preserved as internal moulds, and thecal structure is discernible in the stipe walls. The ends
of the side branches are detectably thinner, consisting of a smaller bundle of thecae than the main
stipe (although considerably more than the three or four thecae mentioned in Bulman’s original
diagnosis). Where a pair of stipes are united by unidirectional lateral branches they act as one
branch, and are quite often found superimposed on another pair.

Coremagraptus kalfiisi Boucek, 1957
Plate 20, fig. 1

1931 Callograptus sp. Boucek, pi. 22, fig. la.

1957 Coremagraptus kalfusi Boucek, p. 124, text-fig. 51 b\ pi. 26, fig. 3.

Diagnosis. See Boucek (1957, p. 124).

Material. One specimen on RSM GY. 1985.28. 1 . Only one other specimen is known, the holotype figured by
Boucek (1957) from Lochkov, Bohemia.

Description. The specimen shows distinctive secondary thickening in the basal regions, up to three times
normal stipe thickness of 0-4-0-6 mm. New stipes are formed by both bifurcation and lateral branches. Most
main branches form dichotomously; most side branches form laterally. Six to seven side branches occur in
10 mm and the angle of bifurcation is 30-40°. Side branches reach a mean length of 1 mm before terminating
and are more common proximally; they often have blunt ends, and it appears that all thecae ceased to grow
when a certain length of branchlet had been attained. Anastomosis is more common distally with fenestellae
of a fairly constant sized, elongated oval (2-7 x 10 mm). Overall rhabdosome size, 60 x 40 mm.

Discussion. This specimen and a specimen of C. plexus are on slab RSM GY. 1985.28.1 which has
been mislabelled " Acanihograptus cf. multispinus (Bassler)’. It is well preserved; such features as

132

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 13. Coremagraptus imperfectus Y,.ra.h.
12093, showing unilateral anastomosis, the superim-
posing of neighbouring stipes, and the presence of
short side branches. Scale bar 10 mm.

blunt branchlet ends are unlikely to be the product of damage to the colony during or after
deposition.

Coremagraptus imperfectus Kraft, 1982
Plate 19; text-fig. 1 3

1982 Coremagraptus imperfectus Kraft, p. 90, pi. 11, fig. 1; pi. 12, fig. 2.

Emended diagnosis. After Kraft (1982, p. 90). Rhabdosome conical, of medium to large size, up to
120 x70 mm. Stipes irregularly curved but continuous and 0-4-0-8 mm thick. Stipes occasionally
bifurcating, lateral branching at 45-50° common, forming both new stipes and thinner branches.
Six to nine stipes in 10 mm. Anastomosis frequent; other branches thin (thecae cease to form),
terminate with rounded ends, and are up to 2 mm long. Fenestellae irregular ovoids pointed at each
end, 1-0- 1-5 mm wide and 2-5-5 0 mm long. Thecae tubular, adnate throughout their length with
mean thickness of 0 08 mm.

Material. RSM GY. 1985.28.8, 1985.28.11 and eight unregistered specimens, all located in ‘Drawer 62’ of the
RSM; GSE 12093 and its counterpart. One other specimen is known, the holotype figured by Kraft (1982)
from the Zelkovice Formation, Hyskov, Czechoslovakia.

Discussion. Kraft’s original figured specimen is incompletely preserved but is unmistakably con-
specific with the material described here. Both have long sections of stipe between each branching
or anastomosing episode, and both show a tendency for the lateral branches of one stipe to arise in
one direction. These may then join (anastomose) with another stipe that is not at that point
producing any branches, so that a ‘step-ladder’ appears, which is often free to move independently
and can be found superimposed over other stipes (text-fig. 13).

C. imperfectus is fairly similar to C. spectabilis Boucek, 1957 (a common species of Coremagraptus
from the Lochkov beds of Lejskov, Czechoslovakia). C. spectabilis has much stouter, thicker stipes,
which do not thin towards the ends of the branches, and C. imperfectus does not show secondary
thickening. Hence C. spectabilis is not an adult form of C. imperfectus.

EXPLANATION OF PLATE 19

Figs. 1-4. Coremagraptus imperfectus Kraft, 1982. Reservoir Formation, Gutterford Burn (see text-fig. 2);
crenulata Zone, upper Llandovery. 1, RSM GY. 1985.28.8, distal region of large, fan-shaped rhabdosome,
X 1. 2, RSM GY. 1985.28. 1 1, sheared, incomplete fragment, x 1. 3, GSE 12093, showing anastomosis,
side branches, and traces of thecal structure, x 10. 4, GSE 12093, showing anastomosis and many terminat-
ing side branches, x2-5.

PLATE 19

BULL, Coremagraptus

134

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 14. Coremagraptus plexus Voctdi. GY .\9%5.2'i.\p,
showing regular anastomosis and many short side branches, as-
sociated with a short section of Monograptus spiralis Geinitz. Scale
bar 10 mm.

Coremagraptus plexus (Pocta, 1894)

Text-fig. 14

1894 Desmograptus plexus Pocta, pp. 188-189, pi. 5, figs. 1 and 2 (non figs. 3 and 4).

1894 Desmograptus textorius Pocta, pi. 4, fig. 9 (non fig. 8, 8fl).

1957 Coremagraptus plexus (Pocta); Boucek, pp. 117-1 19, text-fig. 54a-J; pi. 27, figs. 2 and 3; pi. 28,
fig. 1; pi. 33, fig. 3.

Material. One well-preserved specimen, RSM GY. 1985.28.1, a fragment of a larger form.

Description. This species is distinguished by the presence of many short processes on the stipe walls, or very
many short side branches 0-2-0-4 mm long. These appear to be bundles of thecae turning away from the stipe
together when they have nearly reached total length; this feature is very much more common in C. pseudoplexus
Boucek, 1957, but the present specimen is from a much older horizon.

EXPLANATION OF PLATE 20

Fig. 1. Coremagraptus kalfusi Boucek, 1957. Reservoir Formation, Gutterford Burn (see text-fig. 2); crenulata
Zone, upper Llandovery. RSM GY. 1985.28.1, showing basal secondary thickening, a distal increase in
anastomosis, and branches abruptly truncated; associated with Monograptus spiralis Geinitz, 1 842, x 2.

Figs. 2-4. Palaeodictyota pergracilis Hall and Whitfield, 1872. Location as fig. 1. 2, GSE 14163, showing
shape of fenestellae and distortion of mesh, x 3. 3, GSE 14164, x 3. 4, showing thecal structure and thecae
terminating as pores on the surface of stipes, x 10.

PLATE 20

BULL, Coremagraptus, Palaeodictyota

136

PALAEONTOLOGY, VOLUME 30

The specimen figured (text-fig. 14) is on the same mislabelled slab as a specimen of C. kalfusi (see above). It
is a broken fragment but still quite large (20x50 mm). Most branches form through bifurcation of others.
Anastomosis is quite common, although some branches were superimposed during compression to appear
anastomosed.

Genus palaeodictyota Whitfield, 1902

Type species. P. anastomotica (Ringueberg, 1888) from Ordovician and Silurian of New York, Ontario, and
Kentucky, figured by Ruedemann (1947, p. 269).

Diagnosis. ‘Rhabdosome of anastomosing branches composed of numerous tubular thecae which
are adnate throughout their length and open as pores on the surface of the branches’ (Bulman
1938).

Discussion. Bulman’s diagnosis is short but adequate, and requires only certain points to be emphas-
ized. The presence of tubular thecae, a characteristic of the Acanthograptidae, distinguishes Palaeo-
dictyota from the otherwise morphologically similar Desmograptus. Palaeodictyota represents an
‘end-member’ of the Family Acanthograptidae (see above), being characterized by complete
anastomosis with no branches terminating separately, except at the outer edge of the rhabdosome.
Ruedemann (1947) separated this genus from the Acanthograptidae and referred it to ‘Dendroidea
incertae sedis' .

Occurrence. Most common in the Wenlock to upper Ludlow of Bohemia; recorded from Ordovician to
uppermost Silurian rocks.

Palaeodictyota pergracilis (Hall and Whitfield, 1872)

Plate 20, figs. 2-4; text-figs. 15 and 16

1872 Dictyonema pergracile Hall and Whitfield, p. 181, figured.

See Ruedemann (1947, p. 272) for 1888-1896 synonymy.

1908 Desmograptus pergracilis Ruedemann, p. 2.

1915 Desmograptus pergracilis Bassler, p. 403.

1947 Palaeodictyota pergracilis (Hall and Whitfield); Ruedemann, p. 272, pi. 23, fig. 10; pi. 30, figs.
10 and 1 1.

Emended diagnosis. Rhabdosome of medium to small size, 30 x 40 mm, narrowly conical, extending
from a short stem. Branches thin and continuous, 0-2-0-5 mm. Bifurcation rare, lateral branching
common. Anastomosis complete, branches do not terminate separately. Branches of constant
thickness, forming a flexible meshwork with sixteen to eighteen stipes in 10 mm. Fenestellae rounded
rectangular, T7-4 0 mm long by 0-3- 1-6 mm wide. Thecae tubular, adnate throughout, opening as
pores on stipe surface.

Material. GSE 14163-14165, and 12040; unregistered RSM specimens in ‘Drawer 62’ of the museum.

Description. Stem about 3 mm wide and 4 mm long, with stipes branching out from it (text-fig. 15). Branches
usually remain straight but are pulled around fenestellae as if it were more important for the fenestellae to
maintain a constant shape. The meshwork is also prone to being preserved stretched out at the edges, or
compressed (text-fig. 1 5). Branching and anastomosis follows no regular pattern. Branches maintain a remark-
ably constant thickness. Thecae open as pores on the surface (text-fig. 16) but the openings are not seen in
profile.

Discussion. This species is very similar in general morphology to P. undulatum (Pocta, 1894) (see
Boucek 1957, p. 131), but is distinguished by its much finer branehes (less than 0-5 mm, compared
with over 10 mm). P. raymondi Ruedemann, 1947, has similar meshwork dimensions but a mueh
smaller rhabdosome. P. raymondi also encompasses extreme variation in mesh shape and size, and
shows some secondary thickening, whieh is not seen in P. pergracilis. GSE 14164 and GSE 12040
were identified on specimen labels by O. M. B. Bulman as Koremagraptus sp. A (undated).

BULL: LLANDOVERY DENDROID GRAPTOLITES

137

B

TEXT-FIG. 15. Palaeodictyota pergracilis Hall and Whitfield. A, GSE 14163, showing the mesh stretched at
one edge but continuous elsewhere, b, GSE 14165, specimen greatly compressed on deposition, causing
mesh to close up. c, GSE 12040, showing usual distribution of mesh, extending from stem region. Scale

bar 10 mm.

TEXT-FIG. 16. Thecal structure of Palaeodictyota pergracilis Hall and
Whitfield. GSE 14163, showing thecae terminating as pores on surface

of stipes, X 15.

138

PALAEONTOLOGY, VOLUME 30

Acknowledgements. I thank Dr E. N. K. Clarkson for supervision, guidance, and encouragement throughout
the project, and also for critically reading a draft of the manuscript. Invaluable assistance was gained by
discussion with Dr R. B. Rickards, and I thank Dr P. R. Crowther and Dr I. Strachan for their comments.
Many specimens were kindly loaned by Mr W. Baird of the Royal Museum of Scotland and Mr P. Brand of
the British Geological Survey. Mrs D. Baty helped with photography, Mrs B. Robins typed the manuscript,
and Dr J. B. Robins gave encouragement. Finally, I thank the staff of the Grant Institute of Geology,
Edinburgh, for their help and permission to use the facilities of the Department.

REFERENCES

BARRANDE, J. 1850. GiaptoUtes de Boheme, vi + 74 pp., 4 pis. The author, Prague.

BASSLER, R. s. 1915. Bibliographic index of American Ordovician and Silurian fossils. Bull. US natn. Mus. 92,
1521 pp.

BJERRESKOV, M. 1971. The stratigraphy of the Llandovery Series on Bornholm. Bull. geol. Soc. Denm. 21, 34-
50.

BLUCK, B. J. 1983. Role of the Midland Valley of Scotland in the Caledonian orogeny. Trans. R. Soc. Edinh.:
Earth Sci. 74, 119-136.

1984. Pre-Carboniferous history of the Midland Valley of Scotland. Ibid. 75, 275-295.

bouCek, b. 1931. Deux contributions a la connaissance de la Paleontologie et de la Stratigraphic des Zones a
graptolites du Gothlandien de la Boheme. Vest. st. geol. Ust. csl. Repub. 7, 1-8.

1953. Biostratigraphy, development and correlation of Zelkovice and Motol beds in the Silurian of

Bohemia. Sb. Ustred. Ust. Geol. 20,473-480.

1957. The dendroid graptolites of the Silurian of Bohemia. Rozpr. ustred. Ust. geol. 23, 1-294, pis. 1-39.

BRONN, H. G. 1835. Letliaea geognostica, 1, 544 pp. Stuttgart.

1849. Index palaeontologicus. B, Enumerator palaeontologicus, 1 106 pp. Stuttgart.

BROWER, J. c. 1975. Silurian crinoids from the Pentland Hills, Scotland. Palaeontology, 18, 631-656.

BULMAN, o. M. B. 1927. Koremagraptus, a new dendroid graptolite. Ann. Mag. nat. Hist. 19, 344-347.

1927-1967. British dendroid graptolites. Palaeontogr. Soc. [Monogr.], 1927, Pt. I, pp. 1-28, pis. 1, 2;

1928, Pt. II, pp. i-xxxii + 29-64, pis. 3-6; 1934, Pt. Ill, pp. xxxiii-lx + 65-92, pis. 7-10; 1967, Pt. IV, pp. Ixi-
lxiv + 93-97.

1932. On the graptolites prepared by Holm, 3. The structure of some Monograpti from the Retiolites

Shales of Stygforsen, Dalarne. Ark. Zool. 24A (9), 10-15.

1933. On the graptolites prepared by Holm, VI. Structural characters of some Dictyonema and Desmo-

graptus species from Ordovician and Silurian rocks of Sweden and the East Baltic regions. Ibid. 26A (5),
1-52.

1938. Graptolithina. In schindewolf, o. h. (ed.). Handbuch der Paldozoologie, 2D, 1-92. Borntraeger,

Berlin.

1950. Rejuvenation in a rhabdosome of Dictyonema. Geol. Mag. 87, 351-352.

1955. Graptolithina. In moore, r. c. (ed.). Treatise on Invertebrate Paleontology, Part V, xvii+ 101 pp.

Geological Society of America and University of Kansas Press, New York and Lawrence, Kansas.

1970. Graptolithina. In teichert, c. (ed.). Treatise on Invertebrate Paleontology, Part V (2nd edn.),

xxxii + 163 pp. Geological Society of America and University of Kansas Press, Boulder, Colorado, and
Lawrence, Kansas.

BURGESS, I. c., RICKARDS, R. B. and STRACHAN, I. 1970. The Silurian strata of the Cross Fell area. Bull. geol.
Surv. GtBr. 32, 167-182.

CHAPMAN, A. J. and RICKARDS, R. B. 1982. Pcridcrmal (cortical) ultrastructure in Dictyonema cf. rhinanthiforme
Bulman, and the significance of its bithecae. Palaeont. Z. 56, 217-227.

CLARKSON, E. N. K., ELDREDGE, N. and HENRY, J. L. 1977. Some Phacopina (Trilobita) from the Silurian of
Scotland. Palaeontology, 20, 119-142, pis. 18-20.

and HOWELLS, Y. 1981. Upper Llandovery trilobites from the Pentland Hills. Ibid. 24, 507-536, pis.

77-82.

COCKS, L. R. M., HOLLAND, c. H., RICKARDS, R. B. and STRACHAN, I. 1971. A Correlation of Silurian rocks in the
British Isles. J. geol. Soc. 127, 103-136.

and TOGHiLL, p. 1973. Biostratigraphy of the Silurian rocks of the Girvan District. Ibid. 129, 209-243.

and ZIEGLER, A. M. 1970. Stage names within the Llandovery series. Geol. Mag. 107, 79-87.

DAVIDSON, T. 1868. The Silurian Brachiopoda of the Pentland Hills and of Lesmahagow in Lanarkshire. Trans,
geol. Soc. Glasg., Pal. Ser. 1, 1-24, pis. 1-3.

BULL; LLANDOVERY DENDROID GRAPTOLITES

139

ELLES, G. L. and WOOD, E. M. R. 1901-1918. Monograph of British graptolites. Palaeontogr. Soc. [Monogr.],
clxxi + 539 pp, 52 pis.

FRECH, F. 1897. Lethaea geognostica; Theil 1, Lethaea palaeozoica, 1. Bd., Graptolithiden, pp. 544-684.
Leipzig.

GEINITZ, M. B. 1842. Ueber Graptoliten. Neues Jh. Miner. Geol. Paldont. 1842, 697-701.

IFALL, J. 1851. Description of new, or rare species of fossils from the Palaeozoic series. In foster, j. w. and
WITNEY, I. D. Geology of Lake Superior. Part 2. Washington.

and WHITFIELD, R. p. 1872. Description of new species of fossils from the vicinity of Louisville, Ky. and

the falls of the Ohio. Ann. Rep. NY St. Cab. Nat. Hist. 24, 181-200.

HASWELL, G. C. 1865. On the Silurian formation in the Pentland Hills. Nimmo, Edinburgh, 47 pp., 4 pis.
HENDERSON, J. 1874. On some Silurian fossils found in the Pentland Hills. Trans. Edinb. geol. Soc. 2, 373-375.

1880. On some recently discovered fossiliferous beds in the Silurian rocks of the Pentland Hills. Ibid. 3,

353-356.

and BROWN, d. j. 1869. On the Silurian rocks of the Pentland Hills. Ibid. 1, 266-272.

HOLLAND, c. H., RICKARDS, R. B., SKEVINGTON, D., STRACHAN, I. and TARLO, L. B. H. 1967. Chordata: Hemichor-
data (including Graptolithina), Pogonophora, Urochordata and Cephalochordata. In harland, w. b. (ed.).
The fossil record, pp. 601-61 1. Geological Society, London.

HOWELLS, Y. 1982. Scottish Silurian trilobites. Palaeontogr. Soc. [Monogr.], 76 pp., 15 pis.

KEMP, A. E. s. 1985. The later (Silurian) sedimentary and tectonic development of the Southern Uplands
accretionary terrain. Ph.D. thesis (unpublished). University of Edinburgh.

and WHITE, D. E. 1985. Silurian trench sedimentation in the Southern Uplands, Scotland: implications of

new age data. Geol. Mag. 122, 275-277.

KiER, p. M. 1973. A new Silurian echinoid genus from Scotland. Palaeontology, 16, 651-663, pis. 80-83.

KRAFT, J. 1982. Dendroid graptolites of Llandoverian age from Hyskov near Beroun (Barrandian). Sb. geol.
Ved. Praha, Paleont. 5, 83-95.

1984a. Post Lochkovian dendroid graptolites from the Devonian of Central Bohemia. Cas. Miner. Geol.

29(2), 115-132.

\9%Ab. The occurrence of Dictyonema (Dictyonema) delicatulum Lapworth, (Graptolithina, Dendroidea)

in the Silurian of Central Bohemia. Ibid. 29 (4), 401-408.

LAMONT, A. 1943. Scottish Dragons. Quarry Managers J. 31, 531-535.

1947. Gala-Tarannon beds in the Pentland Hills near Edinburgh. Geol. Mag. 84, 193-208.

1952. Ecology and correlation of the Pentlandian— A new division of the Silurian system in Scotland.

Rep. 18th hit. Geol. Congr. 10, 27-30.

1954. New lamellibranchs from the Gutterford Burn Flagstones (Gala-Tarranon) of the Pentland Hills,

near Edinburgh. Proc. R. Soc. Edinb. 65, 871-884.

1978. Pentlandian miscellany. Scott. J. Sci. 1, 245-302.

LAPWORTH, c. 1874. Notes on the graptolites found by J. Henderson in the Silurian shales of Habbies Howe,
Pentland Hills. Trans. Edinb. geol. Soc. 2, 375-377.

1881. On the Cladophora (Hopk.) or dendroid graptolites collected by Prof. Keeping, etc. Q. Jl geol.

Soc. Lond. 37, 191-197.

LAURIE, M. 1892. Some eurypterid remains from the Upper Silurian rocks of the Pentland Hills. Trans. R. Soc.
Edinb. 37, 151-164.

1898. The eurypterid-bearing rocks of the Pentland Hills— final report, etc. Rep. Br. Ass. Advmt Sci.

557-558.

1899. On a Silurian scorpion and some additional eurypterid remains from the Pentland Hills. Trans. R.

Soc. Edinb. 39, 575-596.

LEGGETT, J. K., MCKERROW, w. s. and EALES, M. H. 1979. Southern Uplands of Scotland: a lower Palaeozoic
accretionary prism. J. geol. Soc. Lond. 136, 755-770.

MYKURA, w. and smith, j. d. d. 1962. Ordovician and Silurian. In Mem. Geol. Surv. 3rd edn., pp. 10-22.
NICHOLSON, H. A. 1872. Monograph of the British Graptolitidae, x+133 pp. Blackwood, Edinburgh and
London.

PEACH, B. N. and HORNE, J. 1899. Silurian rocks of Britain. 1, Scotland. Mem. geol. Surv. U.K. pp. 589-603.
poCta, p. 1894. Bryozoaires, Hydrozoaires et partie des Anthozoaires. In barrande, j. (ed.). Systeme Silurien
du Centre de la Bolmne, 8(1), 230 pp., 21 pis. Prague.

RAMSKOLD, L. 1984. Silurian odontopleurid trilobites from Gotland. Palaeontology, 27, 239-264, pis. 26-31.
RICKARDS, R. B. 1970. The Llandovery (Silurian) graptolites of the Howgill Fells, northern England. Palae-
ontogr. Soc. [Monogr.], 108 pp., 8 pis.

140

PALAEONTOLOGY, VOLUME 30

RICKARDS, R. B. 1973. On some highest Llandovery red beds and graptolite assemblages in Britain and Eire.
Geol. Mag. 110, 70-72.

1976. The sequence of Silurian graptolite zones in the British Isles. Geol. J. 11, 153-188.

HUTT, j. E. and BERRY, w. B. N. 1977. Evolution of the Silurian and Devonian graptoloids. BuU. Br. Mas.

nat. Hist. (Geol.), 28, 1-123.

RiNGUEBERG, E. N. s. 1888. Some new species of fossils from the Niagara Shales of western New York. Proc.
Philad. Acad. Nat. Sci. 131 pp.

ROBERTSON, G. 1985. Palaeoenvironmental interpretation of the Silurian rocks of the Pentland Hills. Ph.D.
thesis (unpublished). University of Edinburgh.

ROLFE, w. D. I. 1960. The Silurian inlier of Carmichael, Lanarkshire. Trans. R. Soc. Edinb. 64, 245-258.
RUEDEMANN, R. 1904. Graptolites of New York, Part 1. Mem. NY St. Mas. 7, 457-803, pis. 1-17.

1908. Graptolites of New York, Part 2. Ibid. 11, 1-583, pis. 1-31.

1925. The Utica and Lorraine Formations of New York. 1. Stratigraphy. Bull. NY St. Mas. 258, 175

pp., 7 pis.

1947. Graptolites of North America. Mem. geol. Soc. Am. 19, 652 pp., 92 pis.

SPENCER, J. w. 1878. Graptolites of the Niagara Formation. Canad. Nat., Ser. 2, 8, 457-463.

SPENCER, w. K. 1914-1940. The Palaeozoic Asterozoa. Palaeontogr. Soc. [Monogr.].

TIPPER, J. c. 1976. Stratigraphy of the North Esk Inlier, Midlothian. Scott. J. Geol. 12, 15-22.

TORNQUIST, s. L. 1881. Om nagra graptolitarter fran Dalarne. Geol. For. Stockh. Fork. 5, 434-445.

WATERSTON, c. D. 1979. Problems of functional morphology and classification of stylonuroid eurypterids
(Chelicerata, Merostomata), with observations on Scottish Silurian Stylonuroidea. Trans. R. Soc. Edinh.
70, 251-322.

WHITFIELD, R. p. 1902. Notice of a new genus of marine algae, fossil in the Niagara Shale. Bull. Am. Mus. nat.
Hist. 16, 399-400, pi. 53.

ELIZABETH E. BULL
Grant Institute of Geology
University of Edinburgh
King’s Buildings

Typescript received 8 November 1985 West Mains Road

Revised typescript received 18 March 1986 Edinburgh EH9 3JW

A NEW REPORT OF A THEROPOD DINOSAUR FROM

SOUTH AFRICA

by N. J. MATEER

Abstract. The first unequivocal theropod dinosaur is reported here from the late Jurassic-early Cretaceous
of South Africa. The specimens are a third pedal ungual (claw) and two teeth. The ungual is from the Sundays
River Formation and the teeth are from the Enon Formation. The age of these formations is discussed.

Dinosaurs from the Jurassic and Cretaceous of South Africa are rare and are known only from
fragmentary specimens which are mostly from the non-marine Kirkwood Formation (Jurassic-
Cretaceous) in the Algoa Basin surrounding Port Elizabeth in the eastern Cape Province. They
comprise principally remains of the sauropod Algoasaurus (Broom 1904) and the stegosaur Paran-
thodon africanus (Atherstone 1857, but see Galton and Coombs 1981); fragmentary remains from
other specimens may or may not pertain to these genera. The iguanodontid Kangnasaurus (Flaugh-
ton 1915) from Namaqualand in the north-west Cape Province is the only other dinosaur known
from outside this southern region of South Africa. Until recently (Rich et al. 1983), no theropod
dinosaurs had been reported from South Africa, but during field-work in 1978, Rich et al. (1983)
found a fragmentary, though promising, terrestrial fauna from the Kirkwood Formation which
contained, almost certainly, three teeth of a theropod, though they remain unidentified as such.
During a recent visit to the South African Museum, I noted a theropod ungual. It was apparently
collected from the Sundays River Beds. Two theropod teeth from the Enon Formation are also
described from the same collection.

STRATIGRAPHY

The Jurassic-Cretaceous sequence of the Cape Province is concentrated along the southern edge of
South Africa in a number of fault-controlled basins stretching throughout the Cape Fold Belt, from
about 70 km east of Cape Town to Port Elizabeth in the eastern Cape Province (text-fig. 1 ). The
Algoa Basin is the largest and best developed of these basins and was initiated in close association
with the splitting of southern Africa and South America, and the Falkland Plateau during the latest
Jurassic. Following the split, marine, paralic and non-marine sedimentation has been continuous
until the present day, although the rate of sedimentation was more rapid during basin subsidence
in the Jurassic (McLachlan and McMillan 1979).

The non-marine sequence has yielded few fossil vertebrates and plants, and has generally not
thought to have been promising palaeontologically. However, the recent report of a diverse terres-
trial fauna (Rich et al. 1983) of freshwater fish, testudines, crocodiles, sphenodontids and ornithisch-
ian and saurischian dinosaurs has changed this. This latter fauna is from the Upper Kirkwood
Formation which is thought to be Tithonian to early Valangian (late Jurassic to early Cretaceous)
in age (McLachlan and McMillan 1976, 1979), but Winter (1979, fig. 2) regarded it as only Tithonian.
The Sundays River Formation is probably early Cretaceous (?Berriasian to Barremian) and is
predominantly paralic, thus it is unlikely to contain the theropod in question. The locality given for
the theropod ungual is vague, simply ‘Kirkwood’; this town lies on the Sundays River Formation,
although the Kirkwood and Enon formations outcrop very close by. According to McLachlan
and McMillan (1976, p. 205), all three formations (together forming the Uitenhage Group) have

[Palaeontology, Vol. 30, Part 1, 1987, pp. 141-145.) © The Palaeontological Association

142

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1. Location of Jurassic-Cretaceous sedimentary basins in the Cape Fold Belt, South Africa.
The Enon and Kirkwood formations are cross-hatched, the Sundays River Formation is stippled.
Modified after McLachlan and McMillan (1976).

yielded reptile fossil fragments (an imperfect lacertilian femur, and a dinosaur centrum and tooth
are reported from the Sundays River Formation). Owing to the complex stratigraphic intertonguing
between the terrestrial and marine units during this regressionary-transgressionary period, the
Jurassic and Cretaceous boundary within these formations is imprecisely known thus leaving the
age of these specimens as either latest Jurassic or earliest Cretaceous (text-fig. 2).

PALAEONTOLOGY

The ungual specimen reported here (SAM K1475 in the South African Museum) is well preserved
except for the proximal articular facet which is missing (text-fig. 3u, b). The ungual is 9-4 cm long
and is gently curved and moderately robust. The large size would indicate that this is a pedal
ungual. The lateral grooves are low proximally, rising upward distally, a pattern which seems to be
characteristic of the third terminal phalange rather than the second or fourth. Comparison with
other theropods (e.g. Ostrom 1969, p. 152) shows this specimen to be too robust, large, and gently
curved to have been a coelurosaur despite the missing proximal end, thus it should be placed within
the carnosaurs. This group has a number of late Jurassic-early Cretaceous genera to which this
specimen may belong, the most likely candidate being within the Allosauridae or Megalosauridae
(grouped in one family, the Megalosauridae, by Romer (1966)), on account of the gentle curvature
of this large ungual. This specimen is identical with a similar ungual in Allosaurus fragilis (see
Madsen 1976, pi. 54, fig. 3,). The other two carnosaur families known from the late Jurassic-early
Cretaceous, Ceratosauridae (Marsh 1884) and Spinosauridae (Stromer 1915), do not have the same

MATEER: THEROPOD DINOSAUR FROM SOUTH AFRICA

143

TEXT-FIG. 2. Stratigraphic relationships of Upper Jurassic and Lower Cretaceous strata of the
Algoa Basin, South Africa. Modified after Winter (1979).

ungual form. Ostrom (1976) noted variability in the ungual curvature of Deiuonychus, the type
specimen being more pronounced than a second specimen he described. Such variability has not
been noted in other carnosaurs, but the diagnostic value of unguals should be treated with utmost
caution.

In addition to this ungual, two theropod teeth (SAM 643, 649) are in the same collection (text-
fig. 3c, d) although there is no record of their stratigraphic provenance; they were collected from
the area around Oudtshoorn (about 350 km west of Port Elizabeth). According to McLachlan
and McMillan (1976, fig. 1), only the Enon Conglomerate is exposed in this region, thus indicating
a late Jurassic age for these teeth, and possibly older than those reported by Rich et al. (1983).
Specimen SAM 643 (text-fig. 3r/) is nearly complete with serrations on the posterior side only and
is 2-8 cm in length (crown to anterior lower point). The crown is unworn and quite sharply recurved.
Specimen SAM 649 (text-fig. 3c) is also serrated on the posterior side and is similar in size (3-2 cm,
although the base is partially missing), though the crown is more worn than SAM 643. Both
specimens have poorly preserved serrations but reveal an estimated eighteen to twenty posterior
serrations per 5 mm. It was not possible to count the anterior serrations. It would appear that these

144

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 3. Specimens of theropod dinosaurs from South Africa, a, ?third ungual of a carnosaur
pes (SAM 1475) from the Sundays River Formation, Kirkwood, South Africa; b, same as a, opposite
side; c, theropod tooth (SAM 649) from Oudtshoorn (?Enon Conglomerate), South Africa; d,
theropod tooth (SAM 643) from Oudtshoorn (?Enon Conglomerate), South Africa. All scale

bars = 1 cm.

teeth are not from Allosaurus since they do not have vertical striations noted in Madsen (1976,
fig. 9). These teeth reported are quite diflferent from those described by Rich et al. (1983, fig. 6)
as possibly being theropod.

DISCUSSION

The carnosaurs had a wide geographic range during the late Jurassic and early Cretaceous with
specimens reported from the Western Interior of North America (e.g. Madsen 1976), Argentina
(Bonaparte 1980), Australia (Molnar et al. 1980), Europe (Taquet and Welles 1977), northern and
eastern Africa (Stromer 1931, respectively Janensch 1925), and China (Dong et al. 1978; Zhen et al.
1985). Valid early Cretaceous carnosaurs are comparatively rare comprising Acrocanthosaurus
(Stovall and Langston 1950) from Oklahoma, 1 Allosaurus from Australia (Molnar et al. 1980,
Car char odontosaurus (Stromer 1931) from North Africa, Erectopus (Huene 1923) from France,
and IMegalosaurus from Wyoming (Ostrom 1970). Despite these several forms that could have
been present in South Africa during the early Cretaceous, Theropoda incertae sedis is the most
accurate assignment that can be given to these specimens. The presence of theropods in southern
Africa is not unexpected in view of the abundant dispersal routes at that time with close continental
connections (Gallon 1977, 1980)

Acknowledgements. I wish to thank Dr Michael Cluver of the South African Museum, Cape Town, for
allowing me to study this, and other specimens. Drs M. Cluver, H. Klinger, R. Molnar, and Professor J.
Ostrom kindly reviewed this paper. Thanks are also due to Barney and Margaret Newman for showing me
the field outcrops of the Sundays River Eormation.

MATEER: THEROPOD DINOSAUR FROM SOUTH AFRICA

145

REFERENCES

ATHERSTONE, w. G. 1857. Geology of Uitenhage. East. Prov. Mon. Mag. 1, 518-532, 579-595.

BONAPARTE, J. F. 1980. Jurassic tetrapods from South America and dispersal routes. In Jacobs, l. l. (ed.).
Aspects of Vertebrate History, 406 pp. Museum of Northern Arizona Press, Flagstaff.

BROOM, R. 1904. On the occurrence of an opisthocoelian dinosaur (Algoasaurus baueri) in the Cretaceous beds
of South Africa. Geol. Mag. 1,445-477.

DONG, z., CHANG, Y., LI, X. and ZHOU, s. 1978. Note on a new carnosaur (Yangchuanosaurus shangyouensis)
from the Jurassic of Yangchuan District, Szechuan Province. Kexue Tongbao, 23, 302-304. [In Chinese.]

GALTON, p. M. 1977. The ornithopod dinosaur Dryosaitrus and a Eaurasia-Gondwanaland connection in the
Upper Jurassic. Nature, 268, 230-232.

1980. Dryosaunts and Camptosaunis, intercontinental genera of Upper Jurassic ornithopod dinosaurs.

Mem. Soc. geol. Fr. ns, 139, 103-108.

and COOMBS, w. p. 1981. Paranthodon africanus (Broom) a stegosaurian dinosaur from the Lower

Cretaceous of South Africa. Geobios, 14, 299-309.

HAUGHTON, s. A. 1915. On some dinosaur remains from Bushmanland. Trans. R. Soc. S. Afr. 5, 259-264.

HUENE, F. VON. 1923. Camivorous Saurischia in Europe since the Triassic. Bull. geol. Soc. Am. 34, 449-457.

JANENSCH, w. 1925. Die Coelurosaurier und Theropoden der Tendaguruschicten Deutsch-Ostafrikas.
Palaeontographica, Supplement, 7, 1-99.

MCLACHLAN, I. R. and MCMILLAN, I. K. 1976. Review and stratigraphic signihcance of southern Cape Mesozoic
palaeontology. Trans, geol. Soc. S. Afr. 79, 197-212.

1979. Microfaunal biostratigraphy, chronostratigraphy and history of Mesozoic and Cenozoic

deposits on the coastal margin of South Africa. Geol. Soc. S. Afr., Spec. Pubis, 6, 161-181.

MADSEN, J. H., JR. 1976. Allosaurus fragilis, a revised osteology. Bull. Utah geol. miner. Surv. 109, 1-163.

MARSH, o. c. 1884. Principal characters of American Jurassic dinosaurs. Part VIII: the order Theropoda. Am.
J. Sci. 35, 329-340.

MOLNAR, R. E., FLANNERY, T. F. and RICH, T. H. v. 1980. An allosaurid theropod dinosaur from the Early
Cretaceous of Victoria, Australia. Alcheringa, 5, 141-146.

OSTROM, J. H. 1969. Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous
of Montana. Bidl. Peabody Mus. nat. Hist. 30, 1-165.

1970. Stratigraphy and paleontology of the Cloverly Formation (Lower Cretaceous) of the Bighorn

Basin area, Wyoming and Montana. Ibid. 35, 1-234.

1976. On a new specimen of the Lower Cretaceous theropod dinosaur Deinonvchus antirrhopus. Breviora,

439, 20 pp.

RICH, T. H. v., MOLNAR, R. E. and RICH, PATRICIA v. 1983. Fossil vertebrates from the Late Jurassic or Early
Cretaceous Kirkwood Formation, Algoa Basin, southern Africa. Trans, geol. Soc. S. Afr. 86, 281-291.

ROMER, A. s. 1966. Vertebrate Paleontology, 468 pp. Univ. of Chicago Press, Chicago.

STOVALL, J. w. and langston, w., jr. 1950. Acrocanthosaurus atokensis, a new genus and species of Lower
Cretaceous Theropoda from Oklahoma. Am. Midi. Nat. 43, 696-728.

STROMER, E. 1915. Wirbeltier Reste der Barharije Stiife. 3. Das original des Theropoden Spinosaurus aegyptiacus
nov. gen., nov. spec. Abh. bayer. Akad. Wiss., Math.-phys. Kl. 28 (3), 1-32.

1931. Wirbeltier Reste der Barharije Strife (untertes Cenoman) 10. Ein skelett-Rest von Carcharodonto-

saurus. Ibid, nf, 9, 1-23.

TAQUET, p. and WELLES, s. p. 1977. Redescription du crane de dinosaure theropode de Dives (Normandie).
Annls Paleont. 63, 191-206.

winter, h. de la r. 1979. Application of basic principles of stratigraphy to the Jurassic-Cretaceous interval
in southern Africa. Geol. Soc. S. Afr., Spec. Pubis, 6, 183 -196.

ZHEN, s., ZHEN, B., MATEER, N. J. and LUCAS, s. G. 1985. The Mesozoic reptiles of China. In lucas, s. g. and
matter, n. j. (eds.). Studies of Chinese Fossil Vertebrates. Bull. geol. Instn Univ. Upsala, ns, 11, 133-150.

NIALL J. MATEER

1467 N. 17th

Laramie, Wyoming 82070
USA

Typescript received 7 January 1986
Revised typescript received 29 March 1986

•w-

-•N

1.

J, '

\ :

t

.'flr

THE CRETACEOUS DIMITOBELID AE
(BELEMNITIDA) OF THE
ANTARCTIC PENINSULA REGION

by PETER DOYLE

Abstract. The Dimitobelidae is a Cretaceous belemnite family that was restricted to within the 30° S.
palaeolatitude for its Aptian-Maastrichtian range. The family is described in detail from the Antarctic
Peninsula region for the first time, and the three component genera Peralobelus Whitehouse, Tetrabelus
Whitehouse, and Dimitobelus Whitehouse are revised. Species of Peratobelus (Aptian) and Dimitobelus
(Aptian/Albian-Maastrichtian) from Antarctica closely resemble those from Australia, and include the new
species D. praelindsayi sp. nov. A new species from Australia, D. dayi sp. nov., is also described. Species of
Tetrabelus (Aptian-Albian) are unknown from Australasia, but the single form T. seclusus (Blanford) occurs
in both the Antarctic Peninsula and southern India. Two new species, T. willeyi sp. nov. and T. whitehousei
sp. nov., apparently endemic to the Antarctic Peninsula, are described.

The Dimitobelidae constitute a belemnite family that was restricted to within the 30° S. Cretaceous
palaeolatitude as part of the marine Austral Realm (Stevens 1965, 1973; Doyle 1985/?). Despite its
biostratigraphical potential in the Southern Hemisphere (e.g. Day 1967, 1969; Taylor et al. 1979) a
revision of the Dimitobelidae has not been attempted since that of Stevens (1965). The family was
first described by Whitehouse (1924) who identified four genera; Dimitobelus, Peratobelus, Tetra-
belus, and Cheirobelus, all characterized by their uniquely paired ventrolateral alveolar grooves.
Later authors have reduced this number (Glaessner 1945, 1957; Stevens 1965; Doyle 1985(?) and
three genera are now recognized as valid: Dimitobelus Whitehouse (= Cheirobelus Whitehouse),
Peratobelus Whitehouse, and Tetrabelus Whitehouse.

The purpose of this paper is to revise the dimitobelid genera further, and describe in detail the
Antarctic species. The dimitobelids of the Antarctic Peninsula region are of great importance, as it
is only here that all three genera occur together. The family was first described from this region in
a preliminary study by Willey (1972). Since this work, further Dimitobelidae from the Antarctic
Peninsula have been recognized by the author in the collections of the British Antarctic Survey.
Elsewhere, Dimitobelidae are well known from Australasia (Whitehouse 1924; Glaessner 1945;
Stevens 1965, 1973) and to a lesser extent southern India (Whitehouse 1924; Doyle 19856). The
Antarctic dimitobelids have great affinities to those from Australia {Peratobelus, Dimitobelus) and
southern India {Tetrabelus). Thus, for greater comparison, a number of Australasian specimens
were examined and, where appropriate, these are listed with the species described below. In addition,
a single new species of Dimitobelus from Australia is described.

GEOLOGICAL SETTING

Dimitobelids occur locally towards the top (c.T2 km) of the approximately 4 km-thick Fossil Bluff
Formation in south-eastern Alexander Island. The formation ranges from the Upper Oxfordian-
Kimmeridgian to the Albian and consists of mainly mudstones, tuffaceous sandstones, and con-
glomerates that were deposited in a fore-arc environment (Taylor et al. 1979; Thomson et al. 1983).
The relatively few belemnites that have been described from the formation have come from the
reasonably well-known Ablation Valley (79° 49' S., 68° 28' W., to the north of the area in text-
fig. 1) and Keystone Cliffs areas (text-fig. 1; Willey 1972, 1973; Taylor et al. 1979). However, many

(Palaeontology, Vol. 30, Part 1, 1987, pp. 147-177, pis. 21-23.|

© The Palaeontological Association

148

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1 . Locality and geological map of the south-eastern coast of Alex-
ander Island (see inset).

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

149

T

1

1 64°S-

(t

j

JAMES ROSS ISLAND J

-1

58°W

1

0 , 25

58“15'

58°00'W

57M5'

• Fossil locality
D 8212

HOLLUSCHICKIE BAY

- 64°00’S

KOTICK POINT
D 8403
D 8420

LOST VALLEY i

D 8423[/ ^ 8422/
GIN COVET
D 8412

TUMBLEDOWN

TEXT-FIG. 2. Locality and geological map of the north-eastern coast of James Ross Island
(see upper inset). Lower inset; locality map of the Antarctic Peninsula region.

of the belemnites described below were obtained from isolated nunataks to the south of Keystone
Cliffs (text-fig. 1), where the only data available for correlation are faunal records of ammonites
and bivalves (Table 1).

Similar belemnites have been found in a series of varied clastic sediments deposited in a back-arc
setting that is exposed on James Ross Island (text-fig. 2; Thomson et al. 1983; Farquharson et al.
1984; Ineson et al. 1986). The James Ross Island Cretaceous succession may well be in excess of
4-5 km thick (Crame 1985; Ineson et al. 1986), all of which, until recently, was thought to be
Campanian in age (Bibby 1966). However, recent field-work by geologists of the British Antarctic
Survey has revealed that the lowermost 2-4 km (Gustav Group) has an approximate range of
Barremian to Santonian (Crame 1981, 1983; Thomson 1982, 1984a, 6; Ineson et al. 1986) while the
topmost part (Marambio Group) is of Campanian-Maastrichtian age (Thomson 19846; Ineson et
al. 1986). The majority of belemnites from James Ross Island described below were collected from
a 3 0 km sequence measured in the Gin Cove and Brandy Bay areas (text-fig. 2), presented as a
composite section in text-fig. 3.

TABLE 1 . Belemnites and associated age-diagnostic fauna from the Cretaceous of Alexander Island. Details of fauna other than belemnites compiled

150

PALAEONTOLOGY, VOLUME 30

ce

"O

JD

<

a.

<

<

a

<

£

<

s ^ ^ s s

^2 -,S S -S a

C-. c^.

< <

< <

X)

<

D-

<

X

d

a

c

3

C

3

c

I ^

ON

r-

ON

2 X

■S ^

O Cl

^ I I I I I ^

3

3

H

r-

ON

G

O

c/i

6

o

X

H

r-

ON

Cs

S) -1
2 =
5 id

'2 I

tti

^ =3
2 §■

SJ ^

i: Co i

^ S3 ^ ^

^

5
1 .2
I ^e: ^

= -2 ^
^

- s: ^ ^ 'c:
*3 -5* kJ ^

- S;^

=0

-«j<‘

§,§

•>* 3
0‘ <3

3

C

0

1 -

i b

^ s

1 .1 I ^

S S ■§

t>:i

3 ^

“I d

^ > >

•!C 00

C C
3 . .

-w ^ -t-J Cl CL

1) ^ >U c/5 CO

•O .3 "O •:- •:-
C > C ^ ^

dJ Cij

• 3 • 33 3s

CL CL

CO Cl ^ ^ ^

ci q Q

u ,

>
O ■
G

> ^ >

CL 'L
co^-3
3

ci Q tS;

CO CO ^ ^ ^ ^

q q q q q q

a

ONOr-irir-
O — — <N »r5
VO X NO NO X

00 m
to NO
NO X

-r^Nor-ON —
00 00 — — <N
NO NO r- r-

1737 D. sp. indet. — — ?Aptian-Albian

1746 D. sp. indet. ITropaeum Birostrinal ci. concentrica Albian

1748 D.diptychus,D.sp.no\.B.D.sp.no\.C — Birostrinal ci. concentrica Albian

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

151

2600-1

I D. sp. nov.aff. diptychus

— 2500

— 1900

UJ LU ^

Q ^ S

5 ^ 1800-

• D. diptychus

Q.

Z)

o

□c

>
<
I —
CP
3
CJ3

O

I-

<

cr

O

u.

<

CO

>

(P

I

1600-

0D. cf. day/ sp. nov

1400-

Birostrina concentrica

1200 -

•

I

I T. whitehousei sp. nov.

I

• Neohibolites

• ParahiboHtes

1000-

CL

o

H

O

g

H

<

CC

o

U-

800-

700-

I T. seclusus

D. stimulus

D. cf. stimulus var. extremis
D. praelindsayi sp.nov.

m

TEXT-FIG. 3. Diagrammatic composite stratigraphic section from the Gin Cove and
Brandy Bay areas of James Ross Island, with the approximate ranges of important
belemnites. The inoceramid bivalve Birostrina concentrica is indicative of the mid-late

Albian (Crame 1985).

SYSTEMATIC DESCRIPTIONS

Many of the specimens described are imperfect and fragmentary. New species have been formally
named only where there is a sufficient number of well-preserved specimens available. For other
material, a system of open nomenclature has been employed. Forms closely resembling pre-existing
species have been designated by the prefix ‘cf.’, while new forms with affinities to pre-existing species
have been designated by the prefix ‘aff.’ Other forms not resembling already named examples are
designated by a letter (e.g. a, b, c).

In the descriptions, approximate size (length) ranges are given by the terms small ( < 40 mm),
medium (40-70 mm), and large ( > 70 mm). Where possible, the following measurements are

152

PALAEONTOLOGY, VOLUME 30

bO

X»

u < e u

S C
o E
c .2

(U X)

B

o

c

o

U

G cQ

C3

§■ b
o sg

Q-§

^ §■
§ §
2-g

C/5

C/3 . <-

O 'O

Qi ^

O

c s;
.

CL ^

c/3 {3

53 s:

C i:
^ 03

i 2

'L) Cy

^ Oc,

— J Cij

° o

^ C/3

G <D
O C

O;

O . ^
cj ^

^'Z
^ e

■t: ?3

E U
o ..
-c

03 ?f
c ^
3 00

tiS

.y c

§ i

bO O
c3 X

^ H

i) S

^ O

^ .tH

T3

C

cd

d>

PQ

G

O

6

a

<

S '3

C) ^

K -2

X

^ ^ I

a

>

o

G I

S2

2 a

a

g’

a

bo

a

a

■S 2

^ oj
0^

. s-^'

^ ^ 3

t s CO

^ C--

0)

^ J-i

O

.Oi, >

">>* c/3

3

: 3

) <3

.

a .

<5 “a

2 I
.1 ?

o c
bo g ^
a 2 Sy

>
o

Q ^

r a

Cu

2 a
^ ?s a

■s'I'l.
^ §,2:

X, ^ s:

^ CO ^

•Si 's ;

a

I -I
s g

S ci e2; ci

5!^

o s

a ^ 2

.§

^ Q ^

Q ci h;

3

>) a

h. Q

a

fN r- (N
r- (N r<3
(N (N (N <N
*cf 00 00 00

o — —

<N rn
Tj- Tt
00 00

DOYLE; ANTARCTIC CRETACEOUS DIM ITOBELIDAE

153

recorded (in mm): L, total preserved length; X, length from apex to D max (position of maximum
inflation); Dv, dorsoventral diameter at the protoconch; Dvmax, maximum dorsoventral diameter
at position of maximum inflation; Dvmin, minimum dorsoventral diameter; Dl, lateral diameter at
the protoconch; Dlmax, maximum lateral diameter at position of maximum inflation; Dlmin,
minimum lateral diameter. In addition to the short glossary of terms given by Doyle (1985a),
reference to the following definitions may prove useful:

Apical canal. Open canal along apical line of rostrum formed either in-life (by resorption?) or by
post-mortal diagenesis. Opens at apex with an apical foramen.

Doppellinien. Paired, parallel, and finely spaced lateral lines.

Nadelspitze. Axial projection in centre of pseudalveolus left by decay of alveolar region.

Pseudalveolus. Secondary conical cavity in anterior of rostrum produced by decay of the alveolar
region.

All the specimens described below are housed in the collections of the British Antarctic Survey,
Cambridge, unless otherwise stated. Specimens prefixed by KG and E come from Alexander Island
and those prefixed by D from James Ross Island. Other repositories are abbreviated as follows:
BM, British Museum (Natural History), London; GSSA, Geological Survey of South Australia,
Eastwood; HM, Hunterian Museum, Glasgow; MUGD, Melbourne University Geology De-
partment, Melbourne; NMV, National Museum of Victoria, Melbourne; SAM, South Australian
Museum, Adelaide; SM, Sedgwick Museum, Cambridge.

Class CEPHALOPODA Cuvier, 1794
Subclass COLEOIDEA Bather, 1888
Order belemnitida Zittel, 1895
Suborder belemnopseina Jeletzky, 1965
Eamily dimitobelidae Whitehouse, 1924

Type genus. Whitehouse, 1924.

Diagnosis. Belemnopseina with rostra bearing pair of ventrolateral alveolar grooves underlain by
splitting surfaces, without apical grooves or unpaired ventral or dorsal alveolar grooves or slits
(Whitehouse 1924, p. 410; Jeletzky 1966, p. 147).

Range. Aptian to Maastrichtian of Australasia, Antarctica, and southern India.

Genus peratobelus Whitehouse, 1924
Type species {by original designation). Belemnites Tenison-Woods, 1884.

Diagnosis. Large, robust, elongate cylindrical, cylindriconical to subhastate Dimitobelidae with
weak to moderate depression. Outline symmetrical, cylindriconical to subhastate, apex acute to
moderately obtuse. Profile asymmetrical to nearly symmetrical, cylindriconical to cylindrical. Venter
generally flattened. Transverse sections subcircular to subquadrate, generally only weakly depressed
or compressed. Two pronounced, straight, undeflected ventrolateral alveolar grooves extend adapi-
cally one half to two thirds the length of the rostrum. Lateral lines are poorly developed. The
phragmocone penetrates one third to one half of the rostrum and is ventrally deflected. Alveolus
normal. Apical line weakly cyrtolineate, apical canal common.

Range. Aptian of Australia and Antarctica.

Discussion. The genus Peratobelus was erected by Whitehouse (1924, p. 410) to encompass ‘cylin-
drical or clavate [belemnites] with ventro-lateral [alveolar] grooves only. These grooves extend for
about half the length of the guard. Alveolus normal.’ Most later authors concur with this diagnosis
although restricting this genus to cylindriconical, non-hastate (non-clavate) species (e.g. Glaessner

154

PALAEONTOLOGY, VOLUME 30

1957; Stevens 1965). However, the type species {P. oxys (Tenison-Woods)) and related forms {P. cf.
oxys, P. australis (Phillips)) all display some hastation, and that restriction is here considered invalid.

Peratobelus is easily distinguished from Dimitobelus by the latter’s depressed transverse section
(text-figs. 5-8) and dorsally deflected ventrolateral grooves, and from Tetrabelus which has a
compressed transverse section (text-fig. 4) and ventrally deflected ventrolateral grooves.

Species included: P. oxys (Tenison-Woods), P. australis (Phillips), P. bauhinianus Skwarko ( =
(l)Dimitobelus youngensis Skwarko). Whitehouse (1924, p. 411) referred to a single large belemnite
from the Upper Aptian of South Australia (BM.C.5309) as ‘an unnamed species of Peratobelus' .
This specimen was later identified with Belenmites selheimi Tension-Woods by Woods (1961, p. 3),
even though this species was originally described from phragmocones only (Tenison-Woods 1883,
p. 250, pi. 7, fig. 1). Later authors have since listed B. selheimi as a species of Peratobelus (e.g. Day,
1967, 1969; Hill et al. 1968) or even Dimitobelus (Ludbrook 1966), even though it is extremely
difficult to assign separated phragmocones to any particular rostrum (cf. Etheridge 1902a, p. 50),
as they are known to protrude some distance from the rostra (see Naef 1922, fig. bid, e), and their
alveolar angles are often undiagnostic (Schwegler 1961). The name B. selheimi should therefore be
restricted to Tenison-Wood’s type, which cannot be definitely assigned to Peratobelus. Whitehouse’s
(1924) unnamed specimen resembles P. oxys, and may be an aberrant form of this species.

Peratobelus cf. oxys (Tenison-Woods, 1884)

Plate 21, figs. 1 and 2

cf. *1884 Belenmites oxys Tenison-Woods, p. 236, pi. xiii, figs. 1-3.

cf 19026 Belenmites oxys Tenison-Woods; Etheridge, p. 48, pi. vi, figs. 4-6; pi. vii, figs. 5-7; pi. viii,
figs. 4-6.

V 1972 Dimitobelus sp. aff. Dimitobelus macgregori (Glaessner); Willey, p. 32, fig. 3a, b non c.
v 1972 Peratobelus oxys (Tennison-Woods); Willey, p. 37, fig. Ad, e.

Material. Two rostra from the Upper Aptian of Alexander Island: KG. 10.61, BM.C.46251 [E. 148.1] (text-
fig. 1).

Dimensions.

KG. 10.61
BM.C.46251

L X Dvmax Dlmax

62-6 36 0 11-4 12-3

93-7 44-7 - 10-3

Description. Large, slender, subhastate to cylindrical Peratobelus. The outline is symmetrical and subhastate
to cylindriconical, and the apex is very acute. The profile is nearly symmetrical and similar to the outline.
Transverse sections weakly depressed (Dlmax ; Dvmax IT) and elliptical to subcircular in the stem region,
becoming subquadrate in the alveolar region.

Two long, straight ventrolateral alveolar grooves extend adapically for approximately one half of the
rostrum (in BM.C.46251). The apparent shortness of the grooves in KG. 10.61 is due to the loss of the alveolar
region of this specimen. The grooves are undeflected and apparently independent of the poorly defined lateral
lines. The phragmocone penetrates one third of the rostrum. The apical line is weakly cyrtolineate, and an
apical canal is present in one of the specimens (BM.C.46251).

Discussion. Willey (1972) recognized the affinity of one of these specimens (BM.C.46251) to P. oxys,
although misidentifying the second specimen (KG. 10.61) as Dimitobelus sp. aff. D. macgregori
(Glaessner). This latter specimen possesses the straight, undeflected alveolar grooves typical of
Peratobelus, quite unlike the dorsally deflected grooves of D. macgregori (Glaessner) (see Stevens
1965, text-fig. 29), and greatly resembles the stem and apex of specimen BM.C.46251. Both resemble
P. oxys, but are much more slender than is typical, and cannot be assigned with any more certainty
to this species.

Occurrence. In Australia, P. oxys is well known from the Upper Aptian of the Rolling Downs Group
(Whitehouse 1924; Day 1969). In Alexander Island (text-fig. 1), P. cf oxyj occurs in unit A2 of Succession
Cliffs (Taylor et al. 1979) with an ammonite fauna of Eotetragonites, Australiceras, and Aconeceras suggestive
of an Upper Aptian or Lower Albian age (Table 1; Willey 1972; Taylor et al. 1979).

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

155

Peratobelus(l) sp.

Plate 21, fig. 7

V 1972 Peratobelus aff. australis (Phillips); Willey, p. 38, fig. 4/.

Material. One natural mould. Upper Aptian, Waitabit Cliff's, Alexander Island: KG. 103. 148 (text-fig. 1).

Discussion. The single specimen KG. 103. 148 was described by Willey (1972) as P. aff. australis
(Phillips). It has a medium-sized relatively robust, hastate rostrum with an almost subcircular
transverse section, and two long, apparently straight ventrolateral alveolar grooves. The apical
region is missing. This specimen is considerably more compressed and somewhat more hastate than
the lectotype of P. australis designated by Day (1967) (Phillips 1870, pi. xvi, figs. 1 and 2) and is
not obviously referable to this species. Some of the specimens described by Skwarko (1966, pi. 15,
figs. 1 and 2) as "Dimitobelus' (= Peratobelus) youngensis sp. nov. resemble KG. 103. 148 in the
form of their rostra and grooves, but the Antarctic specimen cannot be definitely assigned to this
species, which is based on poorly preserved material. P.(?) sp. also approaches D. praelindsayi
sp. nov., but is distinguished by its undeflected ventrolateral grooves and subcircular transverse
alveolar section.

Occurrence. The single specimen of P.(l) sp. was collected from a mudflake conglomerate in unit Tn of
Taylor et al. (1979) at Waitabit Cliffs (text-fig. 1) below an Upper Aptian Eotetragonites ammonite fauna
(Willey 1972; Thomson 1974) and above ammonites indicative of the Lower Aptian (Eulytoceras aff.
polare, Emericiceras(1) sp., Sanmartinoceras patagonicum) (Table 1; Willey 1972; Thomson 1974; Taylor
et al. 1979).

Genus tetrabelus Whitehouse, 1924
Type species (by original designation). Belemnites seclusus Blanford, 1861.

Diagnosis. Small, moderately robust to slender, hastate to cylindriconical, compressed Dimitobel-
idae. Outline symmetrical, subhastate to cylindriconical. Apex acute. Profile asymmetrical to nearly
symmetrical, hastate to cylindriconical. Venter commonly inflated to very inflated. Transverse
sections compressed to very compressed, subquadrate to pyriform. Two pronounced ventrolateral
alveolar grooves with splitting surfaces extend adapically for one third of the rostrum. Ventrolateral
grooves straight in alveolar region, becoming shallow and curving ventrally (a feature reduced in
some species). Dorsolateral depressions present. Lateral lines (Doppellinien) close to and parallel
with the dorsum, extending adapically from the dorsolateral depressions. Fine striae may be present
on venter and dorsum. Phragmocone penetrates one quarter to a third of rostrum, ventrally
deflected. Apical line (?)goniolineate to (?)cyrtolineate.

Range. Aptian to Albian of Antarctica, southern India, and possibly the Malagasy Republic.

Discussion. Tetrabelus was originally erected by Whitehouse (1924, p. 413) to encompass ‘Clavate
[hastate] belemnites provided with dorso-lateral lines, but having, in addition, independent ventro-
lateral grooves’. The validity of this genus has been questioned by Glaessner (1958) and Stevens
(1965), but Doyle (19856) has recently illustrated its independence from Dimitobelus. The dorsola-
teral ‘grooves’ of Whitehouse (1924) are more properly referred to as ‘depressions’, and the diagnos-
tic features of true Tetrabelus are its compressed transverse section (text-fig. 4) and ventrally curving
grooves, while Dimitobelus has a depressed section (text-figs. 5-8) and dorsally curving grooves,
and Peratobelus has straight, undeflected grooves, and a robust transverse section.

Species included: T. seclusus (Blanford), T. willeyi sp. nov., and T. whitehousei sp. nov. Species ex-
cluded: ‘T.’ kleini (Gurich) (= Dimitobelus kleini) and ‘T.’ macgregori Glaessner (= D. macgregori),
see discussion in Doyle 19856.

156

PALAEONTOLOGY, VOLUME 30

Tetrabelus seclusus (Blanford, 1861)

Plate 21, figs. 3-6

* 1861 Belemnites seclusus Blanford, p. 4, pi. 1, figs. 43-51; pi. 11, fig. 8.

1865 Belemnites seclusus Blanford; Stoliczka, p. 202.

1910 Belemnites seclusus Blanford; Spengler, p. 153, pi. xiv, fig. 7.

1920 xec/i/xH.? (Blanford); Bulow-Trummer, p. 166.

1924 Tetrabelus seclusus (Blanford); Whitehouse, p. 413, fig. 4.

V 1985Z) (Blanford); Doyle, p. 29, fig. 5a.

Type specimen. Lectotype (designated Doyle 19856), the original of Blanford (1861, pi. 1, fig. 44), Lower
Uttattur Group, Uttattur, southern India.

Material. Five complete and three fragmentary rostra from the Kotick Point and Whisky Bay formations
(Albian) of James Ross Island: D. 8412. 49, D. 8420. 37, 40a, b, d, g, D. 8422. 107, 1 19 (text-figs. 2 and 3).

Diagnosis. Small, hastate Tetrabelus. Outline symmetrical, subhastate to hastate. Venter inflated.
Transverse sections compressed elliptical to rounded subquadrate, pyriform in alveolar region.
Ventrolateral alveolar grooves display pronounced ventral curvature in stem region. Dorsolateral
depressions broad.

L

X

Dvmax

Dlmax

Dvmin

Dlmin

D.8420.37

44-5

19-9

8-9

7-0

-

6-4

D.8420.40A

37-4

18-5

7-1

5-5

6-8

4-9

D.8420.40B

34-9

10-3

5-9

4-8 -

4-5

3-5

D.8420.40G

25-9

10-4

61

50

—

—

D. 8412.49

33-8

17-6

6-3

4-6

—

—

Description. Small, relatively robust, compressed hastate Tetrabelus. Total length approximately five times
Dv. The outline is symmetrical, subhastate (accentuated by the dorsolateral alveolar ‘pinching’), and the apex
is acute. The profile is asymmetrical and hastate to subhastate depending on the ventral or dorsal inflation of
the apical and stem regions. The stem and apical sections are compressed (Dlmax : Dvmax 0-8; text-fig. 4)
and rounded subquadrate to elliptical, becoming pyriform in the alveolar region (at Dvmin).

Two deep, straight, ventrolateral alveolar grooves with splitting surfaces are developed, reaching adapically
for one third of the rostrum. The grooves shallow adapically, curving sharply on to the venter in the stem
region. Dorsolateral broad depressions occur (at Dvmin) as ‘pinches’. Lateral lines (Doppellinien) extend
adapically from these pinches, remaining close to the dorsum. The phragmocone commonly penetrates one
third to a quarter of the rostrum, and is slightly ventrally displaced. A pseudalveolus may be developed. The
apical line is (?)cyrtolineate.

Discussion. B. seclusus was the first dimitobelid species to be described (Blanford 1861) and was
designated type species of Tetrabelus by Whitehouse (1924). T. seclusus differs from T. willeyi sp.

EXPLANATION OF PLATE 21

Figs. 1 and 2. Peratobelus cf. oxys (Tenison-Woods). Aptian, Alexander Island. 1, right profile and ventral
outline. KG. 10.61, xl. 2, silicone rubber cast, right profile and dorsal outline. BM.C. 46251 x 1.

Figs. 3-6. Tetrabelus seclusus (Blanford). Albian, James Ross Island. 3 and 6, right profiles and ventral
outlines. D. 8420. 37, D. 8420.40b, x 1. 4 and 5, left profiles and ventral outlines. D. 8422. 107, D. 8420. 40a,

X 1 .

Fig. 7. P.(?) sp. Aptian, Alexander Island. Latex cast, right profile and ventral outline. KG. 103. 148, x 1.

Figs. 8-11. T. willeyi sp. nov. Aptian, Alexander Island. 8-10, left profiles and ventral outlines. 8, holotype,
silicone rubber cast. KG. 103.29, x 1. 9, paratype, silicone rubber cast. KG. 1625. 11, x 1. 10, paratype,
latex cast. KG. 1657.1 1, xl. 1 1, long section. KG. 103.24, xl.

Figs. 12-14. T. whitehousei sp. nov. Albian, James Ross Island. 12, alveolar fragment, left profile, and ventral
outline. D. 8412. 85, x 1. 13, right profile and ventral outline. D. 8227.1, xl. 14, holotype, right profile and
ventral outline. D. 8412. 98, x 1.

PLATE 21

DOYLE, Peratobelus, Tetrabelus

158

PALAEONTOLOGY, VOLUME 30

Dl max
mm

7 '

• T. sedusus
A T whitehousei sp nov

TEXT-FIG. 4. Relationship of maximum diameters
in Tetrabelus sedusus (Blanford) and T. whitehousei
sp. nov.

A

-J 1 1 1 1 ■ ■

3456789 Dv max

mm

nov. which has a much more conical form, and from T. whitehousei sp. nov. which is extremely
compressed, with reduced ventral curvature of its ventrolateral grooves.

Bulow-Trummer (1920) considered B. sedusus a species of Parahibolites (see also Stolley 1919),
presumably because of its close association with species of this genus (see Doyle 1985^), and its
degree of compression. However, its paired ventrolateral grooves clearly distinguish it from species
of Parahibolites, which possess only a single ventral groove.

Occurrence. T. sedusus occurs in the Lower Uttattur Group of southern India (Splengler 1910), which ranges
from the Albian to Lower Cenomanian (Bhalla 1983). This species has also been listed from the Malagasy
Republic (Lemoine 1906). In James Ross Island, T. sedusus occurs at the 915-1215 m levels in the composite
section (text-fig. 3) with ammonites indicative of the Albian {Anagaudryceras buddha, Labeceras{l), Ptychoceras,
and Pachydesmoceras) (Thomson 1984f>; Ineson et al. 1986).

Tetrabelus willeyi sp. nov.

Plate 21, figs. 8-11

V 1972 Peratobelus sp. (?)nov. Willey, p. 34, fig. 4a-c.

V \9S5b Tetrabelus sp. nov. A Doyle, p. 29, fig. 5c.

Type specimens. Holotype: KG. 103.29, Lower Aptian, Waitabit Cliffs, Alexander Island. Parafypes:
KG. 1657.41, Apfian, Hyperion Nunataks; KG. 1625. 11, Apfian, Mounf Phoebe, Alexander Island (text-
fig. 1).

Other material. Two fragmentary natural moulds: KG. 1657.25, Lower Aptian, Hyperion Nunataks, Alexander
Island; KG. 1680.27, (?)Albian, Keystone Cliffs, Alexander Island. A single rosfral long secfion: KG. 103.24,
Aptian, Waitabit Cliffs, Alexander Island (text-fig. 1).

Derivation of name. In recognition of Dr L. E. Willey who first identified dimitobelids from Alexander Island.

Diagnosis. Small, cylindriconical Tetrabelus. Outline symmetrical and cylindriconical. Profile sym-
metrical, cylindriconical to subhastate. Transverse sections compressed elliptical to subquadrate.
Ventrolateral alveolar grooves extend for two thirds of rostrum, with gentle ventral curvature.
Dorsolateral depressions reduced.

Description. Small to medium-sized cylindriconical Tetrabelus. Total length approximately seven times Dv.
The outline is symmetrical and cylindriconical, and the profile is similar, being symmetrical and cylindriconical
to subhastate. Transverse sections of the rostrum are compressed elliptical, becoming subquadrate where
dorsum or venter are flattened.

Ventrolateral alveolar grooves are well developed, and curve gently towards the venter in the stem region,
continuing parallel to the venter for two thirds to three quarters of the length of the rostrum. Lateral lines are
poorly developed, dorsally positioned, and independent of the ventrolateral alveolar grooves. The dorsolateral

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELID AE

159

depressions are reduced. The phragmocone (in specimens KG. 103.24, 29) is centrally positioned and penetrates
one third of the rostrum. The alveolus is normal, and the apical line is weakly cyrtolineate.

Discussion. T. willeyi sp. nov. was previously described as Per at obelus sp. (?)nov. by Willey (1972),
but not formally identified. Further material was collected by Dr M. R. A. Thomson in January
1973 and this has confirmed the unique characters of the species. T. willeyi sp. nov. is clearly a
species of Tetrabelus, as it possesses the compressed transverse section and ventrally curving alveolar
grooves characteristic of this genus. The great length of these grooves in T. willeyi sp. nov. probably
misled Willey (1972) who assigned this form to Per at obelus.

T. willeyi sp. nov., the earliest known Tetrabelus, can be distinguished from the later T. seclusus
(Blanford) which is hastate and inflated with sharply curved grooves, and from T. whitehousei sp.
nov. which is characterized by its extreme compression and its weakly deflected grooves.

Occurrence. T. willeyi sp. nov. occurs exclusively in Alexander Island. Two specimens (including the holotype)
from Waitabit Cliffs (text-fig. 1) occur in unit Te of Taylor et al. (1979) with a distinct Lower Aptian fauna
(Eulytoceras, Emericiceras, and Sanmartinoceras) below the conglomerate containing Peratobelus{l) sp. (Willey
1972; Thomson 1974; Taylor et al. 1979). At Mount Phoebe and Hyperion Nunataks (text-fig. 1) T. willeyi sp.
nov. occurs with ammonites of a general Aptian (or older) age (Table 1), while at Keystone Cliffs a fragment
of this species was found with the Albian ammonites Phyllopachyceras aureliae, Antarcticoceras, and Silesites
(Table 1; Thomson 1974).

Tetrabelus whitehousei sp. nov.

Plate 21, figs. 12-14

v 1985fi sp. nov. B Doyle, p. 29, fig. 56.

Type specimen. Holotype: D. 8412. 98, Albian (Whisky Bay Formation), Lost Valley, James Ross Island (text-
figs. 2 and 3).

Other material. Five fragmentary rostra: D. 8412. 70, 78, 84, 85, Albian (Whisky Bay Formation), Lost Valley,
James Ross Island; D.8227.1, Albian (Whisky Bay Formation), Whisky Bay, James Ross Island (text-figs. 2
and 3).

Derivation of name. In recognition of the work of Dr F. W. Whitehouse, original author of the Dimitobelidae.

Diagnosis. Medium sized, extremely slender Tetrabelus. Outline symmetrical, cylindrical to slightly
subhastate. Profile asymmetrical and cylindrical. Transverse sections very compressed. Ventrolateral
alveolar grooves with slight dorsal deflection, becoming ventrally curved in stem region. Dorso-
lateral alveolar depressions developed as ‘pinches’.

Dimensions.

L Dv D1
D.8412.98 55-3 4-8 2-9

Description. Medium sized, cylindrical Tetrabelus. Total length approximately 11 -5 times Dv. The outline is
symmetrical and cylindrical, becoming slightly subhastate in the alveolar region. For the most part the laterals
remain parallel until very near the apex. The profile is asymmetrical and cylindrical, quite often with a
somewhat sinuous dorsum and venter. Transverse sections are extremely compressed (Dlmax : Dvmax 0-6;
text-fig. 4) and elliptical, becoming slightly pyriform in the alveolar region.

Two well-developed, incised ventrolateral alveolar grooves are slightly deflected dorsally, but a distinct
ventral curvature is detectable at their apical-most point. Lateral lines are poorly developed, but Doppellinien
can be observed confined to a dorsal position on the laterals. Two short, dorsolateral ‘pinches’ are present in the
alveolar region. The phragmocone is relatively central, with a ventrally deflected protoconch, and penetrates up
to one quarter of the rostrum. The alveolus is normal. The form of the apical line was not observed.

Discussion. T. whitehousei sp. nov. is a distinctive species, characterized by its very compressed
transverse section. Although the ventrolateral alveolar grooves slope gently towards the dorsum
before curving ventrally, they are unlike the sharply dorsally deflected grooves characteristic of

160

PALAEONTOLOGY, VOLUME 30

Dimitobelus. Thus although Dimitobelus sp. nov. C (see below) resembles T. whitehousei sp. nov. in
its alveolar region, it differs because of its depressed apical region and its dorsally deflected grooves.
T. whitehousei sp. nov. may be distinguished from all other species of Tetrabelus by its extreme
compression (text-fig. 4) and distinct groove form.

Occurrence. T. whitehousei sp. nov. is known only from James Ross Island, where it occurs approximately
25 m below the Birostrina concentrica bio-horizon indicative of the mid-late Albian, at the 1225-1350 m level
in the composite section (text-fig. 3). It occurs with the Albian ammonites Anagaudryceras buddha and
Labecerasil) (Table 2; Ineson et al. 1986).

Genus dimitobelus Whitehouse, 1924
= C/;e;>ofic/n.s Whitehouse, 1924, p. 414

Type species (by original designation). Belemnites canhami Tate, 1880 ( = Belemnitella diptycha M'Coy, 1867).

Diagnosis. Small to large, robust to slender, subhastate to hastate Dimitobelidae that are often
markedly depressed. Outline symmetrical, subhastate to hastate, apex obtuse to mucronate. Profile
asymmetrical to nearly symmetrical, subhastate to cylindriconical. Venter often flattened. Trans-
verse sections depressed elliptical to subcircular in stem and apical regions, subquadrate to pyriform
in alveolar region. Two short, pronounced ventrolateral alveolar grooves with splitting surfaces
extend adapically for one quarter to one third of the rostrum, where they are dorsally deflected,
joining lateral lines. Lateral lines (Doppellinien) well developed, ventral lines joining alveolar
grooves, dorsal lines extended as dorsolateral alveolar depressions. The phragmocone is slightly
dorsally eccentric but with a ventrally incurved protoconch, and penetrates one quarter to a third
of the rostrum. Pseudalveolus and Nadelspitze common. Apical line weakly cyrtolineate, apical
canal commonly developed with apical foramen.

Range. Albian to Maastrichtian of Australia, New Zealand, Antarctica, and New Guinea.

Discussion. Dimitobelus was originally described by Whitehouse (1924, p. 412) to encompass ‘clavate
[hastate] belemnites provided with dorso-lateral [ventro-lateral] grooves and lateral lines, both of
which may be straight or somewhat curved. The alveolus is normal, but generally a pseudalveolus
with axial projection is developed. A ventro-lateral [dorso-lateral] groove may be formed by the
furcation of the lateral lines, but it becomes isolated.’ The true position of the incised grooves has
been the subject of some debate (Glaessner 1957; Stevens 1965). Whitehouse (1924), in common
with earlier authors (e.g. Tate 1880; Etheridge 1902a) favoured a dorsolateral position, whilst
Woods (1917), Glaessner (1957, 1958), and Stevens (1965) have proved the ventrolateral position
of these grooves by their relationship to the siphuncle. The present study confirms the views of
these later authors.

Cheirobelus Whitehouse was considered by Glaessner (1957) and Stevens (1965) to be a synonym
of Dimitobelus. Whitehouse (1924) erected this genus solely on the ‘isolation’ of its ventrolateral
(dorsolateral of Whitehouse) grooves, but Stevens (1965) subsequently determined that these
grooves were connected to the lateral lines in its type species B. lindsayi Hector. Examination of
specimens of this species (SM.B. 2916-2918) confirms this view, and the genus Cheirobelus should
be considered a synonym of Dimitobelus.

Both Glaessner (1958) and Stevens (1965) have considered Tetrabelus Whitehouse to be only a
subgenus or even a synonym of the genus Dimitobelus. However, recent work by Doyle (1985Z?) has
confirmed the generic validity of Tetrabelus, which is a markedly compressed genus (text-fig. 4, cf.
text-figs. 5-8) with ventrally curving ventrolateral grooves. Peratobelus is also distinct from Dimito-
belus in possessing long, straight, undeflected ventrolateral grooves.

Species included: D. diptychus (M‘Coy) ( = D. ca«/iam/ Tate), D. stimulus Whitehouse, D. stimulus
var. extremis Whitehouse, D. lindsayi (Hector), D. kleini (Giirich), D. macgregori (Glaessner), D.
hectori Stevens, D.(l) ongleyi Stevens, D. praelindsayi sp. nov., D. dayi sp. nov. The species B.
liversidgei Etheridge has also been assigned to Dimitobelus (Hill et al. 1968, pi. KII). However, this

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

161

species was originally described as possessing a ‘ventral surface bearing a groove extending for
about half the guard length. . . (Etheridge, in Jack and Etheridge 1892, p. 491), suggesting
belemnopseid rather than dimitobelid affinities. This form may possibly be a species of Neohibolites
(Doyle 1985^).

Dimitobelus diptychus (M‘Coy, 1867)

Plate 22, figs. 1,6-10

V* 1867 Belemnitella diplycha M‘Coy, p. 356.
p 1870 Belemnites australis Phillips, p. 258, pi. xvi, figs. 3 and 4 non 1 and 2.

V 1880 Belemnites canhami Tate, p. 104, pi. iv, fig. 2.

1892 Belemnites canhami Tate; Etheridge (in Jack and Etheridge), p. 490, pi. 35, figs. 3-5, 7-9, 12-14.
1902a Belemnites canhami Taif, Etheridge, p. 49.

V 1902a Belemnites eremos Tate; Etheridge, p. 51, pi. vii, figs. 18-21.

19026 Belemnites canhami Tate\ Etheridge, p. 45, pi. viii, figs. 8 and 9; pi. ix, fig. 2.

V 1924 Dimitobelus canhami (Tate); Whitehouse, p. 412, figs. 2 and 3.

v 1925 Dimitobelus canhami (Tate); Whitehouse, p. 35, pi. 1 1, figs. 1-7, 9-1 1.

V 1966 Dimitobelus diptychus (M‘Coy); Eudbrook, p. 191 , pi. 27, figs. 1-11.

71966 Dimitobelus canhami (Tate); Skwarko, p. 124, pi. 15, figs. 13 and 14.

Type specimen. Holotype: NMV. P2177, Albian, Walker’s Table Mountain, Queensland, Australia.

Material. Eleven fragmentary rostra from the (?)Upper Albian, Alexander Island: KG. 1716.4, 1717.3, 5-7,
8a-d, 1719.24a, b, 1748.15, 16c. Two fragmentary rostra from the Albian-Cenomanian (Whisky Bay Forma-
tion), James Ross Island D. 8403. 35a, D. 8410. 64 (see text-figs. 1 and 2 for localities). Also the following from
Australia: HM.S. 8430-8477, S. 5580-5588, Upper Albian, Woodduck Creek, S. Australia; BM.C. 59263-59264,
Albian, S. Australia; GSSA.M. 2473-2476, Albian, Reedy Springs, S. Australia.

Diagnosis. Large, hastate Dimitobelus. Outline symmetrical, hastate, apex mucronate. Profile asym-
metrical, cylindriconical, venter commonly arched. Transverse sections depressed elliptical in stem
region, subquadrate in alveolar region. Ventrolateral alveolar grooves extend for one third of the
rostrum, lateral lines (Doppellinien) well developed, dorsally placed.

Dimensions.

L

X

Dvmax

Dlmax

KG. 1717.7

85-3

47-5

140

18-3

D. 8410.64

—

—

141

18-3

NMV.P.2177

92-7

47-5

13-5

18-5

GSSA.M.2473

76-9

36-9

10-2

14-3

M.2476

79-3

33-9

11-3

16-2

HM.S. 5581

54-5

260

8-4

11-8

S.5582

53-2

18-6

7-3

10-6

S.5583

47-7

23-3

6-9

9-8

S.5584

48-1

18-5

6-4

8-8

Description. Large, hastate Dimitobelus. Total length of the rostrum is approximately seven times Dv. The
outline is symmetrical and hastate with Dlmax at approximately the mid-point of the rostrum, and the apex is
mucronate. The profile is asymmetrical and cylindriconical, the venter commonly being flattened and arched,
with a corresponding curvature of the dorsum. Where not arched the profile appears symmetrical. Transverse
sections are depressed (Dlmax : Dvmax 1-4; text-fig. 5) and elliptical in the stem and apical regions, the venter
and dorsum being flattened. The alveolar section is more compressed, approaching subquadrate.

Two deep, ventrolateral alveolar grooves extend adapically for one quarter to one third of the rostrum,
curving dorsally to join the ventral-most lines of the well-defined Doppellinien. The dorsal-most lines commonly
expand into dorsolateral alveolar depressions. The phragmocone is central to slightly dorsal in position, with
a ventrally incurved protoconch, and penetrates one quarter of the rostrum. A pseudalveolus is common. The
apical line is weakly cyrtolineate, and an apical canal is commonly developed.

Discussion. D. diptychus is described here for the first time outside Australia. It most closely
resembles D. stimulus Whitehouse (from Australia and Antarctica) and D. superstes (Hector) (from

162

PALAEONTOLOGY, VOLUME 30

New Zealand). However, D. stimulus is more slender, less hastate, and weakly depressed (text-fig.
6) and D. superstes is more inflated with a pyriform alveolar section, although of a similar size.

D. diptychus was originally described as a species of Belemnitella by M‘Coy (1867) who mistook
its ventrolateral grooves for the dorsolateral depressions of Belemnitella. Belemnites canhami, a
species described by Tate (1880), types: SAM.T.1324a-c, T.1326), and the type species of Dimito-
helus, has subsequently been identified as a junior subjective synonym of D. diptychus (type specimen
figured by Dorman and Gill 1959, pi. 8, figs. 1 and 2). Other specimens referable to D. diptychus
include one example figured as B. australis (a species of Perat obelus) by Phillips (1870, pi. xvi, figs.
3 and 4 non 1 and 2) and those referred to B. eremos Tate (a species of uncertain affinities) by
Etheridge 1902«, pi. vii, figs. 18-21: SAM.T.1311, 1312). The forms described by Skwarko (1966)
from the Albian of the Australian Northern Territories are too poorly preserved to be assigned
with certainty to this species.

Occurrence. D. diptychus is common in the late Albian of the Australian Rolling Downs Group (Whitehouse
1925; Ludbrook 1966), and has also been reported from the Cenomanian of the West Australian Carnarvon
Basin (Stevens 1965). D. diptychus also occurs in both Alexander and James Ross islands. In Alexander Island
it is known from a number of stations grouped around Pagoda Ridge (KG. 1717-1719; text-fig. 1). Few other
fossils have been found in association, but nearby at station KG. 17 18 ammonite fragments of the genus
Lechites occur, indicative of the Upper Albian. D. diptychus is also known from north Succession Cliff's
(KG. 1748; text-fig. 1) where it occurs with the Albian bivalve Birostrma(l) cf. concentrica (Crame 1985). In
James Ross Island, D. diptychus occurs as a derived fossil at the 1705 m level (text-fig. 3) with ammonites
indicative of the Cenomanian (Eucalycoceras of Thomson 1982, 19846), occurring some 180 m above Lechites
similar to the Alexander Island examples.

Dimitobelus sp. nov. aff. diptychus (M‘Coy, 1867)

Plate 22, figs. 2-5

aff. V* 1867 Belemnitella diptycha M‘Coy, p. 356.

aff. V 1925 Dimitobelus canhami (Tate); Whitehouse, p. 35, pi. II, figs. 1-7, 9-1 1.
aff. V 1966 Dimitobelus diptychus (M‘Coy); Ludbrook, p. 191, pi. 27, figs. 1-11.

Material. Twenty-two rostra and rostral fragments; D.8232.96a-f, 1 14a-f, 125a, b, 136a-g, and (?)D.427.1a-
c, Campanian (Marambio Group), James Ross Island (text-fig. 2).

L

X

Dvmax

Dlmax

D.8232.136A

47-0

21-7

8-9

11-9

D.8232. 125b

44-7

17-4

100

12-5

D.8232. 136b

41-7

180

7-8

91

D.8232.136C

431

231

81

100

D.8232. 136d

36-2

210

7-3

9-3

D.8232.125A

46-5

14-5

9-6

—

D.8232.96A

44-7

18-2

11-9

—

Description. Medium sized, hastate Dimitobelus. The outline is symmetrical and hastate, with a mucronate or
obtuse and rounded (e.g. D. 8232. 125b, 1 14a) apex. The profile is almost symmetrical and hastate to subhastate,
with some bulbous individuals displaying an inflated venter and dorsum. Transverse sections are depressed
(Dlmax ; Dvmax 1-3; text-fig. 5) and elliptical, becoming very depressed close to the alveolar region, where
the venter is flattened. Ventrolateral alveolar grooves are not preserved in these specimens, although the
Doppellinien are well developed and dorsally placed. The apical line is weakly cyrtolineate, and an apical canal
is commonly developed.

Discussion. The description is based on twenty-two rostra, none of which have their alveolar regions
preserved. A single alveolar fragment (D.427.1a; PI. 22, fig. 5) was found in association with stem
and apical fragments of juveniles similar to those of D. sp. nov. aff. diptychus from locality D.8232,
west of Lachman Crags (text-fig. 2). This fragment has a similar transverse section to the juveniles
in its stem region, becoming attenuated adorally. It possesses two ventrolateral alveolar grooves

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

163

TEXT-FIG. 5. Relationship of maxi-
mum diameters in Dimitobelus dip-
tychus (M'Coy) and D. sp. nov. aff.
diptychus (M‘Coy).

D| max X HOLOTYPE

• GREGORY COLLECTION. HM
19 • ■ LUDBROOK COLLECTION. GSSA
A ANTARCTIC SPECIMENS

18 -

A D sp nov aff. diptychus
17 •

D diptychus

16 •

15 -

13 -
12 -

•• ^

&

A

11 -
10 -
9 -

8 •

•• •

...■a .•

^ A

• A

• A

5 6 7

8 9 10 11 12 13 14 15 Dv max

mm

that parallel the venter before curving dorsally at the stem region. The rostra described above greatly
resemble the Albian-Cenomanian D. diptychus (M‘Coy) in shape and form, although differing by
possession of a more depressed anterior. If the alveolar fragment from Lachman Crags proves
representative of the alveolar region of D. sp. nov. aff. diptychus, it will be necessary to separate
this form from the true D. diptychus. However, until further collecting can be completed, this cannot
be done with certainty.

The alveolar fragment found at Lachman Crags closely resembles that of D. hectori Stevens from
the Maastrichtian (Haumurian) of New Zealand (Stevens 1965, pi. 25, fig. 21). However, the rostrum
of this species differs from D. sp. nov. aff. diptychus in possessing an almost circular, undepressed
transverse section.

Occurrence. D. sp. nov. aff. diptychus is unique to James Ross Island and occurs at the 2560-2600 m level in
the composite section (text-figs. 2 and 3) in associaton with the ammonites Anapachy discus, Gaudryceras
varagurense var. patagonicum, Baculites, and Bostrychoceras. A general age of Campanian has been suggested
for this fauna (Olivero 1984; Thomson 19846). The alveolar fragment from Lachman Crags (D.427.1a; text-
fig. 2) was found in association with the Campanian ammonite Maorites seymourianus (Spath 1953, p. 24).

Dimitobelus stimulus Whitehouse, 1925
Plate 22, figs. 11-15

1924 Dimitobelus simulus Whitehouse, p. 412.
v* 1925 Dimitobelus stimulus Whitehouse; Whitehouse, p. 35, pi. II, figs. 8, 12-17.
v 1966 xt/mw/wx Whitehouse; Ludbrook, p. 192, pi. 27, figs. 12-21.

164

PALAEONTOLOGY, VOLUME 30

Type specimens. Holotype: HM.S5592. Paratypes: HM.S. 5589-5591, 5593, 5594. Upper Albian, Woodduck
Creek, Lake Eyre, S. Australia.

Material. One natural mould; KG. 1680.55, Albian, Alexander Island. Three rostra: D. 8212. 253, 8403.46,
8422.94, Albian (Kotick Point Formation), James Ross Island (text-figs. 1 and 2). Also the following from
Australia: HM.S. 8478-8532, Upper Albian, Woodduck Creek, Lake Eyre, S. Australia; GSSA.M.2484, 2485,
2488, Albian, Reedy Springs, S. Australia.

Diagnosis. Medium sized, slender, subhastate Dimitobelus. Outline symmetrical, subhastate. Profile
symmetrical, subhastate. Transverse sections elliptical. Lateral lines {Doppellinien) well developed,
centrally placed.

KG. 1680.55

L

53-4

X

25-6

Dvmax

Dlmax

7-5

D. 8403-46

39-8

22-2

7-0

8-4

D.8212.253

64-7

32-2

6-3

7-4

GSSA.M.2485

59-0

30-6

9-1

10-2

M.2484

60-5

27-2

7-6

9-1

M.2488

55-3

26-3

6-9

8-2

HM.S. 5592

49-7

19-5

6-2

7-2

S.5591

50-8

17-8

6-9

8-2

S.5590

67-1

25-4

8-4

10-0

S.5589

72-7

25-7

9-5

—

S.5593

48-7

21-2

5-9

7-2

S.5594

47-1

18-6

5-0

6-0

Description. Medium sized, subhastate Dimitobelus. The total length of the rostrum is approximately nine
times Dv. The outline is symmetrical and subhastate, with Dlmax approximately at the mid-point, or nearer
the apex of the rostrum. The apex is non-mucronate and acute. The profile is symmetrical and subhastate,
with dorsum and venter slightly inflated. Transverse sections weakly depressed and almost perfectly elliptical,
with no perceptible flattening (Dlmax : Dvmax 1-2; text-fig. 6). However, the alveolar region may be more
depressed.

Where present, two ventrolateral grooves extend for one quarter of the rostrum, but quite often the grooves
are lost through formation of the pseudalveolus. The grooves curve dorsally to join the ventral-most line of
the centrally positioned Doppellinien. The dorsal-most line may expand into a shallow, dorsolateral depression,
but this is also lost in specimens possessing a pseudalveolus. The form of the phragmocone is difficult to assess
due to the frequent development of a pseudalveolus and Nadelspitze in this species, but it apparently penetrates
one quarter of the rostrum. The apical line is weakly cyrtolineate, and an apical canal is common.

Discussion. Dimitobelus stimulus commonly occurs in some numbers with D. diptychus (M‘Coy)
in the Australian Great Artesian Basin, and is here described for the first time outside Australia.

EXPLANATION OF PLATE 22

Figs. 1, 6-10. Dimitobelus diptychus (M‘Coy). 1, 6-8, right profiles and ventral outlines. 1, Albian-?Cenoman-
ian, James Ross Island. D. 8410. 64, x 1. 6, plaster cast of holotype, Albian, Queensland, Australia.
NMV. P.2177, X 1. 7, Albian, Alexander Island. KG. 1717.7, x 1. 8, Albian, South Australia.

GSSA.M.2476, x 1. 9, silicone rubber cast, ventral outline, Albian, Alexander Island. KG. 1748. 16c, xl.
10, juvenile, silicone rubber cast, (?)ventral outline, Albian, Alexander Island. KG. 1 748. 15, x 1 .

Figs. 2-5. D. sp. nov. aff. diptychus (M‘Coy). Campanian, James Ross Island, left profiles and ventral outlines.

2, D. 8232. 136a, xl. 3, 0.8232. 125b, x 1. 4, D.8232. 1 36c. x 1. 5, D.427.U, x 1.

Figs. 11-15. D. stimulus Whitehouse. 11, specimen with pseudalveolus, left profile and ventral outline, Albian,
James Ross Island. D. 8403.46, x 1 . 12, holotype, right profile and ventral outline, Albian, South Australia.
HM.S. 5592, X 1. 13, Silicone rubber cast, (?)ventral outline, Albian, Alexander Island. KG. 1680.55, x 1.
14, left profile and ventral outline, Albian, James Ross Island. D. 8212.253, x 1. 15, right profile and ventral
outline, Albian, South Australia. HM.S. 5590, x 1.

Fig. 16. D. stimulus var. extremis Whitehouse. Albian, South Australia. Lectotype, right profile and ventral
outline. HM.S.5595, x 1.

PLATE 22

DOYLE, Dimitobelus

166

PALAEONTOLOGY, VOLUME 30

Dl max
mm

X HOLOTYPE

• GEGORY COLLECTION, HM
■ LUDBROOK COLLECTION GSSA
▲ ANTARCTIC SPECIMENS
O GREGORY COLLECTION, HM

A antarctic specimens

\

' D stimulus

J

D stimulus var extremis
D cf stimulus var extremis

11 -
10 -

• • *

9 -

8 -
7 ■

6 -

OP

O

O

TEXT-FIG. 6. Relationship of maxi-
mum diameters in Dimitobelus
stimulus Whitehouse and D. stimulus
var. extremis Whitehouse.

1 J- i I L.

3 4 5 6 7

8

9

10

Dv max mm

D. stimulus resembles juveniles of D. diptychus, but can easily be distinguished by its elliptical
transverse section (text-figs. 5 and 6) and subhastate profile. Most other Dimitobelus species are
more robust and hastate than D. stimulus.

Occurrence. D. stimulus is known from the Upper Albian of the Australian Rolling Downs Group (Whitehouse
1925; Ludbrook 1966). In Antarctica it is known from both Alexander and James Ross islands. In Alexander
Island, D. stimulus has been found at Keystone Cliffs with Albian ammonites {Phyllopacliyceras, lAntarctic-
oceras, and Silesites) and bivalves (Birostrinal cf. concentrica, Inoceramus sp. aff. comancheanus, I. cf. anglicus
elongatus) (Thomson 1974; Crame 1985). In James Ross Island it occurs at the 920-950 m level in the composite
section with similar Albian ammonites and bivalves (text-fig. 3; Table 2).

Dimitobelus cf. stimulus var. extremis Whitehouse, 1925

Plate 22, fig. 16; Plate 23, figs. 1 and 2

cf. 1924
cf V* 1925
cf 1966
V 1972

Dimitobelus extremus Whitehouse, p. 412.

Dimitobelus stimulus var. extremis Whitehouse, p. 35, pi. II, figs. 18-20.
Dimitobelus stimulus var. extremis Whitehouse; Ludbrook, p. 192.
Dimitobelus macgregori (Glaessner); Willey, p. 33, fig. hd.

Material. Two rostra from the late Aptian-Albian, Alexander Island: KG. 10.41, 1721.24. Two rostra from
the Albian (Kotick Point Formation), James Ross Island: D. 8403. 35b, 8212.132 (see text-figs. 1 and 2 for
localities).

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

167

Dimensions.

KG. 10.41

L

61-5

X

22- 1

Dvmax

Dlmax

5-4

D.8403.35B

68- 1

20-7

4.9

5-7

D.8212.132

—

—

5-9

6-5

HM.S. 5595

54-6

24-0

51

5-8

HM.S.5596

—

—

4-5

4-8

HM.S.5597

—

—

40

4.4

Description. Medium sized, very slender, subhastate Dimitobeliis. The total length of the rostrum is approxim-
ately ten times Dv. The outline is symmetrical and subhastate with Dlmax in the apical third of the rostrum,
and the apex is acute. The profile is symmetrical (in stem and apical regions) and subhastate to cylindrical,
but in specimens KG. 10.41 and D. 8403. 35b there is a marked dorsal curvature of the alveolar region, giving
the rostrum the general appearance of an antique firearm. Transverse sections are elliptical to subcircular
(Dlmax : Dvmax M; text-fig. 6) in stem and apical regions, approaching subquadrate in the alveolar region.

Two deep ventrolateral alveolar grooves join the ventral-most lateral lines of the Doppellinien. Some
dorsolateral alveolar flattening may be present, extending from the dorsal-most lateral line. The Doppellinien
are well defined and centrally positioned. In the two best preserved specimens, KG. 10.41 and D. 8403. 35b,
information is not available concerning the position of the alveolus. However, D. 8212. 132 displays a ventrally
deflected apical line.

Discussion. Few specimens of Dimitobelus stimulus var. extremis have been described. Whitehouse
(1925) based his description on three syntypes (HM.S. 5595-5597) and four other fragmentary
rostra (HM.S. 8433-8436), and as such was unwilling to separate specifically these slender, almost
cylindrical forms from D. stimulus s.s. The Antarctic specimens are very similar to the specimen
designated as lectotype by Ludbrook (1966, p. 192: HM.S. 5595), but they differ in the possession
of a complete alveolar region, as all previously described specimens possess only a pseudalveolus.
This alveolar region is somewhat unusual, with a distinctive curvature and subquadrate section,
and assignment of these specimens with greater certainty to D. stimulus var. extremis, and the
possible revision of its taxonomic status will have to await the discovery of more complete topotypes
in Australia.

Specimen KG. 10.41, previously considered by Willey (1972) as a typical D. Anucg/-cgc»/-/(Glaessner)
differs from the holotype of this species (MUGD.1876), which is much larger with a more hastate
outline and depressed section (Glaessner 1945, pi. vi, fig. 12). The Antarctic specimen resembles
much more closely the cylindrical, slightly depressed lectotype of D. stimulus var. extremis.

Occurrence. D. stimulus var. extremis occurs with D. stimulus Whitehouse in the Upper Albian of the Australian
Rolling Downs Group (Whitehouse 1924; Ludbrook 1966). Like D. stimulus, the related form D. cf stimulus
var. extremis occurs in both Alexander and James Ross islands. However, D. cf stimulus var. extremis is
known from lower levels occurring with Peratohelus cf and the late Aptian or early Albian ammonites
Eotetragonites, Australiceras, and Aconeceras in Unit A2 of Taylor et al. (1979) at Succession Cliffs (text-fig.
1) (Willey 1972; Thomson 1974; Taylor et al. 1979). In James Ross Island, D. cf stimulus var. extremis was
found at the 810-930 m levels in the composite section (text-fig. 2) with Albian ammonites (Anagaudryceras
buddlia, ' Pseudothurmannia', and Silesites) and bivalves (I. cf sutherlandi) (Ineson et al. 1986).

Dimitobelus praelindsayi sp. nov.

Plate 23, figs. 3-7

Type specimens. Holotype: D. 8420.1 1a, Albian (Kotick Point Formation), Gin Cove area, James Ross Island.
Paratypes: D. 8212. 265, Albian (Kotick Point Formation), Gin Cove area, James Ross Island; KG. 1610.3,
Albian (Fossil Bluff Formation), Hyperion Nunataks, Alexander Island (text-figs. 1 and 2).

Other material. Ten fragmentary rostra from the Albian (Kotick Point and Whisky Bay formations), James
Ross Island: D. 8403. 35a, c-f, 48, 8420.56, 8, 31 (also (?)D.8420.7, 11b). Two fragmentary rostra from the
Albian, Alexander Island: KG. 1610.2, 1663.10 (text-figs. 1 and 2). Also a single specimen from the Lower
Cretaceous of Cootanorina Station, S. Australia: BM.C.531 1.

168 PALAEONTOLOGY, VOLUME 30

Derivation of name. Prae (Latin), before; lindsayi, a Campanian species of Dimitobelus that the new form
resembles.

Diagnosis. Medium sized, hastate Dimitobelus. Outline symmetrical, hastate, Dlmax close to apex.
Profile asymmetrical, cylindrical to subhastate. Transverse sections depressed elliptical (stem and
apex) to pyriform (alveolar region). Long, straight ventrolateral alveolar grooves deflected dorsally
in stem region.

L

X

Dvmax

Dlmax

D.8420.1 lA

50-2

14-2

8-4

9-5

D.8212.265

48-7

16-0

8-0

9-0

D.8420.5

33-9

15-3

7-2

8-3

KG. 1610.3

59-8

—

—

—

BM.C.531 1

59-2

25-2

9-9

11-2

Description. Medium sized, hastate Dimitobelus. Total length approximately six times Dv. The outline is
symmetrical and hastate, with Dlmax close to the apex, producing a bulbous apical region. The profile is
nearly symmetrical, and cylindrical to subhastate. Transverse sections of the rostrum are weakly depressed
(Dlmax : Dvmax IT; text-fig. 7) and elliptical in the stem and apex, becoming more compressed and pyriform
in the alveolar region, where the venter is flattened.

Two deep, straight ventrolateral grooves parallel the venter for the length of the alveolar region, curving
dorsally at the stem region to join the Doppellinien. The ventrolateral grooves extend adapically for one half
to two thirds the length of the rostrum. The lateral lines (Doppellinien) are faint. Pronounced dorsolateral
alveolar flattened areas or depressions occur for the length of the alveolar region, and may also join the
Doppellinien. The phragmocone is central or slightly dorsal in position, with a ventrally incurved protoconch,
and penetrates one quarter of the rostrum. A pseudalveolus may be developed. The apical line is weakly
cyrtolineate, and an apical canal is present in specimen BM.C.531 1 .

Discussion. Dimitobelus praelindsayi sp. nov. is relatively common in the Albian of the Antarctic
Peninsula region, and, although possessing long straight ventrolateral grooves similar to those of
Peratobelus{l) sp., is clearly a species of Dimitobelus due to its distinct depression and dorsally
deflected grooves.

D. praelindsayi sp. nov. most closely resembles the New Zealand Campanian species D. lindsayi
(Hector). Stevens (1965, p. 117) noticed this similarity when he suggested that specimen BM.C.531 1,
considered here a typical D. praelindsayi sp. nov., was identical to the New Zealand species.
However, D. praelindsayi sp. nov. differs from D. lindsayi in possessing much longer ventrolateral

D1 max
mm

11 -

X HOLOTYPE

© ANTARCTIC SPECIMENS
■ AUSTRALIAN SPECIMEN

10 -

9 -

7 ‘

D. praelindsayi sp nov

TEXT-FIG. 7. Relationship of maximum diameters in Dimito-

9 10 Dv max ^

mm helus praelmdsavi sp. nov.

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

169

grooves and alveolar depressions, in addition to a squatter stem and apical region. D. praelindsayi
sp. nov. is unlike most other species of Dimitohelus in possessing a much longer, pyriform alveolar
region than is typical.

Occurrence. A single specimen of this species is known from the ‘Lower Cretaceous’ of South Australia. It
occurs in some number in both James Ross and Alexander islands, where it occurs with Albian ammonites
and bivalves (Tables 1 and 2); Thomson 1974; Crame 1985; Ineson et al. 1986.

Dimitohelus dayi sp. nov.

Plate 23, figs. 8-10

Type specimens. Holotype: BM.C. 35019. Paratypes: BM.C. 35010, 35002, 35000. Lower Cretaceous (?Albian)
of the Hughenden area. North Queensland, Australia.

Other material. Thirteen rostra: BM.C. 35001, 35003-35009, 35011-35018, 35020-35027. Lower Cretaceous
(?Albian) of Hughenden, North Queensland, Australia.

Derivation of name. In recognition of Dr R. W. Day, who has discussed this species in manuscript.

Diagnosis. Medium sized, robust, hastate Dimitohelus. Outline symmetrical, hastate. Profile sym-
metrical, hastate. Apical region bulbous. Transverse sections elliptical to subcircular.

L

X

Dvmax

Dlmax

BM.C.35019

60-7

20- 1

12-2

15-5

C.35010

57-7

22-2

13-5

160

C.35002

57-8

20-5

18-6

20-8

C.35000

65-4

28-9

17-4

19-2

C.35003

52-5

19-6

14-2

160

C.35004

5L2

79-8

15-2

18-3

C.35005

64-4

23-1

14-3

16-8

Description. Medium sized, hastate Dimitohelus. The outline is symmetrical and hastate, with Dlmax in the
posterior third of the rostrum. In more robust, rounded individuals, the hastation may be reduced. In some
cases the apex may become attenuate, but it is usually obtuse. The profile is symmetrical to slightly asymmet-
rical, and hastate. The venter and dorsum are inflated to the same degree, but in some cases the venter is
flattened. The transverse sections are elliptical (Dlmax : Dvmax 1-2; text-fig. 8), becoming more depressed
adorally.

A pseudalveolus is often developed, so that the alveolar grooves are often lost. However, the holotype
(BM.C. 35019) displays dorsally deflected ventrolateral grooves that join the Doppellinien. This specimen also
shows evidence of dorsolateral depressions continued from the dorsal-most lateral lines. The Doppellinien are
often very incised (e.g. BM.C. 35010). Because of the presence of a pseudalveolus it is difficult to comment on
the phragmocone of this species. However, the apical line is cyrtolineate, and quite often an apical canal is
developed.

Discussion. Dimitohelus dayi sp. nov. is based on specimens from the Wilkins Collection at the
British Museum, from the Lower Cretaceous of the Hughenden region of Queensland. Dr R. W.
Day of the Geological Survey of Queensland had previously identified this species as a variant of
D. diptyclms (M‘Coy) from the Albian of Queensland (1969; R. W. Day, pers. comm. 1985), where
it characterizes the base of the Albian Tambo fauna at various Queensland localities (R. W. Day,
pers. comm. 1985). D. dayi sp. nov. is easily distinguished from D. diptyclms as it possesses a much
more inflated, bulbous rostrum, as compared with the depressed rostrum of the latter (text-figs. 5
and 8).

Occurrence. D. dayi sp. nov. is known from Australia, principally the Hughenden district of Queensland,
where it was previously identified by Day (1969). Day recognized that this form occurs at the base of the
Albian Tambo fauna in this region, occurring 12-15 m above the Aptian ammonite Australiceras irregulare
(R. W. Day, pers. comm. 1985). The related form D. cf. dayi sp. nov. occurs in James Ross Island (see below).

170

PALAEONTOLOGY, VOLUME 30

Dl max
mm

21 -

X HOLOTYPE

• WILKINS COLLECTION, BM

20 -

19 -
18 -
17 -
16 -
15 -
14 ■

TEXT-FIG. 8. Relationship of maximum diameters in Di-
mitobelus dayi sp. nov.

13 -
12 -

D dayi sp nov.

11 ■

_i I I i I I 1 ■ I

11 12 13 14 15 16 17 18 19

Dv max mm

Dimitobelus cf. dayi sp. nov.

Plate 23, fig. 1 1

Material. One rostrum: D.8430.15, Albian (Whisky Bay Formation), James Ross Island (text-fig. 2).

Discussion. Specimen D.8430.15 resembles the holotype of D. dayi sp. nov. in possessing a moder-
ately sized, robust, bulbous rostrum, with an obtuse to mucronate apex. It is slightly tectonically
deformed, but possesses clear ventrolateral grooves (the position of which is confirmed by the
preserved siphuncle) that are dorsally deflected to eroded lateral lines. The alveolar section is
quadrate, similar to D. dayi sp. nov., and the specimen shows evidence of the development of a

EXPLANATION OF PLATE 23

Figs. 1 and 2. Dimitobelus cf. stimulus var. extremis Whitehouse. 1, left profile and ventral outline, Albian,
James Ross Island. D. 8403. 35b, xl. 2, silicone rubber cast, oblique ventral outline, ?Aptian-Albian,
Alexander Island. KG. 10. 41, x 1.

Figs. 3-7. D. praelindsayi sp. nov. 3-5, 7, left profiles and ventral outlines. 3, Albian, South Australia.
BM.C.5311, X 1. 4, holotype, Albian, James Ross Island. D. 8420. 11a, x 1. 5, paratype, Albian, James
Ross Island. D. 82 12.265, x 1. 6, paratype, silicone rubber cast, oblique dorsal outline, Albian, Alexander
Island. KG. 1610.3, x 1. 7, alveolar fragment, Albian, Alexander Island. KG.1610.2, x 1.

Figs. 8-10. D. dayi sp. nov. Albian, Queensland, Australia, right profiles and ventral outlines. 8, paratype.

BM.C.35010, X 1. 9, holotype. BM.C. 35019, x 1. 10, paratype. BM.C.35002, x 1.

Fig. II. Z). cf. dayi sp. nov. Albian-(?)Cenomanian, James Ross Island, left profile and ventral outline.
D.8430.15, X 1. ’

Figs. 12 and 13. D. sp. nov. A. Albian, Alexander Island. Silicone rubber casts. 12, left profile and ventral
outline. KG.1663.13, x 1. 13, right profile. KG.1663.14, x 1.

Figs. 14 and 15. D. sp. nov. B. Albian, Alexander Island. Silicone rubber casts. 14, left profile and ventral
outline. KG. 1663.1 1, x 1. 15, right profile and ventral outline. KG. 1748.20, x 1.

Fig. 16. D. sp. nov. C. Albian, Alexander Island. Silicone rubber cast, left profile. KG. 1658.16, x 1.

PLATE 23

DOYLE, Dimitobelus

172

PALAEONTOLOGY, VOLUME 30

pseudalveolus. However, owing to the slightly tectonized nature of the rostrum, and the position of
Dlmax very close to the apex, this specimen cannot be definitely assigned to D. dayi sp. nov.

D. cf. dayi sp. nov. also resembles the more bulbous specimens of D. sp. nov. aff. diptychus
(M'Coy), but differs by possessing a compressed alveolar section.

Occurrence. The single specimen of D. cf. dayi sp. nov. occurs in James Ross Island at the 1540 m level in the
composite section (text-figs. 2 and 3) with inoceramids possibly indicative of a later Albian age (Table 2;
Crame 1985; Ineson el al. 1986).

Dimitobelus sp. nov. A

Plate 23, figs. 12 and 13

Material. Four natural moulds: KG. 1663. 13, 14, 1680.69, 1728.16, Albian, Alexander Island (see text-fig. 1).

Description. Medium sized, slender, subhastate Dimitobelus. The outline is symmetrical and subhastate, with
very regular, straight laterals anterior to the apex. The profile is almost symmetrical, and cylindrical to
subhastate. Transverse sections are elliptical in the stem and apical regions, becoming pyriform in the alveolar
region, with a broad flat venter.

The ventrolateral grooves are deep and well developed, extending to just past the mid-point of the rostrum
before being deflected towards the dorsum. Lateral lines are difficult to distinguish on casts, and in specimen
KG. 1663. 13 the deflection of the ventrolateral grooves is barely perceptible, so that it resembles P. cf. oxys
(Tenison-Woods). The form of the phragmocone and apical line are unknown in these casts and moulds.

Discussion. D. sp. nov. A is a slender conical form that closely resembles P. cf. oxv^. However, for
the most part D. sp. nov. A has a clearly defined dorsal curvature of its grooves, and the depressed
section characteristic of Dimitobelus. It is easily distinguised from the other Albian Dimitobelus due
to its flat venter and distinctive anterior.

Occurrence. D. sp. nov. A is restricted to Alexander Island where it occurs with D. praelindsayi sp. nov. and
Albian ammonites and bivalves at Stephenson Nunatak (text-fig. 1; Table 1; Taylor et al. 1979; Crame 1985).

Dimitobelus aff. sp. nov. A

Material. A single natural mould with a rather more robust profile than is normal in D. sp. nov. A: KG. 1721. 27,
Mount Phoebe, Alexander Island, occurring with the Albian bivalve Birostrinal cf. concentrica (Table 1;
Crame 1985).

Dimitobelus sp. nov. B
Plate 23, figs. 14 and 15

Material. One fragmentary rostrum: KG. 1748.20, one natural mould: KG. 1663.1 1. Albian, Alexander Island
(text-fig. 1).

Description. Medium sized, robust, subhastate to cylindrical Dimitobelus. The outline is symmetrical and
subhastate to cylindrical, and the profile is generally similar. The transverse sections are slightly depressed
and elliptical to subcircular. Two deep ventrolateral alveolar grooves extend for approximately one half of
the length of the rostrum, where they are deflected dorsally close to the apex of the rostrum. Doppellinien are
not well developed, but there is evidence of some shallow dorsolateral depressions. The apex of specimen
KG. 1663. 11 exhibits some striations, but this may be due to the erosion of the apex. The phragmocone
penetrates an estimated one third of the rostrum, and the apical line is ventrally displaced.

Discussion. Dimitobelus sp. nov. B is a robust form, most closely resembling D. kleini (Gurich) from
the White Cliffs of W. Australia (Gurich 1901, p. xix, fig. 2, and topotypes BM.C. 12086 and 20248)
in the length of its grooves and its conical apex. However, D. sp. nov. B differs from D. kleini in
possessing a depressed transverse section, and from most other species of this genus due to its
robust form.

Occurrence. Like D. sp. nov. A, D. sp. nov. B occurs with Albian bivalves and ammonites at Stephenson
Nunatak, Alexander Island (Table 1; Thomson 1974; Crame 1985).

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELIDAE

173

Dimitobelus sp. nov. C
Plate 23, fig. 16

Material. One natural mould: KG. 1658.16, Albian, Alexander Island (text-fig. 1).

Dimensions. L, 47-9; X, 18-7; Dvmax, approx. 4 0; Dlmax, 4-5.

Description. Medium sized, slender, subhastate Dimitobelus. The outline is symmetrical and subhastate, with
Dlmax close to the apex, and with straight laterals anterior to the apical region. The profile is asymmetrical
and cylindriconical to subhastate. The transverse sections are slightly depressed (Dlmax : Dvmax approx. 11),
elliptical in the apical region, becoming compressed and subquadrate in the stem and alveolar regions.

Two rather sinuous ventrolateral alveolar grooves extend for approximately one half of the rostrum, where
they curve gently towards the dorsum, joining faint lateral lines.

Discussion. D. sp. nov. C is unlike most other species of Dimitobelus as it is slender, with a compressed
alveolar section. D. stimulus var. extremis Whitehouse approaches it in slenderness, but does not
possess the long compressed alveolar region of D. sp. nov. C. T. whitehousei sp. nov. is also similar,
but this species does not have the depressed apical region characteristic of D. sp. nov. C.

Occurrence. D. sp. nov. C occurs with the biostratigraphically undiagnostic ammonites Lytoceras and Hypo-
phylloceras at Hyperion Nunataks, Alexander Island (text-fig. 1; Table 1). However, the related form D. all',
sp. nov. C. has been found with the Albian bivalve B.l cf. concentricaat Succession Cliffs (text-fig. 1; Table 1 ).

Dimitobelus aff. sp. nov. C

Material. A single natural mould with a markedly pyriform alveolar region, although as slender as D. sp. nov.
C; KG. 1746.1, Alexander Island, occurring with the Albian bivalve B.l cf. concentrica (Table 1; Crame 1985).

Dimitobelus(l) spp.

Material. Twelve otherwise unidentifiable alveolar fragments from Alexander Island; KG. 10.28, 1687.22,
1721.25, 26, 1610.1, 1748.1, 17, 1681.1, 1658.13b, 1609.7, 1617.3, 1737.21. One alveolar fragment with a
markedly quadrate transverse section and dorsally deflected ventrolateral grooves: KG. 1657.28. Localities are
given in text-fig. 1, and associated faunas are given in Table 1.

RANGE AND DISTRIBUTION OF THE DIMITOBELIDAE

The Dimitobelidae may have evolved from a southern endemic stock of the Tethyan belemnopseid
Hibolithes in the later early Cretaceous, as this genus possesses well-developed Doppellimen, and an
often reduced single ventral alveolar groove (see Doyle (in press) for further discussion). The family
remained endemic to within the 30° S. palaeolatitude, and ‘counterbalanced’ the Belemnitellidae
which were restricted to the Boreal Realm in the north (this family may also have evolved from
endemic northern stocks of Hibolithes, Doyle, in press).

The Dimitobelidae first appeared in the Aptian, with Peratobelus and Tetrabeius. Peratobelus is
known from Australia and Antarctica (it has also been listed from Southern Mozambique (Wachen-
dorf 1967)), and both regions share forms referable to P. o.viw (Tenison-Woods), while Tetrabeius
is only known in Antarctica with the single species T. willeyi sp. nov. (Stevens 1973; Doyle 19856)
(text-fig. 9). The supposed late Neocomian beds that have yielded dimitobelids (principally Perato-
belus) in the Australian Northern Territory (Skwarko 1966) are now considered to be of Aptian
age on other faunal evidence, principally the occurrence of the ammonite Australiceras (Day 1969,
p. 154).

By the Albian, Peratobelus was replaced by Dimitobelus, the first representatives of which prob-
ably appeared in the late Aptian or early Albian. Dimitobelus is recorded from Antarctica, New
Zealand, Australia, New Guinea, and possibly South America (Whitehouse 1924, 1925; Glaessner
1945, 1958; Stevens 1965; Willey 1972; Pettigrew and Willey 1975; Doyle 19856). Faunal links

174

PALAEONTOLOGY, VOLUME 30

between Australia and Antarctica at this time are indicated by the presence of D. diptychus (M‘Coy),
D. stimulus Whitehouse, D. praelindsayi sp. nov., and D. dayi sp. nov. in both regions (text-fig. 9).
Although apparently close to possible migration routes, New Zealand Dimitohelus are distinct and
probably endemic, remaining so from the Albian to the Maastrichtian (Stevens 1965) (text-fig. 9).
Tetrabelus survived into the Albian, and increased its range and diversity, occurring in Antarctica,
southern India, and possibly the Malagasy Republic (Blanford 1861; Lemoine 1906; Spengler 1910;
Doyle 1985Z)). Close faunal links between Antarctica (James Ross Island) and southern India (also
possibly the Malagasy Republic), are indicated by the occurrence of the single species T. seclusus
(Blanford) in these regions (Doyle 19856) (text-fig. 9). Therefore, in the Albian, two distinct distri-
butions became apparent, indicated by the occurrence of Dimitohelus (circum-Gondwanian) and
Tetrabelus (trans-Gondwanian) (Doyle 19856).

ANTARCTICA NEW ZEALAND

JAMES ROSS Is ALEXANDER Is

EASTERN

AUSTRALIA

MAASTRICHTIAN

CAMPANIAN

SANTONIAN

TURONIAN

CENOMANIAN

NEW GUINEA SOUTHERN

INDIA

D. hectori

D sp nov aff

D. lindsayi

diptychus

?D ongleyi

D superstes

D diptychus

D diptychus

D diptychus

D. macgregori

D cf dayi sp nov

D stimulus

T whitehousei sp nov

D macgregori

T seclusus

D praelindsayi sp nov

D praelindsayi sp nov

D stimulus

D praelindsayi sp nov

D stimulus

D dayi sp nov

P cf oxys

P oxys

T willeyi sp nov

P australis

TEXT-FIG. 9. Distribution of the Dimitobelidae. Diagram not to scale.

Few dimitobelids have been described from post-Albian sediments, but circum-Gondwanian links
were maintained at least into the Cenomanian, and Dimitohelus is known from Antarctica, New
Zealand, Australia, and possibly New Guinea (Stevens 1973; Doyle 19856). However, no further
trans-Gondwanian dimitobelids are known after the last appearance of Tetrabelus in the late
Albian-Cenomanian (Doyle 19856). In the Coniacian-Santonian interval, Dimitohelus is known
only from New Zealand (Stevens 1965), but in the Campanian, a single form is known from James
Ross Island (Z). sp. nov. afif. diptychus (M‘Coy)) in addition to D. lindsayi (Hector) from New
Zealand (text-fig. 9). The last dimitobelid to appear is D. hectori Stevens from the Maastrichtian
(Haumurian) of New Zealand (Stevens 1965).

DOYLE: ANTARCTIC CRETACEOUS DIMITOBELID AE

175

CONCLUSIONS

1. The family Dimitobelidae consists of three genera; Peratobelus Whitehouse (Aptian), Tetra-
beliis Whitehouse (Aptian-(?)Cenomanian), and Dimitobelus Whitehouse (= Cheirobeliis White-
house) ((?)late Aptian/Albian-Maastrichtian) which were all restricted to within the 30° S.
Cretaceous palaeolatitude.

2. In the Aptian-Albian fore-arc and back-arc sediments of the Antarctic Peninsula region a
relatively diverse fauna of Peratobelus, Tetrabelus, and Dmitobelus occur. In no other Gondwanian
region do all three genera occur together.

3. Peratobelus is known from the Aptian sediments of Alexander Island and Australia, and both
regions share similar species (e.g. P. oxys (Tenison-Woods)).

4. Dimitobelus from the Albian sediments of James Ross Island have considerable affinities to
those of Alexander Island and Australia. However, they do not resemble the New Zealand Dimito-
belus which were probably endemic to this region. D. sp. nov. aff. diptyclms (M‘Coy) from James
Ross Island and D. lindsayi (Hector) from New Zealand are the only Campanian Dimitobelus so
far known.

5. Tetrabelus is known from Aptian sediments only in Alexander Island, where T. willeyi sp.
nov. occurs. In the Albian the single species T. seclusus (Blanford) is known from James Ross Island
and southern India, but not from Alexander Island. T. whitehousei sp. nov. is apparently endemic
to James Ross Island in the Albian.

Acknowledgements. This work was carried out during a period appointment (1984-1985) at the British
Antarctic Survey, Cambridge. I am indebted to Drs. J. A. Crame, J. R. Ineson, and M. R. A. Thomson for
access to their material and stratigraphic data, and P. J. Howlett for discussion. Professor M. F. Glaessner,
Dr R. W. Day, Dr H. C. Klinger, and Dr G. R. Stevens provided helpful advice. I thank also the follow-
ing for supplying specimens or casts from, or allowing access to, the collections in their care: D. Phillips,
M. K. Howarth (BM); J. M. Lindsay (GSSA); W. D. I. Rolfe (HM); G. W. Quick (MUGD); D. J.
Holloway (NMV); N. Pledge (SAM); and D. Price (SM). Gill Clarke typed the manuscript and Chris Gilbert
took the photographs.

REFERENCES

BHALLA, s. N. 1983. India. In moullade, m. and Nairn, a. e. m. (eds.). The Phanerozoic Geology of the World.
II. The Mesozoic, B., 305-352. Elsevier, Amsterdam.

BiBBY, J. s. 1966. The stratigraphy of part of north-east Graham Land and the James Ross Island Group.
Scient. Rep. Br. Antarct. Siirv. 53, 1-37.

BLANFORD, H. F. 1861-1865. The fossil Cephalopoda of the Cretaceous rocks of southern India: Belemnitidae-
Nautilidae. Mem. geol. Siirv. India Palaeont. indica, 1, 1-400.

BULOW-TRUMMER, E. VON. 1920. FossUium Catalogus 1: Animalia. Pars II: Cephalopoda dibranchiata, 313 pp.
Junk, Berlin.

CRAME, J. A. 1981. Upper Cretaceous inoceramids (Bivalvia) from the James Ross Island group and their
stratigraphical signihcance. Bull. Br. Antarct. Siirv. 53, 29-56.

1985. Lower Cretaceous inoceramid bivalves from the Antarctic Peninsula region. Palaeontologv, 28,

475-525.

DAY, R. w. 1967. Marine Lower Cretaceous fossils from the Minmi Member, Blythesdale Formation, Roma-
Wallumbilla area. Pubis geol. Surv. Qd, 335, Palaeont. paper 9, I -30.

1969. The Lower Cretaceous of the Great Artesian Basin, 140-173. In Campbell, k. s. w. (ed.). Stratigraphy

and Palaeontology: Essays in honour of Dorothy Hill. Australian National University Press, Canberra.
DORMAN, F. H. and GILL, E. D. 1959. Oxygen isotope palaeotemperature measurements on Australian fossils.
Proc. R. Soc. Viet., NS, 71, 73-98.

DOYLE, p. 1985a. Sexual dimorphism in the belemnite Yoimgibelus from the Lower Jurassic of Yorkshire.
Palaeontology, 28, 133-146.

19856. ‘Indian’ belemnites from the Albian (Lower Cretaceous) of James Ross Island, Antarctica. Bull.

Br. Antarct. Surv. 69, 23-34.

(in press). The belemnite family Dimitobelidae in the Cretaceous of Gondwana. Int. Un. geol. Sci.,

Series A.

176

PALAEONTOLOGY, VOLUME 30

ETHERIDGE, R. JR. 1902fl. The Cretaceous mollusca of South Australia and the Northern Territory. Mem. R.
Soc. S. Aust. 2, 1-54.

\9Q2b. A monograph of the Cretaceous invertebrate fauna of New South Wales. Mem. geo!. Surv.

N.S.W., Palaeont. Ser. 11, 1-98.

FARQUHARSON, G. w., HAMER, R. D. and iNESON, j. R. 1984. Proxiiual volcaniclastic sedimentation in a
Cretaceous back-arc basin, northern Antarctic Peninsula, 219-229. In kokelaar, b. p. and howells, m. f.
(eds.). Marginal Basin Geology. The Geological Society and Blackwell Scientific Publications, London.
GLAESSNER, M. F. 1945. Mesozoic fossils from the Central Highlands of New Guinea. Proc. R. Soc. Viet., ns,
56, 151-168.

1957. Cretaceous belemnites from Australia, New Zealand and New Guinea. Aust. J. Sci. 20, 88-89.

1958. New Cretaceous fossils from New Guinea. Rec. S. Aust. Mus. 13, 199-226.

GURiCH, G. 1901. Jura- und Devon-Fossilien von White Cliffs, Australien. Neues Jb. Miner. Geol. Paldont.
Beil. Bd. 14, 484-518.

HILL, D., PLAYFORD, G. and WOODS, J. T. (eds.). 1968. Cretaceous fossils of Queensland, 35 pp. Queensland
Palaeontographical Society, Brisbane.

INESON, J. R., CRAME, J. A. and THOMPSON, M. R. A. 1986. Lithostratigraphy of the Cretaceous strata of west
James Ross Island, Antarctica. Cret. Res. 7, 141-159.

JACK, R. L. and ETHERIDGE, R. JR. 1892. The geology and palaeontology of Queensland and New Guinea, 768 pp.
J. C. Beal, Brisbane.

JELETZKY, J. A. 1966. Comparative morphology, phylogeny and classification of fossil Coleoidea. Paleont.
Contr. Univ. Kans. Mollusca, art. 7, 162 pp.

LEMOINE, p. 1906. Etudes geologiques dans le Nord de Madagascar, 520 pp. Hermann, Paris.

LUDBROOK, N. H. 1966. Cretaceous biostratigraphy of the Great Artesian Basin in South Australia. Bull. geol.
Surv. S. Aust. 40, 1-223.

m'coy, f. 1867. On the occurrence of Icthvosaurus and Plesiosaurus in Australia. Ann. Mag. nat. hist. 19,
353-356.

NAEF, A. 1922. Die fossilien Tintenfische, 322 pp. Junk, Berlin.

OLiVERO, E. B. 1984. Nuevos ammonites Campanianos de la Isla James Ross, Antartida. Ameghiniana, 21,
53-84.

PETTIGREW, T. H. and WILLEY, L. E. 1975. Belemnite fragments from Annenkov Island. Bull. Br. Antarct. Surv.
40, 33-36.

PHILLIPS, J. 1870. In MOORE, c. Australian Mesozoic geology and palaeontology. Q. Jl geol. Soc. bond. 26,
226-260.

SCHWEGLER, E. 1961. Revision der Belemniten des Schwabischen Jura, teil 1. Palaeontographica, A116,
59-103.

SKWARKO, s. K. 1966. Cretaceous stratigraphy and palaeontology of the Northern Territory. Bull. Bur. Miner.
Resour. Geol. Geophys. Aust. 73, 1-135.

SPATH, L. F. 1953. The Upper Cretaceous cephalopod fauna of Graham Land. Sclent. Rep. Falkld Isl. Depend.
Surv. 3, 1 -60.

SPENGLER, E. 1910. Die Nautiliden und Belemniten des Trichinopolydistrikts. Beitr. Paldont. Geol. Ost.-Ung.
23, 125-157.

STEVENS, G. R. 1965. The Jurassic and Cretaceous belemnites of New Zealand and a review of the Jurassic and
Cretaceous belemnites of the Indo-Pacific region. Pal. Bull. geol. Surv. N.Z. 36, 1-283.

1973. Cretaceous belemnites. In hallam, a. (ed.). Atlas of Palaeobiogeography, 259-274. Elsevier,

Amsterdam.

stoliczka, f. 1866. Cretaceous fauna of Southern India, Ammonitidae with a revision of the Nautilidae, etc.

Mem. geol. Surv. India Palaeont. indica, 3, 1-216.

STOLLEY, E. 1919. Die Systematik der Belemniten. Jber. niedersdehs. geol. Ver. 11, 174-191.

TATE, R. 1880. Description of a new species of belemnite from the Mesozoic strata of central Australia. Trans.
R. Soc. S. Aust. 3, 104-105.

TAYLOR, B. J., THOMSON, M. R. A. and WILLEY, L. E. 1979. The geology of the Ablation Point-Keystone Cliffs
area, Alexander Island. Scient. Rep. Br. Antarct. Surv. 82, 1-65.

TENisoN- WOODS, J. E. 1883. On some Mesozoic fossils from the Palmer River, Queensland. J. R. Soc. N.S.W.
16 (for 1882), 147-154.

1884. On some Mesozoic fossils from Central Australia. Proc. Linn. Soc. N.S. W. 8, 235-242.

THOMSON, M. R. A. 1974. Ammonite faunas of the Lower Cretaceous of south-eastern Alexander Island. Scient.
Rep. Br. Antarct. Surv. 80, 1-44.

DOYLE; ANTARCTIC CRETACEOUS DIMITOBELIDAE

177

1982. A comparison of the ammonite faunas of the Antarctic Peninsula and Magallanes Basin. J. geol.

Soc. Lond. 139, 763-770.

1984o. Report on Antarctic Field Work. Preliminary ammonite zonation of the mid-Cretaceous rocks

of James Ross Island. Bull. Br. Antarct. Surv. 64, 85-91.

1984^. Cretaceous ammonite biostratigraphy of western James Ross Island, Antarctica, 308-314. In

PERRILLIAT, M. DE c. (ed.). Memoria, III Congreso Latinoamericano de Paleontologia, Mexico, 1984. Universi-
dad Nacional Autonoma de Mexico, Instituto de Geologia.

PANKHURST, R. J. and CLARKSON, p. D. 1983. The Antarctic Peninsula— a late Mesozoic-Cenozoic arc

(Review), 289-294. In Oliver, r. l., james, p. r. and jago, j. b. (eds.). Antarctic Earth Science. Cambridge
University Press.

WACHENDORF, H. 1967. Zur Unterkreide-stratigraphie von Siid-Mo^ambique. Neues Jb. Geol. Paldont. Ahh.
129, 272-303.

WHiTEHOUSE, F. w. 1924. Dimitobelidae— a new family of Cretaceous belemnites. Geol. Mag. 61, 410-416.

1925. On Rolling Downs fossils collected by Prof. J. W. Gregory. Trans. R. Soc. S. Aust. 49, 27-36.

WILLEY, L. E. 1972. Belemnites from south-eastern Alexander Island I. the occurrence of the family Dimito-
belidae in the Lower Cretaceous. Bull. Br. Antarct. Surv. 28, 29-42.

1973. Belemnites from south-eastern Alexander Island II. The occurrence of the family Belemnopseidae

in the Upper Jurassic and Lower Cretaceous. Ibid. 36, 33-59.

WOODS, H. 1917. The Cretaceous faunas of the north-eastern part of South Island of New Zealand. Pal. Bull,
geol. Surv. N.Z. 4, 1-41.

WOODS, J. T. 1961. Mesozoic and Cainozoic sediments of the Wrotham Park area. Pubis, geol. Surv. Ad.
304, 1-6.

Typescript received 27 November 1985
Revised typescript received 27 March 1986

PETER DOYLE

Department of Palaeontology
British Museum (Natural History)
Cromwell Road, London SW7 4BD

Note added in proof. Since completion of this paper. Dr J. A. Crame has collected fourteen additional specimens
of T. willeyi sp. nov. (KG.3401.223, 111, 618, 619, 621, KG. 3403.218-221, 224, 225, 229, 230, 232) of possible
Aptian age from Spartan Cwm, Alexander Island (71° 04' S., 68° 25' W., 10 km north-west of Succession
Cliffs). Dimensions of KG.3403.229: L, 38-4; Dv, 5-7; Dl, 5-2.

THE CALLOVIAN (MIDDLE JURASSIC) MARINE
CROCODILE METRIORHYNCHUS FROM
CENTRAL ENGLAND

by SUSAN M. ADAMS-TRESMAN

Abstract. For many years the taxonomy of the Callovian marine crocodile genus Metriorhynchus has been
in a state of confusion. Bivariate and principal coordinate analyses are used in an attempt to identify cranial
characters for discriminating species. Many of the characters used previously to define eight species of
Metriorhynchus are shown to be individually variable or continuously variable. Only two Callovian species
can now be identified on the basis of their skull proportions: M. superciliosus incorporates specimens previously
assigned to M. supercilioswn de Blainville, 1853, M. moreli E. E. Deslongchamps, 1867, M. leedsi Andrews,
1913, and M. laeve Andrews, 1913; M. hrachyrhynchus includes M. brachyrhynchus E. E. Deslongchamps,
1867, Suchodus durobrivense Lydekker, 1890, M. cultridens Andrews, 1913, and M. casamiquelai de Gasparini
and Diaz, 1977.

The current classification of the Callovian marine crocodiles of the genus Metriorhynchus was
established largely by E. E. Deslongchamps (1863-1869) and Andrews ( 1913). Deslongchamps gave
the first detailed descriptions and figures of Metriorhynchus, emending von Meyer’s (1830) earlier
generic descriptions and de Blainville’s (1853) specific description of M. supercilioswn. He created
new species for the material collected around Caen in Normandy and three of these (M. supercili-
osum, M. moreli, and M. brachyrhynchus) include the metriorhynchids found in the English Lower
Oxford Clay and contained in the Leeds Collection at the British Museum (Natural History). The
characters used to diagnose the species included skull proportions, ornament of cranial bones,
relationships of cranial bones, and numbers and types of teeth.

Lydekker (1890) compared an imperfect skull and mandible from Peterborough with Deslong-
champs’s figures, and concluded that it proved the existence of a separate genus Suchodus, with one
new species S. durobrivense. Eraas (1902) extended the descriptions of M. brachyrhynchus,
M. supercilioswn, and M. moreli and erected a new species M. blainvillei (shown by Wenz (1968) to
be synonymous with M. supercilioswn). Schmidt (1904) created M.jaekeli for a specimen from the
Oxford Clay which was subsequently also shown by Leeds (1907), Andrews (1913), and Wenz
(1968) to be synonymous with M. supercilioswn.

Andrews (1913) erected three new species for specimens collected from the Peterborough district,
which he distinguished from species created by E. E. Deslongchamps, Lydekker, and Eraas. His
work was based on the metriorhynchids in the Leeds Collection. He recognized seven species in all
(his text-fig. 73) and defined them as follows:

(a) Forms in which the surface of the cranial bones is without sculpture:

M. laeve— "A small species with a narrow skull, teeth numerous and close set, upwards of thirty on each
side of the mandible.’

M. 'Skull broader and more massive than in [M. laeve]. Teeth large and close set, about thirty-six in

each maxilla.’

(b) Forms in which the surface of the cranial bones is more or less sculptured with pits and grooves:

(i) Narrow skulled forms:

M. supercilioswn — ‘A narrow skulled form in which the surface of the frontal is sculptured with sharply
defined pits. The frontal extends forwards nearly to the level of the anterior angle of the prefrontals, and its

[Palaeontology, Vol. 30, Part 1, 1987, pp. 179-194.]

© The Palaeontological Association

180

PALAEONTOLOGY, VOLUME 30

length in front of the orbits of the temporal fossae is considerably greater than the least width between the
orbits. About twenty eight teeth in each maxilla.’

M. moreIi—"A. narrow skulled form in which the frontal bears a sculpture of shallow and, as it were, partly
obliterated pits, its anterior angle does not extend forwards to the level of the anterior angle of the prefrontals,
and its length in front of the temporal fossae is about equal to the least width between the orbits. About
twenty-six teeth in each maxilla.’

(ii) Broad skulled forms:

M. cultridens—'SikuW with comparatively short rostrum in which the nasals are separated from the premax-
illae by a distance about equal to a quarter of their own length. Supraorbital notch an open continuous curve,
teeth smooth and with strongly compressed crowns, about twenty teeth in the maxilla.’

M. brachyrhynchus — 'Skull with short rostrum in which the nasals meet or nearly meet the premaxillae.
The supraorbital notch forms a sharp angle, and a line joining the outer angles of the prefrontals passes
behind the posterior angle of the nasals. About twenty-one teeth in each maxilla.’

M. chirobrivense— '’Skull broad with a short rostrum in which the nasals do not quite reach the premaxillae.
A line joining the outer angles of the prefrontals passes through the hinder end of the nasals. About sixteen
teeth in each maxilla.’

Wenz ( 1 968), working with Callovian specimens from France, concluded that the metriorhynchids
should be arranged in two groups: (i) those species with a narrow, long snout— M. superciliosum,
M. teedsi, and M. laeve (in the definition of these species ornament is important); (ii) those species
with large skulls and short snouts — M. cultridens, M. durohrivense, and M. brachyrhynchus.

Wenz ( 1 968 ) noted the degree of individual and age variation amongst the Callovian metriorhyn-
chids, by analogy with variation in living crocodiles (Mook 1921; Kalin 1955). She doubted the
value of certain taxonomic criteria employed by previous authors, and later (Wenz 1970) favoured
the provisional retention of the species M. cultridens, M. durohrivense, and M. brachyrhynchus.

De Gasparini and Diaz (1977) created M. casamiquelai for a Metriorhynchus skull from the
Callovian of Northern Chile, and included it with the second of Wenz’s (1968) groups.

To summarize, the currently accepted species of Metriorhynchus from the Callovian are M.
superciliosum de Blainville, 1853, M. moreli E. E. Deslongchamps, 1867, M. durohrivense (Lydekker,
1890), M. brachyrhynchus E. E. Deslongchamps, 1867, M. laeve Andrews, 1913, M. leedsi Andrews,
1913, M. cultridens Andrews, 1913, and M. casamiquelai de Gasparini and Diaz, 1977. The criteria
used in the division of the genus by Deslongchamps (1863-1869) and Andrews (1913) were: A, size
and proportions of the skull— broad or narrow; B, sculpture of the dorsal surface of the cranial
bones— with or without sculpture; C, numbers of teeth; D, relationship between frontal and pre-
frontal bones; E, distance between the nasals and premaxillae; and F, development of the frontal
bone— measured as the projection of the frontal anterior to the supratemporal fenestrae compared
to the least width between the orbits. Deslongchamps’s and Andrews’s specific diagnoses were based
on the typological species concept, in that each Metriorhynchus species was viewed as being virtually
invariable, and small morphological variations were considered to have taxonomic significance.
The taxonomy rested upon small samples or individual specimens because morphological standards
were established in the type specimen, so that additional specimens were not thought to be relevant
to the diagnosis of that species.

Wenz (1968, 1970), de Gasparini and Diaz (1977), and Buffetaut (1977) have all expressed doubts
in recent years as to the validity of these taxonomic criteria, but have not suggested any formal
revision of the taxonomy because they did not have access to sufficient material on which such a
revision could be based.

MATERIAL AND METHODS

The results presented here are based on an extensive study of Callovian Metriorhynchus specimens belonging
to the Leeds Collection, together with a range of supplementary material. These crocodiles occur principally
in the jason and coronation zones of the Lower Oxford Clay and were found in a small area of north
Cambridgeshire, near Peterborough.

A variety of cranial characters was measured, including those which had been used to diagnose species by

ADAMS-TRESMAN: CALLOVIAN CROCODILE METRIORHYNCHUS

181

TEXT-FIG. 1 . The pattern of dorsal cranial bones in Metriorhyncims, showing the measurements taken during
morphological analysis (see Table 1). Abbreviations: boc, basioccipital; f, frontal; j, jugal; mx, maxilla;
n, nasal; p, parietal; pmx, premaxilla; pof, post frontal; prf, prefrontal; q, quadrate; sq, squamosal; A, length
in mid-dorsal line; A-l-B, length from occipital condyle to tip of snout; C, length from the anterior end of
frontal to tip of snout; D, length of frontal anterior to supratemporal fenestrae; E, distance between nasals
and premaxillae; F, length of nasals; G, least width between orbits; H, width between outer angles of
prefrontals; I, width between outer angles of quadrates.

Deslongchamps, Andrews, and de Gasparini and Diaz, so that information about dissociated metriorhynchid
material could be synthesized. In this way the sample size available for analysis was substantially increased
(numbers in parentheses) beyond that analysed by Andrews (1913): M. laeve, 2 (4); M. leedsi, 2 (4); M. moreli,
4 (7); M. superciliosum, 4 (26); M. cultridens, 2 (2); M. brachyrhynchus, 2 (4); M. durobriveme, 2 (4); and M.
sp. (34).

The features A-F (listed above) which are currently accepted as having taxonomic validity are evaluated
here. A, C, E, and F can be examined quantitatively, B and D qualitatively. Some of these are not, in fact,
independent characters. The variation in cranial characters is assessed in the light of the neontological species
concept (Newell 1956) bearing in mind the probable levels of individual variation which might occur in a
crocodile population (Mook 1921; Cott 1961; Dodson 1975). Where the current criteria are shown to be
invalid, appropriate revisions are suggested. The cranial measurements taken are shown in text-fig. 1, and the
data so obtained set out in Table 1.

Ranked statistics derived from Table 1, a similarity matrix, and nearest neighbour scores have been tabulated
for all specimens allocated a computer number; these tables have been deposited with the British Library as
Supplementary Publication No. 14029 (7 pages). It may be purchased from the British Library, Lending
Division, Boston Spa, Wetherby, Yorkshire LS23 7BQ, UK. Prepaid coupons for such purposes are held by
many technical and university libraries throughout the world.

CHARACTER ANALYSES

Bivariate plots

Combinations of two and three cranial measurements were plotted and their compatibility with the
present taxonomic interpretations tested (text-figs. 2 and 3). Text-fig. 3 indicates that two groups of
metriorhynchids can be determined, where the groups are based on the relationship between the
width of the skull (measured across the prefrontals) (text-fig. 3a) and the separation of the nasals and
premaxillae (text-fig. 3b). These results are not entirely compatible with the present classification.

182

PALAEONTOLOGY, VOLUME 30

Tt

■ ' x> Jd

*= n ^

T3 ^

cU ^

O ^ ^

>2 e ^

2^ £
0 I =
£

^oc

2 M i£

5 ==

<D (L)

« 00 C

5s 2

- c/u p

3 t;

^ t3 g

^ 3_
2 u ^

•-G c/T
.22 ^
cc '55
Jr > >»

“ r ■s

9 ° §

Z § s
:§ 3 a

=Q 9 E

O ^
' ^

C3 (D C3

g z s ^

B-’S

l:^' E-
2 5 E

Xi ^

2.6

>A C/5

— <u

^t.l

. o U

X O E

»3 S 3

aj u

g

^ g

S 6 o

“r** 3 <D
~ (U D.

> 00 c/5

9'..2 <u

^ c/5

.C) o

1^2

czi Cj c3

G L/' O

<D ^ —

£ o 15
S w)

V- c/5 (L»

G cd »-

30 ^

Egg

^ 3 *1^

c3 aj "B

■£ 3 E

2 S c

^ 3 -

• p ^

' 'p p
W B
m S E
H X O

smpiiStiJ.Upvuq 6e6£>l (HN)Wa
snqM.upAqjvjq /y 669£a (HN)lAia
smpuAip.Upojq /y 00/.£a (HN)N3
suapuiiiw '/y (HN)lAa

suapui/nj iM yo8£a (HN)lMa
tpjow Pi ^fOc^ (HN)PMa
umsoi/pjadns -p/ sso^a (HN)lAia
lunsotipjadns pj itrora (HN)Pia
lunsoqpjadns pi igOca (HNlWa
umsoijpjadm p/ gsOc^ (HN)JMa
lunsotpDjadns pi 0£0£a (HN)Wa
lunsoqpjadns -pi ygoca (HN)Wa
ipjow Pi yyoca (HN)Wa
Hajow Pi y-soca (HN)Wa
tspaai Pi oysey (HN)IAia
ispaai Pi 668ea (HN)Pia
aAav) Pi i£0ca (HN)Wa
lunsofipjadm p[ £9/.ya (HN)Wa
a^avi -pi t7ioea (HN)Pia
a^aoi Pi 5 1 oca (HN)Wa

'>0 m

rj rr^

—

00 SO
<N — Tt
— fN

00 — <

— —

•— O >/T CN ON
ON NO CN — 00

n —

oOTj-m'^rj-
r-- (N 00
ON —

C3 _ O

— CQ
cd .9

t a

B V

c s

B V

C V
>

s i

I

c ^ ^ ^

a js o -c

B w) "Sb c ob •

o c c c -

U S ^

< CQ J

+

<

o o .'d

"B "B ^

M ti y5

C — V5 C C fl3
^ "O —

&< <J

S c

ca

W U< ' O X

width between outer angles of pre-

frontals 114 90* 111 109 138 146 148 130 107 150 130 117 114 114 255 166 213 224

width between outer angles of quadrates 133 180 178 185 185 156 135 254 247

mandible length 492 720 635 720 560 670 823

length of symphysis 360 270 305 270

Table 1 {Cent.)

ADAMS-TRESMAN: CALLOVIAN CROCODILE M ET RIO RH Y N C HU S

183

OO

On

O

ds ff

I86A WH

—

—

*

■ds n

8/.6A lAlH

O

ON

ON

rj

o

o

oo

ON

00

NO

o

On

ON

’d"

00

O

O'

NO

in

rn

r-4

—

\ ds 7Y

SZ.6A IMH

On

O

NO

''d'

rn

00

*

*

■ds -/V

e/.6A WH

ON

r-

oo

NO

o

C4

r-

NO

o-

o

rn

oo

00

ijdjolu 7Y

I/.6A WH

ON

oo

On

in

00

rf

oo

00

m

»n

Tj-

O'

O'

oo

NO

oo

00

NO

rn

ni

—

*

♦

ds ‘/Y

996A IMH

00

o

rn

00

o

»n

00

nj

NO

ON

NO

r 1
ri

o

ON

rn

NO

NO

rn

—

“

*

■»

ds /V

S96A WH

r-

r-

—

’d-

Tf

GO

OO

ON

m

P!

NO

rn

C4

—H

umsoiiiDJddns /y

L96A IMH

r-

o

NO

ON

m

in

O

ON

O

04

O'

oo

m

04

O'

«r»

m

n

umsoijpAddns /y

e96A lAIH

NO

rn

m

ON

OO

OO

NO

''d-

o

«

■ds ■/v

i:96A WH

NO

NO

NO

NO

o

NO

ON

NO

O

f*)

rn

r4

ds /-V

3 WS

ON

m

r-

nj

o

NO

r-

m

00

m

■*»

rn

“

m

ds ■/v

a IMS

00

m

ri

r-

On

in

O'

'd*

On

00

■ds iM

V IMS

r-

NO

ON

r4

NO

o

fN

00

Tt

rn

00

o

ON

On

rn

04

O'

m

NO

Tt

—

—

ds IM

6L\D AVIMN

600*

638*

o

r-

m

oo

00

r*-

O

04

O'

O'

O'

•"d-

NO

o

*

*

lunsoijpjsdns ■/y

oi:a diMD

NO

o

m

r"

n

rj

r*-

On

r-

On

00

oo

O'

oo

04

ON

00

fS

UD

in

m

rN

—

uinsoijpjadns jaj

8ia dIM3

in

ON

ON

t"

rn

m

ON

O'

oo

rn

r4

umsoi]i.JJ3dns

08Id dIM3

00

•n

00

*n

r-

NO

oo

■'d’

On

3SU3,\uqomp y/ 8I9cy (HN)lAia

425

00

m

ra

o

00

NO

rvj

oo

Tf

04

O'

ON

•'d*

asudAuqomp yj aeOca (HN)IMa

m

C4

r-

00

04

On

04

““

04

3su3,\uqomp y/ [jeea (HN)Wa

M

rn

On

O

oo

NO

On

ON

ON

04

04

U, Cu

•:5 a rz:

M U S

H ^

5 B
c o

.£ ^ ’

S <

ftO W) :

< CQ
+
<

CC

u

o S
& n

0 s

c &x>

ou ^

c E

D. 5

E c

?? ffl

ex)-s

c c —

^ '-B

o o .s!

•5 -5 ^
^ ob to
c G cd

lU <u 4>

g ^ -

,c o js ^ -5

^ i g S'
s s s "

UJ U. 'OK

184

PALAEONTOLOGY, VOLUME 30

c

o

U

u

aa

<

H

wnsoijpj^dns py 3(7^ dPV3
lunsoiipjsdns ■ py /.I'ddJMO

wnsoiipjadns ■ py 01>l dlAID
lunsoijpuadns py gy dWD
isp33i py t^3-a dIMO
ssu3Auqojnp py 6Id dPID
snqjuAqj,\q3Vjq py 9/.I-adIAlD

l/86mi JMH
■ds py 000 1 A PiH
ds py (700 1 A NH
ipjow /V fl7llA WH
ds py Zt-IIAPIH
lunsoqpjadns py Ot"!! A NH
wnsoqpuadns py SIOIA PIH
ds py 866A WH
lunsoqpjadns py 686A PJH
ds py /.86A WH
lumoqpjadns py sg6A NH
ds /v eS6APlH
ds py 386A PIH

—

O 0^ ^

m m \o —

« fy, ^

ro —

'sO -o
O ‘O
(N fN

(N
— (N

00 'C' — o

<N -H

00 Tj* <N — Tt
On 00 CN <N
r4 —

^ _ O ««

s ■?

•p E

S '£

Oil oc

P ^

CJ

< PQ

' 3 ^ ^

2 1 's '

o O. o

■S ^-5 £ ^
00-- 00 <u ,2
c ^ C 2

S S -6 ,

S 5
>

•o .

O O .ii

■S ^ j=
00 00 to

C C ?3 ^

II

<i: T3

W u, I o X

<

mandible length 636

length of symphysis 294

ADAMS-TRESMAN; CALLOVIAN CROCODILE M ETRIO RH YNC HU S

185

130 ■

120 •

I 10

— 100 •

E

E

w 90 -

80 -

70 -

■ M. laeve

# M. moreli

A M. superciliosum
0 M. cultridens
O M. brachyrhynchus
□ M. durobrivense
A M. leedsi

• M. sp.

• 10

38,

• 20^* /1:2067 «2|9,

340 4|«45 0

^7

V983

.44

^I8

2048

36

.40

V942

■^A2I

2036

^32

i33

^29

^V993
□ 35

^26

^R8

^30

a23

■5 60 -

- 50 -

40 -

J529

48

2065

^24

-| I I I I I I I I

50 60 70 80 90 lOO 110 120 130

length of frontal anterior to supratemporal fenestrae (mm)

TEXT-FIG. 2. Bivariate plot of least width between orbits and length of frontal anterior to supratemporal
fenestrae. Specimens identified by computer number (see Table 1) or, when no computer number exists, by
registration number, as follows: V, Hunterian Museum, University of Glasgow; NMW, National Museum of
Wales, Cardiff; four digit numbers without prefix, British Museum (Natural History), London. Same species
symbols and specimen numbering are used in text-figs. 3-7.

Although a division into two groups seems reasonable on the basis of these characters (the groups
correspond to Andrews’s ‘broad’ and ‘narrow’ skulled species), there is no evidence for further
taxonomic subdivision.

The two features compared in character F by Andrews are not independent. The length of the
frontal bone anterior to the supratemporal fenestrae and the width of this bone between the
orbits together characterize the development of the frontal bone. Furthermore, the degree of this
development might be expected to vary between individuals.

A further range of six character combinations for the two groups suggested from text-fig. 3a
and B were then examined by ranked ratio and percentage values (Table 2). Two groups are again
apparent, defined on the basis of characters 2 and 6, where Group 1 includes M. leedsi, M.
superciliosum, M. moreli, and M. laeve, and Group 2 includes M. cultridens, M. brachyrhynchus,
and M. durobrivense. All other characters measured (see Table 2) show greater ‘within group’ than
‘between group’ variation. Thus, characters 2 and 6 may be valid specific characters. Together they
quantify an aspect of snout growth, character 2 being a measure of the ‘width component’ of growth
and character 6 an indication of the amount of anteroposterior elongation in relation to broadening
that occurred during growth.

Principal coordinate analysis

The quantitative analysis of cranial characters was then extended by the use of principal coordinate
analysis, a multivariate technique. In this method, each specimen is defined by a particular array of

186

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 3. Bivariate plots showing a, skull length/width of skull between outer angles of prefrontals against
skull length, and b, distance between nasals and premaxillae as a percentage of skull length against total skull

length.

TABLE 2. Ranges in Groups 1 and 2 (see text-fig. 4) of ranked statistics derived from cranial measurements of

Metriorhynchus.

Character

Character

number

Range in Group 1

Range in Group 2

Ratio of width of skull (outer
angles of quadrates) to length
of skull

1

2-80-3-55: 1

2-83-2-90: 1

Ratio of width between outer
angles of prefrontals to length
of skull

2

3-97-5 00: 1

2-83-3-85: 1

Distance between premaxillae and
nasals as % of total length of
skull

3

7'91-22-18%( = 14-27%)

0-00-7-85 %( = 7-85%)

Length of nasals as % of total
length of skull

4

Shows continuous range of over-
lap between Groups 1 and 2

Shows continuous range of over-
lap between Groups 1 and 2

Length of mandibular symphysis
as % of length of mandible

5

Shows continuous range of over-
lap between Groups 1 and 2

Shows continuous range of over-
lap between Groups 1 and 2

Distance between nasals and pre-
maxillae as % of length of na-
sals.

6

19-53-7L73%(= 52-21 %)

0-00-23-01 %( = 23-01 %)

ADAMS-TRESMAN: CALLOVIAN CROCODILE M ET RIO RH Y N C HU S

187

values (the measured cranial characters). The vectors describing the collection of metriorhynchid
specimens can be represented by points in //-dimensional space, where n equals the number of
metriorhynchids in the sample minus 1 . The distance between any pair of specimens represents their
similarity in terms of all the characters measured. The metric expression of these measures of
similarity is the similarity coefficient, and coefficients expressing the relationships between every
pair of metriorhynchids were combined together to form a similarity matrix. Projecting the coordi-
nates of the vectors on to the first and second principal coordinates or axes (eigenvectors) highlights
the morphological differences that exist between members of the sample, since the first and second
axes maximize the variance of the measured characters. The results of the multivariate analysis are
shown in text-fig. 4, which illustrates a reasonably convincing clustering into three groups (the
relationship of specimens 28 and 34 to Groups 2 and 3 is problematical. These spatial relationships
must be a representation of the cranial characters possessed by the specimens.

TEXT-FIG. 4. Principal coordinate analysis plot showing spatial relationships on the first and second
axes, and similarity values. Specimens denoted by their computer numbers (see Table 1).

Text-fig. 4 shows that a large percentage of Group 2 are linked internally by similarity coefficient
values greater than 950 (out of 1000). Marginally lower similarity coefficients link specimens 49,
18, 46, 44, and 42 with each other; these individuals appear to represent a small divergence from
the character states possessed by the specimens clustered around the ‘norm’ for this group. Values
between 900 and 950 occur between most of the specimens within Group 3 and, with a similar
frequency, between end members of Groups 3 and 2. Group 1 shows an apparently lower level of
similarity between its members. This is because there are far fewer individuals to compare with
each other in this group, with the resultant smaller likelihood of attaining high levels of similarity
(i.e. there is less chance for specimens to approximate to the mean for a given character). The levels
of similarity between each specimen and the six ‘most similar’ specimens to them are shown in the
nearest neighbour score charts (deposited with the British Library, see above).

SECOND

188

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 6. An indication of skull ‘shape’ using a plot of width index against ranked lengths.
Specimens denoted by their computer numbers (see Table 1).

ADAMS-TRESMAN: CALLOVIAN CROCODILE M ET RIO RH YNC H US

189

The characters which seem to be most significant in separating Groups 1 from Groups 2 and 3
are width of the skull and, to some extent, length of the skull. Text-fig. 5 shows that the most
significant factor separating Groups 2 from Group 3 is length of the skull. These apparently decisive
characters of width and length of the skull were examined by calculating a ‘width index’ for each
specimen: W/L, where W is the width between the outer angles of the prefrontals, and L is the total
length in the mid dorsal line. Where possible, L was estimated for incomplete skulls.

Width index is plotted against the rank values (from text-fig. 5) in text-fig. 6 which confirms that
skull width is independent of skull length, i.e. broad and narrow skulls occur at both ends of the
ranked series, despite the fact that the majority of broad-skulled specimens available for analysis
and which plot in Group 1 (text-fig. 4) are large specimens. The grouping between and within
Groups 2 and 3 (text-fig. 4) is therefore a function of size (length), not shape (width). This explains
why the similarity values of the links between many of the end members of these two groups are as
high as 940-949. Although it is reasonable to assume that size is an indication of individual growth,
there is no evidence from skull proportions or suture closure (Mook 1921) to suggest a juvenile
status for any of the smaller skulls. Thus, Groups 2 and 3 include metriorhynchids of variable size,
but with skulls of similar widths.

Text-fig. 6 identifies specimens 28 and 34 (M. durohrivense) as ‘broad-skulled’ metriorhynchids.
Their relationship to the other specimens in Group 1 is complicated by the fact that they are
significantly smaller. Since it has already been established that text-fig. 4 measures skull size (length)
and skull width, it follows that a marked difference in either character (in this case skull length) will
increase the distances between 28 and 34 and Group 1, because distance is, in this sense, a measure
of morphological difference.

TAXONOMIC VALIDITY OF CHARACTERS

If skull width is truly independent of skull size, and not merely an allometric index in the larger
skulls, it is a base on which a classification could be erected. This is in partial agreement with
Andrews (1913), who subdivided his narrow-skulled and broad-skulled groups:

Narrow-skulled metriorhynchids

Andrews (1913) used four characters in his subdivision:

1. Relationship between length of frontal bone anterior to supratemporal fenestrae and its least
width between orbits. Text-fig. 2 shows that the amount of variance of this index between individuals
said to belong to the same species is considerable. Andrews justified using this character on the
basis of the sample illustrated in text-fig. 7, but it is now clear that these individuals account for
only a small percentage of the total variation.

2. Relationship between frontal and prefrontal bones. The degree of individual variation shown in
text-fig. 8 indicates that this feature should not be considered a valid taxonomic character.

TEXT-FIG. 7. Bivariate plot of least width between orbits and length of
frontal anterior to supratemporal fenestrae in specimens of Metriorhyn-
chus superciliosum and M. moreli analyzed by Andrews (1913).

length of frontal anterior to
supratemporal fenestrae (mm)

190

PALAEONTOLOGY, VOLUME 30

3. Ornament of bones of dermal skull roof. Andrews devised a classification of the ‘narrow-skulled’
metriorhynchids based on whether the roofing bones were ornamented or smooth. Wenz (1968,
p. 31) stated, ‘It seems right that among the long muzzle forms we can recognize ornamented forms
(M. superciliosum and M. moreli) and forms without ornament (M. laeve and M. leedsi) but to push
the specific determination further seems hardly satisfactory ... in fact on most of the specimens
examined the variations in ornamentation are minimal . . . moreover there exists a range of
intermediaries between the two extremes’. My examination of narrow-skulled metriorhynchids,
however, shows that there is continuous variation between ‘smooth’ and ‘faintly ornamented’ skulls
(see particularly specimens CMP R7 and R72), and those showing dense ornamentation. It is
impossible to subdivide this spectrum in any taxonomically useful way.

4. Number of teeth. I have found considerable variation in overall numbers of teeth between
specimens assigned to the same species by Andrews (1913) and subsequent workers. Variation also
occurs between the right and left sides or upper and lower jaw of the same individual. Similar
variation was recorded by Wenz (1968).

Kalin (1933) concluded that tooth number varied by only one or two within the various species
of living crocodilians which he examined. It is also clear, however, that tooth number was never the
sole characteristic used to recognize a species — it always occurred in conjunction with morphological
differences in the skull.

There are no differences in tooth form among the Callovian metriorhynchids. All specimens have
longitudinally ridged dental enamel with anterior and posterior carinae and pointed crowns. The
numbers of teeth are: M. superciliosum, 28-30; M. moreli, c. 26; M. laeve, c. 30; and M. leedsi, 30-
36. These variations are in excess of those recorded by Kalin (1933), but it has been shown above
that there are no convincing additional differences in either skull form or ornament of the skull
bones that can be used to separate these specimens any further.

Broad-skulled metriorhynchids

Andrews (1913) used five characters in his subdivision:

1. Degree of separation of nasals and premaxillae. Text-fig. 3b indicates that specimens with the
broadest snouts (defined by the distance across the outer angles of the prefrontals) show the least
distance between nasals and premaxillae. The separation of these bones is an expression of the
degree of antero-posterior growth contrasted with widthways expansion of the rostrum. In the
widest skulls, growth has been apparently more concentrated in the expansion of the skull, with
consequently less elongation, and since the expression of antero-posterior elongation is the extension
of the maxillae, this greater expansion is signified by the reduced distance between the nasals and
premaxillae.

Within the broad-skulled metriorhynchids, however, the distance separating the nasals and pre-
maxillae is too variable to be taxonomically useful (see text-fig. 4) and the group does not conform
to the relationships expressed by Andrews (1913).

2. Number of teeth. The variation is as follows: M. cultridens, 20-21; M. brachyrhynchus, 18-21;
M. durobrivense, 16; and M. casamequalai, 21. There is no significant or systematic variation in
tooth form between these species. The dental enamel is raised into extremely fine longitudinal ridges
with smooth pointed crowns. There are, in all cases, fewer but larger teeth in the jaws of the broad-
skulled as opposed to the narrow-skulled crocodilians; Mcllheny (1976) demonstrated this feature
in living crocodilians. It is the result of the reduced development of the maxillae which are the main
tooth-bearing elements.

If one argues that all the broad-skulled specimens should be classified as one species, the differ-
ences in tooth number are larger than that seen in living crocodilians (Kalin 1933), but there are no
other significant morphological differences of the skull which vary together with tooth number (see
below) to further subdivide the broad-skulled group. Such a subdivision would be based on tooth
number alone.

3. Length of rostrum. Assessment of the taxonomic validity of this feature is difficult because the
relevant part of the skull is preserved in only a few specimens. The available data suggest that it is

ADAMS-TRESMAN: CALLOVIAN CROCODILE M ET RIO RH Y NC H US

191

variable within species (Table 1)— as it is in the narrow-skulled metriorhynchids, for which there is
a more comprehensive sample size (Table 1). Also, this character is not independent: it is determined
by the amount of antero-posterior elongation, as expressed mainly in the length of the maxillae
which, in turn, determines the separation of the nasals and premaxillae— and, indeed, the number
of teeth (character 2).

4. Relationship between frontal and prefrontal hones. Frontal/prefrontal relationships are shown
in text-fig. 8, which illustrates their variability and, hence, unsuitability for taxonomic use.

TEXT-FIG. 8. Variation in form of frontal/nasal suture. Distance sepa-
rating the most anterior point of projection of frontal from pre-
frontals is indicated, a-g, Metriorhynchus superciliosum; A, R2053
(27 mm); b, R1529 (19 mm); c, R2051 (2 mm); d, R6859 (38 mm); e,
R1665 (10 mm); f, R2041; G, R2036 (12 mm); h, R1666 (22 mm), i,
M. moreli, R2044 (31 mm), j-l, M. laeve\ j, R2031 (1 1 mm); k, R4762
(18 mm); l, R3015 (22 mm), m, M. leedsi, R3899 (6 mm), n, M. leedsi,
R3540 (19 mm), o, M. cultridens, R3541 (28 mm), p, M. durobrivense,
R26I8 (24 mm), q, M. durobrivense, R3321 (18 mm), r, M. brach-
yrhynchus, R3700 (44 mm), s, M. brachyrhynclms, R3999 (25 mm).

All specimens in BM(NH).

5. Ornament of dorsal skull bones. Andrews attached relatively minor importance to this character
in his type descriptions of M. cultridens, M. brachyrhynclms, and M. durobrivense, he concluded
that its development varied continuously between the three broad-skulled Metriorhynchus species
he recognized, between individual specimens, and even between different cranial bones for a given
specimen.

The ornament on BM(NH) R3804 (M. cultridens) is of particular interest; it does not appear to
be as strongly developed as Andrews concluded (1913, p. 196) and represents the least degree of
development of ornament recorded to date in a European specimen. In this context one of the
characters used by de Gasparini and Diaz (1977, p. 357) to define their new species M. casamiquelai
needs to be re-examined: ‘There is no ornamentation (all species in group (b) Wenz (1968) have
different degrees of ornamentation)’.

192

PALAEONTOLOGY, VOLUME 30

M. casamiquelai

The type specimen of M. casamiquelai is in South America, and comparisons here are based upon
personal communication with de Gasparini about aspects of morphology.

1. Ornament. There is a gradational development of ornament between M. casamiquelai and M.
cultridens (R3804); both show lateral pitting of the maxillae and faint striae on the nasals. Unfortu-
nately a portion of the frontal and prefrontal of M. casamiquelai has been restored, thereby reducing
the amount of morphological detail which can be retrieved from this vital area of the skull. M.
casamiquelai appears to support the concept of a continuous range of variation of ornament and
provides an example of the very beginnings of the development of that ornament (as do CMP R72
and R7 for the narrow-skulled metriorhynchids).

2. Cranial sutures. One other character used by de Gasparini and Diaz (1977, p. 357) merits close
examination: That the suture separating the premaxillae and maxillae has the shape of an “M”, all
known species have the shape of a “V” ’. The known variability of other cranial sutures suggests
that this character is not taxonomically useful. My work has revealed that although some specimens
do have an ‘M’-shaped suture (text-fig. 9: a, ‘broad-skulled’; b, c, ‘narrow-skulled’), the ‘V’-shaped
suture occurs very much more frequently.

All of the characters used by de Gasparini and Diaz (1977) to define M. casamiquelai are
taxonomically invalid because they exhibit a wide and continuous range of variation.

TEXT-FIG. 9. Form of premaxillary/maxillary suture, a, M. brachyrhyn-
chus, R3699. b, M. moreli, R6860. c, M. superciliosum, R2058. All speci-
mens in BM(NH).

INTERPRETATION OE SKULL WIDTH DIMORPHISM

There are two feasible interpretations of the facts presented above. First, that two species are
represented by the specimens studied, corresponding to the ‘broad-skulled’ and ‘narrow-skulled’
groups, and that the majority of ‘broad-skulled’ specimens occupy the top end of the spectrum of
skull sizes (lengths). Niche partitioning between these two species would then be a function of their
different dietary requirements which, in the fossil specimens, are reflected by a progressive disparity
in skull and jaw size and breadth (text-fig. 6; highlighted by specimens 38 and 24 in relation to 35,
29, 30, and 27). The resulting contrast in jaw mechanics would facilitate the taking of a different
range of prey (see Cott 1961, table 19 and fig. 34, for different categories of food taken by crocodiles
in relation to their size).

The second interpretation is that the ‘broad-skulled’ and ‘narrow-skulled’ dimorphism is due to
intraspecific variation in skull shape and proportions. Sexual dimorphism in crocodiles has been
reported by Cott (1961, p. 252, fig. 16), Neill (1971, pp. 243, 250, fig. 85), and Mcllheny (1976,
p. 64); their records show that males are, in general noticeably larger than females. This might also
be expected to apply to fossil crocodiles, as measured by skull size (Greer 1974). If so, female
metriorhynchids would correspond to those specimens which plot on the bottom and mid-left-hand
area of text-figs. 4 and 5 (Group 3), while males would correspond to the remainder (Groups 1 and
2); the line between males and females would, of course, be somewhat ill-defined. The explanation

ADAMS-TRESMAN: CALLOVIAN CROCODILE METRIO RH YNC HUS

193

of the broad-skulled group in this instance would be that some crocodiles (presumably males?), on
reaching a certain skull length, broaden out rather than elongate. Neill (1971, p. 180) documented
this feature for living crocodilians, and the evidence in the literature always refers to its occurrence
in males (but with no statement that it is exclusive to this sex).

SYSTEMATIC PALAEONTOLOGY

Using the principal criterion of skull width, two species of Metriorhynchus have been defined, in which the
contrasted growth profiles required to produce the segregation are reflected in characteristic dimensions of
and interrelationships between the dermal roofing bones (principally the prefrontals, nasals, premaxillae, and
maxillae) and tooth number.

Genus metriorhynchus von Meyer, 1830, emend. E. E. Deslongchamps, 1867
Type species. Metriorhynchus geoffroyi von Meyer, 1 832.

Metriorhynchus superciliosus de Blainville, 1853, emend. E. E. Deslongchamps, 1867

1853 Metriorhynchus superciliosum de Blainville, MS Conybeare, figured by Deslongchamps, 1867,
pi. 20, fig. 2a, b, c.

1867 Metriorhynchus moreli E. E. Deslongchamps, p. 320, pi. 21, figs. 4 and 5; pi. 22, figs. 1 and 2.
1913 Metriorhynchus leedsi Andrews, p. 178, text-fig. 73a.

1913 Metriorhynchus laeve Andrews, p. 178, text-fig. 73b.

Type data. Location of type specimen from the collection of de Blainville is unclear in Deslongchamps (1867,
pp. 306-319, pi. 20, fig. 2a, b, c) and subsequent publications.

Diagnosis. Narrow skull (where ratio of width between outer angles of prefrontals and length
of skull falls approximately within 3-97-5-00: 1) with many close-pointed teeth (c. 26-28, but
exceptionally more than 30 in each maxilla). Frontal and prefrontal bones show extensive range in
development of sculpture and in relationship between their most anterior points of projection.
Nasal/frontal suture displays a variety of forms. Development of frontal bone, as illustrated by
relationship between its width between orbits and its length in front of supratemporal fenestrae, is
individually variable. Nasal and premaxillary bones separated to a variable degree, but never in
contact.

Metriorhynchus brachyrhynchus E. E. Deslongchamps, 1867

1867 Metriorhynchus brachyrhynchus E. E. Deslongchamps, p. 333, pi. 23, figs. a-d.

1890 Suchodus durobrivense Lydekker, p. 285, figs. 2 and 3.

1913 Metriorhynchus durobrivense Andrews, p. 179, text-fig. 73c.

1913 Metriorhynchus cultridens Andrews, p. 179, text-fig. 73e.

1977 Metriorhynchus casamiquelai de Gasparini and Diaz, p. 426, no figure.

Type data. The type material figured by Deslongchamps (1867, p. 333, pi. 23, figs, a-d) was destroyed in the
Second World War.

Diagnosis. Broad skull (where ratio of width between outer angles of prefrontals and length of skull
falls approximately within 2-83-3-85: 1) with fewer teeth than M. superciliosus (16-20 in each
maxilla) but with similar pointed crowns. Varying amounts of sculpture on frontal and prefrontal
bones, whose relationship between their most anterior point of projection also varies. Nasals and
premaxillae in contact or separated by up to 33 mm (but always less than in M. superciliosus).

Acknowledgements. The initial research for this paper was carried out during tenure of a NERC grant which
is gratefully acknowledged. Facilities were provided by the Department of Palaeontology, British Museum
(Natural History). I thank Professor C. B. Cox for critically reading an earlier draft of the manuscript.

194

PALAEONTOLOGY, VOLUME 30

REFERENCES

ANDREWS, c. w. 1913. A descriptive catalogue of the marine reptiles of the Oxford Clay, 2, 206 pp. British
Museum (Natural History), London.

BUFFETAUT, E. 1977. Sur un Crocodilien marin, M. superciliosus de I’Oxfordien superior (Rauracien) de Tile
de Re (Charente-Maritime). Annls Soc. Sci. nat. Charente-Marit. 6 (4), 252-266.

COTT, M. B. 1961. Scientific results of an enquiry into the ecology and economic status of the Nile Crocodile
(Crocodilus niloticus) in Uganda and Northern Rhodesia. Trans, zool. Soc. Lond. 29, 21 1-356.

DE BLAINVILLE, H. M. 1853. Lettres sur les Crocodiles vivants et fossiles. Mem. Soc. linn. Normandie, 9, 109-

120.

DE GASPARiNi, z. B. and DIAZ, c. 1977. Metriorhynchus casamiquelai n. sp. (Crocodilia, Thalattosuchia), a
marine crocodile from the Jurassic (Callovian) of Chile, South America. Neues Jb. Geol. Paldont. Mh. 153,
341-360.

DESLONGCHAMPS, E. E. 1863-1869. Notes Paleontologiques, 392 pp. Caen and Paris.

DODSON, p. 1975. Functional and ecological significance of relative growth in Alligator. J. Zool., Lond. 175,

FRAAS, E. 1902. Die Meer-Krocodilier (Thalattosuchia) des oberen Jura unter specieller Berucksichtigung von
Dacosaurus und Geosaurus. Palaeontographica, 49, 1 -72.

GREER, A. E. 1974. On the maximum total length of the Salt Water Crocodile {Crocodvlus porosus). J. Herpet.
8,381-384.

K.ALIN, J. 1933. Beitrage zur vergleichenden Osteologie des Crocodilidenschiidels. Zool. Jb. Abt. Anat. 57, 535-
714.

1955. Crocodilia. In piveteau, j. (ed.). Traite de paleontologie, 5, 695-784. Masson, Paris.

LYDEKKER, R. 1890. On a crocodilian jaw from the Oxford Clay of Peterborough. Q. Jl geol. Soc. Lond. 46,
284-288.

MciLHENNY, c. A. 1976. The alligator's life history, 117 pp. The Society for the Study of Amphibians and
Reptiles.

MOOK, c. c. 1921. Individual and age variations in the skulls of recent Crocodilia. Bull. Am. Mus. nat. Hist.
44, 51-66.

NEILL, w. T. 1971. The last of the Ruling Reptiles: alligators, crocodiles and their kin, 486 pp. Columbia
University Press, New York.

NEWELL, N. D. 1956. Fossil populations. In sylvester-bradley, p. c. (ed.). The species concept in palaeon-
tology. Pubis Syst. Ass. 2, 63-82.

SCHMIDT, w. E. 1904. Uber Metriorhynchus jaekeli nov. sp. Z. dt. geol. Ges. 56, 97-108.

WENZ, s. 1968. Contribution a I’etude du genre Metriorhynchus. Crane et moulage endocranien de Metriorhyn-
chus superciliosus. Annls Paleont. (Vert.), 54, 149-183.

1970. Sur un Metriorhynchus a museau court du Callovien des Vaches Noires (Calvados). Bull. Soc. geol.

Fr. 12, 390-397.

315-355.

SUSAN M. ADAMS-TRESMAN

Open University (East Anglian Region)
Cintra House, 12 Hills Road
Cambridge

Typescript received 28 February 1985
Revised typescript received 13 June 1986

Home address:
‘Lowood’, Grove Road
Cranleigh, Surrey GU6 7LH

THE CALLOVIAN (MIDDLE JURASSIC)
TELEOSAURID MARINE CROCODILES FROM
CENTRAL ENGLAND

by SUSAN M. ADAMS-TRESMAN

Abstract. For many years the taxonomy of the Callovian marine crocodiles of the genera Steneosawus and
Mycterosuchus has been in a state of confusion. Bivariate and principal coordinate analyses are used in an
attempt to identify cranial characters for discriminating species. Many of the characters used previously to
define five species of Steneosaunis and one of Mycterosuchus are shown to be individually variable. Only two
Callovian species can now be identified on the basis of their skull proportions and numbers and form of their
teeth: S. leedsi incorporates specimens previously assigned to S. leedsi Andrews, 1909, and Mycterosuchus
nasutus Andrews, 1913; S. durohrivensis includes S. durobrivemis Andrews, 1909, S. obtusidens Andrews, 1909,
S. hulkei Andrews, 1913, and S. depressus Phizackerley, 1951.

The current classification of the Callovian marine crocodiles of the genera Steneosaunis and
Mycterosuchus was established largely by E. E. Deslongchamps (1863-1869) and Andrews (1909,
1913). Deslongchamps gave the first detailed descriptions and figures of Steneosaunis, emending
Geoffroy Saint-Hilaire’s (1825) generic description, creating many new species for the material
collected around Caen in Normandy and emending other specific diagnoses. He established the
type specimen of Steneosaunis as that figured by Cuvier (1824, pi. 8) and named S. megistorhynchus
by Geoffroy Saint-Hilaire (1825); he also emended the diagnosis of the species (1869, p. 217,
pi. 15).

Andrews (1909, 1913) worked with crocodilian material of the Leeds Collection housed in
the British Museum (Natural History). These mesosuchian crocodiles occur in the jason and
coronatum zones of the Oxford Clay, and were found close to Peterborough, Cambridgeshire.
Andrews (1913) described a range of additional material, extended his earlier descriptions (Andrews,
1909), and erected a new species S. hulkei. He established the new genus Mycterosuchus in 1913 to
accommodate the specimen described in 1909 as S. nasutus. In distinguishing between the genera
Steneosaurus and Mycterosuchus Andrews (1913) used the following cranial characters possessed
by Mycterosuchus:

(i) greatly elongated snout, sharply marked off from the cranial region of the skull;

(ii) slender teeth;

(hi) temporal fossae relatively smaller and shorter than in the typical steneosaur.

The characters he used to determine the four species of Steneosaurus from the Oxford Clay were as
follows:

S. leedsi— grzai length and slenderness of rostrum (i.e. preorbital length 73 % of total skull length); mandibu-
lar symphysis 58 % of mandible length; great distance between nasals and premaxillae; number and form of
teeth.

S. short rostrum (preorbital length less than 60 % of total skull length); nasals separated from

premaxillae by a shorter distance than in S. durohrivensis: anterior angle of frontal blunt and far behind
anterior angle of prefrontals; mandibular symphysis 40 % of mandible length; large teeth; scutes shallow,
transversely elongate and bearing widely separated pits.

S. durohrivensis— ros>ir\xm of moderate length (preorbital length 61 % of total skull length); very large
temporal fenestrae; frontals terminate in very obtuse angle a little way in front of anterior rim of orbits;
mandibular symphysis 44 % of mandible length; number and form of teeth; scutes large with shallow pits but
no tendency of elongation.

S. obtusidens— short rostrum (preorbital length 52 % of total skull length); species distinguished mainly by
form and number of teeth, and by relationship of frontal and prefrontal bones.

I Palaeontology, Vol. 30, Part 1, 1987, pp. 195-206.|

© The Palaeontological Association

196

PALAEONTOLOGY, VOLUME 30

Thus, in the division of this genus, Andrews used the following criteria:

A, length of preorbital region of skull and its percentage of the skull as a whole;

B, degree of robustness or massiveness of cranial features, particularly the snout;

C, form and relationship of frontal and prefrontal bones;

D, separation of premaxillae and nasals;

E, length of mandibular symphysis;

F, tooth form and number;

G, scute form.

In 1951 Phizackerley diagnosed a further Callovian steneosaur species, S. depressus, on the basis of
the ‘delicate construction of the skull and narrow alveolar region’, with rostrum 64 % of total skull
length, and mandibular symphysis 48 % of mandible length.

These generic and specific diagnoses were all based on the typological species concept; each
crocodile species was viewed as being virtually invariable, so that small morphological variations
were considered to have specific significance. Most of the current taxonomic problems within
Callovian steneosaurs have arisen from this methodology, which was applied uncritically by sub-
sequent workers, such as Phizackerley (1951) and Mateer (1974).

MATERIAL AND METHODS

Mook (1921) was the first author to identify types of osteological variation seen in living crocodiles which
had been previously interpreted as having taxonomic significance in fossil forms. Mook’s study was an attempt
to determine the ‘value’, in taxonomic terms, of the variations observed. He identified two sources of variation:
age and individual— e.g. proportional relations of length and breadth of the skull, of preorbital and postorbital
regions, shape of snout, size and number of teeth, variation in the form of certain sutures, and ornamentation
of cranial bones.

A variety of cranial characters was measured, including those which had been used to diagnose species by

TEXT-FIG. 1. The pattern of dorsal cranial bones in Steneosaurus, defining the measurements taken during the
morphological analysis (see Table 1). Abbreviations: f, frontal; j, jugal; 1, lachrymal; mx, maxilla; n, nasals;
p, parietal; pmx, premaxilla; pof, post frontal; prf, prefrontal; q, quadrate; qj, quadrato jugal; sq, squamosal;
A, total length of skull (occipital condyle to tip of snout); B, preorbital length; C, width between outer angles
of quadrates; D, length of supratemporal fenestra; E, width of supratemporal fenestra; F, width at anterior
end of nasals; G, width at anterior rim of orbits; H, distance between nasals and premaxillae; I, long diameter

of orbit; J, transverse diameter of orbit.

TABLE 1. Cranial measurements in Steneosaurus (see text-fig. 1). Abbreviations: BM(NH) British Museum (Natural History); CMP,
City Museum, Peterborough; HM, Hunterian Museum, Glasgow; OUM, Oxford University Museum; SM Sedgwick Museum, Cam-
bridge. OUM specimens J1401, J1403, J29850, J29851 are Bathonian steneosaurs which were not included in the computer analysis.
Computer numbers were allocated to those specimens sufficiently complete for computer analysis. All measurements in millimetres

(* denotes estimated value).

ADAMS-TRESMAN: CALLOVIAN TELEOSAURID CROCODILES

snjnsvu jai \ PMS
suapismqo ’S’ 8Z.1>I dNO
sisudAuqounp 'S’ S/. I'd dWO
ispsais 6/.IddP13
ispaai s OSIUWH
snssdAdap s OZVU IMHO
xuiauaw s 1^8631 IMHO
xuiauaut s 058631 fMflO
uatpmoq s eObir IMHO
uaipmoq S’ lOLir WHO
sninsvu fM (HN)IAIO

suapisniqo s 89I£>I (HNlWH
sisuaAuqounp s g983d (HN)Ne
sisuaAuqomp s I03£d (HNlIMH
;ay/«¥ S L303>I (HN)fVa
6193d (HN)lAia
ispaai s 908£d (HN)lAia
jspaaj s 03eed (HN)Wa

rn C'J

O

O

ITi ^

00 oo
os r-

—

r- m —

r-. in

— »n

rj —

00 O so
— ' so rn

rn so

m so O

in »n o so ^

C4 00 Tf

O 00

o
oo in

<^i —

in o

— < CM

00

SO

o o
o cn
o

»n O O
m o r-4
Tt — < so

in os
m 00
Os »/A

o o
r-- 00
rn in

(N O
OS sn r^
— 00 rn

O O
O i/~i

so rn

00 O Os

r4 00 in Tj-
— r- nj

O —
OS rn

m m
in >
oo so

r-4 o 00
r- so o
(N Os m

o o

^ Os
OO m

■rfoomr-
so Os ON »n rn
00 ICS

T3

C

cd

O 3

-i-J 43 O
cn •«_>
bC C

(L) 0>

■S~ ^

bO cd
o ^

o

a

6

P o

2 S

0) 3

2 ^
00 O
C

V, o
c -c

2 •S

■S ^
c

Cd C/5

c ^

_ o ^

O o'o 8
'22 6 g

o> bJi)
cd c
33 >U

V5 o

0)

D, ° e
E te .2

£ 2

emu

w m O

198

PALAEONTOLOGY, VOLUME 30

Andrews, so that information about dissociated steneosaur material (belonging mainly to the Leeds Collection
but including some important additional specimens) could be synthesized. In this way the sample size available
for analysis was substantially increased (numbers in parentheses) beyond that analysed by Andrews; S. leedsi,
2 (5); S. hulkei, remains at 1; S. durohrivensis, 2 (6); S. ohtusidens, 2 (4); and M. nasutus, 1 (2). The majority of
the taxonomic criteria used in Andrews’s species diagnoses were also quantified, thus facilitating both an
objective critique of the existing taxonomy and an investigation into characters which may have taxonomic
significance.

The cranial measurements taken on specimens of Sleneosaunis and Mycterosuchus are shown in text-fig. 1 .
Mandible length and mandibular symphysis were also measured. The data obtained are shown in Table 1.
Combinations of two or three measurements were plotted and their compatability with the present classifica-
tion tested.

Ranked statistics derived from Table 1, a similarity matrix, and nearest neighbour scores have been tabulated
for all specimens allocated a computer number; these tables have been deposited with the British Library as
Supplementary Publication No. 14030 (5 pages). It may be purchased from the British Library, Lending
Division, Boston Spa, Wetherby, Yorkshire LS23 7BQ, UK. Prepaid coupons for such purposes are held by
many technical and university libraries throughout the world.

CHARACTER ANALYSES

Bivariate plots

The length of the preorbital region of the skull, degree of robustness of the snout, and separation
of nasals and premaxillae in both Steneosaurus and Mycterosuchus are quantified in text-fig. 2a-e.
None of these plots shows groups that might imply the presence of two genera, with the possible
exception of text-fig. 2d. It seems that differences can be detected in the length and robustness of
the snout between specimens known at present as S. leedsi and M. nasutus (Group 1) and the
remainder of the Callovian steneosaurs, S. hulkei, S. durohrivensis, S. ohtusidens, and S. depressus
(Group 2). But this is not compatible with the present recognition of two genera: Mycterosuchus
(one species) and Steneosaurus (five species).

Some additional cranial measurements for the two groups suggested by text-fig. 2d were then
examined by ranked ratio and percentage values as shown in Table 2. The data illustrate the wide
range of overlap of values between species and a high level of individual variation within species.
With the exception of character 1 (Table 2: preorbital region as a percentage of the total length of
the skull), the variation within Groups 1 and 2 is greater than that which exists between them. The
pattern shown in text-fig. 2d only becomes apparent when three, rather than two, cranial characters
are compared simultaneously.

TABLE 2. Ranges in Groups 1 and 2 (see text-fig. 4) of ranked statistics derived from cranial measurements of

Steneosaurus.

Character

Character

number

Range in Group 1

Range in Group 2

Preorbital region as % of total length of skull

1

7L63-74-90%(= 3-27%)

58-33-64-26 %( = 5-93%)

Ratio of width of skull (outer angles of
quadrates) to length of skull

2

4-00-4-50:l

2-72-3-73:1

Length of supratemporal fenestrae as % of total
length of skull

3

13-95 18-64% (= 4-45%)

19-69-25-70 %( = 6-01 %)

Length of supratemporal fenestrae as % of
preorbital length

4

18-61-25-32%(= 6-71 %)

30-64-46-47 %( = 15-83%)

Width of supratemporal fenestrae as % of width
between outer angles of quadrates

5

34-80 42-00 %( = 7-20%)

33-49-37-82 %( = 4-33%)

Width of snout at anterior end of nasals as
% of preorbital length

6

7-28-9-62 %( = 2-34%)

9-43-15-42 %( = 5-99%)

Width of skull opposite anterior rim of orbits
in relation to preorbital length

7

17-79-22-87 % ( = 5-08 %)

22-20-30-57% ( = 8-37%)

Distance between premaxillae and nasals as
% of length of rostrum

8

Marked overlap between
members of Groups 1 and 2

Marked overlap between
members of Groups 1 and 2

ADAMS-TRESMAN: CALLOVIAN TELEOSAURID CROCODILES

199

total skull length (mm) pre orbital length (mm)

pre orbital length (mm)

total skull length (mm)

pre orbital length (mm)

TEXT-FIG. 2. Bivariate plots showing: a, relationship between preorbital length and
total skull length; b, relationship between width of snout at anterior end of nasals
and preorbital length; c, relationship between width of snout at anterior rim of
orbits and preorbital length; d, relationship between width of snout at anterior
end of nasals/preorbital length, and total skull length; and e, relationship between
distance separating nasals and premaxillae, and preorbital length. Key to symbols
used in text-figs. 2-4: o Steneosaurus ohtusidens', □. S. durobrivensis\ •, S. leedsi., ▲,
S. hidkei\ A, S. depressus; ■ Mycterosuchus nasutus.

Principal coordinate analysis

A multivariate technique, principal coordinate analysis, was used in order to test further for
differentiation of species. All of the measured cranial characters were analysed together, and the
specimens (i.e. the vectors representing the specimens) represented by points in n-dimensional space,
where n = number of crocodiles in the population minus 1, and the distance between any pair of

200

PALAEONTOLOGY, VOLUME 30

a

z

o

(_)

liJ

t/>

5(600)

A j^6(740)

,4(598)

A 16(792)

3 (744)

* 1(810)

^2(853)

Q 10(1030)

,.,7(935)

° 9(1160)
O

8(1112)

n

, I I (1000)
,17(1003)

FIRST

TEXT-FIG. 3. Principal coordinate analysis plot showing spatial
relationships between members of the genera Steneosaurus and
Mycterosuchus on the first and second axes. Specimens denoted
by computer number (see Table 1); total skull lengJhs(mm)
shown in parentheses alongside each computer number.

specimens is a measure of their overall similarity in terms of all the characters measured. Projecting
the coordinates of the crocodiles along the first and second principal coordinate highlights the
morphological dilferences that exist between members of the population because these axes are the
two directions along which variance (of measured characters) is greatest. The results of the multivari-
ate analysis are shown in text-fig. 3.

TAXONOMIC VALIDITY OF CHARACTERS

Cranial dimensions

The first axis in text-fig. 3 corresponds to measures of length or size. The spatial relationships
between specimens illustrated in text-fig. 3 could, therefore, be explained by either a similar morpho-
logical form at different scales, or morphological differences which might be used to define species.

Using principal coordinate analysis the effect of size on the other measured characters can be
eliminated. Size (defined here as total length in dorsal mid-line) is the first of ten variables measured,
i.e. variable 1. This is divided into all the other variables and subsequently excluded from the
analyses. The results of this procedure produced a similarity matrix and the new pattern in text-fig.
4. If the specimens in text-fig. 4 are ranked according to size, where 1 denotes the smallest specimen
and 1 3 the largest (ranked numbers are shown in parentheses), no size trend is apparent— in contrast
to text-fig. 3. The similarity values and spatial relationships between the specimens illustrated
in text-fig. 4 must, therefore, depend on morphological characters that are not independent of
size; the way in which they group together reinforces that seen already (text-fig. 2d; Character 1
of Table 2). Text-fig. 4 shows two groups linked internally by similarity values approaching, or in
excess of, 80 %: Group 1 (on the left of the graph) includes all specimens of S. leedsi and M.
nasutus; Group 2 contains the remaining Callovian steneosaur species. The value of the similarity
coefficient linking the two ‘most similar’ end-members of these two groups reaches 77-7 %. The
levels of similarity linking the end-members of these two groups can be seen clearly in text-fig. 4.
Interestingly enough, the maximum value of 883 defines the morphological similarity between

ADAMS-TRESMAN: CALLOVIAN TELEOSAURID CROCODILES

201

TEXT-HG. 4. Principal coordinate analysis plot showing similarity values between members of the
genera Steneosaurus and Mycterosuchus. Rank values appear in parentheses alongside computer

numbers.

specimens 6 and 9, which were previously defined as the type specimens of S. durobrivensis and S.
obtusidens respectively.

The fact that it is so difficult to further subdivide these two groups of specimens is strong evidence
that only two species exist.

Number of teeth

Specimens of S. leedsi are characterized by large numbers of teeth, c. 45-46 on each side of the
upper jaw and 43-44 in the lower jaw. M. nasutus has c. 38-40 teeth in the upper jaw and 42 in the
lower jaw. The remaining Callovian steneosaur species all have far fewer teeth; a minimum of 24 in
the upper jaw and 26 in the lower jaw of S. hulkei; c. 28 in the lower jaw of S. obtusidens', and
approximately 32 in the upper jaw and 30 in the lower jaw of S. durobrivensis.

If we assume that S. leedsi and M. nasutus belong to one species, and the other Callovian
steneosaurs to another, then it must be acknowledged that this degree of variation exceeds, to some
extent, the amount recorded by Kalin (1933) in his study of living crocodilians. It is clear from his
study, however, that tooth numbers never provide the sole basis on which species are recognized,
rather they occur together with clear morphological differences in the skull, and it is on the basis of
the latter that different species are distinguished. No such differences in cranial morphology can be
detected within the two groups of steneosaurs defined by the present study, and it would seem
unwise to split them on the basis of Kalin’s comments about tooth number in isolation from other
aspects of cranial morphology.

Qualitative characters

Three of the criteria used by Andrews (1913) to distinguish species of Steneosaurus (and reiterated
by Phizackerley 1951) are qualitative in nature— the form and relationship of the frontal and
prefrontal bones, tooth form, and scute form:

1. Frontal and prefrontal bones. The range in form of the nasal/frontal suture and its relationship
with the prefrontals is illustrated diagrammatically in text-fig. 5. The figure shows that this feature

202

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 5. Variation in form of frontal/nasal suture in Steneosaurus. Distance separating the most anterior
point of projection of frontal from prefrontals is indicated, a, S. durobrivensis, R3701 (30 mm); b, S.
durobrivensis, R2073 (19 mm); c, S. durobrivensis, R2865 (27 mm); D, S. hulkei, R2074 (20 mm); e, S. leedsi,
R3806 (4 mm); f, S. leedsi, R3320 (9 mm); g, S. obtusidens, R3168 (31 mm); h, S. depressus, OUM J1420

(32 mm). All R numbers are BM(NH) specimens.

is highly variable in its intricacy and shape, in the length of the anterior projection of the frontal,
and its level of termination relative to the most anterior point of the prefrontals. Andrews (1913,
p. 122) attached great importance to the above character and at times used it as a major criterion
by which species were distinguished, e.g. between S. hulkei and S. durobrivensis and between S.
edwardsi and S. hulkei. The first example suggests that Andrews did not take account of the
individual variation that occurs in this character in specimens of S. durobrivensis. The similarity in
form of the frontal/nasal suture of 5. hulkei and S. durobrivensis, BM(NH) R2865, in text-fig. 5 is
notable, but if R2865 is compared with R3701 and R2073 (both 5. durobrivensis) there are sharp
contrasts between all specimens. Andrews, of course, would only have been able to compare S.
hulkei with S. edwardsi by comparing the type of S. hulkei with a figure of the type of S. edwardsi.
On the basis of the above evidence, the synonymy of S. hulkei, S. edwardsi, and S. durobrivensis
seems probable.

2. Form of teeth. In S. leedsi and M. nasutus the teeth are slender with sharply pointed crowns
and the enamel is sculptured into a series of very fine longitudinal ridges. The pattern of tooth
development in the steneosaurs forming Group 2 shows a progressive change in crown form, i.e.
teeth become increasingly blunt as the size of the snout increases. This trend is complicated by
individual differences in tooth character between individuals with similar sized rostra (text-fig. 6).
There are no differences in the sculpture of the enamel between these specimens and S. leedsi and
M. nasutus.

Andrews (1913) diagnosed S. obtusidens mainly by the form of the teeth, which he stated were
blunt and rounded at the tips. He noted (1913, pp. 130-131), however, ‘that some of the replacing
teeth in the type skull of S. durobrivensis [R3701] are somewhat similar in form [to S. obtusidens] . . .
and although other differences between that species and the present one exist, the possibility that
the specimens on which the latter is based may be very large and old individuals of S. durobrivensis
cannot be ignored’. The two type specimens were widely separated in terms of ‘massiveness of the
rostra’ (see text-fig. 2b) with an accompanying disparity in tooth crown form. BM(NH) R2075

ADAMS-TRESMAN: CALLOVIAN TELEOSAURID CROCODILES

203

TEXT-FIG. 6. Variation in form of tooth crowns (arrowed) in Steneosaurus obtusidens. a, CMP R39; b, CMP

R178.

A B

TEXT-FIG. 7. A, B, form of tooth crowns (arrowed) in Steneosaurus durobrivensis, BM(NH) R2075.

(text-fig. 7) and some of the replacement teeth in BM(NH) R370I indicate the change in tooth
form. Also important in this context is the National Museum of Wales specimen 19 96 G12r/
showing tooth crowns intermediate in form between S. durobrivensis and S. obtusidens.

3. Form of scutes. Among the lines of evidence used by Andrews ( 191 3) to distinguish S. obtusidens
from S. durobrivensis (and the genera Mycterosuchus and Steneosaurus) is the form of the dorsal
scutes. In S. obtusidens Andrews said these were shallow and elongated— arranged in lines radiating
from the middle of the keel and sometimes almost running together to form shallow grooves. Few
scutes are preserved in the type specimen of S. obtusidens, BM(NH) R3618 (Andrews figured only
one), so reference was made to scutes of another specimen, BM(NFI) R3169, labelled as S. obtusidens
by Andrews, so that a more accurate assessment of their variance could be measured; both specimens
were compared with the many scutes preserved in BM(NH) R3701 (the type of S. durobrivensis).

204

PALAEONTOLOGY, VOLUME 30

B

TEXT-FIG. 8. Variation in form of dorsal scutes, a, R3169 Steneosaurus obtusidens, BM(NH) R3169; B, Y.

durobrivensis, BM(NH) R2865. Both x 0-5.

The considerable amount of variation between individual scutes is illustrated by text-fig. 8 which
shows that the elongate pits so characteristic and notable in the figured S. obtusidens scute (Andrews
1913, pi. 8, fig. 6) are neither universally present nor, indeed, representative of other S. obtusidens
scutes. Examination of modern crocodilian scutes reveals a high degree of variation in form,
dependent on the position of the scute on the body. The figured scute from the type specimen of S.
obtusidens is an atypical representative of those available for analysis illustrating only one aspect of
a highly variable feature and is, therefore, taxonomically invalid.

Additional characters used to establish Mycterosuchus

In distinguishing between the genera Steneosaurus and Mycterosuchus, Andrews (1913) used a
combination of cranial and post-cranial characters. The results of the analysis of the cranial
characters for the type specimen of Mycterosuchus, BM(NH) R2617 (specimen 11 on text-fig. 4)
and another, SMI (specimen 17), show that both exhibit high levels of similarity to specimens of S.
leedsi. Four post-cranial characters were noted by Andrews (1913) as having taxonomic importance:

1 . Size of fore-limb. Andrews described this as being less reduced than in Steneosaurus. There is
a limited amount of material from which to derive relevant data (Table 3) but the measurements
taken appear to indicate that fore-limb size in Mycterosuchus is broadly comparable with that of
steneosaurs of similar size (e.g. CMP R 178 in Table 3).

2. Degree of development of fore-limb condyles. Andrews described both distal condyles of the
fore-limb as being well developed in Mycterosuchus. The nature of preservation of all Oxford Clay
crocodiles renders this kind of evidence unreliable and of little importance.

3. Form of caudal vertebrae. Andrews noted that neural spines in the middle and posterior caudal
regions were notched anteriorly and posteriorly in Mycterosuchus. Caudal notching was indeed
found to be present in the specimens analysed. This feature was not seen in the steneosaur specimens
examined in the course of this work.

ADAMS-TRESMAN: CALLOVIAN TELEOSAURID CROCODILES

205

TABLE 3. Forelimb measurements in Steneosaurus and Myctero-
suclnts. All measurements in millimetres (* denotes estimated
value). Abbreviations: see Table 1; MB, private collection of
M. Bishop.

Specimen number

Species

Humerus

Radius

Ulna

BM(NH) R3806

S. leedsi

128

67

88

BM(NH) R3701

S. durobrivensis

122

71

89

CMPR175

S. durobrivensis

186

101

99

CMP R178

S. obtusidens

197

94

125

SM 1

M. nasutus

188

102

116

BM(NH) R2617

M. nasutus

211

113

142

MB 1

M. nasutus

189*

106*

—

4. Size of dorsal scutes. Andrews concluded that the dorsal scutes were more massive than in
Steneosaurus. The scutes from S. leedsi and M. nasutus show an increase in size which corresponds
to the increase in the overall size of the specimens (analogous to that seen between S. durobrivensis
and S. obtusidens).

The analysis of an albeit limited range of post-cranial characters exhibited by specimens currently
classified as Mycterosuchus has failed to reveal any characters which dilfer markedly from those of
Callovian steneosaurs (with the possible exception of the caudal vertebrae) and which could, by the
nature of their variance, be considered taxonomically significant.

SYSTEMATIC PALAEONTOLOGY

Using the results of the analyses of the cranial dimensions of Callovian steneosaurs, two species are
distinguished by the differences they show in the length and robustness of the rostral portions of the skull,
and the type and number of teeth they possess.

Genus steneosaurus Geoffroy Saint-Hilaire, 1825, emend. E. E. Deslongchamps, 1867
Type species. Steneosaurus megistorhynchus Geoffroy Saint-Hilaire, 1825.

Steneosaurus leedsi Andrews, 1909

1909 Steneosaurus leedsi Andrews, p. 300, pi. 8, fig. 1.

1909 Steneosaurus nasutus Andrews, pp. 308-309, pi. 9, fig. 1.

1913 Mycterosuchus nasutus Andrews, pp. 136-140, pi. 8, figs. 1-10; text-figs. 51-54.

Type data. Holotype, BM(NH) R3320.

Diagnosis. Elongated, slender rostrum; preorbital length 72 % or more of total length of skull.
Considerable variation in degree of separation of nasals and premaxillae; this distance accounts for
45-62 % of length of rostrum. Teeth slender, with sharply pointed crowns, forty or more in each
maxilla. Mandibular symphysis c. 58 % of length of jaw.

Steneosaurus durobrivensis Andrews, 1909

1867 Steneosaurus edwardsi E. E. Deslongchamps, p. 239, pi. 17, figs. 1-3.

1909 Steneosaurus durobrivensis Andrews, p. 304, pi. 8, fig. 2.

1909 Steneosaurus obtusidens Andrews, pp. 308-309, pi. 9, fig. 2.

1913 Steneosaurus hulkei Krv3iVQSN?,,Tp. 122.

1951 Steneosaurus depressus Phizackerley, p. 1 190, fig. 10a, b.

206

PALAEONTOLOGY, VOLUME 30

Type data. On the basis of the evidence presented here the synonymy of S. edwardsi is proposed. Characters
isolated from E. E. Deslongchamps’s (1867) original descriptions by Andrews (1913) as being taxonomically
significant (i.e. those which could be used to distinguish between S. edwardsi and his Callovian species) have
been shown to be invalid. Because of the destruction (during the Second World War) of material from which
the diagnosis of S. edwardsi was made, it is not possible to make a direct comparison between the type
specimen of S. edwardsi and the Callovian steneosaurs from England. There are no records of suitable material
from similar stratigraphic horizons in France, which could be designated as lectotype, and one cannot,
therefore, be absolutely certain that the proposed synonymy is correct. Although its relationship with the
steneosaurs examined in this work remains problematical, it is important to record the existance of this senior
Callovian steneosaur species S. edwardsi. The next available name is S. durohrivensis Andrews, 1909, whose
holotype is BM(NH) R3701 .

Diagnosis. Short rostrum; preorbital length c. 60 % or less of total length of skull. Nasals and
premaxillae separated by 36-56 % of total length of rostrum. Teeth blunt, rounded at tips; crowns
become increasingly blunt as size of skull increases; twenty-eight to thirty teeth in each maxilla.
Mandibular symphysis c. 40 % of length of jaw.

Acknowledgements. The initial research for this paper was carried out during tenure of a NERC grant which
is gratefully acknowledged. Facilities were provided by the Department of Palaeontology, British Museum
(Natural History). I thank Professor C. B. Cox for critically reading an earlier draft of the manuscript, and
Mr A. Howard for photographic work.

REFERENCES

ANDREWS, c. w. 1909. On some new steneosaurs from the Oxford Clay near Peterborough. Ann. Mag. nat.
Hist. SER. 8, 3, 299-308.

191 3. T descriptive catalogue of the marine reptiles of the Oxford Clay, 2, 206 pp. British Museum (Natural

History), London.

CUVIER, G. 1824. Recherches sur les ossemens fossiles. (2eme edition) Tome 5, 2eme, serie, 547 pp. Dufour et
D’Ocagnes, Paris.

DESLONGCHAMPS, E. E. 1863-1869. Notes Paleontologiques, 392 pp. Caen and Paris.

DODSON, p. 1975. Functional and ecological significance of relative growth in Alligator. J. Zool., Lond. 175,
315-355.

GEOFFROY SAiNT-HiLAiRE, E. 1825. Recherches sur I’organisation des gavials, sur leurs affinites naturelles
desquelles resulte la necessite d’une autre distribution generique, Gavialis, Teleosaurus et Steneosaurus.
Man. Mus. natn. Hist. nat. Paris. 12, 97-155.

KALIN, J. 1933. Beitrage zur vergleichenden Osteologie des Crocodilidenschadels. ZooL Jh. Abt. Anat. 57, 535-
714.

MciLHENY, c. A. 1976. The alligator s life history, 1 17 pp. The Society for the Study of Amphibians and Reptiles.
MATEER, N. J. 1974. Three Mesozoic crocodiles in the collections of The Palaeontological Museum, Uppsala.
Bull. geol. Instn Univ. Upsala, 4.

PHiZACKERLEY, p. H. 1951. A revision of the Teleosauridae in the Oxford University Museum, and the British
Museum (Natural History). Ann. Mag. nat. Hist. ser. 12, 4, 1 169-1 192.

SUSAN M. ADAMS- TRESMAN

Open University (East Anglian Region)
Cintra House, 12 Hills Road
Cambridge

Typescript received 28 February 1985
Revised typescript received 13 June 1986

Home address:
‘Lowood’, Grove Road
Cranleigh, Surrey GU6 7LH

i '

i*

•’ ■n;.? V . '• ■ ■(’„

4>.r

....

:i*^

fh

<f-

>

:5|.?

/t'

/« I?'

v_.

- ■>>

-■j

EV^‘

1-

e;<

ir-

ar::;

,1 <'■

i

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform in
style to those already published in this journal, and should be sent to Dr. Dianne Edwards, Department of Plant Science,
University College, P.O. Box 78, Cardiff CFl IXL, who will supply detailed instructions for authors on request (these are
published in Palaeontology 1985, 28, pp. 793-800).

Special Papers in Palaeontology is a series of substantial separate works conforming to the style of Palaeontology.

SPECIAL PAPERS IN PALAEONTOLOGY

In addition to publishing Palaeontology the Association also publishes Special Papers in Palaeontology. Members may
subscribe to this by writing to the Membership Treasurer; the subscription rate for 1987 is £33 (U.S. S58) for Institutional
Members, and £16-50 (U.S. S30) for Ordinary and Student Members. A single copy of each Special Paper is available to
Ordinary and Student Members only, for their personal use, at a discount of 25% below the listed prices. Non-members
may obtain copies, but at the listed prices, from Marston Book Services, P.O. Box 87, Oxford 0X4 ILB. England.

RECENT PALAEONTOLOGICAL ASSOCIATION

PUBLICATIONS

Special Papers in Palaeontology

Numbers 1-19 are still in print and are available (post free) together with those listed below:

20. (for 1977): Fossil Priapulid Worms, by s. conway morris. 155 pp., 99 text-figs., 30 plates. Price £16 (U.S. Jf24).

21. (for 1978): Devonian Ammonoids from the Appalachians and their bearing on International Zonation and Correlation.
by M. R. HOUSE. 70 pp., 12 text-figs., 10 plates. Price £12 (U.S. S18).

22. (for 1978, published 1979); Curation of Palaeontological Collections. A joint Colloquium of the Palaeontological
Association and Geological Curators Group. Edited by m. g. bassett. 279 pp., 53 text-figs. Price £25 (U.S. $38).

23. (for 1979): The Devonian System. A Palaeontological Association International Symposium. Edited by m. r. house, c. t.
SCRUTTON and m. g. bassett. 353 pp., 102 text-figs., I plate. Price £30 (U.S. $45).

24. (for 1980): Dinoflagellate Cysts and Acritarchs from the Eocene of Southern England, by j. p. bujak, c. downie, g. l.
EATON andG. L. williams. 100 pp., 24 text-figs., 22 plates. Price £15 (U.S. $23).

25. (for 1980): Stereom Microstructure of the Echinoid Test, hv a. b. smith. 81 pp., 20 text-figs., 23 plates. Price £15
(U.S. $23).

26. (for 1981): The Fine Structure of Graptolite Periderm, by p. r. crowther. 119 pp., 37 text-figs., 20 plates. Price £25
(U.S. $38).

27. (for 1981): Late Devonian Acritarchs from the Carnarvon Basin. Western Australia, by G. playford and R. s. dring.

pp.. \Q text-figs., \9 plates. Price £1 5 (U.S. $23).

28. (for 1982): The Mammal Fauna of the Early Middle Pleistocene cavern infill site of Westbury-sub-Mendip, Somerset, by
m. j. bishop. 108 pp., 47 text-figs., 6 plates. Price £25 (U.S. $38).

29. (for 1982): Fossil Cichlid Fish of Africa, by j. a. h. van couvering. 103 pp., 35 text-figs., 10 plates. Price £30 (U.S. $45).

30. (for 1983): Trilobites and other early Arthropods. Edited by d. e. g. briggs and p. d. lane. 276 pp., 64 text-figs.. 38 plates.
Price £40 (U.S. $60).

31 . (for 1984): Systematic palaeontology and stratigraphic distribution of ammonite faunas of the French Coniacian, by w. J.
KENNEDY. 160 pp., 42 text-figs., 33 plates. Price £25 (U.S. $38).

32. (for 1984): Autecology of Silurian organisms. Edited by m. g. bassett and j. d. lawson. 295 pp., 75 text-figs., 13 plates.
Price £40 (U.S. $60).

33. (for 1985): Evolutionary Case Histories from the Fossil Record. Edited by j. c. w. cope and p. w. skelton. 202 pp., 80 text-
figs., 4 plates. Price £30 (U.S. $45).

34. (for 1985): Review of the upper Silurian and lower Devonian articulate brachiopods of Podolia, bv o. i. Nikiforova,
T. L. modzalevskaya and m. g. bassett. 66 pp., 6 text-figs.. \(> plates. Price £10 (U.S. $15).

35. (for 1986): Studies in palaeobotany and palynology in honour of N. F. Hughes. Edited by d. j. batten and
D. E. G. BRIGGS. 178 pp., 29 plates. Price £30 (U.S. $50).

36. (for 1986): Campanian and Maastrichtian ammonites from northern Aquitaine, France, by w. i. Kennedy. 145 pp.. 43
text-figs., 23 plates. Price £20 (U.S. $35).

Field Guides to Fossils

1. (1983): Fossil Plants of the London Clay, by M. E. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $12).

Other Publications

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $30).

1985. Atlas of Invertebrate Macrofossils. Edited by j. w. Murray. Published by Longman in collaboration with the
Palaeontological Association. xiii-p241 pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24.95.

© The Palaeontological Association, 1987

Palaeontology

VOLUME 30 • PART 1

CONTENTS

Mass extinction: a commentary

D. M. RAUP 1

Two new specimens of Anthracosaurus (Amphibia: Anthraco-

sauria) from the Northumberland Coal Measures

J. A. CLACK 15

Basal Turonian ammonites from west Texas

W. J. KENNEDY, C. W. WRIGHT and J. M. HANCOCK 27

The Scandinavian Middle Ordovician trinucleid trilobites

A. W. OWEN 75

A new cyclocystoid from the Lower Ordovician of Oland,
Sweden

V. BERG-MADSEN 105

Upper Llandovery dendroid graptolites from the Pentland
Hills, Scotland

E. E. BULL 1 17

A new report of a theropod dinosaur from South Africa

N. J. MATEER 141

The Cretaceous Dimitobelidae (Belemnitida) of the Antarctic
Peninsula region

P. DOYLE 147

The Callovian (Middle Jurassic) marine crocodile Metriorhyn-
chus from central England

S. M. ADAMS-TRESMAN 179

The Callovian (Middle Jurassic) teleosaurid marine crocodiles
from central England

S. M. ADAMS-TRESMAN 195

Printed in Great Britain at the University Printing House, Oxford
by David Stanford, Printer to the University

ISSN 0031-0239

Published by

The Palaeontological Association London
Price £23 00

THE PALAEONTOLOGICAL ASSOCIATION

The Association was founded in 1957 to promote research in palaeontology and its allied sciences.

COUNCIL 1987-1988

President: Dr. L. R. M. Cocks, Department of Palaeontology. British Museum (Natural History), Cromwell Road,

London SW7 5BD

Vice-Presidents: Dr. D. E. G. Briggs, Department of Geology, University of Bristol, Bristol BS8 IRJ
Dr. L. B. Halstead, Department of Geology, University of Reading, Reading RG6 2AB
Treasurer: Dr. M. Romano, Department of Geology, University of Sheffield, Sheffield SI 3JD
Membership Treasurer: Dr. A. T. Thomas, Department of Geological Sciences, University of Aston, Birmingham B4 7ET
Institutional Membership Treasurer: Dr. A. W. Owen, Department of Geology, The University, Dundee DDl 4HN
Secretary: Dr. P. W. Skelton. Department of Earth Sciences, The Open University, Milton Keynes MK7 6AA
Circular Reporter: Dr. D. J. Siveter, Department of Geology, University of Hull, Hull HU6 7RX
Marketing Manager: Dr. V. P. Wright, Department of Geology, University of Bristol, Bristol BS8 IRJ
Public Relations Officer: Dr. M. J. Benton, Department of Geology, The Queen’s University of Belfast. Belfast BT5 6FB

Editors

Dr. M. J. Benton, Department of Geology, The Queen’s University of Belfast, Belfast BT5 6FB
Dr. P. R. Crowther, City of Bristol Museum and Art Gallery, Bristol BS8 IRL
Dr. D. Edwards, Department of Plant Science, University College, Cardiff CFl IXL
Dr. T. J. Palmer, Department of Geology, University College of Wales, Aberystwyth SY23 2AX
Dr. C. R. C. Paul, Department of Geology, University of Liverpool, Liverpool L69 3BX
Dr. P. A. Selden, Department of Extra-Mural Studies, University of Manchester, Manchester M13 9PL

Other Members

Dr. H. A. Armstrong, Newcastle upon Tyne Dr. G. B. Curry, Glasgow

Dr. M. E. CoLLiNSON, London Professor B. M. Funnell, Norwich

Dr. J. A. Crame, Cambridge Dr. P. D. Taylor, London

Overseas Representatives

Australia: Professor B. D. Webby, Department of Geology, The University, Sydney, N.S.W., 2006
Canada: Dr. B. S. Norford, Institute of Sedimentarv and Petroleum Geology, 3303-33rd Street NW., Calgary, Alberta
Japan: Dr. I. Hayami. University Museum, University ofTokyo, Hongo 7-3-1, Bunkyo-Ku, Tokyo
New Zealand: Dr. G. R. Stevens, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt
U.S.A.: Dr. R. J. Cuffey, Department of Geology, Pennsylvania State University, Pennsylvania 16802
Professor A. J. Rowell, Department of Geology, University of Kansas, Lawrence, Kansas 66045
Professor N. M. Savage, Department of Geology, University of Oregon, Eugene, Oregon 97403
South America: Dr. O. A. Reig, Departamento de Ecologia, Universidad Simon Bolivar, Caracas 108. Venezuela

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1987 arc:

Institutional membership
Ordinary membership
Student membership
Retired membership

£50-00 (U.S. S79)
£21-00 (U.S. $38)
£11-50 (U.S. $20)
£10-50 (U.S. $19)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr. A. W. Owen,
Department of Geology, The Lmiversity, Dundee DDl 4HN. Student members are persons receiving full-time instruction at
educational institutions recognized by the Council. On first applying for membership, an application form should be obtained
from the Membership Treasurer, Dr. A. T. Thomas, Department of Geological Sciences, University of Aston, Aston Triangle,
Birmingham B4 7ET. Subscriptions cover one calendar year and are due each January; they should be sent to the Membership
Treasurer. All members who join for 1987 will receive Palaeontology, Volume 30, Parts 1-4. Back numbers still in print
may be ordered from Marston Book Services, P.O. Box 87, Oxford 0X4 ILB, England.

Cover: Pedicle valve of the brachiopod Strophonella euglypha (Dalman, 1828) from the Wenlock Limestone of Dudley,
West Midlands; x 2. Photography by Harry Taylor of the British Museum (Natural History) Photographic Studio. One of
the specimens illustrated in the Atlas of Invertebrate Macrofossils published by the Association.

IDAMEAN (LATE CAMBRIAN) TRILOBITES FROM
THE DENISON RANGE, SOUTH-WEST TASMANIA

by }. B. JAGO

Abstract. Fourteen species of trilobites are described and figured from three faunas within the clastic
submarine fan sequence of the Singing Creek Formation, Denison Range, south-west Tasmania. It is suggested
that all faunas fall within the top three Idamean (early Late Cambrian) zones of Proceratopyge ayptica,
Erixaniwn sentum, and Stigmatoa diloma. The genus Denagnostus gen. nov., its type species D. corhetti sp.
nov., and Pseudagnostus idalis denisonensis subsp. nov. are erected. Proceratopyge is reviewed and its constitu-
ent species split into two broad groups based on cranidial characteristics. Proceratopyge gordonensis sp. nov.,
Aphelaspis cantori sp. nov., and Pseiidoyuepingia vanensis sp. nov. are erected.

The Upper Cambrian trilobites from the Singing Creek Formation of the Denison Range, south-
west Tasmania, are the first Cambrian fossils to be described from the Adamsfield Trough. The
Middle Cambrian to Lower Ordovician stratigraphy of the Denison Range area (Corbett 1975)
may be summarized as follows:

Denison Supergroup

Squirrel Creek Formation
Reeds Conglomerate
Great Dome Sandstone
Singing Creek Formation

600 m
1,560 m
510m
720 m

angular unconformity

Trial Ridge Beds 500 m

The upper part of the Trial Ridge Beds contains late Middle Cambrian {Lejopyge laevigata
Zone) fossils, including the agnostoid trilobites Tasagnostus, Hypagnostiis, Clavagnostiis, and Ptych-
agnostus (Jago 1979). The Great Dome Sandstone is a shallow marine-deltaic-fluvial sequence
which contains abundant trace fossils, rare inarticulate brachiopods, and a gastropod similar to
Kohayashiella (Corbett 1975).

The Singing Creek Formation comprises 720 m of quartz wacke turbidites interbedded with
fossiliferous siltstone, siliceous conglomerate, and slump sheets deposited as a submarine fan com-
plex in a fault controlled basin (Corbett 1972, 1973, 1975). In the Denison Range, fossils are found
over three stratigraphic intervals (text-fig. 1); the trilobites are described herein. Fossils from
stratigraphic equivalents of the Singing Creek Formation found elsewhere in the Adamsfield Trough
will be described in later papers.

The specimens from each fossiliferous interval were collected in 1967 and 1968 by K. D. Corbett
as bulk samples, rather than bed by bed, because of the nature of the outcrop and the difficulties of
collection in this rather inaccessible area. However, with the exception of the trilobites described
below as Leiostegiacea gen. et sp. indet., all of the relatively common species from each fossiliferous
interval occur throughout the range of available lithologies. Leiostegiacea gen. et sp. indet. is
restricted to a slightly coarser siltstone than the other fossils.

The ‘bottom fauna’ (c. 185-240 m above the base of the Singing Creek Formation) contains
trilobites (Micragnostus sp. 2, Pseudagnostus idalis denisonensis sp. nov., Denagnostus corbetti gen.
et sp. nov., Agnostoid gen. et sp. indet., Eugonocare sp., Dokimocephalidae gen. et sp. indet., and
Proceratopyge sp.) and brachiopods {Lingulella{l) sp., an acrotretid, and Billingsella sp.); only
Pseudagnostus idalis denisonensis and Billingsella sp. are reasonably abundant.

[Palaeontology, Vol. 30, Part 2, 1987, pp. 207-231, pis. 24-27}

© The Palaeontological Association

208

PALAEONTOLOGY, VOLUME 30

Gradational contact with
Great Dome Sandstone

Top fauna

Middle fauna

Bottom fauna

Unconformity on
Trial Ridge Beds

TEXT-FIG. 1. Stratigraphic position of the faunas from the Sing-
ing Creek Formation, Denison Range, south-west Tasmania.
Lithologies after Corbett (1975, fig. 2). The location of Denison
range is shown below.

0 100km

Eg

1 Sandstone, conglomerate
1 and slump sheets

□

Siltstone with isolated
1 sandstone beds

The Ttiiddle fauna’ (c. 410-430 m above the base of the Singing Creek Formation) contains
trilobites {Denagnostus corbetti gen. et sp. nov., Aphelaspis cantori sp. nov., Proceratopyge gordonen-
sis sp. nov., P. sp., Pseudoyuepingia vanensis sp. nov.), trilobite tracks, hyolithids (gen. et sp.
indet.), and brachiopods {Lingulella{l) sp. and Obolus{l) sp.); Proceratopyge gordonensis, P. sp., and
Pseudoyuepingia vanensis are common.

The ‘top fauna’ (c. 540-610 m above the base of the Singing Creek Formation) is by far the
richest and contains trilobites {Micragnostus sp. 1, Pseudagnostus idalis denisonensis subsp. nov., P.
cf. /. sagittus, P. sp., D. corbetti gen. et sp. nov., A. cantori sp. nov., Leiostegiacea gen. et sp. indet.,
Proceratopyge gordonensis sp. nov., P. sp., and a cranidium gen. et sp. indet.), trilobite tracks,
hyolithids (gen. et sp. indet.), and brachiopods {Obotus(i) sp., two other species of unassigned
Obolidae, a different species of acrotretid to that found in the bottom fauna, and Billingsella sp.).

Correlation

None of the species found in the Denison Range faunas has been recorded elsewhere, so an exact
zonal age cannot be determined. However, the presence of Eugonocare sp. in the ‘bottom fauna’
and that of a new subspecies of Pseudagnostus idalis in both the ‘bottom fauna’ and ‘top fauna’
suggest, by comparison with the range charts given by Henderson (1976) and Shergold (1982), that
all faunas are of Idamean age. This is supported by the presence of Proceratopyge gordonensis sp.
nov. and P. sp., both of which (particularly P. sp.) are similar to the Idamean species P. lata.

In Queensland, neither Pseudagnostus idalis nor any of its subspecies range up into the Irvingella
tropica Zone, the lowest zone of the post-Idamean. Although an exact correlation is not possible,
this suggests that the fossils described herein fall within the top three Idamean zones, i.e. the
Proceratopyge cryptica, Erixanium sentum, and Stigmatoa dilonia zones.

JAGO: LATE CAMBRIAN TRILOBITES

209

Probably the main reason that there are no species in common between Queensland and Tasmania
is that the faunas from the two areas occur in sediments of contrasting depositional environments.
The Queensland faunas are found in shallow water carbonate sequences while those of Tasmania
are found in a more offshore, clastic submarine fan sequence.

Faunal affinities

The faunas described herein show affinities with other Late Cambrian faunas of Australia, northern
Victoria Land (Antarctica), China, Korea, Kazakhstan, Alaska, and the Siberian Platform.

Pseudoyuepingia Chien is here described from Australia for the first time, although previously
reported from China (Chien 1961; Lu and Lin 1980) and, as Iwayaspis, from Korea (Kobayashi
1962) and Alaska (Palmer 1968). The closely related genus Yuepingia is known from China (Lu
19566; Lu and Lin 1980), Alaska (Palmer 1968), and Queensland (Henderson 1976).

Aphelaspis cantori sp. nov. is most closely related to A. australis from Queensland (Henderson
1976) and lA. sp. alf. A. australis from western New South Wales (Jell in Powell et al. 1982).
Kobayashiella problematica of Ivshin (1962) from Kazakhstan is closely related to A. cantori.

Proceratopyge is a widespread Late Cambrian genus. P. gordonensis sp. nov. and P. sp. are best
compared with the species of Proceratopyge from western Queensland described by Whitehouse
(1939), Opik (1963), Henderson (1976), and Shergold (1982), and with P. cf. P. lata of Shergold et
al. (1976) from northern Victoria Land, Antarctica. Among other species of Proceratopyge, the
Chinese form P. fenghwangensis Hsiang appears to be closest to the Tasmanian species.

As noted by Shergold (1982, p. 38) Eugonocare is known from Queensland, Victoria, China, and
the Siberian Platform. Pseudagnostus idalis denisonensis sub sp. nov. is a subspecies of P. idalis
from Queensland (Opik 1967; Shergold 1982).

Material and methods

All Tasmanian Cambrian fossils have undergone tectonic distortion to some extent. The terminology
used herein with respect to distortion is the same as that used by Jago (1976), and is based on
Henningsmoen (1960). The trilobites from the Denison Range area, however, are among the least
distorted of Tasmanian Cambrian faunas. All trilobites from these localities are preserved as internal
and external moulds in weathered siltstone or very fine sandstone. For description, silicone rubber
casts of the external moulds were prepared and then photographed after whitening with magnesium
oxide. The terminology used for agnostoid trilobites is essentially that of Robison (1982); that used
for polymeroid trilobites is after Harrington et al. (1959). All specimens are housed in the collection
of the Geology Department, University of Tasmania (UT).

SYSTEMATIC PALAEONTOLOGY

Order miomera Jaekel, 1909
Superfamily agnostacea M‘Coy, 1849
Family agnostidae M‘Coy, 1849
Subfamily agnostinae M‘Coy, 1849
Genus micragnostus Howell, 1935

Type species. Agnostus calvus Lake, 1906, p. 23, pi. 2, fig. 18.

Micragnostus sp. 1
Plate 24, fig. 1

Material. A moderately well-preserved internal mould of a cephalon (UT 88515) and an associated partial
pygidium comprising only the posterior border area.

Description. Cephalon slightly longer than wide. Very wide border furrow; narrow border. At anterior, width
of border almost 0-2 that of cephalon. Unconstricted acrolobe tapers markedly to anterior. Preglabellar

210

PALAEONTOLOGY, VOLUME 30

median furrow absent; genae smooth. Well-developed axial furrows. Glabellar length c. 0-6 that of cephalon;
glabella tapers markedly to broadly rounded anterior. Well-developed transverse glabellar furrow curves
gently rearwards. Details of posteroglabella poorly preserved.

All that can be seen of the associated pygidium is that the posterior border is very wide.

Discussion. Fortey (1980) and Shergold and Sdzuy (1984) discussed Micragnostus, Geragnostus, and
related genera.

Micragnostus sp. 2
Plate 24, figs. 2, 3

Material. Two poorly preserved cephala (UT 89510) and a poorly preserved pygidium (UT 89516) are included
together in a species referred to herein as Micragnostus sp. 2. The preservation is such that no formal
description or discussion is warranted.

Family diplagnostidae Whitehouse, 1936
Subfamily pseudagnostinae Whitehouse, 1936
Genus pseudagnostus Jaekel, 1909
Subgenus pseudagnostus Jaekel, 1909
Synonymy. See Shergold (1977, pp. 98-100).

Type species. Agnostus cyclopyge Tullberg, 1880, p. 26, pi. 2, fig. 15o, c.

Diagnosis. See Shergold (1977, p. 92).

Pseudagnostus {Pseudagnostus) idalis Opik, 1967
Pseudagnostus {Pseudagnostus) idalis denisonensis subsp. nov.

Plate 24, figs. 4-12

Diagnosis. A subspecies of P. {P. ) idalis with a centroposteriorly placed glabellar node and a very
wide cephalic border.

Holotype. The cephalon, UT 88519 (PI. 24, fig. 6) is designated as holotype.

Material. Over twenty cephala and pygidia are available (including UT 88353, 88366, 88371, 88376, 88381,
88382, 88494, 88499, 88519).

Description. Gently convex cephalon slightly wider than long. Border widens markedly to anterior. Border
furrow very wide and moderately deep. Unconstricted acrolobe length c. 0-85-0-90 that of cephalon. Well-
defined preglabellar median furrow shallows anteriorly. Genae smooth. Glabella has elongated oval shape;
length 0'65-0-70 that of cephalon; at transverse glabellar furrow, glabella width c. 0-30 that of cephalon.

EXPLANATION OF PLATE 24
Fig. 1 . Micragnostus sp. 1, UT 88515, cephalon, internal mould, x 7-5.

Figs. 2 and 3. M. sp. 2, from ‘bottom fauna’. 2, UT 89510, cephalon, internal mould (intermediate distortion),
X 9. 3, UT 89516, pygidium, external mould, x 9.

Figs. 4-12. Pseudagnostus idalis denisonensis subsp. nov. 4, UT 88376, cephalon, external mould, x 10. 5,
UT 88381, cephalon, internal mould, x 10. 6, UT 88519, holotype cephalon, external mould, W form,
X 10. 7, UT 88353, cephalon, external mould, L form, x 10. 8, UT 88366, pygidium, external mould, L
form, X 10. 9, UT 88499, cephalon, internal mould, x 10. 10, UT 88382, pygidium, external mould, x 10.
1 1, UT 88371, pygidium, internal mould, W form, x 10. 12, UT 88494, pygidium, external mould, W form,

X 10.

Figs. 13-19. Denagnostus corbetti gen. et sp. nov. 13, UT 88463, holotype cephalon, external mould, x 9. 14,
UT 88394, cephalon, external mould, x 9. 15, UT 88389a, cephalon, external mould, x 9. 16, UT 88513,
pygidium, external mould, x9. 17, UT 89424, from ‘middle fauna’, cephalon, internal mould, x 9. 18,
UT 89443, from ‘bottom fauna’, pygidium, internal mould, W form, x 9. 19, UT 88495, cephalon, internal
mould, X 9.

All specimens from ‘top fauna’ (see text-fig. 1) unless otherwise stated.

PLATE 24

JAGO, Micragnostiis, Pseudagnostus, Denagnostus

212

PALAEONTOLOGY, VOLUME 30

Glabella bounded by wide, shallow axial furrows which shallow slightly to anterior. Small, simple basal lobes
linked by narrow connective band. Faintly outlined, V-shaped, transverse glabellar furrow. Pair of faintly
developed lateral glabellar furrows just anterior of midpoint of glabella. Elongated node on centroposterior
part of glabella.

Gently convex pygidium slightly wider than long.

Wide, elevated border widens posteriorly; wide, shallow border furrow. Narrow, elevated, strongly genicu-
late shoulders. Wide, shallow, articulating furrow arched posteriorly; short (sag.), elevated, articulating half-
ring. Unconstricted or very slightly constricted acrolobe.

Anteroaxis outlined by shallow axial furrows which converge gently to the F2 furrow. M2 about twice as
long (sag.) and slightly narrower than Mj. Prominent elongated node on M2 extends across F2 and just on to
posteroaxis. Fi furrow almost obsolete; F2 furrow shallow and directed inwards and slightly to posterior
from either end.

Accessory furrows fade posteriorly; if continued across border, they would strike pygidial margin a little
behind posterolateral spines. Short posterolateral spines lie just forward of line drawn across rear of deuter-
olobe. Small terminal axial node visible on some specimens. Narrow, smooth pleural areas.

Discussion. The specimens clearly fall within the P. idalis species complex as described and discussed
in some detail by Shergold (1982); he noted that this complex should be investigated at the subspecific
level. P. i. denisonensis subsp. nov. is close to P. i. idalis Opik, 1967, as discussed by Shergold (1982).
The cephalon of P. i. denisonensis differs from that of P. i. idalis and that of P. i. sagittus in that the
glabellar node of denisonensis is placed further to the posterior than that of the other two subspecies.
The cephalic border of P. i. denisonensis is wider than that of either P. i. idalis or P. i. sagittus. The
pygidia of P. i. idalis and P. i. denisonensis appear to be identical, but the pygidial spines of P.i.
sagittus are placed further to the posterior than those of P. i. denisonensis. The slightly different
appearance of the pygidium figured in Plate 24, fig. 1 1 is thought to be due to the fact that it is
preserved in shale, whereas all other figured specimens are preserved in siltstone.

Pseudagnostus (Pseudagnostus) cf. idalis sagittus Shergold, 1982
Plate 25, fig. 3

Material. One pygidium (UT 88478) associated with P. (P.) idalis denisonensis sp. nov.

Description. Length (including axial half-ring), 3-6 mm; width, 3-7 mm. The pygidium has similar axial features
to P. i. idalis, P. i. sagittus, and P. i. denisonensis. However, its pygidial margins are markedly tapered, and
the border spines are set closer together than those of the other three subspecies. The spines are placed level
with the end of the deuterolobe, rather like those of P. i. sagittus.

Discussion. The similarity of axial characteristics and the fact that there is only one known specimen
of this type raises the possibility that it is an aberrant pygidium of P. i. denisonensis. However, due
to the retral position of the spines, it is referred to P. (P.) cf. i. sagittus Shergold.

Pseudagnostus sp.

Plate 25, fig. 5

Material. An internal mould of a cephalon (UT 88517).

Description: Gently convex cephalon, 4 mm in length, about as wide as long. Narrow border; wide, shallow
border furrow. Unconstricted acrolobe; smooth genae. Well-defined preglabellar median furrow. Markedly
tapering glabella has length about two-thirds that of cephalon. Shallow axial furrows. Very shallow, almost
straight transverse glabellar furrow. Pair of faintly developed lateral glabellar furrows at midpoint of glabella.
Small circular node placed between lateral glabellar furrows. Small, simple basal lodes.

Discussion. The combination of a markedly tapering glabella and a wide border distinguish this
cephalon from most species of Pseudagnostus. It may belong in an undescribed species of Pseudagno-
stus, but without an associated pygidium a new species cannot be erected. Other species of Pseudag-
nostus which show this combination include the cephalon illustrated by Bell and Ellinwood (1962,
pi. 36, fig. 11) as P. communis, although Palmer (1968, p. 30) considered this specimen not to

JAGO: LATE CAMBRIAN TRILOBITES

213

belong in communis. P. chinensis (Dames) shows a similar combination of markedly tapering glabella
and a wide border, as illustrated by Schrank (1974, pi. 1, figs. 1-7), although its glabella is shorter
and its genae are faintly scrobiculate, a feature which is not apparent in the Tasmanian specimen.

Genus denagnostus gen. nov.

Type species. Denagnostus corbetti sp. nov.

Diagnosis. Almost effaced, gently convex cephalon with subovoid outline and straight posterior
margin. Unconstricted acrolobe. Faintly outlined glabella; rounded glabellar rear; very faintly
outlined V-shaped transverse glabellar furrow; spectaculate; small centrally placed glabellar node.
Pair of large anterolateral lobes immediately anterior of node. Wide anterior border narrows
posteriorly and disappears about halfway to posterior margin. Pygidium slightly more convex than
cephalon; pygidial acrolobe slightly constricted. Wide border with very small posterolateral spines
placed well forward of acrolobe posterior. Faint ridge around centre of posterior border, indicating
that pygidium is slightly zonate. Faintly outlined axis has two small anterior segments and long
slightly expanded posteroaxis which reaches acrolobe posterior; small terminal axial node. Low
elevated node on second axial segment.

Discussion. The effaced nature of Denagnostus hinders its classification, but it appears to be most
closely related to Rhaptagnostus Whitehouse, a member of the Pseudagnostinae. Similarities between
the two genera include the shape and essentially effaced nature of the cephala and pygidia, the
slightly constricted pygidial acrolobes, and the presence of very small pygidial posterolateral spines
placed well forward of the acrolobe posterior. In addition, the transverse glabellar furrow of
Denagnostus is V-shaped like that of Rhaptagnostus. Denagnostus shows no clearly defined deuter-
olobe, but neither do many of the species of Rhaptagnostus illustrated by Shergold (1975, 1977,
1980).

Denagnostus differs from Rhaptagnostus in that it is spectaculate rather than papilionate, i.e. the
axial glabellar node of Denagnostus lies to the rear of the anterolateral glabellar lobes rather than
between them (see Shergold 1975, 1977). It could be argued that Denagnostus is simply a spectaculate
species of Rhaptagnostus and that the diagnosis of Rhaptagnostus given by Shergold should be
expanded to allow for this. Flowever, as discussed by Shergold (1977), the position of the axial
glabellar node is important, from the viewpoint of both anatomy and classification; hence Denagno-
stus should not be included in Rhaptagnostus.

The slightly zonate nature of the pygidial border separates Denagnostus from all other known
members of the Pseudagnostinae which have simplimarginate borders. Denagnostus differs from all
previously described agnostoid genera in the way in which the very wide anterior cephalic border
narrows markedly to the posterior and disappears about half-way around the cephalon.

Apart from the features noted above, Denagnostus differs from Neoagnostus, which has a specta-
culate glabella, in the shape of the shields: those of Neoagnostus are generally subquadrate whereas
those of Denagnostus are subovoid. The pygidial border spines of D. corbetti are placed much
further forward than those of any species of Neoagnostus. The anterior part of the pygidial axis of
many species of Neoagnostus show three segments in that part of the axis outlined by axial furrows
(Shergold 1977), whereas in D. corbetti only two such segments are outlined. Denagnostus differs
from many species of Pseudagnostus in that it shows no clearly defined deuterolobe. D. corbetti has
a faint V-shaped transverse glabellar furrow, whereas those in Pseudagnostus are either straight or
gently curved to the posterior.

Denagnostus corbetti sp. nov.

Plate 24, figs. 13-19; Plate 25, figs. 1 and 2; text-fig. 2
Diagnosis. See generic diagnosis.

Holotype. UT 88463 (PI. 24, fig. 13) is selected because it is the best preserved cephalon.

214

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 2. Reconstruction of Denagnostus corbetti gen. et sp. nov. Cephalon based on UT
88463 (PI. 24, fig. 13); pygidium based on counterparts UT 88511 and UT 88495 (PI. 25,
figs. 1 and 2), x 5.

Material. One reasonably well-preserved cephalon (UT 88463), two partially preserved pygidia as external
moulds (UT 88511, 88513), and several reasonably preserved internal moulds of both cephala and pygidia
(including UT 88495, 89424, 89443).

Description. Gently convex subovoid cephalon, about as wide as long, with almost straight posterior margin.
Margins of cephalon diverge anteriorly up to a point just under half-way to anterior of cephalon; from this
point, margins converge to give anterior margin a subelliptical outline.

Border absent in posterior half of cephalon, except for very short posterolateral spines which are separated
from acrolobe by narrow, shallow posterior border furrows. Almost flat border appears about half-way along
cephalic margins and widens markedly to anterior of cephalon, where it is quite wide. Narrow, shallow border
furrow. (The apparent very narrow border, placed posterolaterally on right side of UT 89424 (PI. 24, fig. 17)
is an artefact of preparation.)

Glabella only faintly outlined at posterior; is markedly convergent to anterior and fades out in that direction;
length about two-thirds that of cephalon. The basal lobes are of moderate size (PI. 24, fig. 13); rear of glabella
rounded. To anterior of basal lobes, glabella tapers to broadly rounded front, only seen faintly in some speci-
mens (PI. 24, fig. 14) and not at all in others. Very faintly outlined V-shaped transverse glabellar furrow. Small
node at about centre of glabella. Spectaculate. Pair of large anterolateral lobes immediately anterior of node.

Pygidium a little longer than wide, and slightly more convex than cephalon. Acrolobe slightly constricted
in some specimens (best seen on PI. 24, fig. 18).

Wide, shallow border furrow; border wide and almost flat at posterior, becoming narrow and more elevated
anteriorly. Very small posterolateral spines placed well forward of acrolobe posterior. Around centre of
posterior border, paralleling acrolobe margin, is a low ridge which meets margin a little anterior of border

EXPLANATION OF PLATE 25

Figs. 1 and 2. Denagnostus corbetti gen. et sp. nov. 1, UT 88511, pygidium, external mould, L form, x9. 2,
UT 88495, pygidium, internal mould, L form (counterpart of UT 8851 1), x 9.

Fig. 3. Pseudagnostus {Pseudagnostus) cf. idalis sagittus Shergold, 1982, UT 88478, pygidium, external mould,
x9.

Fig. 4. Agnostoid gen. et sp. indet., UT 89438b, from ‘bottom fauna’, cephalon, external mould, L form, x 10.

Fig. 5. P. sp., UT 88517, cephalon, internal mould, x 9.

Figs. 6-16. Aphelaspis cantori sp. nov. 6, UT 88521, holotype cranidium, external mould, W form, x 4. 7, UT
89406, from ‘middle fauna’, cranidium, external mould, L form, x 4. 8, UT 88520a, librigena, internal
mould, X 4. 9, UT 88520b, cephalon and part of thorax, external mould, x3. 10, UT 88393, pygidium
and most of thorax, internal mould, x4. 11, UT 88532, cranidium and anterior part of thorax, internal
mould, W form, x 3. 12, UT 88533, almost complete cephalon, internal mould, W form, x4. 13, UT
88393a, cranidium, internal mould, L form, x 4. 14, UT 88538, specimen showing most of cranidium, part
of a librigena, and most of thorax and pygidium, internal mould, x 3. 15, UT 88522, cranidium, external
mould, W form, x 4. 16, UT 89424, from ‘middle fauna’, cranidium, internal mould, x 4.

Fig. 17. Dokimocephalid gen. et sp. indet., UT 89508, from ‘bottom fauna’, cranidium, internal mould, x 4.

All specimens from ‘top fauna’ (see text-fig. 1) unless otherwise stated.

PLATE 25

JAGO, Denagnostus, Pseudagnostus, Aphelaspis

216

PALAEONTOLOGY, VOLUME 30

spines. This ridge indicates that pygidium is slightly zonate; it appears to reflect outline of cephalic border
and probably represents position of anterior margin of cephalon during enrolment. Narrow, shallow shoulder
furrows; narrow, elevated, convex shoulders. Neither facets nor fulcra seen. Articulating device nowhere
completely preserved; articulating furrow of moderate depth and width, with shallow articulating recess.

At anterior, axis width c. 045 that of pygidium. Anterior pair of axial segments outlined by narrow, shallow
axial furrows. Both segments short (sag.); more posterior of pair slightly narrower (tr.) than other. Very faint
traces of transverse axial furrow between anterior segments and between second segment and posteroaxis.
Low, elongated node at centre of second segment.

Long posterior axial lobe only very faintly outlined, slightly expanded, and just reaches posterior border
furrow. Small terminal node at posterior of posterior lobe. Some suggestion of internotular axis, but preser-
vation not good enough to be certain. Pleural areas smooth.

Agnostoid, gen. et sp. indet.

Plate 25, fig. 4

Material. Several cephala (including LIT 89438b).

Description. Cephalon slightly longer than wide. Gently convex anterior border; wide shallow border furrow.
Lateral borders not visible. Acrolobe appears unconstricted; smooth genae. Posteriorly directed spines arise
from posterolateral corners; spine length cannot be determined but spines at least of moderate length. Deep,
wide preglabellar median furrow flares forwards. Parallel sided glabella length c. 0-7 that of cephalon. Deep,
wide axial furrows. Broadly rounded glabellar front. Straight, shallow transverse glabellar furrow very shallow
at centre. Anteroglabella length about one-third that of glabella. No distinct glabellar node, although this
may be function of preservation. Pair of faint lateral glabellar furrows placed well forwards on posteroglabella.
Very broadly rounded glabellar rear; small, simple basal lobes connected by a wide connecting ring. Trace of
cephalo-thoracic aperture visible.

Discussion. These cephala cannot be placed with certainty in any previously described genus or
species and are hence left in open nomenclature.

Order ptychopariida Swinnerton, 1915
Superfamily ptychopariacea Matthew, 1887
Family pterocephaliidae Kobayashi, 1935
Subfamily aphelaspidinae Palmer, 1960

Discussion. The classification of the Aphelaspidinae and related genera is difficult due to the
considerable number of genera which show, or may be derived from, a basic aphelaspidine mor-
phology. The concepts of Aphelaspis and the Aphelaspidinae were based originally largely on
North American material by Palmer (1960, 1962, 1965) who described and figured numerous
species.

The species of Aphelaspis accepted by Shergold (1982, p. 37) encompass considerable morphologi-
cal variation. They include species such as A. australis Henderson, 1976, from Queensland, with a
much shorter preglabellar field than the type species of Aphelaspis, A. walcotti Resser, 1938.
A. australis shows similarities to some aphelaspidine-like trilobites from Siberia (see discussion
on A. cantori sp. nov. below).

As noted by Shergold (1982), it is possible that quite a number of genera from the Siberian
Platform were derived from a basic aphelaspidine morphology. These include Apheloides Ivshin,
Elegantaspis Ivshin, Kobayashella Ivshin, Nganasanella Rosova, Tamaranella Rosova, Kuraspis
Chernysheva, Pedinocephalites Rosova, Maduiya Rosova, Anwrphella Rosova, Ketyna Rosova,
Kujandaspis Ivshin, Nyaya Rosova, Kaninia Walcott and Resser, Monosulcatina Rosova, Graciella
Rosova, and Acrocephalaspina Ergaliev. Various species of these genera were described and dis-
cussed by Ivshin (1962), Rosova (1963, 1964, 1968, 1977), and Appollonov and Chugaeva (1983).
Comments on some of these genera are made below.

Ivshin (1962, p. 80) erected Elegantaspis, with type species E. elegantula, plus one other species.

JAGO: LATE CAMBRIAN TRILOBITES

217

E. beta. An inspection of the features of these two species illustrated by Ivshin (1962, pi. 5) reveals
no significant differences between them; hence, E. beta is a junior synonym of E. elegantula.

Nganasella, with type species N. nganasanensis Rosova, 1963 (p. 10, pi. 1, fig. 2), is a distinctive
genus with a markedly tapering glabella, the length of which varies within the different species of
the genus. I suggest that N. interminata Rosova, 1964, (p. 74, pi. 8, figs. 1-3, 5-11) be placed in
synonymy with N. tavgaensis, since the specimens figured by Lazarenko and Nikiforov (1968,
p. 1 3) appear to be conspecific.

TamaraneUa is based on T. bella Rosova, 1963 (p. 10, pi. 1, fig. 5); the holotype and two other
specimens were figured by Rosova (1964, pi. 18, figs. 12-15). I ascribe all these specimens to
Nganasella', hence, TamaraneUa is a junior synonym of Nganasella. Lazarenko and Nikiforov (1968)
placed TamaraneUa in Apachia Frederickson, but Nganasella is more appropriate. The species
described by Lazarenko and Nikiforov (1968, p. 41, pi. 4, figs. 11-13) as A. plana should also
be placed in Nganasella, as probably should A. sima Lazarenko and Nikiforov, 1968 (p. 42,
pi. 7, figs. 18-26) but the poor preservation of the latter makes definite generic assignment
difficult.

The type species of Maduiya, M. maduensis Rosova, 1963 (p. 11, pi. 1, hg- 1 1) is based on a single
rather incomplete cranidium refigured by Rosova (1968, pi. 4, figs. 10-12). Rosova (1968) referred
two other species to Maduiya, i.e. M. sibirica Rosova, 1963 and M. composita (Rosova, 1963); the
latter’s original assignment to Idahoial was adhered to by Lazarenko and Nikiforov (1968). The
holotype cranidium of M. composita, as illustrated by Rosova (1968, pi. 4, figs. 17-19), falls well
within the range of morphologies illustrated by Rosova (1968, pi. 4, figs. 1-9) for M. sibirica', hence,
composita should be regarded as a junior synonym of sibirica. Due to the incomplete nature of the
holotype cranidium of the type species, M. maduensis, it is not clear whether sibirica and maduensis
should be placed in the same genus. It is possible that the concept of Maduiya should be restricted
to M. maduensis and that sibirica belongs in Idahoia or a related genus.

The genera Amorphella Rosova, 1963, with type species A. modesta Rosova, 1963 (p. 14, pi. 2,
figs. 1 and 2), Ketyna, with type species K. ketiensis Rosova, 1963 (p. 16, pi. 2, fig. 7), and
Acrocephalaspina Ergaliev, 1980, with type species A. insueta Ergaliev, 1980 (p. 130, pi. 14, figs. 13-
15) appear to be closely related in that they all have a glabella which tapers slightly forwards, a
bluntly rounded glabellar front, well-developed palpebral lobes, well-developed axial and border
furrows, and a median swelling in the preglabellar field. It is arguable that Acrocephalaspina should
be placed in synonymy with Amorphella, although the well-developed eye ridges and distinctly
shorter glabella in the species of Acrocephalaspina illustrated by Ergaliev (1980, pi. 14) suggest that,
for the time being at least, it may be better to keep the genera separate. The figures of A. insueta, A.
insueta spinosa, A. magna, and A. longa illustrated by Ergaliev (1980) show no significant differences
and I regard them all as A. insueta.

Ketyna Rosova, 1963 is similar to both Amorphella and Acrocephalaspina but the cranidium of
Ketyna has much smaller posterolateral limbs, as shown by K. ketiensis (type species) and K. glabra
figured by Rosova (1968, figs. 40 and 41). It should be noted, however, that the various species of
Ketyna illustrated by Apollonov and Chugaeva (1983) suggest that Acrocephalaspina could be
accommodated in Ketyna.

The generic position of Amorphellal magna, as described by Rosova (1968, 1977), is not clear,
but I suggest that not all the specimens figured by Rosova (1977, pi. 8, figs. 1-15) belong in one
species. For example, the length of the palpebral lobes of one cranidium (Rosova 1977, pi. 8,
fig. 1 1) is about half that of another (ibid., fig. 4).

The genus Jingxiana Chien, 1974, from China, was erected by Lu et al. (1974) with J. beigongliensis
as type species. Three other species, J. zhuangliensis Chien, J. tangcunensis Chien, and J. traversa
Chien were erected by Lu et al. (1974). There seems almost no difference between the specimens
illustrated, and it is probable that zhuangliensis, tangcunensis, and transversa are junior synonyms
of beigongliensis. This is partly confirmed by the specimens of tangcunensis shown by Qiu (1984, pi.
3, figs. 1-3) which are indistinguishable from the holotype of beigongliensis (Lu et al. 1974, pi. 4,
fig. 13).

218

PALAEONTOLOGY, VOLUME 30

Genus aphelaspis Resser, 1935
Type species. Aphelaspis walcotti Resser, 1938, p. 59, pi. 13, fig. 14.

Diagnosis. See Palmer (1965, p. 58).

Discussion. Palmer (1960, 1962, 1965) discussed Aphelaspis in some detail. The species he included
show considerable variation in length of glabella, length of preglabellar field, and width of cranidial
border. Australian species are A. australis Henderson, 1976 (p. 342, pi. 49, figs. 5-7), lA. sp. aff. A.
australis of Jell in Powell et al. (1982, p. 142, fig. 10, 7-8), and A. sp. undet. of Shergold (1982,
p. 37, pi. 17, figs. 5 and 6). As noted by Shergold (1982, p. 37), lA. sp. B of Opik (1963, p. 76) may not
belong in Aphelaspis. The poor preservation of ?^. sp. B makes a generic assignment inappropriate.

The new species of Aphelaspis described below, A. cantori, is similar to both A. australis and I A.
sp. alT. A. australis in that it has a short preglabellar field, deeply impressed axial furrows, and long
palpebral lobes. As noted by Jell in Powell et al. (1982, p. 142), the exclusion of australis (and hence
cantori) from Aphelaspis can be argued by virtue of the short preglabellar field, well-impressed axial
furrows, and relatively long palpebral lobes. Although such species might form the basis of a new
genus, there are already numerous genera with a basic aphelaspinid morphology (as noted above in
the discussion of the subfamily) and I prefer to follow Jell and Henderson and assign cantori, along
with australis, to Aphelaspis.

Both A. cantori and A. australis are similar to the single cranidium described by Ivshin (1962,
p. Ill, pi. 7, fig. 11) as Kobayashella problematica gen. et sp. nov., which probably belongs in
Aphelaspis. If so, then Kobayashella is a junior synonym of Aphelaspis, but with only the one
figured partial cranidium available it is not possible to make a meaningful comparison between A.
cantori, A. australis, and K. problematica, particularly as Ivshin’s (1962, p. 112, fig. 29) figure of
problematica shows a much larger preglabellar field than is suggested by his pi. 7, fig. 11.

Al kazachstanica Lisogor, 1977 (p. 217, pi. 30, figs. 4 and 5) from Kazakhstan may also be close
to A. cantori, although the glabella of cantori is longer. The pygidium assigned to kazachstanica is
clearly different to that of cantori, but until more and better material of kazachstanica is figured a
detailed assessment of the species cannot be made.

Aphelaspis cantori sp. nov.

Plate 25, figs. 6-16

Diagnosis. Cranidium markedly wider than long. Glabella tapers gently forwards. Very gently
impressed Ip furrows; other lateral glabellar furrows almost effaced. Deep axial furrows; fossulae
present. Very short preglabellar field. Wide deep border furrow; wide anterior border. Prominent,
long, centroanteriorly placed palpebral lobes separated from fixigenae by well-developed palpebral
furrows. Thorax of thirteen segments. Small, transversely elliptical pygidium with axial length about
three-quarters that of pygidium. Pygidium has narrow, shallow border furrow and very narrow
border.

Holotype. Cranidium, UT 88521 (PI. 25, fig. 6).

Material. Two specimens in which at least part of the cephalon, thorax, and pygidium are present (UT 88538);
several isolated librigenae (UT 88520a); two cranidia with a few attached thoracic segments (UT 88532); two
cranidia with attached librigenae; ten individual cranidia (including UT 88393a, 88521, 88522, 89406, 89424);
and one specimen with a pygidium and twelve thoracic segments (UT 88393). Preservation varies from poor
to reasonable.

Description. Surface ornament lacking on all specimens. Cranidium markedly wider than long. Length of
gently convex glabella (including occipital ring) c. 0-7-0-75 that of cranidium; between the palpebral lobes
glabellar width 0-4-0'5 that of cranidium. Glabella margins almost parallel up to Ip furrows, from where
glabella tapers gently to almost straight glabella anterior. Axial and preglabellar furrows deeply impressed;
distinct fossulae present. Very short, gently convex preglabellar field; wide, deeply impressed anterior border
furrow about same width as gently convex border. Moderately impressed occipital furrow shallowest at centre.

JAGO: LATE CAMBRIAN TRILOBITES

219

Low, centrally placed occipital node. Lateral glabellar furrows almost effaced. Pair of very shallow, posteriorly
directed Ip furrows about one-third of way along glabella. Faint traces of 2p and 3p furrows on some
specimens (PI. 25, fig. 13). Wide palpebral furrows deepen at either end; long, well-developed, gently curved
palpebral lobes opposite centroanterior part of glabella. Well-developed, slightly curved eye ridges. Preocular
sections of facial suture diverge slightly up to border furrow from where they converge. Postocular sections of
facial suture diverge markedly.

Preocular areas of fixigenae slope down markedly to border furrow. Palpebral and posterior areas of
fixigenae gently convex; fixigenae slope gently to broad, moderately impressed, posterior border furrow which
widens abaxially. (Apparent differences in shape of illustrated cranidia (cf. PI. 25, figs. 1 1 and 7) caused by
slight tectonic distortion of enclosing sediments.)

Moderately convex librigenae. Gently impressed, wide border furrow. Almost flat border extends into
narrow genal spine which reaches fourth thoracic segment.

Thorax of thirteen segments, each about twelve times as wide as long. Width of axis about one-third that of
each segment. Moderately impressed pleural furrows wide up to geniculation, from where they narrow and
are curved gently to posterior. Rounded pleural extremities.

Small pygidium with transversely elliptical outline. Gently impressed axial furrows. Axis width about three-
quarters that of pygidium. Axial details not preserved. Slightly elevated pair of pleural ribs near anterior
margin of pygidium; remainder of pleural areas almost smooth. Narrow, shallow border furrow; very narrow
border.

Discussion. When compared with previously described species of Aphelaspis, A. cantori is closest to
A. ciustralis. However, australis has a more rounded glabellar front, better developed lateral glabellar
furrows, shallower axial furrows, and a narrower anterior cranidial border than cantori. The fact
that cantori is closest to australis lends support to the suggestion of Jell in Powell et al. (1982) that
some Australian aphelaspidines may belong to a lineage distinct from the Aplielaspis of North
America.

Of the various North American species of Aplielaspis, A. cantori is closest to A. brachyphasis
Palmer, 1962 (p. 33, pi. 4, figs. 1-19) by virtue of the latter’s short preglabellar field. The posterior
part of the thorax and the pygidium of brachyphasis, as illustrated by Palmer (1962, pi. 4, fig. 14),
are very similar to those of cantori, as illustrated herein (PI. 25, figs. 10, 14). The palpebral lobes of
brachyphasis are shorter than those of cantori, the dorsal furrows of brachyphasis are shallower
than those of cantori.

Genus EUGONOC ARE Whitehouse, 1939

Type species. Eugonocare tessellatum Whitehouse, 1939, p. 226, pi. 23, figs. 15, 17 (non figs. 16, 18); pi. 25, fig.
Ih (fide Henderson 1976).

Eugonocare sp.

Plate 26, fig. 12

Material. One internal mould of a partial cranidium (UT 88361 ).

Discussion. As noted by Henderson (1976) and Shergold (1982) the cranidia of the various species
of Eugonocare are essentially indistinguishable. Hence, as no pygidium is available this specimen is
simply referred to Eugonocare sp.

Superfamily dikelocephalacea Miller, 1889
Family dokimocephalidae Kobayashi, 1935
Dokimocephalid, gen. et sp. indet.

Plate 25, fig. 17

Material. The external and internal (UT 89508) moulds of a partial cranidium.

Discussion. This cranidium is placed in the Dokimocephalidae because of the combination of
palpebral lobes placed close to the glabella, a short preglabellar field, and the bifurcating nature of
the Ip glabellar furrows. In addition, the posterior branch of the Ip furrow has a sigmoidal shape,
in common with many members of the Dokimocephalidae.

220

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 3. A-D, F, Leiostegiaccan gen. et sp. indet. a, UT 88490, partial cranidium showing base of occipital
spine, internal mould, x 3. b, UT 88490, pygidium, internal mould, x 2. c, UT 88482, pygidium, internal
mould, X 2. D, UT 88411, pygidium, external mould, x 2. f, UT 88503, pygidium, internal mould, x 2. e,
Pseudoyuepingia vanensis sp. nov., UT 89415, from 'middle fauna’, pygidium and two thoracic segments,
internal mould, x 2. All specimens except E from 'top fauna’ (see text-fig. 1).

Superfamily leiostegiacea Bradley, 1925
Leiostegiaeean, gen. et sp. indet.

Text-fig. 3a-d, f

Material. One partial cranidium (UT 88490) and ten partial pygidia (including UT 88411, 88482, 88490,
88503).

Description. Gently convex cranidium with almost effaced glabella which extends to almost straight anterior
border. Axial and preglabellar furrows shallow. Well-developed occipital furrow; base of large occipital spine.
Eye ridges not showing; smooth fixigenae. Moderately deep border furrow; narrow border. Large, moderately
convex pygidium, probably slightly wider than long. Axis has length c. 0-8 that of pygidium; it is outlined by
moderately deep axial furrows which shallow posteriorly. Axis comprises fourteen or fifteen axial rings plus
terminus; tapers evenly to posterior, apart from slight constriction near twelfth axial ring. Pleural areas
strongly convex in adaxial areas but abaxially they slope markedly down to pygidial margin. Pleural areas
comprise thirteen progressively smaller ribs separated by well-defined furrows. Ribs and furrows best defined
in adaxial part of pleural areas, but can be traced clearly across flatter outer part of pleural areas almost to
pygidial margin. Terrace lines on wide doublure seen where outer part of pleural areas not fully preserved
(text-fig. 3b). Anterior rib much larger than other ribs and, in contrast to other ribs, widens abaxially and
extends into broad, posteriorly directed macropleural spine of unknown length. Anterior pleural furrow
considerably wider than other furrows; extends abaxially into small flat area near anterolateral corner of
pygidium. Between spines, posterior margin evenly curved, except behind axis where it is deflected to anterior
and also elevated slightly.

JAGO: LATE CAMBRIAN TRILOBITES

221

Discussion. The cranidium and the pygidia described above are included with some hesitation in
one species. They occur within a distinctive, slightly micaceous coarse siltstone to fine sandstone,
with the cranidium described above being the only one which is big enough to be matched with the
quite large pygidia. The shape of the cranidium suggests affiliation with the Leiostegiidae; it is
similar to the cranidium described by Lu and Qian (1983) as Chiiangia {Leptochuangia) benxiensis.
However, the pygidia described above have more pleural ribs than previously described species of
the Leiostegiidae and appear to be more closely related to the Kaolishaniidae than the Leiostegiidae.
Since the Leiostegiidae and Kaolishaniidae belong to the Leiostegiacea, the specimens are left in
open nomenclature within that superfamily.

Superfamily ceratopygacea Linnarsson, 1869
Family ceratopygidae Linnarsson, 1869

Subfamily proceratopyginae Wallerius, 1895
Genus proceratopyge Wallerius, 1895

Svnonvmy. See Palmer (1968, p. 53), to which should be added Proceratoprge (Henderson 1976, p. 332;
Shergold et al. 1976, p. 281; Lisigor 1977, p. 254; Yin and Lee 1978, p. 547; Yang 1978, p. 65; Shergold 1982,
p. 49; Rushton 1983, p. 131), Lopnorites (Yang 1978, p. 67), and Proceratopvge (Sinoproceralopvge) (Lu and
Lin 1980, p. 128).

Type species. Proceratopyge conifrons Wallerius, 1 895, p. 57, pi. 1 , hg. 6.

Discussion. Palmer (1968), Henderson (1976), and Shergold (1982) discussed Proceratopyge and its
possible subgeneric groupings. However, at present there is no general agreement on the validity or
otherwise of subgeneric divisions. Shergold et al. (1976) and Shergold (1982) followed Opik (1963)
in recognizing at least two subgenera, i.e. Proceratopyge {Proceratopyge) with five or less pygidial
axial annulations, and P. {Lopnorites) with more than six such annulations. Henderson (1976, p.
333) regarded such subdivisions as valueless, while Yang (1978, p. 67) maintained that Lopnorites
should retain full generic status. Lu and Lin (1980, p. 129) not only recognized P. {Proceratopyge)
and P. {Lopnorites) but also erected a third subgenus P. {Sinoproceratopyge), with P. kiangshanensis
Lu as type species. However, none of the six species placed in Sinoproceratopyge by Lu and Lin is
particularly well known and certainly do not justify the erection of a new subgenus. I follow
Rushton (1983) in regarding it as a synonym of Proceratopyge. Two species placed by Lu and Lin
(1980, p. 129) with Sinoproceratopyge, i.e. P. latilimbatus (recte latimbata) Zhou (see Zhou et al.
1977, p. 232, pi. 70, figs. 11-13) and P. latirhachis Zhou (ibid., figs. 14-16) appear from their figured
material to be synonyms.

Henderson (1976, p. 333) noted that Kogenium Kobayashi is of uncertain status, and that it
should probably be regarded as a synonym of Proceratopyge, a move followed by Rushton (1983)
and supported herein. Although Henderson (1976) suggested that useful subgeneric groupings are
not yet apparent, I believe that it is possible to split the species described under Proceratopyge and
Lopnorites into at least two broad groupings based on cranidial characters. The first group comprises
species which have small palpebral lobes placed well forwards, large posterolateral limbs, and
preocular sections of the facial suture which diverge only slightly; species include P. conifrons
Wallerius, 1895, P. niagnicauda Westergard, 1948, P. similis Westergard, 1947, P. nathorsti Wester-
gard, 1947, P. aff. nathorsti of Rushton (1983), P. rectispinata (Troedsson, 1937), P. cf. rectispinata
of Rushton (1983), P. fragilis (Troedsson, 1937), P. cylindrica Chien, 1961, and P. taojiangensis
Zhou, 1977. The type species of Proceratopyge and Lopnorites, respectively conifrons and rectispi-
nata, are included in this group, thus supporting Henderson’s view that Lopnorites is a valueless
taxon.

The second and much larger grouping comprises species possessing a palpebral lobe with a
semicircular outline which is generally placed centrally or centroposteriorly in relation to the
glabella. This group generally has strap-like posterolateral limbs and preocular sections of the facial
suture which diverge considerably; species include P. tullbergi Westergard, 1922, P. lata Whitehouse,
1939, P. nectans Whitehouse, 1939, P. gracilis Lermotova, 1940, P. liaotungensis Kobayashi and

222

PALAEONTOLOGY, VOLUME 30

Ichikawa, 1955, P. cf. P. liaotungensis of Shergold and Cooper (1985), P. asiatica Ivshin, 1956, P.
chuhsiensis Lu, 1956<2, P. cf. P. chuhsiensis of Palmer (1968); P. tenuita Lazarenko, 1966, P. capitosa
Lazarenko, 1966, P. fenghwangensis Hsiang, 1963, specimens figured as P. conifrons by Jegorova et
al. (1963, pi. 10, figs. 11, 12), P. constricta Lu, 1964 (see Lu et al. 1965, pi. 115, fig. 1), P.
kiangshenensis Lu, 1964 (ibid., fig. 3), P. cryptica Henderson, 1976, P. orthogonialis (Yang, 1978),
P. latilimbata Lee in Yin and Lee, 1978, P. cf. lata of Shergold et al. (1976), and P. sp. of Shergold
(1982). The two species described below, P. gordonensis sp. nov. and P. sp., belong to this group.
However, within the group there is considerable variation with respect to development of plectral
lines, length of preglabellar field, and pygidial characteristics. This group presumably corresponds
in part to a grouping noted by Shergold et al. (1976, p. 283) of species ‘characterized by rather
widely diverging facial sutures, well-developed plectral lines, long (exag.) palpebral lobes with strap-
like posterolateral limbs, and a pauci-furrowed pygidium’.

Species which cannot be placed in either group include P. truncata Yang in Zhou et al., 1977, P.
corrugis Romanenko, 1977, P. triangula Ivshin, 1962, P. longispina Ivshin, 1962, P. latilimbata Lee
in Yang, 1978, P. capitosa Lazarenko, 1966, P. latilimbata Zhou in Zhou et al., 1977, P. latirhachis
Zhou in Zhou et al., 1977, and P.1 brevirhacliis Zhou in Zhou et al., 1977. The species described by
Troedsson (1937) as Lopnorites grabaui may not belong in Proceratopyge.

Proceratopyge gordonensis sp. nov.

Plate 26, figs. 1-10; Plate 27, figs. 1-8

Diagnosis. Gently tapering glabella; short (sag.) concave preglabellar field; Ip lateral glabellar fur-
rows represented by pair of elongated pits, 2p and 3p furrows almost effaced. Semicircular palpebral
lobes placed close to glabella. Narrow strap-like posterolateral limbs. Librigenae with wide borders.
Pygidial axis with six or seven axial rings plus terminus. Very wide pygidial border. Pygidial pleural
areas with traces of three segments. Anterior segment extends into long straight rearwardly directed
spines.

Holotype. UT 88350a (PI. 26, fig. 1).

Material. One almost complete specimen (UT 88350a); several isolated incomplete cranidia (including UT
88447, 88521); one specimen with a complete thorax and attached pygidium, librigena, and hypostome (UT
88351); several specimens with a thorax or partial thorax with attached cranidium and/or pygidium (UT
88389b, 88486, 89405, 89412); and about twenty pygidia (including UT 88350b, 88353, 88449, 88501, 88522).

Description. Cephalon wider than long. Length of gently convex glabella (including occipital ring) c. 0-8 that
about half of cranidium. Between palpebral lobes, glabella width that of cranidium. Glabella tapers gently
forwards to broadly rounded anterior. Axial furrow moderately impressed. Very gently impressed preglabellar
furrow. Short, concave preglabellar field. Very gently impressed occipital furrow. Ip furrows represented by
pair of shallow pits just forward of posterior of palpebral lobes; 2p and 3p furrows faintly developed near

EXPLANATION OF PLATE 26

Figs. 1-10. Proceratopyge gordonensis sp. nov. 1, UT 88350a, holotype, almost complete specimen, external
mould (intermediate distortion), x 2. 2, UT 88389b, pygidium, most of thorax plus posterolateral limb of
cranidium, external mould, x 2. 3, UT 88351, pygidium, most of thorax, librigena and hypostome (see PI.
27, fig. 8), internal mould, x 1-5. 4, UT 88521, cranidium, external mould, x3. 5, UT 88389a, internal
mould of ventral side of librigena showing faint radiating caecal pattern, x 2. 6, UT 88392, librigena,
external mould, x 2. 7, UT 88486, specimen with pygidium, thorax, and most of cranidium, x 2. 8, UT
88447, partial cranidium, external mould, x 2. 9, UT 88350b, pygidium, internal mould, L form, x 2. 10,
UT 89412, from ‘middle fauna’, cranidium and anterior part of thorax, external mould, x 2.

Fig. 1 1. Cranidium gen. et sp. indet., UT 88487, internal mould, x 3.

Fig. 12. Eugonocare sp., UT 88361, from ‘bottom fauna’, cranidium, internal mould, x 3.

All specimens from ‘top fauna’ (see text-fig. 1) unless otherwise stated.

PLATE 26

JAGO, Proceratopygi\ Eugonocare

224

PALAEONTOLOGY, VOLUME 30

anterior of palpebral lobes. Small, circular, centroposteriorly placed glabellar node. Shallow border furrow
merges into narrow, very gently convex border. Semicircular, centrally placed, narrow, very slightly elevated
palpebral lobes close to glabella; very shallow palpebral furrow. Small palpebral and anterior areas of fixigenae
very narrow with moderately impressed posterior border furrow. Preocular sections of facial suture diverge
up to border furrow from where they converge; postocular sections of facial suture diverge very markedly.

Gently convex librigena with faint caecal pattern radiating away from eye socle. Gently impressed border
furrow; wide, almost flat border with faint terrace lines extends into narrow genal spine which also exhibits
terrace lines.

Hypostome with convex median body and well developed elongated maculae.

Thorax of nine segments; moderately impressed axial furrows. Axial width c. 0-25 that of segment. Each
segment about thirteen times as wide as long. Abaxial part of segments extend into spines which are directed
strongly to the posterior. Wide pleural furrows narrow at geniculation and extend well along spine.

Large pygidium, length just under 0-3 that of entire carapace. Pygidium almost twice as wide as long
(excluding axial half-ring). Convex axis outlined by moderately impressed axial furrows which shallow pos-
teriorly. Axis comprises six or seven axial rings plus terminus. Axis tapers evenly with a slight constriction at
third axial ring. Axial posterior bluntly rounded. Axis extends just on to border with narrow low ridge
extending to posterior of axis. Very wide flat border; very wide doublure with terrace lines. At posterior,
border has length (sag.) 0-25-0'3 that of pygidium. Three pairs of pleural furrows and three pairs of interpleural
furrows present, best seen on smaller pygidia (PI. 27, fig. 5; PI. 26, fig. 7); become more effaced in larger
pygidia (PI. 26, fig. 1). The first pleural segment extends into a pair of long, broad, straight spines bearing
terrace lines; spine length c. 1-3 that of pygidium. Posterior margin of pygidium broadly and evenly rounded.

Discussion. See discussion of P. sp. below.

Proceratopyge sp.

Plate 27, figs. 9-11

Material. Three specimens with most of the cranidium, thorax, and pygidium present (UT 88407, 89421,
89448). Two of these possess at least part of a librigena. A fourth specimen comprises the pygidium and part
of eight thoracic segments.

Description. Cephalon wider than long. Gently convex glabella tapers slightly forwards to broadly rounded
anterior. Axial furrow moderately impressed; very gently impressed preglabellar furrow. Short, concave
preglabellar field. Very gently impressed occipital furrow shallows abaxially. Lateral glabellar furrows almost
entirely effaced. Very small posteriorly placed node. Details of anterior border area nowhere well preserved.
Semicircular, centrally placed, narrow palpebral lobes close to glabella; palpebral furrows so shallow that it is
difficult to distinguish palpebral lobes from flat, small palpebral areas of fixigenae. Narrow posterior areas of
fixigenae with shallow border furrow. Preocular sections of facial suture diverge slightly; postocular sections
of facial suture diverge markedly.

Gently convex librigenae; gently impressed border furrow merges with almost flat border. Terrace lines on
both furrow and border. Border extends into gently curved spine which bears terrace lines and extends to
level of sixth thoracic segment.

EXPLANATION OF PLATE 27

Figs. 1-8. Proceratopyge gordonensis sp. nov. 1, LIT 89405, from ‘middle fauna’, pygidium and four thoracic
segments, internal mould, x 2. 2, UT 88501, pygidium with widely divergent spines, external mould, W
form, X 2. 3, UT 88350b, pygidium, internal mould, W form, x 2. 4, UT 88353, pygidium, external mould,
L form, X 2. 5, UT 88449, pygidium, external mould, L form, x 3. 6, UT 88522, pygidium, internal mould,
L form, X 5. 7, UT 88385, hypostome, external mould, x 5. 8, UT 88507, hypostome, internal mould
(counterpart of UT 88351; see PI. 26, fig. 3), x 5.

Figs. 9-11. Proceratopyge sp. 9, UT 89448, from ‘middle-fauna’, pygidium and posterior part of thorax,
external mould, x 2. 10, UT 89421, from ‘middle fauna’, almost complete specimen, external mould, x 2.
1 1, UT 88407, specimen with most of cephalon, partial thorax, and partial pygidium, external mould, x 2.

Figs. 12-14. Pseudoyuepingia vanensis sp. nov., from ‘middle fauna’, 12, UT 89414, pygidium, internal mould,
X 2. 13, UT 89433, pygidium, internal mould, x 3. 14, UT 89415, holotype, internal mould, x 2.

All specimens from ‘top fauna’ (see text-fig. 1) unless otherwise stated.

PLATE 27

JAGO, Proceratopyge, Pseudoyuepingia

226 PALAEONTOLOGY, VOLUME 30

Impression of hypostome crushed under glabella seen in PI. 27, figs. 10 and 1 1, but no separate hypostome
available for description.

Thorax of nine segments; moderately impressed axial furrows. Each segment about thirteen times as wide
as long. Abaxial parts of segments appear to extend into spines, but details not clear. Pleural furrows
moderately to gently impressed.

Pygidium length c. 0-2 0-25 that of entire specimen. Pygidium about twice as wide as long (excluding axial
half-ring). Axial furrows moderately impressed. Axis comprises five axial rings, plus terminus; length 0-75-
0-80 that of pygidium (excluding axial half-ring). Moderately shallow border furrow, with terrace lines and
gently convex border. Two pairs of pleural furrows and two pairs of interpleural furrows clearly visible.

Pair of gently curved, long, thin spines emerge from near posterior of anterior pleural segment and extend
past posterior of pygidium. Spines deflected outwards at point where they leave pygidial margin; at their
anterior they diverge slightly before becoming slightly convergent to posterior. Only two specimens show
posterior pygidial margin; that in PI. 27, fig. 9 is more sharply rounded than PI. 27, fig. 10.

Discussion. As discussed below, the specimens described can be differentiated from previously
described species of Proceratopyge, but their preservation is such that the erection of a new species
is not warranted. Proceratopyge sp. is quite close to P. gordonensis sp. nov. The lateral glabellar
and palpebral furrows of P. sp. are more effaced than those of gordonensis, and the cephalic border
of gordonensis is wider than that of P. sp. The two differ more clearly in pygidial characteristics:
the axis of gordonensis has six or seven axial rings, plus a terminus, while that of P. sp. has only
five, plus a terminus. The pleural details are clearer in gordonensis', the pygidial spines of gordonensis
are straight and broad, while those of P. sp. are thinner and curved. Both P. gordonensis and P. sp.
belong in the second species group noted in the generic discussion, hence the latter will be compared
only with species in this grouping.

Compared with previously described Australian species of Proceratopyge, P. gordonensis differs
from P. nectans, P. cryptica, and P. lata in not having a distinct plectrum, although this is a rather
variable feature (e.g. cf. cranidia of P. lata figured by Henderson 1976, figs. 5 and 8). The preglabellar
details of P. sp. are too poorly preserved to allow comparison of the plectral details. The glabella
of gordonensis is larger than that of nectans and cryptica, and longer than those in many speci-
mens of P. lata illustrated by Henderson (1976, pi. 48) and Shergold (1982, pi. 16). However, the
cranidia of P. lata illustrated by Henderson (1976, pi. 48, figs. 4, 10) and Shergold (1982, pi. 16, figs.
1 and 2) have a glabella of similar length to that of gordonensis. The pygidial spines of P. lata are
much finer than those of P. gordonensis. The extremities of the palpebral lobes of both P. gordon-
ensis and P. sp. are closer to the glabella than in lata, cryptica, nectans, or P. sp. of Shergold (1982).
The preglabellar field of P. cf. cliuhsiensis Lu of Opik (1963) is longer than that of either Tasmanian
species. The glabella of P. cf. lata Whitehouse illustrated by Shergold et al. (1976, pi. 40, fig. 1)
from northern Victoria Land, Antarctica, is shorter than that of either gordonensis or P. sp.; it is
less effaced than that of P. sp.

The glabella of the Swedish P. tullbergi is shorter than that of either Tasmanian species. Of the
various Chinese species of Proceratopyge, P. fenghwangensis Hsiang is probably closest to gordonen-
sis and P. sp. in cranidial characters. It differs, however, in having a distinct plectrum; the pre-
glabellar fields of both gordonensis and P. sp. are shorter than that of fenghwangensis', the base
of the pygidial spines of fenghwangensis is bigger than those of P. sp., and the pygidial border of
fenghwangensis is narrower than that of gordonensis.

Genus pseudoyuepingia Chien, 1961

Synonymy. Pseudoyuepingia Chien, 1961, p. 106, Lu et al. (1965, p. 506), Yin and Lee (1978, p. 534), Lu and
Lin (1980, p. 127). Iwavaspis Kobayashi, 1962, p. 122, Palmer (1968, p. 53), non Lazarenko in Datsenko et al.
(1968, p. 184).

Type species. Pseudoyuepingia modesta Chien, 1961, p. 106, pi. 5, figs. 5-7.

Diagnosis. Semicircular cranidium with long, very slightly tapered to parallel sided glabella which
has bluntly rounded anterior. Poorly developed to effaced lateral glabellar furrows. Shallow occipital

JAGO: LATE CAMBRIAN TRILOBITES

227

furrow. Short preglabellar field. Small, centro-anteriorly placed, semicircular palpebral lobes close
to glabella. Preocular sections of facial suture slightly divergent; postocular sections of facial suture
diverge markedly enclosing subtriangular posterior areas of fixigenae. Almost flat librigenae with
long genal spines.

Thorax of eight or nine segments; spinose pleurae. Large semicircular pygidium with low axis of
five or six axial rings plus terminus; axis extends to border. Pleural furrows in anterior part of
pleural areas better defined than at posterior where shallow or effaced. Narrow, flat pygidial
border.

Discussion. Pseudoyuepingia belongs in a group of trilobites which show characteristics of both the
Asaphidae and Ceratopygidae. As noted by Shergold (1982, p. 52), this group includes Yuepingia
Lu, 1956/), Iwayaspis Kobayashi, 1962, Eoasaphus Kobayashi, 1936, Norinia Troedsson, 1937,
Charchaqia Troedsson, 1937, Haniwoides Kobayashi, 1935, and Aplotaspis Henderson, 1976. To
this group should be added Metayuepingia Liu in Zhou et ai, 1977, Cerniatops Shergold, 1980, and
Yuepingioides Lu and Lin, 1984. Various authors have assigned the above genera to different
subfamilies, both within the Asaphidae or Ceratopygidae. However, I prefer to leave them in a
single group within the Ceratopygidae, as was done by Palmer (1968), Henderson (1976), and
Shergold (1982).

I follow Lu and Lin (1980, p. 127) in placing Iwayaspis (type species I. asaphoides Kobayashi,
1962, p. 122, pi. 6, figs. 1-10; pi. 8, fig. 24) in synonymy with Pseudoyuepingia, although the latter’s
type species, P. modesta, is not particularly well preserved (see Chien 1961, pi. 5, figs. 5-7, and Lu
et cd. 1965, pi. 103, figs. 1-3 where the type material is refigured). My diagnosis of Pseudoyuepingia
is based on P. modesta, P. asaphoides, P. zhejiangensis Lu and Lin, 1980, and P. vanensis sp. nov.
described below.

Pseudoyuepingia is close to Yuepingia but, as noted by Palmer (1968, p. 56), the palpebral lobes
of Yuepingia are placed further to the posterior than those of Pseudoyuepingia [= Iwayaspis in
Palmer] and the shapes of the posterolateral limbs of the cranidia are different. It can be argued
that such differences are of specific rather than generic importance, in which case Pseudoyuepingia
becomes a junior synonym of Yuepingia. However, Yuepingia is neither particularly well known
nor, with the exception of Y. glabra Palmer, 1968 (p. 56), particularly well illustrated, so I prefer to
treat Yuepingia and Pseudoyuepingia as separate genera.

Lazarenko in Datsenko et al. (1968, pp. 184-185) described two new species of Iwayaspis, I.
caelata and /. curta. Only two pygidia were figured for I. curta, so a detailed comparison with other
taxa is not possible. I. caelata has a relatively short glabella, large preglabellar field, small circular
palpebral lobes, and strap-like posterolateral areas of fixigenae; hence, it would appear to belong in
Yuepingia rather than Pseudoyuepingia. Metayuepingia was erected by Liu in Zhou et al. (1977)
with M. angustilimhata Liu in Zhou et al., 1977 (p. 216, pi. 64, figs. 1-3) as type species. Two other
species, M. intermedia and M. latilimbata, were erected by Liu in Zhou et cd. (1977, pp. 216-217);
M. intermedia appears to be a synonym of M. angustilimbata and it is possible that M. latilimbata
is so too.

Pseudoyuepingia vanensis sp. nov.

Plate 27, figs. 12-14; text-fig. 3e

Diagnosis. Semi-elliptical cephalon wider than long. Long glabella tapers slightly to broadly rounded
glabellar anterior. Gently impressed axial and preglabellar furrows. Shallow occipital furrow. Short,
almost fiat preglabellar field; almost fiat border. Lateral glabellar furrows effaced. Centro-anteriorly
placed, semicircular, palpebral lobes close to glabella. Large triangular posterolateral areas of
fixigenae. Preocular sections of facial suture diverge slightly up to border furrow, from where they
converge markedly; sinuous postocular sections of facial suture diverge markedly. Gently convex,
smooth, wide librigenae. Thorax of nine segments. Large semicircular pygidium with low axis
comprising five or six axial rings plus terminus. Axial rings become poorly defined to posterior.

228

PALAEONTOLOGY, VOLUME 30

Poorly defined pleural furrows on anterior part of pleural areas; posterior part of pleural areas
smooth. Narrow, flat pygidial border.

Holotype. UT 89415 (PI. 27, fig. 14).

Material. One incomplete specimen (UT 89415), one incomplete cranidium, and five pygidia (including UT
89414, 89415, 89433).

Description. Cephalon has semicircular outline; wider than long. Length of gently convex glabella (including
occipital ring) c. 0-8 that of cranidium. Between palpebral lobes, glabella has width about half that of
cranidium. Glabella tapers very slightly forwards to broadly rounded anterior. Gently impressed axial and
preglabellar furrows. Short, almost flat preglabellar field slopes slightly down to shallow border furrow.
Narrow, slightly elevated border. Shallow occipital furrow. Lateral glabellar furrows effaced. Presence or
absence of glabellar node not determined. Semicircular, centro-anteriorly placed, slightly elevated palpebral
lobes close to glabella; very shallow palpebral furrow. Small, almost flat palpebral and anterior areas of
fixigenae with shallow border furrow. Preocular sections of facial suture diverge slightly up to border furrow,
from where they converge markedly; sinuous postocular sections of facial suture diverge markedly.

Gently convex, smooth, wide librigena with very shallow border furrow and narrow flat border. Presence
or absence of genal spines not determined. Hypostome unknown.

Thorax of nine segments; moderately deep axial furrows. Axial width c. 0-25 that of segment. Shallow
pleural furrows deepen and narrow abaxially. Pleurae appear to extend into short spines.

Large, gently convex, semicircular pygidium; length c. 0-3 that of entire carapace; wider than long. Low,
gently convex axis; moderately deep axial furrows. Axis comprises five or six axial rings, plus terminus; only
anterior three axial rings clearly distinguished. Axis has slight centro-posteriorly placed constriction. Up to
three poorly defined pleural furrows distinguished in anterior part of pleural areas; posterior part of pleural
areas smooth. Shallow border furrow; narrow flat border. Posterior margin evenly curved.

Discussion. Pseudoyuepingia vanensis differs from P. modesta and P. asaphoides in that it has a more
effaced glabella, which is also narrower than those of modesta, asaphoides, and P. zhejiangensis Lu
and Lin, 1980 (p. 127, pi. 2, figs. 8 and 9). The palpebral lobes of vanensis are smaller than those of
other species of Pseudoyuepingia, and placed closer to the glabella than those of modesta. P.
zhejiangensis has eight thoracic segments whereas both P. vanensis and P. modesta have nine. The
pleural furrows of vanensis are shallow and become effaced towards the posterior, as do those of
modesta and zhejiangensis; those of asaphoides are better developed.

cranidium, gen. et sp. indet.

Plate 26, fig. 1 1

Material. Partial cranidium, UT 88481.

Description. This very poorly preserved specimen is figured for completeness.

Acknowledgements. The fossils were collected by Dr K. D. Corbett (Geological Survey of Tasmania); their
cataloguing, and transport from Hobart to Adelaide, was arranged by Dr M. R. Banks (University of
Tasmania). I thank Drs J. H. Shergold and J. Laurie (Bureau of Mineral Resources, Canberra) and Professor
R. J. Henderson (James Cook University, Townsville) for reviewing a draft of this paper for me. The work
was supported by a grant from the Australian Research Grants Scheme.

REFERENCES

APPOLLONOV, M. K. and CHUGAEVA, M. N. 1983. Some trilobites from the Cambrian -Ordovician boundary,
Batyrbaisai Valley, Maly Karatau. In appollonov, m. k., bandaletov, s. m. and ivshin, n. k. (eds.). The
Lower Palaeozoic stratigraphy and palaeontology of Kazakhstan, 66-90, pis. 7-10. Akad. Nauk Kazakh.
SSR, Alma-Ata. [In Russian.]

BELL, w. c. and ellinwood, h. l. 1962. Upper Franconian and Lower Trempealeuan Cambrian trilobites and
brachiopods, Wilberns Formation, Central Texas. J. Paleont. 36, 385-423, pis. 51-64.

JAGO: LATE CAMBRIAN TRILOBITES

229

BRADLEY, J. H. 1925. Trilobites of the Beekmantown in the Philipsburg region of Quebec. Can. Fid Nat. 39, 7-8.

CHiEN Yi-YUAN. 1961. Cambrian trilobites from Sandu and Duyun, southern Kweichow. Acta palaeont. sin. 9,
91-139, pis. 1-5. [In Chinese with English summary.]

CORBETT, K. D. 1972. Features of thick-bedded sandstones in a proximal flysch sequence. Upper Cambrian,
south-west Tasmania. Sedimentology, 19, 99-114.

1973. Open-cast slump sheets and their relationship to sandstone beds in an Upper Cambrian flysch

sequence, Tasmania. J. sedim. Petrol. 43, 147-159.

1975. The Late Cambrian to Early Ordovician sequence on the Denison Range, southwest Tasmania.

Pap. Proc. R. Soc. Tasm. 109, 111-120.

DATSENKO, V. A., ZHURAVLEVA, I. T., LAZARENKO, N. P., POPOV, YU N. and CHERNYSHEVA, N. E. 1968. Biostratigra-
phy and fauna of the Cambrian deposits of the north-west part of the Siberian Platform. Trudy nauchno-
issled. Inst. Geol. Arkt. 155, 213 pp., 23 pis. [In Russian.]

ERGALIEV, G. KH. 1980. Trilobites of the Middle and Upper Cambrian of the Maly Karatau, 211 pp., 20 pis.
Akad. Nauk Kazakh. SSR, Alma-Ata. [In Russian.]

FORTEY, R. A. 1980. The Ordovician trilobites of Spitsbergen. III. Remaining trilobites of the Valhallfonna
Formation. Skr. norsk Polarinst. 171, 1-163, pis. 1-25.

HARRINGTON, H. J. et al. 1959. Arthropoda 1. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part
O, 560 pp. Geological Society of America and University of Kansas Press, New York and Lawrence,
Kansas.

HENDERSON, R. A. 1976. Upper Cambrian (Idamean) trilobites from western Queensland, Australia. Palaeon-
tology, 19, 325-364, pis. 47-51.

HENNiNGSMOEN, G. 1960. The Middle Ordovician of the Oslo region, Norway 13. Trilobites of the family
Asaphidae. Norsk, geol. Tidsskr. 40, 203-257, pis. 1-14.

HOWELL, B. F. 1935. Cambrian and Ordovician trilobites from Herault, southern France. J. Paleont. 9, 222-
238, pis. 22 and 23.

IVSHIN, N. K. 1956. Upper Cambrian trilobites of Kazakhstan, part /, 1 19 pp., 9 pis. Akad. Nauk Kazakh. SSR,
Alma-Ata. [In Russian.]

1962. Ibid., part 2, 412 pp., 21 pis. Akad. Nauk Kazakh. SSR, Alma-Ata. [In Russian.]

JAEKEL, o. 1909. liber die Agnostiden. Z. dt. geol. Ges. 61, 380-401.

JAGO, J. B. 1976. Late middle Cambrian agnostid trilobites from north-western Tasmania. Palaeontology, 19,
133-172, pis. 21-26.

1979. Tasmanian Cambrian biostratigraphy— a preliminary report. J. geol. Soc. Aust. 26, 223-230.

JEGOROVA, L. L, HSIANG, L. w., LEE, s. c., NAN, J. s. and KUO, c. M. 1963. The Cambrian trilobite faunas of
Kueichou and western Hunan. Spec. Pap. Inst. Geol. Min. Resour., Peking, {B), Strut. Palaeont. 3(1), 117
pp., 15 pis.

KOBAYASHi, T. 1935. The Cambro-Ordovician formations and faunas of South Chosen. Palaeontology. Part
III. The Cambrian faunas of South Chosen with a special study on the Cambrian trilobite genera and
families. J. Fac. Sci. Tokyo Univ. (sect. 2), 4, 49-344, pis. 1-24.

1936. Three contributions to the Cambro-Ordovician Faunas. I. The Dikelokephalininae (nov.), its

distribution, migration, and evolution. Jap. J. Geol. Geogr. 13, 163-184.

1962. The Cambro-Ordovician formations and faunas of South Korea, part IX, Palaeontology VIII. J.

Fac. Sci. Tokyo Univ. (sect. 2), 14, 1-152, pis. 1-12.

and ICHIKAWA, T. 1955. Discovery of Proceratopyge in the Chuangia Zone in Manchuria, with a note on

the Ceratopygidae. Trans. Proc. palaeont. Soc. Japan (ns), 19, 65-72, pi. 11.

LAKE, p. 1906-1946. British Cambrian trilobites. Palaeontogr. Soc. (Monogr.), 350 pp., 47 pis.

LAZARENKO, N. p. 1966. Biostratigraphy and some new trilobites from the Upper Cambrian of the Olenek
Uplift and Karaulakh Mountains. Uchen. Zap. nauchno-issled. Inst. geol. arkt., Paleont. Biostratigr. 11, 33-
78, pis. 1-7. [In Russian.]

and NIKIFOROV, h. i. 1968. Trilobite assemblage from Upper Cambrian deposits of the Kulyumbe River

(north-western Siberian Platform). Ibid. 23, 20-80, pis. 1-15. [In Russian.]

LERMONTOVA, E. 1940. Arthropoda. In vologdin, a. et al. (eds.). Atlas of the leading forms of the fossil faunas of
the U.S.S.R. 1, Cambrian, pis. 1 -49. State Editorial Office for Geological Literature, Moscow. [In Russian.]

linnarsson, j. g. o. 1869. Om Vestergdtlands Cambriska och Siluriska Aflagringar. K. svenska VetenskAcad.
Handl. 8 (2), pis. 1 and 2. [In Swedish.]

LisoGOR, K. A. 1977. Biostratigraphy of Upper Cambrian and Tremadoc trilobites of Maly Karatau (southern
Kazakhstan). In Zhuravleva, i. t. and rosova, a. v. (eds.). Trudy Inst. Geol. Geofiz. sib. Otd. 313, 197-265,
pis. 28-31. [In Russian.]

230

PALAEONTOLOGY, VOLUME 30

LU YEN-HAO. 1956fl. On the occurrence of Lopnorites in northern Anhwei. Ada palaeont. sin. 4, 267-284, 1 pi.
[In Chinese with English summary.]

\956b. An Upper Cambrian faunule from eastern Kueichow. Ibid. 365-380, 1 pi. [In Chinese with English

summary.]

CHANG, w. T., CHU CHAO-LING, CHIEN Yi-YUAN and HSIANG LEE-WEN. 1965. Chinese fossUs of all groups.

Trilobita. 1, 1-362, pis. 1-66; 2, 363-766, pis. 67-135. Science Publ. Co., Peking. [In Chinese.]

et al. 1974. Bio-environmental control hypotheses and its application to the Cambrian biostratigraphy

and palaeozoogeography. Mem. Nanking Inst. Geol. Palaeont. 5, 27-1 16, pis. 1-4. [In Chinese.]

and LIN HUAN-LiNG., 1980. Cambro-Ordovician boundary in western Zhejiang and the trilobites contained

therein. Acta palaeont. sin. 19, 118-134, pis. 1-3. [In Chinese with English summary.]

1984. Late Late Cambrian and earliest Ordovician trilobites of Jiangshan-Changshan area, Zheji-
ang. In Stratigraphy and Palaeontology of Systematic Boundaries in China, Cambrian-Ordovician Boundary,
1, 45-143, pis. 1-19. Anhui Science and Technology Publishing House.

and QIAN YI-YUAN. 1983. New zonation and correlation of the Upper Cambrian Changshanian stage in

North China. Acta palaeont. sin. 22, 235-254, pis. 1-3. [In Chinese with English summary.]

MATTHEW, G. F. 1887. Illustrations of the fauna of the St. John Group, No. IV. Pt. I. Description of new
species of Paradoxides (Paradoxides regina). Pt. II. The smaller trilobites with eyes (Ptychoparidae and
Ellipsocephalidae). Trans. R. Soc. Can. 5, 115-166, pis. 1 and 2.

M'COY, F. 1849. On the classihcation of some British fossil Crustacea, with notices of new forms in the University
collection of Cambridge. Ann. Mag. nat. Hist. (ser. 2), 4, 161-179, 392-414.

MILLER, s. A. 1889. North American geology and paleontology for the use of amateurs, students and scientists,
718 pp., 1,265 figs. Cincinnati, Ohio.

OPiK, A. A. 1963. Early Upper Cambrian fossils from Queensland. Bull. Bur. Miner. Resour. Geol. Geophvs.
64, 1-133, pis. 1-9.

1967. The Mindyallan fauna of north-western Queensland. Ibid. 74, vol. 1, 404 pp.; vol. 2, 167 pp., 67

pis.

PALMER, A. R. 1960. Trilobites of the Upper Cambrian Dunderberg Shale, Eureka District, Nevada. Prof. Pap.
U.S. geol. Surv. 334-C, 53-109, pis. 4-11.

1962. Glyptagnostus and associated trilobites in the United States. Ibid. 374-F, 1-49, pis. 1-6.

1965. Trilobites of the Late Cambrian Pterocephaliid Biomere in the Great Basin, United States. Ibid.

493, 1-105, pis. 1-23.

1968. Cambrian trilobites of East-Central Alaska. Ibid. 559-B, 1-105, pis. 1-15.

POWELL, c. MCA., NEEF, G., CRANE, D., JELL, p. A. and PERCivAL, I. G. 1982. Significance of Late Cambrian
(Idamean) fossils in the Cupala Creek Formation, northwestern New South Wales. Proc. Linn. Soc.
N.S.W. 106, 127-150.

Qiu HONG-AN. 1984. Trilobites from the Upper Cambrian Tangcun Formation in Jingxian, southern Anhui.

Acta palaeont. sin. 23, 329-341, pis. 1-3. [In Chinese with English summary.]

RESSER, c. E. 1935. Nomenclature of some Cambrian trilobites. Smithson, misc. Colins, 93 (5), 1-46.

1938. Cambrian system (restricted) of the southern Appalachians. Spec. Pap. geol. Soc. Am. 15, 1-139,

pis. 1-16.

ROBISON, R. A. 1982. Some Middle Cambrian agnostid trilobites from western North America. J. Paleont. 56,
132-160, pis. 1-6.

ROMANENKO, K. A. 1977. Cambrian trilobites from the Bol'shaya Isha River Section (north-east of the Altai).

In ZHURAVLEVA, I. T. and ROSOVA, A. V. (eds.). Trudy Inst. Geol. Geofiz. sib. Otd. 313, 161-184, pis. 23-25.
ROSOVA, A. v. 1963. Biostratigraphic scheme for subdividing the Upper Cambrian and upper part of the Middle
Cambrian of the northwestern Siberian Platform and new Upper Cambrian trilobites from the Kulyumbe
area. Ibid. 9, 3-19, pis. 1 and 2. [In Russian.]

1964. Biostratigraphy and description of Middle and Upper Cambrian trilobites from the northwest

Siberian Platfonn. Ibid. 1-106, pis. 1-19. [In Russian.]

1968. Biostratigraphy and trilobites of tbe Upper Cambrian and Lower Ordovician of the northwestern

Siberian Platform. Ibid. 36, 1 196, pis. 1-17. [In Russian.]

1977. Some Upper Cambrian and Lower Ordovician trilobites from the Rybnaya, Khantajka, Kurejka

and Letnyaya River Basins. In zhuravleva, i. t. and rosova, a. v. (eds.). Ibid. 313, 53-84, pis. 1-12.
[In Russian.]

RUSHTON, A. w. A. 1983. Trilobites from the Upper Cambrian Olenus Zone in Central England. In briggs,
D. E. G. and LANE, p. D. (eds.). Trilobites and other early arthropods: papers in honour of Professor H. B.
Whittington, F.R.S. Spec. Pap. Palaeont. 30, 107-139, pis. 14-19.

JAGO: LATE CAMBRIAN TRILOBITES

231

SCHRANK, E. 1974. Kattibrische Trilobiten der China-Kollektion v. Richthofen. Z. geol. fViss. Berlin, 2, 617-
643, pis. 1-5.

SHERGOLD, J. H. 1975. Late Cambrian and Early Ordovician trilobites from the Burke River Structural Belt,
Western Queensland, Australia. Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 153, 1-271, pis. 1-58.

1977. Classihcation of the trilobite Pseudagnostus. Palaeontology, 20, 69-100, pis. 15 and 16.

1980. Late Cambrian trilobites from the Chatsworth Limestone, Western Queensland. Bull. Bur. Miner.

Resour. Geol. Geophys. Aust. 186, 1-1 11, pis. 1-35.

1982. Idamean (Late Cambrian) trilobites, Burke River Structural Belt, Western Queensland. Ibid. 187,

1-70, pis. 1-17.

and COOPER, R. A. 1985. Late Cambrian trilobites from the Mariner Group, northern Victoria Land.

B. M. R. Jour. Aust. Geol. Geophys. 9, 91-106.

MACKINNON, D. I. and YOCHELSON, E. L. 1976. Late Cambrian Brachiopoda, Mollusca, and Trilobita

from northern Victoria Land, Antarctica. Palaeontology, 19, 247-291, pis. 38-42.

and SDZUY, k. 1984. Cambrian and early Tremadocian trilobites from Sultan Dag, central Turkey.

Senckenberg. leth. 65, 51-135, pis. 1-8.

swiNNERTON, H. H. 1915. Suggestions for the revised classification of trilobites. Geol. Mag. (dec. 6), 2, 487-
496, 538-545.

troedsson, g. t. 1937. On the Cambro-Ordovician faunas of western Quruq Tagh, eastern T’ien-Shan.

Palaeont. sin. (ns) B 92 (whole ser. 106), 1-74, pis. 110.

TULLBERG, s. A. 1880. Agnostus-nxiQna i de Kambriska aflagringarne vid Andrarum. Sver. Geol. Unders. Afh.
C42, 1-37, 2 pis. [In Swedish.]

WALLERIUS, I. D. 1895. Utidersokninga dfver zonen med Agnostus laevigatas i Vestergdtiand jamte en inledande
ofversikt af Vestgotlands Samtiiga paradoxides-lager. Akad. afhandl., Lund.

WESTERGARD, A. H. 1922. Sveriges Olenidskiffer. Sver. geol. Unders. Afh. Cal8, 1-188 (Swedish), 189-205
(English), pis. 1-16.

1947. Supplementary notes on the Upper Cambrian trilobites of Sweden. Ibid. C489, Arsb. 41(8), 3-34,

pis. 1-3.

1948. Non-agnostidean trilobites of the Middle Cambrian of Sweden. Ibid. C498, Arsb. 42 (7), 3-32, pis.

1-4.

WHiTEHOUSE, F. w. 1936. The Cambrian faunas of north-eastern Australia. Parts 1 and 2. Mem. Qd Mas. 11,
59-112, pis. 8-10.

1939. The Cambrian faunas of north-eastern Australia. Part 3, the polymerid trilobites (with supplement

no. 1). Ibid. 179-282, pis. 19-25.

YANG jiA-LU. 1978. Middle and Upper Cambrian trilobites of Western Hunan and Eastern Guizhou. Prof.

Pap. Stratigr. Palaeont. 4, 1-74, pis. 1-13. [In Chinese with English summary.]

YIN GONG-SHENG and LEE SHAN-Ji. 1978. trilobita, pp. 385-594, pis. 144-192. In Fossils of south-west China'.

Guizhou. Volume 1, Cambrian to Devonian periods, 843 pp., 214 pis. Earth Sci. Press, Beijing. [In Chinese.]
ZHOU TIAN-MEI, LIU Yi-REN, MENG xiAN-SONG and SUN ZHOU-HUA. 1977. Atlas of the palaeontology of central
southern China. 1, Trilobita, pp. 104-266, pis. 36-81. Geology Press, Beijing. [In Chinese.]

ZHURAVLEVA, I. T. and ROSOVA, A. v. (eds.). 1977. Biostratigraphy and fauna of the Upper Cambrian and
boundary strata. Trudy Inst. Geol. Geofiz. sib. Otd. 313, 1-355, pis. 1-31. [In Russian.]

J. B. JAGO

School of Applied Geology
South Australian Institute of Technology

Typescript received 23 August 1985 PO pox 1, Ingle Earm

Revised typescript received 26 June 1986 South Australia 5098

■ ,; ■■ 3:

->

■

' : ■ ,"■■■■.■

.' . I •‘‘.i- , % .- ‘(ifen'M" ■'®!

.i-. - .

•w5>

L

V"

)

A •

'X\

I

0

M 'I i i

1.'^

THE PRESERVATION OF CONIFER WOOD:
EXAMPLES FROM THE LOWER CRETACEOUS

OF ANTARCTICA

by TIMOTHY H. JEFFERSON
{deceased September 1983)

Abstract. The non-marine, upper part of the Fossil Bluff Formation (Lower Cretaceous) in Alexander
Island, Antarctica, contains abundant fossil wood. Fine details of cell wall structures (including those produced
by fungal and bacterial delignification) and relatively coarse cell-fill mineral textures indicate that silicification
took place in two main stages. A silica overgrowth on cellulose microfibrils suggests early cell wall impregnation
involving the precipitation of a very thin (?monomolecular) silica film on these structures. A later and much
slower lumen fill is indicated by centripetal, euhedral quartz crystals and by collophane and apatite within
cells.

Abundant silicified wood is found in the upper part of the Fossil Bluff Formation which is
exposed on the south-east coast of Alexander Island, Antarctica, as a series of cliffs, ridges, and
isolated nunataks (text-fig. 1). This part of the formation is thought to be of Barremian to Albian
age (Taylor et al. 1979; Jefferson 1981 ). The exposures are separated by glaciers up to 10 km across
and from the mainland Antarctic Peninsula by the ice-bound George VI Sound. The formation was
deposited in a fore-arc basin to the west of a calc-alkaline volcanic arc (Taylor et al. 1979; Suarez
1976). During the late Early Cretaceous, delta-top fluvial and lacustrine processes dominated
sedimentation in the south of this basin. Most of the sediment was deposited as channel sands,
overbank-flood sands, and crevasse-splay sands and silts, but finely laminated silt and fine sand
deposits were also important.

These sediments incorporated a large amount of plant material. Although diverse assemblages
of leaves, stems, and seeds were preserved as compressions in mudstones, siltstones, and fine
sandstones (Jefferson 1981, 1982fl), the best preserved wood is found in coarse-grained porous
volcaniclastic sandstones. Silicification depended on the early breakdown of volcanic components
and the mobility of the resultant mineralizing fluids. Groups of trees were silicified in growth
position as fossil forests. Growth rings within these trees are well preserved. The locations and
stratigraphic positions of well preserved fossil wood were given, and the forests and their palaeo-
climatic significance were discussed by Jefferson (1982^).

Although two types of wood mineralization (silicification and calcification) were found in fossil
wood from the Fossil Bluff Formation, calcified wood fragments were found at only one locality in
a marine sequence. This paper is concerned only with the silicified wood from the non-marine rocks
in this formation.

METHODS

Cell structures and preservational textures were studied both by microscopic examination of acetate peels and
thin sections made in the radial longitudinal, tangential longitudinal, and transverse planes (text-fig. 2a), and
by examination under the SEM. Acetate peels were prepared by etching material in 10% hydrofluoric acid
for 3 to 5-5 minutes and using the standard palaeobotanical peel techniques (Taylor 1981 ). Fractured surfaces
were prepared for SEM work by cleaving blocks up to 1 cm^ from specimens. Fragments were then glued to

(Palaeontology, Vol. 30, Part 2, 1987, pp. 233-249, pis. 28-30.|

© The Palaeontological Association

34

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1. Locality maps, a, Antarctica, showing the position of Alexander Island, the Fossil Bluff
Formation (FBF), and the position of text-fig. lb. b, Southeastern Alexander Island showing the

localities mentioned in the text.

JEFFERSON: CRETACEOUS CONIFER WOOD

235

TRANSVERSE

b

_ RADIAL
LONGITUDINAL

Late wood

Growth ring
boundary

TANGENTIAL

LONGITUDINAL

:arly wood^^i^^'

Primary wa

Secondary wal

'inner layer (S3)

middle layer (S2)

outer layer (SI]

tracheids

bordered

pits

TEXT-FIG. 2. The structure of wood, a. The three planes in which fossil wood was sectioned or split in order to
study its structure, b, Diagram of the structure of coniferous wood showing the anatomical features seen in
the Alexander Island fossil woods, c. The constituent layers of the tracheid wall in extant conifers (after

Wardrop and Bland 1959).

aluminium stubs and sputter coated with gold-palladium. Because files of medullary rays run out in a radial
direction from the centre of woody stems (text-fig. 2b), the wood of extant conifers splits in a radial longitudinal
plane. Similarly, the Alexander Island fossil wood cleaved naturally in this plane and produced some excellent
radial surfaces. However, cleavage in transverse and tangential longitudinal planes is ‘against the grain' of
tracheids or rays, and resultant surfaces were irregular and rarely show clear anatomical details. Some
specimens were polished, and other were polished and then etched in 10 % hydrofluoric acid. Polishing and
etching of specimen surfaces did not improve their quality, and destroyed many of the features seen on broken

236

PALAEONTOLOGY, VOLUME 30

radial surfaces. The etching of material not only removed silica from the cell lumen, but also caused collapse
of the cell wall; this is thought to be a result of the intimate mineralization of the wall (see below). Although
there is sufficient residual organic material in the cell wall to allow differential etching for peel preparation,
there is not sufficient to prevent the wall from collapsing when deep etching occurs.

Qualitative energy dispersive analysis by X-ray (EDAX) was carried out on several SEM stubs to establish
the relative quantities of silica and organic material in cell wall structures. This was substantiated with
quantitative EDAX using an electron microprobe on polished and gold palladium-coated thin-sections; this
method was also used to determine the distribution of other minerals filling cell lumina.

PRESERVATION OE THE CELL WALL

Arnold (1941) recognized that, because of the inert nature of cell materials and their very different
reactions with possible solvents, molecular replacement of cells by silica would be extremely unlikely.
He considered that silicification was by a process of infiltration, a view which has been supported
by detailed study of fossil material (e.g. Buurman 1972; Leo and Barghoorn 1976). Experiments on
silicification have resulted in the coating of the cell wall and in the filling of the cell lumen (Leo and
Barghoorn 1976). The present study, however, strongly suggests that infiltration and impregnation
of the cell wall was the first phase of silicification in all the well preserved Alexander Island fossil
wood. In cases in which the cell walls were not impregnated by silica the walls have been reduced
to lines of inclusions within quartz or chalcedony and the wood is poorly preserved. Although in
thin-sections of well-preserved wood the cell wall appears to be continuous and up to 10 /im thick
(PI. 28, figs. 1 and 2), electron microprobe analysis using EDAX indicates that very little of the
dark cell wall area is wholly occupied by carbon. Analyses in the centre of the cell wall show 82 %
to 87 % silica (mean 85-35 %), and other low atomic weight elements (i.e. hydrogen and oxygen in
hydrated silica) may make up part of the remaining 15 %. This clearly indicates that the various
components of the cell wall have been silicified. The dark colouration in the wall is due to the even
distribution of residual organic material throughout the cell wall area. It is thought that this organic
material represents remnant microfibrillar ‘threads’ around which silica grew (see below).

When Alexander Island fossil wood is viewed under the SEM the cell walls can be seen to be
composed of three layers, each with a silica overgrowth (PI. 28, figs. 4-7). These can be compared
with the constituent layers of the tracheid wall in modern gymnosperm wood (text-fig. 2c). Although
the most common and steepest of the fossilized helical structures could have been a result of helical
thickening, checking, or splitting, this is considered unlikely. Helical ‘checks’ follow the orientation
of the S2 layer microfibrils in the reaction wood of living conifers (Cote and Day 1965; Meylan and
Butterfield 1972), but checks do not overlap in the primary walls or SI cell layers, and the S3 layer
is absent. In the Alexander Island fossil woods the primary cell wall, the SI layer, and probably the
S3 layer, bear helical structures very similar to those interpreted as comprising the S2 layer.

EXPLANATION OF PLATE 28

Preservation of the cell wall in silicified conifer wood.

Figs. 1 and 2. Transverse thin sections. 1, KG. 2815. 71, late-wood tracheid with thick cell walls, x 1000. 2,
KG. 1702.2, early wood tracheids with thin cell walls, x 400.

Fig. 3. Tangential longitudinal acetate peel, KG. 1719.3c, continuous middle lamelli within cell walls,
X 250.

Figs. 4-7. Scanning electron micrographs of tangential fracture surfaces of tracheids, KG.28 17.23. 4, the two
outer layers of the cell wall, x 1000. 5 and 6, two cell wall layers, x 250, and detail, x 500. 7, the thickest
layer (steep helix) underlying a thinner layer (flat helix), x 500.

Fig. 8. Scanning electron micrograph of radial fracture surface, KG.28 17.5, helical thickenings in inner wall
layer of tracheid, x 1000.

Figs. 9-13. Radial longitudinal thin sections. 9, KG. 1704. 10, showing steep helical structure, x 1000. 10,
KG. 1704.10, flat helical structures, x400. 11, KG. 1719. 3c, fibrillar structures on bordered pits, x 800. 12,
KG.28 17. 16, fungal hyphae, x400. 13, KG. 1704. 10, ?fungal spores, x400.

PLATE 28

JEFFERSON, silicified conifer wood

238

PALAEONTOLOGY, VOLUME 30

Furthermore, much of the fossil wood from Alexander Island comes from areas of trunks which
show no sign of the development of reaction wood. Helical thickenings on tracheid walls in a few
living conifers are on the same scale as the helical structures in the Alexander Island fossil woods,
but they are considered to be extensions of the S3 layer and do not form the characteristic steep
helices (Meylan and Butterfield 1972). Drying and splitting of cell wall layers parallel to microfibrils
produce helical structures in modern woods which are more irregular and have angular terminations
(Cote and Day 1965). The most likely origin for the fossilized structures is growth of silica on to
the microfibrils and macrofibrils (microfibrillar bundles) within the individual layers of the cell wall.
The difference in scale between the fossilized fibrillar structures and the microfibrils in extant woods
is thought to be related to biogenic decay and is discussed below. Carbonized remnants of the
original cell walls, including the walls around pits, can sometimes be seen in thin section (PI. 28,
figs. 9-11), providing further evidence that the process is one of growth of silica on to and within
the cell wall layers rather than one of replacement.

Although all the silica seen in cell walls was in the form of chalcedony or cryptocrystalline quartz,
silicic acid monomers probably formed the first phase of silica precipitation (Leo and Barghoorn
1976). Subsequent transformation and ordering of silica through Opal CT would be expected to
take place within 10 my to 100 my (Stein 1982).

Morphology of cell wall layers

Although a complete section through all of the cell wall layers has not been seen, most of the layers
can be recognized individually.

Middle lamella. This is generally poorly preserved. Silica has often grown between adjacent cells
exerting little disruptive pressure and leaving the middle lamella intact (PI. 28, fig. 3). Sometimes
the middle lamella is present only as a line of carbon inclusions. This may be because of early fungal
breakdown of the middle lamella as seen in modern woods (Kaarik 1974). In modern woods the
thickness of the middle lamella varies considerably; at simple cell to cell boundaries it is often less
than 0-2 ^m thick and is often difficult to distinguish, whereas at triple point boundaries, or where
cells taper, it may reach 2 pm across (Harada 1965). In the Alexander Island fossil woods the middle
lamella cannot be positively identified under the SEM, although where the tracheids taper and the
inter-cellular space is at a maximum a structureless layer can be seen (PI. 28, fig. 4). This may
represent the middle lamella or an ‘intercellular substance’ (Harada 1965) but is more likely to be
an intercellular space created by post-mortem decay and shrinkage, and filled by silica.

Primary cell wall. Of the three wall layers recognized in Alexander Island fossil wood, the outer
layer is discontinuous, very thin, and possesses an indistinct structure almost perpendicular to the
tracheid axis (PI. 28, fig. 7). This is similar to the regular transverse orientation of the microfibrils
on the internal surface of the primary cell wall in extant gymnosperms (Wardrop and Bland 1959;
Mark 1967). The more irregular arrangement of microfibrils on the internal surface of this wall in
extant gymnosperms may explain the indistinct nature of the structures in the Alexander Island
fossil wood.

Outer layer of the secondary cell wall {SI layer). The second layer seen in Plate 28, figs. 4 and 7 is
composed of distinct helical structures at an angle of 80-85° to the axis of tracheids. This layer is
thought to represent the SI layer and it is often so thin that the next layer is clearly visible underneath
it when viewed with the SEM (PI. 28, fig. 7). When seen in longitudinal thin-section the structures
appear as dark, organic-rich strands spiralling in either direction (PI. 28, fig. 10). Harada (1965)
and Mark (1967) stated that the microfibrils of both the SI layer and the S3 layer in extant
gymnosperm woods form flat helical layers and spiral in alternate directions. Since it is not possible
to determine whether the helical structures are internal or external to the S2 layer, it is not known
whether they represent the SI or the S3 layer. The structures are seen in only 15-20 % of cell walls
in the Alexander Island fossil woods. This frequency in fossils is consistent with the relatively small
thickness of both the SI layer and the S3 layer in the tracheid walls of extant conifers; in Pinus

JEFFERSON: CRETACEOUS CONIFER WOOD

239

densifiora the SI layer comprises only 12-5 %, and the S3 layer only 7 %, of the total wall thickness
(SI +S3 = 19-5 %) (Harada 1965).

Middle layer of the secondary cell wall (S2 layer). When broken surfaces are prepared for SEM
study the outer layers of the cell wall often break off to expose the thickest of the cell wall layers
(PI. 28, fig. 7). This layer is composed of ‘sublayers’ comprising helical strands at 20-30° to the
tracheid axis. Up to three ‘sublayers’ with strands 0-2-1 -5 /an across, which spiral in alternate
directions, can be recognized in thin-section (PI. 28, fig. 9). The S2 layers of tracheids in extant
conifer woods are by far the thickest of the cell wall layers; T93 jim across in Pinus densifiora,
making up 78% of the total cell wall thickness (Harada 1965). Such layers are made up of
microfibrils with a steep helical arrangement which also spiral in alternate directions. It appears
that the S2 layer in the Alexander Island fossil wood was similar in thickness and form to that of
extant woods. Early decay and delignification are thought to have been important in opening up
the cell wall, and promoting silica impregnation. Silica is thought to have grown on to microfibrils,
and particularly on to groups of microfibrils, and to have filled interfibrillar porosity within the cell
wall. The cryptocrystalline quartz seen on and within this wall probably resulted from the later
ordering of the original silicic acid monomers likely to have been involved in this impregnation.

Inner secondary cell wall (S3 layer). In no case can any layer internal to the S2 layer be distinguished,
using SEM techniques, from material filling the cell lumen. However, some of the flat helical fibrillar
structures seen in thin section, and described above, probably represent the S3 layer.

Origin and silicification of fibrillar structures

Harada (1965) stated that the microfibrils making up the cell wall layers are from 0-01 to 0-03 j/m
(100-300 A) across. The helical structures in all the cell wall layers of the Alexander Island fossil
woods are far larger than this, ranging from 0-2- 1-5 /mi across. Although this may be due in part
to the irregular growth of silica, the original microfibrils were probably only 10 % of the width of
the fossilized structures. The relatively large size of the fibrils in the fossil wood is thought to relate
to biogenic delignification of cell walls. Cowling (1965) and Kaarik (1974) showed that enzymatic
activity of white rot fungi (which remove lignin) isolates and separates individual microfibrils and
groups of microfibrils. The growth of silica on to fibrillar bundles produced by fungal delignification
probably led to the preservation cell wall structures in the Alexander Island fossil woods. Fungal
hyphae and lensoid-ovoid organic bodies 9-12 //m long (probably vegetative yeast cells since asco-
mycetes and basidiomycetes rarely produce spores within wood) are common within tracheids (PI.
28, figs. 12 and 13).

Cowling (1965, p. 341) stated that ‘the specificity of microbial organisms and the very mild
conditions under which their reactions proceed, make them potentially ideal reagents for delicate
study of structure’. Fine structures were probably preserved in some of the Alexander Island fossil
woods because of these properties of microbial activity.

Some Tertiary permineralized woods bear fossilized fibrillar structures of the S2 layer on the
same scale as those of extant gymnospermous woods (Buurman 1972, fig. 34). Buurman (1972)
suggested that silicification was by impregnation or replacement of the cell wall layers. In the
Alexander Island fossil woods silicification of the cell wall layers involved the growth of silica on to
the component microfibrils and microfibrillar bundles of at least three, if not four, of the cell wall
layers. The carbonized remnants of the original cell wall layers can be seen in thin section and have
not been replaced. Buurman (1972) also stated that the preservation of fine details in silicified fossil
wood was normally confined to opalized material and that details of those preserved in quartz or
chalcedony were obscured or disrupted by crystal growth. X-ray diffraction (XRD) of bulk samples
of fossil wood and electron microprobe analysis of thin sections indicate an absence of a hydrated
silica phase in Alexander Island fossil woods; silica is in the form of low quartz or chalcedony.
SEM studies show a cryptocrystalline structure to the silica which has impregnated the cell wall
(PI. 28, figs. 4-7). It is likely that all original opaline silica has transformed into chalcedony or
quartz. An increase in the volume of silica, due to ordering and/or subsequent growth, may have

240

PALAEONTOLOGY, VOLUME 30

accompanied a reduction in the volume of organic material as oxygen and hydrogen were given off
over time; silica may have filled this secondary porosity as it developed, or soon after. The cell wall
structures seen in the Alexander Island fossil woods are much coarser than those seen in many
opalized woods and are preserved because of the pre-mineralizational decay which opened up the
cell wall and allowed silicification of delignified fibrillar bundles and interfibrillar space.

Helical thickenings

In addition to, and readily distinguishable from, the fine helical structures of the S2 layer, some
tracheids bear regular helical bands 7-5 /im across and 5 /tm apart, which have the same orientation
(PI. 28, fig. 8). They are clearly not caused by splitting of the cell wall; they are far too regular and
many are rounded at each end. The distribution of these ‘helically thickened’ tracheids has no
apparent pattern; they occur adjacent to non-thickened cells and often isolated from any other
thickened tracheids. Although their relationship with areas of reaction wood cannot be demon-
strated, they probably formed in cells which were growing under stress, in the same way as helical
‘checking’ forms in the tracheids of modern conifer reaction wood. Slit pits often separate these
thickenings and follow their helical shape (PI. 29, fig. 1 ). Despite the fact that they have no apparent
border, the other side of the pit pair is often formed by a normal bordered pit.

Greguss (1967) figured ‘spiral thickenings’ in fossil gymnospermous wood from Hungary. He
considered that these were characteristic of the genus Platyspiroxylon, although he recognized
similar thickenings in members of the Ginkgoaceae and Araucariaceae and rarely in other genera
of the Cupressaceae. These are usually regular, widely spaced (5-8 ^um) and broad (2-5 /rm), and
bear a close resemblance to the structures in the Alexander Island fossil woods described above.
However, some of the structures which Greguss (1967, plate XXXIV, II) figured may well be
artefacts of a decay process and not true anatomical characters. This highlights one of the major
problems encountered in the taxonomy of fossil wood; a number of the structures which have been
used as diagnostic properties may be artefacts of pre-mineralization decay.

Chemical hypothesis for silicification of the cell wall

The most likely silicifying agent involved in permineralization is molecular silicic acid (H4Si04)
which is the only common natural form of soluble silica, and the form released in the devitrification
of volcanic glass and diagenesis of clay minerals (Murata 1940; Sigleo 1979). Leo and Barghoorn
(1976) suggested that it was the potential for hydrogen bonding between silicic acid and holocellu-
losic complexes of the cell wall which led to exact replication of cell wall structures. The process

EXPLANATION OF PLATE 29

Bordered pits in silicified conifer wood.

Fig. 1. Scanning electron micrograph of radial longitudinal fracture surface, KG.2817.16, rare slit pits associ-
ated with helical thickenings (‘normal’ pits are seen on the right), x 250.

Figs. 2 and 3. Tangential longitudinal thin sections through bordered pits. 2, KG. 1704.10, x 250. 3,
KG.2817.16, X 150.

Fig. 4. Radial longitudinal thin section, KG. 1702.6, x 250, showing contiguous bordered pits.

Figs. 5-14. Scanning electron micrograph radial longitudinal fracture surfaces. 5, KG. 28 1 7.23, view of internal
surface of pit chamber, note ‘S’ and ring-shaped structures in the centres of the pits, possibly representing
the collapsed tori, x 1000. 6-8, KG.2817.16, views of concave cast of pits in the silica filling a cell lumen. 7,
detail, x 1000. 8, internal view of pits, in the lower pit fracture through the cell wall exposed the silica-fill of
the pit chamber, in the upper pit ffacture was external to the cell wall, x 1000. 9, KG. 1702.3, internal view
of pit, fracture through the silica-fill of the pit chamber, x2500. 10, KG. 2817. 23, view of concave cast of
pit in silica filling the cell lumen, x 2500. 11 and 12, KG. 2817. 23, external views of pits. 11, x 250. 12,
X 1000. 13 and 14, KG.2817.16, external view of slightly collapsed pits, showing microcrystalline quartz in
the pit aperture. 13, x 1000. 14, detail, x2500.

PLATE 29

JEFFERSON, bordered pits in silicified wood

242

PALAEONTOLOGY, VOLUME 30

proposed for the Alexander Island fossil woods involves perinineralization of the eell wall structures,
rather than replication, and is as follows:

1. Dilute silicic acid infiltrates and permeates the cell wall and forms hydrogen bonds with
hydroxy functional groups in the molecular constituents of the cell wall (Leo and Barghoorn 1976).

H

R

I

C

I

R'

OH

1

Si -OH

i

OH

Ligno-holocellulosic silicic acid

complex

Because of the four active hydroxyl groups per molecule, silicic acid has a high potential for the
formation of hydrogen bonds.

2. Silicic acid monomers build up on cell wall structures and begin to interact and polymerize,
eliminating water in the formation of siloxane bonds.

OH OH OH OH

! i I I

OH- Si -OH + OH- Si -OH -> OH- Si -O Si OH + H?0

OH OH OH OH

3. A film of silica develops on microfibrils and fibrillar bundles, and along cell wall surfaces, and
silica fills interfibrillar space.

4. The molecular film is converted over tens to hundreds of years into an opaline state
(Si02-nH20), or disordered low-cristobalite (opal-CT) may be formed as an intermediate phase
(Stein 1982).

5. Opal is transformed to chalcedony and low quartz over 10^ to 10® years.

The early decay demonstrated in the Alexander Island fossil wood is likely to promote the first
part of this process by increasing cell permeability and surface area and producing chemical entities
with more active sites for hydrogen bonding.

The final stage in the silicification of these woods was a cavity-fill process, in which silica filled
the cell lumen (see below). In parts of some specimens this is the only form of mineralization and
there has been no impregnation of the cell wall. Within such areas, cell structure is poorly preserved
because of disruptive growth of chalcedony or quartz, and cell walls are marked only by inclusions
of organic material.

Preservation of bordered pits. The walls of bordered pits have been silicified by the same infiltration
processes as is outlined above. Pits are seen in thin section as dark, organic-rich walls (the pit
borders) enclosing a pit chamber and possessing a central aperture (PI. 29, figs. 2-4). Using the
SEM, however, the appearance of pits is highly variable and depends on the position of the fracture
when specimens are prepared. Fractures are likely to pass along the cell walls, the planes of least
resistance. The results of a range of fractures are shown in text-fig. 3. Although fractures are
illustrated as passing along the line of the middle lamella between cell walls, there is likely to be
some fracture through cell walls themselves. The process and results are shown in Table 1 (the
numbers below (la-4b) refer to the numbers on this table), and are summarized as follows:

1. When the pit aperture is blocked and the pit chamber is unfilled, fracture between cell
walls will expose the internal surfaces of both halves of the pit pair (la-b, text-fig. 3a-d). The
circular to S-shaped body and the ring of irregularities over or around the pit aperture in Plate 29,
fig. 5, probably represent the remains of the collapsed tori, indicating early silicification before the
degradation of these structures, although there is no evidence of the margos, the delicate membranes
which support the tori.

JEFFERSON: CRETACEOUS CONIFER WOOD

243

TEXT-FIG. 3. Preservation and form of bordered pits (see also Table 1). a. Wood prior to mineralization,
showing two adjacent tracheid walls and their bordered pit pair interconnection. The torus (t) and margo (m)
form the pit membrane, b. Cell wall permineralized and coated by silica (stipple), c-d. Cell lumina filled by
silica; pit chamber unfilled; d. Fracture along cell wall through chamber, e-h. Cell lumina filled by silica; pit
chamber filled; f, Fracture along cell wall, through pit chamber, around chamber-fill; g. Fracture between cell
wall and lumen-fill, then through pit chamber; h. Fracture between cell wall and lumen-fill.

244

PALAEONTOLOGY, VOLUME 30

TABLE 1. Morphology of bordered pits. See text and text-fig. 2 for explanation.

Text

fig-1

Pit

Chamber

Fracture

Pit 1 of pair

Features

Pit 2 of pair

Features

Plate 2
figs.

a-d

no fill

Between

cell

walls,

through

pit

chamber

a-b

e-f

f il 1 ed

a-b

e.g

filled

a-b

e,h

fil 1 ed
(may be
no fill)

Between
cel 1
walls,
around
pit

chamber

Between
wall &

^ umen-
fill as
in 4 but
around
chamber
as in 2

Between
cel 1
wall S
silica
1 umen-
fill

Concave internal
surface

Torus may
be pres-
erved ;
quartz
crystal s
only in
aperture

Concave internal
surface

Torus may
be pres-
erved ,
quartz
crystal s
only in
aperture

Concave internal
surface

Torus
uni i kel y ;
quartz
crystal s
only in
a perture

Convex mold of
pit chamber^

No torus;
quartz
crystal s
possibl e
on surface

Concave internal
surface

Fractured
cell wall
visible
outside &
above pit

Convex mold of
pit chamber

Fractured
cell wall
visible
outside &
below pit;
crystal s
possible
on surface

Concave external
mold of pit

Quartz
crystals
1 i kely
on

surface ;
no

torus

Convex external
surface

Quartz
crystals
uni i kel y
except
in

aperture

6-7

3-9

10

11-

14

2. When the pit chamber is filled, fracture between the cell walls will pass around the silica-
filled chamber. This will produce an internal surface of one pit (2a) although the torus is less likely
to have survived mineralization, and a concave surface of the internal silica mold of the pit chamber
(2b). (This may be difficult to distinguish from an external view of the pit (Table 1, 4a; text-fig. 2h)
unless a coarse crystalline structure is evident.)

3. When the pit chamber is filled, a fracture between a cell wall and the lumen fill may break
through the cell wall in the area of this fill and pass around it producing an internal surface of one
pit beneath the broken cell wall of the adjacent cell (3a), and a convex surface of the internal silica
mold of the chamber, with the fractured cell wall around and beneath it (3b).

4. A fracture external to one of the cell walls (more common when the pit chamber is filled)
will expose an external surface of one of the pits (4a), and a cast of this surface in the silica filling
the lumen (4b). This will be similar to the internal surfaces of pits (la-b) but may show a crystalline
structure.

INFILL OF THE CELL LUMEN

The terminology applied by Storz (1933) and Buurman (1972) to cell-fill crystalline textures will be
used here: (1 ) polyblastic— many crystals in the space of one cell, (2) oligoblastic— one crystal filling

JEFFERSON: CRETACEOUS CONIFER WOOD

245

each cell, (3) hyperblastic— single crystals filling a number of cells by growing through cell walls,
(4) idioblastic— well shaped crystals in any part of the wood.

Although cell wall impregnation was probably in the form of silicic acid monomers which
transformed via opal and disordered cristobalite to microcrystalline quartz, the variety of mineral-
ization textures seen in the cell lumina suggests that the primary growth of silica was in several
forms.

1. Cryptocrystalline chalcedony with a characteristic radial extinction in thin section (PI.
30, figs. 1 and 5) and a granular appearance using the SEM at high magnification. It is usually
polyblastic and post-dates cell wall silicification.

2. Microcrystalline quartz often found in association with chalcedony and consequently
difficult to distinguish from it. It forms as individual crystals 0-5-2 0 /<m across with poorly de-
veloped crystal faces and random orientation (PI. 30, fig. 2).

3. Euhedral to subhedral polyblastic quartz crystals up to 30 jun long. These show character-
istic cavity-fill textures; small anhedral crystals close to the cell wall and large euhedral crystals in
the centre of the lumen (PI. 30, figs. 3 and 4). The intercellular space may also be filled by polyblastic
silica (PI. 30, fig. 5).

4. Euhedral to subhedral hyperblastic quartz crystals which grew through cell walls and
across several cells. Although preservation is inferior, the cell wall is usually preserved as a line of
coaly inclusions and bordered pits are often seen (PI. 30, fig. 6).

5. Single oligoblastic crystals of chalcedony filling whole cells in which the cell wall has been
silicified. These crystals often contain spherical pores up to 10 /.rni across which do not represent
the usual form of fluid inclusions (PI. 30, fig. 7).

Although there are gradations between 1 and 2, and between 2 and 3, in terms of crystal size,
there is no evidence for transformation of any of these forms to any other, or of pseudomorphed
textures.

In the Alexander Island fossil woods the growth of chalcedony and microcrystalline quartz was
usually polyblastic because silica filled cell lumina after the silicification of the cell wall and did not
grow through them. Larger quartz crystals are often polyblastic; they grew away from previously
silicified cell walls into cell lumina. In some specimens in which no cell wall mineralization took
place (possibly those which had undergone extensive decomposition before mineralization), idioblas-
tic silica grew in sub-spherical masses throughout the wood (PI. 30, fig. 8). These masses range from
OT-10 mm in diameter and also occur in small areas of many specimens which are otherwise well
preserved. There is generally little evidence for the confinement of these silica masses to tracheid
lumina.

In at least two specimens (KG.28 15.252 and KG.28 15.254) the first stage of lumen fill in approxi-
mately 1 % of tracheids involved the growth of isolated, euhedral, cubic iron sulphide crystals 20
jum across (PI. 30, fig. 9). The crystals are completely enclosed in the silica which filled the cell
lumen and are adjacent to, and sometimes penetrate, the cell wall. They represent a very early phase
of crystallization under strongly reducing conditions promoted by the early decay of cells or their
contents.

Partial fill of the cell lumen by apatite. When XRD analyses were carried out to determine the
relative importance of the silica phases opal, disordered cristobalite and quartz/chalcedony, three
specimens from the same unit (KG.28 17. 13, 16 and 17) were found to contain up to 70 % apatite
and/or collophane (amorphous, hydrated apatite). This is difficult to distinguish from quartz by
‘normal’ petrographic techniques, particularly when the two minerals are in close association.
Apatite and collophane formed as part of a two-stage lumen fill process in close association with
chalcedony.

At the first stage, silica grew as spheroidal masses of chalcedony with radial extinction, occupying
the full width of the cell, and sometimes coalescing to fill most of the tracheid (PI. 30, fig. 10).
Collophane then grew in the remaining cell lumen space as large masses composed of radiating
needles, and apatite grew as fine crystals (polyblastic fill) (PI. 30, figs. 1 1 - 1 3). In many cases tracheids

246

PALAEONTOLOGY, VOLUME 30

contain no silica but are completely filled by collophane. Series of microprobe analyses across the
dark wall areas between adjacent apatite-filled and collophane-filled cells indicate that the first
stage in mineralization, even in ‘apatitized’ wood, is the impregnation of the cell wall by silica.
The cavity-fill growth of apatite and collophane is a late-stage alternative to the cavity-fill growth
of quartz.

ORIGIN OF MINERALIZING FLUIDS

Silica. Murata (1940) considered that silicification of wood was due to the early decomposition of
volcanic tulfs and ashes leading to the liberation of free silica and its deposition within tracheids.
He demonstrated that permineralization in active volcanic areas in Yellowstone National Park,
Wyoming, started within a few months of wood being surrounded by silica-rich fluids. Hunt (1972)
supported this view and showed that silica in the form of silicic acid was released in the devitrification
of volcanic material and the diagenesis of clay minerals. Sigleo (1979) conducted geochemical
investigations on silicified wood and associated sediments in Petrified Forest National Park, Ar-
izona, and showed that the silica precipitated within wood cells was derived from the breakdown
of volcanic ash in the surrounding sediments. She found evidence in the external parts of the
wood for co-precipitation of authigenic montmorillonite, a product of ash hydrolysis, and silica.
Montmorillonite also constituted up to 15% of the surrounding sediment. Trace and variable-
valence element concentrations showed that silicification took place within the compositional ranges
of modern stream and ground waters, SiOj concentrations less than 140mg/l, and under anoxic,
slightly acid conditions. Sigleo suggested that the most likely preservational environment was in
stagnant pond water or in swamp bottom muds. Dorf (1964) regarded many of the forest beds in
Yellowstone as volcanic debris flows but recent work by Fritz (1980) suggests that most of the trees
were buried by alluvial sandstones and conglomerates (80-90 % water transported air fall ash, and
10-20 % reworked detrital volcaniclastic material).

Similar preservational and mineralizational settings are suggested by the composition of rocks
associated with the Alexander Island fossil forests. The Fossil Bluff Formation sandstones in which
fossil wood is found are coarse-grained, volcaniclastic sandstones with a high porosity. These are
lithic arenites or greywackes when rock fragments are dominant, or arkosic arenites or greywackes
when feldspar is dominant (Pettijohn 1975). The arkosic arenties may be composed of up to 50 %
plagioclase feldspar, mostly andesine. Detrital plagioclase is common even in the lithic arenites.
The prefix ‘tuflfaceous’ should be used for most of these sandstones since fragments of volcanic
glass, pumice, and chloritized volcanic material may constitute up to 60 % of the rock. The matrix
is greatly altered by zeolite formation but is usually made up of chlorite, calcite and sericite.

EXPLANATION OF PLATE 30

Infill of the cell lumen in conifer wood.

Fig. 1. Transverse thin section, showing polyblastic fill, KG. 1702.2a, x 250.

Figs. 2-4. Scanning electron micrographs of transverse fracture surfaces. 2, thick cell wall and microcrystalline
fill, KG. 1702.2a, x 1000. 3, euhedral polyblastic quartz, KG. 282 1.98, x 1000. 4, detail, x 2500.

Figs. 5 and 6. Radial longitudinal thin sections. 5, polyblastic fill in tracheids and between cell walls,
KG. 1702.2a, x 250. 6, idioblastic fill in central area, KG. 1704. 10, x 100.

Fig. 7. Scanning electron micrograph of tangential fracture surface, showing oligoblastic chalcedony with
?fluid vesicles, KG.28 14.252, x 500.

Fig. 8. Transverse thin section, showing hyperblastic masses of silica, KG.28 17.257, x 100.

Fig. 9. Tangential thin section, showing pyrite crystal, KG. 1 704. 10, x 250.

Figs. 10-13. Apatite filling the cell lumen, KG. 2817. 16. 10 and 1 1, tangential thin sections, x250. 12 and 13,
scanning electron micrograph of tangential fracture surface. 12, x 500. 13, x 1000.

PLATE 30

JEFFERSON, silicified conifer wood

248

PALAEONTOLOGY, VOLUME 30

Mineralization of leaves by laumontite, chlorite, and calcite (Jefferson 1982a) suggests that the
following diagenetic reactions took place under conditions of high PCO2 and high PH2O:

clay minerals ^ chlorite + silica + H2O

glass + H2O + CO2 silica + chlorite + calcite

Ca-plagioclase + HjO + CO2 ^ laumontite + calcite

Some of the silica derived from these reactions was probably deposited in wood, early during
diagenesis because of the hydrogen bonding outlined above, which makes the cell walls of wood
optimum sites for the growth of silica. The anoxic conditions necessary for mineralization probably
developed soon after the burial of trees by this fluvial volcaniclastic sediment.

Apatite. Apatite mineralization of fossil wood is little mentioned in the literature, but this may be
because it is difficult to distinguish from silica without chemical analysis. The two phases of mineral
growth within cells clearly indicate a change in pore water composition after the impregnation of
the cell wall but before complete infill of the lumen. The high volcanic content of all the sediments
in the non-marine part of the Fossil Bluff Formation indicates the close proximity of a volcanic
area (Horne and Thomson 1972). Local influxes of ground water rich in phosphates, calcium, and
other salts are common in present day volcanic environments like Yellowstone National Park, and
similar ground waters might be expected to precipitate apatite into constrictions such as tracheid
lumina.

CONCLUSIONS

Observations on Alexander Island fossil wood support the ‘two stage’ theory of wood silicification
(Leo and Barghoorn 1976), but suggest that, in some fossil woods at least, impregnation of the cell
wall, promoted by biogenic degradation, is more important than the subsequent coating or repli-
cation of the cell wall (Leo and Barghoorn 1976) or replacement of it (Buurman 1976).

The first stage in the mineralization of well-preserved wood from Alexander Island involved
impregnation of the cell wall. Early biogenic delignification and decay opened up the eell wall,
isolating bundles of microfibrils, and created interfibrillar porosity. Silica, in the form of a silicic
acid monomer, grew on to these macrofibrils, and filled the inter-fibrillar porosity. The silica
subsequently became ordered into chalcedony and microcrystalline quartz. Early silicification of
the cell wall is thought to relate to the capacity for hydrogen bonding between silicic acid and cell
wall components. Wood is therefore an optimum site for the early precipitation of silica. The second
stage in mineralization involved the filling of the cell lumen by chalcedony, eryptoerystalline quartz,
euhedral-subhedral quartz crystals, and/or apatite and collophane. The presence of apatite within
the lumina of cells whose cell walls had already been silicified further supports the hypothesis that
cell wall silicification occurred at an early stage and that lumen fill occurred later, in some cases
after major changes in pore water composition.

Acknowledgements. The paper was written when the author was a visitor in 1981-1982 to the Department of
Botany, and Institute of Polar Studies, Ohio State University, Columbus, Ohio 43210, USA. It was prepared
for publication by Dr N. F. Hughes with the assistance of Dr J. Francis.

REFERENCES

ARNOLD, c. A. 1941. The petrifaction of wood. Mineralogist, 9, 253-255.

BUURMAN, p. 1972. Mineralization of fossil wood. Scr. Geol. 12, 1-41.

COTE, w. A. and day, a. c. 1965. Anatomy and ultrastructure of reaction wood. In cote, w. a. (ed.). Cellular
ultrastructure of woody plants, 389-418. Syracuse University Press.

COWLING, E. B. 1965. Micro-organisms and microbial enzyme systems as selective tools in wood anatomy. In
COTE, w. A. (ed.). Cellular ultrastructure of w^oody plants, 341-367. Syracuse University Press.

JEFFERSON: CRETACEOUS CONIFER WOOD

249

DORF, E. 1964. The petrified forests of Yellowstone Park. Scieut. Am. 210, 107-1 14.

FRITZ, w. J. 1980. Reinterpretation of the depositional environments of Yellowstone ‘fossil forests’. Geology,
8, 309-313.

GREGUSS, p. 1967. Fossil gyimto.spernums woods in Hungary from the Permian to the Pliocene, 151 pp., 93 pis.
Akadamiai Kiado, Budapest.

HARADA, H. 1965. Ultrastructure and organisation of gymnosperm cell walls. In cote, w. a. (ed.). Cellular
ultrastructure of woody plants, 215-234. Syracuse University Press.

HORNE, R. R. and THOMSON, M. R. A. 1972. Airborne and detrital volcanic material in the Lower Cretaceous
sediments of south-east Alexander Island. Bull. Br. Antarct. Surv. 29, 103-1 1 1.

HUNT, c. B. 1972. Geology of soils, 344 pp. Freeman, San Francisco.

JEFFERSON, T. H. 1981. Palaeobotanical contributions to the geology of Alexander Island, Antarctica. Ph.D.
thesis (unpubk). University of Cambridge.

1982a. Preservation of leaf fossils in volcaniclastic rocks from the Lower Cretaceous of Alexander Island,

Antarctica. Geol. Mag., 119, 291-300.

19826. The Early Cretaceous fossil forests of Alexander Island, Antarctica. Palaeontology, 25, 681-708.

KAARiK, A. A. 1974. Decomposition of wood. In Dickenson, c. h. and pugh, g. j. f. (eds.). Biology of plant
litter decomposition, 1, 129-174. Academic Press, New York.

LEO, R. F. and BARGHOORN, E. s. 1976. Silicification of wood. Bot. Mus. Leafl. Harv. Univ. 25, 1-47.

MARK, R. E. 1967. Cell wall mechanics of tracheids, 310 pp. Yale Univ. Press.

MEYLAN, B. A. and BUTTERFIELD, B. G. 1972. Three dimensional structure of wood, 80 pp. Chapman and Hall.

MURATA, K. 1940. Volcanic ash as a source of silica for the silicification of wood. Am. J. Sci. 238, 586-596.

PETTIJOHN, E. J. 1975. Sedimentary rocks, 736 pp. Harper, New York.

SIGLEO, A. c. 1979. Organic geochemistry of silicified wood. Petrified Forest National Park, Arizona. Geochim.
cosmochim. Acta. 42, 1397-1406.

STEIN, c. L. 1982. Silica recrystallization in petrified wood. J. Sedim. Pet. 52, 1277-1282.

STORZ, M. 1933. Zur Petrogenesis der Kieserholzer Agyptens. Abh. buyer Akad. Wiss. 16, 24-50.

SUAREZ, M. 1976. Plate tectonic model for the southern Antarctic Peninsula and its relation to the southern
Andes. Geology 4, 21 1 -214.

TAYLOR, B. J., THOMSON, M. R. A. and WILLEY, L. E. 1979. The geology of the Ablation Point to Keystone Clift's
area, Alexander Island. Sci. Rep. Br. antarc. Surv. 82, 65 pp., 10 pis.

TAYLOR, T. N. 1981. Paleobotany, 589 pp. McGraw-Hill, New York.

WARDROP, A. B. and BLAND, D. E. 1959. Process of lignification in woody plants. In kratzl, k. and billek, g.
(eds.). Biochemistry of wood, 92-1 16. Pergamon Press.

Enquiries to: amy l. karowe

Woods Hole Oceanographic Institution

Typescript received 16 October 1985
Typescript accepted 1 1 June 1986

Woods Hole
Massachusetts 02543
USA

jr

-'ir

Si.*

■■’a?

1

J

. ■• i >'

-f.

- '

■'i !

<';=’■■■

.. ..., W' ...

EARLY DEVONIAN CONODONT FAUNAS FROM
BUCHAN AND BINDI, VICTORIA, AUSTRALIA

by RUTH MAWSON

Abstract. Conodont faunas of the Taravale Formation, in sections measured at Buchan and Bindi, Victoria
(south-eastern Australia), contain polygnathid conodonts representative of the delusceiis, perhomisjgronbergi,
inversus, and serotinus zones of the Early Devonian (late Pragian to middle Dalejan). New taxa described are
Polygnathus dehiscens abyssus, P. labiosus and P. pseudoserotinus (both belonging to a lineage derived from P.
d. abyssus), P. nothoperbonus, and Ozarkodina prolata. The ammonoid faunas of the Taravale Formation are
shown to be restricted to the dehiscens and perbonus zones. Three species of goniatites, associated with two or
more species of bactritids, occur within the dehiscens zone at Buchan and are thus among the oldest firmly
dated ammonoids in the world.

Study of the sequence of conodont faunas at Buchan and Bindi, Victoria, had several objectives:
to test the applicability in Australia of the zonal system developed mainly from northern hemisphere
conodont sequences, paying special attention to the polygnathids; to commence development of a
framework for the study of dacryoconarid biostratigraphy in this part of the world; and to provide
more precise ages for the evolutionarily important Early Devonian ammonoids previously described
(Teichert 1948; Erben 1964, 1965) from the Taravale Formation. As international correlation for
the late Early Devonian has come to be based primarily on polygnathid conodonts (e.g. Klapper
and Ziegler 1979; Klapper and Johnson 1980) the discussion of polygnathids given herein spreads
wider than consideration of the polygnathids from the Buchan and Bindi areas. An account of the
implications of the conodont work for ages of the Taravale ammonoids is given.

STRATIGRAPHY

The Buchan Group of eastern Victoria is a sequence of carbonates and shales (maximum thickness
c. 3,300 m) which outcrop in at least fifteen discrete areas to the north and east of their most
extensive development in the Buchan-Murrindal area (text-figs. 1 and 2). Some occurrences are
synclinal, as at Buchan and at The Basin, 12 km north-east of Buchan; others, such as that at Bindi
and those at Jackson Crossing and in the headwaters of the Indi River and Limestone Creek, about
75 km to the north, and at Gillingall, 20 km to the north-north-west, owe their preservation to
faulting. In all these occurrences, the basal unit of the Buchan Group, the Buchan Caves Limestone,
rests disconformably or with minor unconformity on the Snowy River Volcanics (as at Buchan,
The Basin, and at the junction of Dead Horse Creek and Limestone Creek in the headwaters of the
Indi River); it is markedly unconformable at Bindi.

Lithologies of the Buchan Caves Limestone (maximum thickness c. 230 m at East Buchan, but
up to 295 m at Bindi) have been described in some detail by Talent (1956) for the Buchan Murrindal
area. There is remarkable uniformity in lithologic and macro-faunal succession within the Buchan
Caves Limestone throughout eastern Victoria, suggesting original deposition on a near-planar
surface, termed the Buchan-Indi-Combienbar Shelf by Talent (1965, 1969). A basal sequence of
pale- to mid-grey dolomites (weathering to buff) is characteristic of all outcrop areas; it reaches a
maximum thickness exceeding 40 m in the Back Creek area of East Buchan where it has been
quarried commercially on a small scale. The remainder of the Buchan Caves Limestone consists
predominantly of calcarenites, a fairly high frequency of algal pisolites, rare crinoidal limestones,

(Palaeontology, Vol. 30, Part 2, 1987, pp. 251-297, pis. 31-41.)

© The Palaeontological Association

252

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1. The Buchan Group in the Buchan-Murrindal-The Basin area, eastern Victoria.

MAWSON: EARLY DEVONIAN CONODONTS

253

and, in the upper parts of the formation, micritic limestones (calcilutites); the micrites aggregate 7-
14 % of any sequence of the formation and are best developed in the upper third, though the last
15-30 m consists predominantly of calcarenites in all three areas (Buchan-Murrindal, Bindi, and
The Basin) where the highest beds of the Buchan Caves Limestone occur in conjunction with the
Taravale Formation. There are only three occurrences of non-carbonate sediments known from
the Buchan Caves Limestone in the areas about Buchan (Talent 1956); one of these, the areally
restricted Cameron Mudstone Member at East Buchan (loc. Mai 2, text-fig. 1) has yielded dacry-
oconarids consistent with a kindlei or perhaps sulcatus age and are therefore the oldest-dated
sediments from within the Buchan Caves Limestone. At Bindi, impure carbonates and mudstones
are well developed within the upper third of the Buchan Caves Limestone, outcropping best in
Bonanza, McAdam’s, and Harding’s Gullies.

TEXT-FIG. 2. Diagrammatic north-south section of the Buchan-Murrindal area, eastern Victoria, showing
relationship of stratigraphic units of the Buchan Group and position of stratigraphic sections relative to the

stratigraphy.

Coral, ostracode, brachiopod, and bivalve faunas of the Buchan Caves Limestone are notable
for their relatively low diversity (Hill 1950; Krommelbein 1954; Talent 1956). Philip (1966) has
already discussed conodonts from three spot-samples from the Buchan Caves Limestone at Buchan.
The present study has not focused on the conodont faunas of the Buchan Caves Limestone apart
from the highest beds at Murrindal (section ‘P’ sample IT of Table 2) and The Basin (section
‘SLO’, lowest sample. Table 3), spot-sampling at Martin Cameron’s Quarry, South Buchan (sample
Ma 9), and a sequence through the upper Buchan Caves Limestone at Bonanza Gully, Bindi
(section ‘BON’, text-fig. 3, Table 4). The last of these passes through an uncharacteristically richly
fossiliferous carbonate sequence which Talent (1967), on the basis of brachiopod faunas, had
concluded to be significantly younger than the highest horizons of the Buchan Caves Limestone in
the Buchan-Murrindal area, i.e. that there was significant diachronism of the top of the Buchan
Caves Limestone between Buchan and Bindi. As will be shown later, conodont evidence is in accord
with this contention; the diachronism is substantial.

The main focus of this study has been the conodont faunas of the Taravale Formation, a
sequence of nodular limestones, shales, and impure limestones conformably overlying the Buchan
Caves Limestone. Outcrops of Taravale Formation are restricted to three areas: major develop-
ments at Buchan and Bindi, and a minor synclinal body (previously overlooked) occupying only
a few hectares at The Basin. Presently available evidence suggests great differences in rates of

254

PALAEONTOLOGY, VOLUME 30

.Blockfellowsl

MtShonahon A'

'f Section 'SB^p

^(South BlQcktellows).

‘A Ml Woterson

J^^ungiIa

fAULT”

’’ArmeftV,;

^Bmdi Homestead'

■^I^^Bonanza Gully)

i^^yBoucher’

to Omeo and
SwiMs Creek

Limestone-

iDryGullyJ^

Section 'SALC‘-^a
( South Arm of ^
^ Limestone Creek/s=

^t'ocfion

I I Alluviols

upconformity

LATE

DEVONIAN

MtTombo Beds - non-marine sondstones,
conglomerates and silistones
^AAAA^ unconformity

Shonohon Limestone
(name to be formalized)

Buchon Coves Limestone (basal dolomitic
I . . . focies not discriminoted )

EARLY

DEVONIAN '"'^'^u^con^rmity

)' ^ ' ^ '\ Snowy RiverVolcanics - ocid, often
^ — ' — ignimbnfic exirusives

^A,^/vv^AA<^AA<vvww unconformity

SILURIAN

Mount Woterson Formotion- conglomerotes
' — ond sondstones

Omeo Metomorphic Complex —
principolly metamorphosed Ordovician ond
perhops some Silunon flyschoid
sondstones and silistones

|"36 brochiopod-corol locoliiies, mainly of the
Torovole Formotion (only lO yielded
Significant conodont founos)
mom roods
minor roods
= = jeep irock

Geological ond topogrophicol bose from mopping by JATaleni(i966tjnpub.i967)

TEXT-FIG. 3. The Buchan Group at Bindi, eastern Victoria.

MAWSON: EARLY DEVONIAN CONODONTS

255

sedimentation between Bindi (c. 3,000 m of Taravale Formation and Shanahan Limestone— pre-
sumed to have been originally interdigitating southwards with the Taravale Formation— spanning
from somewhere in the perbonus Zone to somewhere low in the serotinus Zone), Buchan (e. 700 m
from somewhere presumed low in the dehiscens Zone to somewhere presumed low in the serotinus
Zone), and The Basin where beds correlative with the dehiscens Zone aggregate a mere 60 m or so
before being overlain by beds yielding Polygnathus perbonus.

In the Buchan area, the Taravale Formation grades northwards into the Murrindal Limestone.
Teichert {in Teichert and Talent 1958) divided this unit of up to 250 m of limestones and subordinate
mudstones into two members: well-bedded limestones were referred to as the McLarty Member
and massive limestones, interpreted as being biohermal, were termed the Rocky Camp Member.
The tongue of Taravale Formation lying between the Murrindal Limestone and the Buchan Caves
Limestone to the north was designated the Pyramids Mudstone Member of the Taravale Formation.
Some reservations have been voiced (Philip 1966) as to the utility of dividing the Murrindal
Limestone into members in this way, but no lithofacies analysis subsequent to Teichert’s work has
yet been published. The Murrindal Limestone (up to 230 m in thickness) embraces a rich array of
carbonate lithologies indicative of a wide range of environments: micrites to calcarenites and very
minor rudites, algal limestones (including a very prominent biostromal marker horizon about 2 m
thick, packed with dasycladacean algae, outcropping about 75 m above the base of the formation
in the Rocky Camp-McLarty Ridge area), and intercalations of mudstones, calcareous mudstones,
and nodular limestones; the last are best exposed in road cuttings 0-6 km south-south-west of the
now-abandoned Murrindal State School.

The Pyramids Member of the Taravale Formation outcrops poorly. The best exposures are by
no means extensive and occur in road cuttings near the Murrindal State School (sections ‘M’ and
‘P’, text-figs. 2 and 5), in Dailey Creek downstream from its junction with Rocky Camp Gully (loc.
Ma2, text-fig. 1), and in the gully east of ‘Chisholm’s’ (text-fig. 5) where there are mudstones with
limestone nodules and subordinate developments of flaggy or nodular argillaceous limestones. More
typically, the Pyramids Member does not outcrop; areas underlain by this unit are normally
represented by grassy slopes with occasional fist-sized or larger nodules bearing chonetid brachio-
pods (usually ‘'Chonetes' australis (M‘Coy) or ‘C’. teicherti Gill), and occasionally cephalopods.

Philip (1966) collected eight bulk samples for acid-leaching: three from the Buchan Caves Lime-
stone, one from the Taravale Formation, and four from the Murrindal Limestone (text-fig. 1); all
samples yielded conodonts. Using the information then available on northern hemisphere faunas,
he concluded that the Murrindal Limestone was Early Emsian and the Buchan Caves Limestone
youngest Siegenian.

Because of its combination of ammonoids and dacryoconarids and the assumed considerable
time-equivalence of strata involved, the initial focus of the present study was the Taravale Forma-
tion. A cutting on the disused road to the old Tara homestead (top of Buchan Caves Limestone
and lower beds of Taravale Formation in continuous outcrop), another at the entrance to the
Buchan Caves Reserve (text-figs. 1 and 2) (type localities for Teicherticeras desideratum (Teichert)
and Talenticeras talenti Erben), and, above all, a near-continuous suite of outcrops in road cuttings
for 3-5 km along the Gelantipy Road north of Buchan (text-fig. 4) provided the basis for investi-
gations which involved bed-by-bed collecting for conodonts, dacryoconarids, and associated macro-
fauna. This work was supplemented by sampling of the basal beds of the Pyramids Member of the
Taravale Formation at Murrindal near the abandoned Murrindal State School, and beds outcrop-
ping on the Buchan-Orbost road that are considered on structural grounds to be slightly higher
stratigraphically than the highest beds exposed on the Gelantipy Road. There are few outcrops of
the Taravale Formation away from road cuttings, so little additional sampling was carried out in
the Buchan-Murrindal area. A previously unsuspected area of Taravale Formation was discovered
in the core of the main syncline at The Basin, 12 km north-east of Buchan; poorly outcropping
nodule horizons yielded excellently preserved conodonts. To supplement information gained from
these sections, spot-samples were taken from key areas (text-fig. 1), for example at the top of the
Murrindal Limestone along the access road to the old Rocky Camp (Commonwealth) Quarry,

256

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 4. Location of collecting sequence in Tara vale Formation in road cutting along the Buchan-

Gelantipy Road, eastern Victoria.

above and below the Murrindal Limestone on McLarty’s Ridge, in and above McLarty’s Landslip,
and in the road cuttings about Teichert’s localities 65 and 66 on the East Buchan Road.

The Taravale Formation at Bindi (text-fig. 3), 50 km north-north-west of Buchan appears to be
homoclinal and, as previously mentioned, much thicker (c. 3,000 m) than at Buchan. This great
thickness is puzzling. Exposures are admittedly very poor over most of Bindi, but among the scatter
of outcrops along Old Paddock Creek no reversed facings were found, so isoclinal folding is not
the answer. It is accordingly concluded that the sequence of Taravale Formation at Bindi is much
thicker than at Buchan or The Basin. Some thirty-five spot-samples from restricted areas of outcrop
(text-fig. 3) previously collected for macrofossils were sampled and some re-sampled; though not
many of these yielded conodonts, none of the results (Table 4) suggests any reversal of sequence.

MAWSON: EARLY DEVONIAN CONODONTS

257

TEXT-FIG. 5. Location of stratigraphic sections and localities sampled at Murrindal,

eastern Victoria.

258

PALAEONTOLOGY, VOLUME 30

>

c

c

o

CQ

c

o

*o

03

O

Di

_D-

C

—

1)

o

c

.2

£

Un

O

U.

c3

>

o3

Ui

C3

H

'o-

E

c

o

■o

o

c

o

o

c

_o

3

029£

90 91

^8 8

—

r.)

1 9£

9 1 zr

"9 9

0 09£

b 1 ;r

■8 9

—

ro

9 et-t

2 1 /r

:9 b

9 Zfrt

1 ;r

b 9

9 9b£

£OZI

6 b

0 9b£

£091

'9 9

'T

—

9 iV<L

1 91'

^8 8

lO

ro

9 1 b£

£ 1 9f

2 9

U >

9 0tȣ

2 91'

‘9 bl

—

lO

Z6££

20 91

■| 9

CD

9 iil

10 91

Z 21

*

9 I2£

9 1 9f

‘8 9

(NJ

ro

c\l

e-691

91 6

8 b

—

0 881

(£|0H)82

b 01

U J

fO

to

—

0 Z8I
9 b8l "

Z 21’

'8 9

9 21"

‘9 9

'S’

—

—

ro

0 ZZ 1

£ 21'

'6 b

8 9Z 1

2 21'

■| 9

—

0 9Z 1

1 21

■| 9

rO

—

—

8 2ZI

80-2i'

'0 9

02ZI

90 21"

'2 9

CM

OOZ 1

90 21"

’9 9

b89l

1021

‘8 9

9 Z9I

bl 6 '

■| 9

0Z9I

9 ir

■99

2 191

b II

'6b

OJ

0 091

£ II

Z b

0£9I
“ 9 Z9 1 ”

£1 6

■| 9

—

90 1 r

'8 9

0 9£l

10 II

■9 9

—

9 611

Z 6

'b9

0 811

9 6

8b

2

b8 09

21 ft

.0 £1

CM

—

CM

9 99

1 1 8

ftf 1

h-

f\l

cO

'f

9 b9

6 018

9 9

CM

92 £9
Z 1 £9

8 01 8

b 9

—

Z 01 8

1 9

—

28 29

9 018

9 01

—

—

9Z 29

£ 01 8

601

ro

—

—

Zb 29

2 018

b bl

'n

CD

O-

rO

l£ 29

01 8

6 Z

rO

CT>

—

<\i

iD

cn

8Z 19

£ 6 8

.b 8

ro

OJ

r,

rr

89 19

2 6 8

1 9

r>-

?

rr

<n

£9 1 9

6 8

2 9

<T>

CO

2 1 9

0188

8 9

m

0

—

ZO 19

8 8 8

6 b

—

609

9 8 8

Z b

ro

89 6b
6 9b

91 b88

9 9

'g-

£1 b88

.6 9

89 9b ^

2 1 b 8 ft

.Z 9

9 1 b

6 b 8 8

1 9

220b ^

ft b ft 8

b 9

—

—

6 8£
9Z£

Z b8 8

-13

9 b 8 8

0 9

Z2 9£ _
26 b£ ^

9 b 8 8

Z b ^

b b 8 8

9 b

9 ££

fb 8 8

-b b

82 2£

2 b ft ft

6 b

6 0£

1 b88

J 9

6£ 62

J 1 ft

0 9

92 82

9 Z -8

Z b

V

V

9 1 Z2

97 8

8 b

£ 92

£ J 8

2 9^

—

91 92

Z 9-8

.9 9

—

Z6 b2

9 9 8

2 9

CM

bZ b2

9 9 8

09
1 9

b£ b2
Z 1 b2

2 9 8

—

9 8

1 I

—

28 £2

01 9 8

2 5-

»n

8b £2

9 '?8

Z 9

■g-

92 £2

b 9 8

-b

—

8 22

£ 9 8

6 b

ro

9 22

1 9 8

1 9

—

9b 02

9 ft

0 9

—

0 22

b b ft

8 9

6Z 1 2

£ b ft

b b

19 12

2 b ft

1 b

—

ZO 1 2

i £ ft

Z b

1 2 02

b £ 8

6 b

CM

96 6 1

2 £ ft

1 9

bJf. 1

f ft

8

£ 1 fi 1

f 2ft

8

0-

ftfi ft 1

2 2ft

b

—

ft ft 1

2 ft

8

—

—

ft? ft 1

2 1 ft

2

6 ) 1

i 8

9 9

1 c n

2 02 J

8

C ] 1

02 )

6 b

£6 9 1

61 Z

2 9

<3-

f ft 91

Oi-ftl7

b 9

—

m

CM

ft^ 91

6 81 7

Z 9

'if

9b 91

8 81 Z

•| 9

CTi

—

2 9 1

_60 91

_ 8 91

9 91

9 81 Z_
b 81 Z_
S 1 81 Z

■ O 9 Zl Z

8 9

9 9

CM

Z

to

6

—

1 9

19 b 1
Z9 £ 1
8£ £ 1
bb 1 1
90 01

_ 66

9£ 6

9 ft

o Zl Z

W 1 91 Z_

^ 991 Z

■ Q b bl Z

c 2 bl Z

:o bl Z_

2 £1 Z_
>- 9 21 Z 1

2 9

Z

—

6

1 9

19

‘2 6

Z 9

1

—

bb 8

£ ft

9= 9 21 Z j

1— £ 21 Z

9

m

6 9

99 Z
ft9 J

■ g 1 Zl

Z 9

■ 1 01

—

£9 9

rrl 1 11 z

b 9 ^

'O-

£9 9

O 1 01 z ^

"2 9

OJ

22 9

b 6 z

2 9

b ft z

9 9

£9 b

2 8 Z

1 9

ro

9£ b

y z

'8 8

90 b

’i L L

0 9

—

9 1 £

i 9 Z

6

26 2

9 Z

Z 9

—

R9 2

f 9 Z

8

9 2

2 9 Z

b

b b Z

8£

-

££ 1

2 b Z

2

f 0

£ Z

b

29 0

v; 2 z

Z

£0

I i L

’8 b

Bose of beds (in metres) '
above bose of section j

Somple number

I

1

Weight of sample (kg)

1

S

(S

<£

s

cX

<?

<5?

0^

CO

£

(>!

I

to

Oulodus murrindo/ensis

Po/ygnotf)us spp.
(See Tables 5 ond 6
tor difterenliotion
of Po elements

e

f

Oexcovolo excavolo

0

0 prolala

c?

Belodello devonico
Belodella resima
Belodello triangularis
Drepanodus sp
fbnderodus umcoslo/us
(undifferenfioled)
Ponderodus recurvotus

■©

0

1

MAWSON: EARLY DEVONIAN CONODONTS

259

Limestones in an appreciable section of what appear to be the highest beds outcropping along a
rough road on the left flank of Limestone Creek (also known as Dry Gully) were sampled.

It has been suggested (Talent 1965, 1967) that the highest beds of the Buchan Caves Limestone
at Bindi extend to horizons substantially younger than those at Buchan; to test this, a stratigraphic
section was measured through these beds in the vicinity of McAdam’s and Bonanza Gullies (text-
fig. 3). An additional section was measured through the Shanahan Limestone, a previously over-
looked but substantial (thickness c. 210 m) and well-exposed sliver of limestones, resembling the
Murrindal Limestone of the Buchan-Murrindal area, outcropping adjacent to the Indi Fault north
of Bindi. The conodont evidence adduced here, that it is correlative with the perbonus Zone, suggests
that configurationally it was very similar to the Murrindal Limestone: interfingering southwards
with the Taravale Formation (spanning from somewhere in the perbonus Zone into the serotinus
Zone).

The Taravale Formation typically consists of impure limestone nodules and irregular, discontinu-
ous limestone beds in mudstones. Because of the muddy nature of the limestone sampled, the
conodont yield was only moderate (Tables 1-6) and the leaching process slow but, using generally
large samples (c. 8-10 kg), sufficient yields have been obtained, in my opinion, to verify the proposed
polygnathid lineages (text-fig. 6). The samples from the Buchan and Bindi areas yielded 5,131
conodonts from 1,581 kg of limestone, equivalent to 3-2 conodonts per kg.

EARLY DEVONIAN POLYGNATHID LINEAGES FROM
EASTERN AUSTRALIA

The phylogeny of Early Devonian polygnathid conodonts has been greatly clarified since Klapper
and Johnson (1975) illustrated, with faunas from the Early Devonian at Lone Mountain, Nevada,
the evolutionary sequence from Polygnathus dehiscens to P. laticostatus and from P. aff. P. perbonus
to P. serotinus, the latter sequence being characterized by an interval of stratigraphic overlap of P.
inversus and P. serotinus.

Comparison of the Nevada and other Early Devonian faunas from the northern hemisphere (e.g.
Carls and Gandl 1969; Klapper 1969; Uyeno in McGregor and Uyeno 1972; Perry et al. 1974; Al-
Rawi 1977; Chatterton 1979; Lane and Ormiston 1979; and Uyeno and Klapper 1980) with those
already known from Australia (e.g. Philip 1966; Philip and Jackson 1967; Telford 1975; Fordham
1976; Pickett 1978) indicates that the southern hemisphere polygnathid lineages for this period of
time differ: The P. gronbergi-P. laticostatus lineage is not clearly present in Australian faunas and,
moreover, the P. aff. P. perbonus lineage (herein described as the P. nothoperbonus lineage) appears
to vary in detail in Australia compared with the same lineage in the northern hemisphere. From P.
dehiscens Philip and Jackson, two lineages were identified by Klapper and Johnson (1975):

( 1 ) P. gronbergi Klapper and Johnson ^ P. laticostatus Klapper and Johnson;

(2) P. aff. P. perbonus P. inversus Klapper and Johnson ~> P. serotinus Telford.

Unlike the succession at Lone Mountain, Nevada, the Buchan and Bindi sequences in south-
eastern Australia do not contain the dehiscens gronbergi laticostatus lineage but it appears
that two different paths of evolution from P. dehiscens are represented:

(1) a lineage very similar to the second lineage of Klapper and Johnson (1975);

(2) a lineage not previously described.

Near the base of the sections measured at Buchan and at Bindi (text-figs. 1-3), the first polygnathid
element in the sequence is P. dehiscens (Tables 5 and 6). In some smaller specimens of early P.
dehiscens (see PI. 32, figs. 1 and 2), the adcarinal grooves are only very slightly developed and
ornamentation consists typically of nodes on very short ridges, consistent with evolution from P.
pireneae Boersma (Boersma 1974; Klapper in Ziegler 1977). No specimens definitely identified as
P. pireneae have been found. However, two subspecies of P. dehiscens can be distinguished: P.
dehiscens dehiscens with a large basal cavity forming an open, flat, or very shallow trough posteriorly
and P. dehiscens abyssus with a deeper, V-shaped basal cavity.

260

PALAEONTOLOGY, VOLUME 30

TABLE 2. Distribution of conodont elements in samples principally from the Taravale
Formation at Buchan-Murrindal, eastern Victoria.

Base of beds (in metres)
above base of section

n r

1 r

T r

OO ro ro roro poro
C\|CM ro^

'P' SEC

ION

OoOOO uoc\jouDOr--'d-aoOosjir> cmOOcvj

CM <T>

Sample number

M' SECTION

— po*=r loco

tf

Din

1

0"

RC

S

EC

no

N

cc CM m r- o ro m r- cn cn ^

l_

o

■EB
SECTION

iDrO —CM

CDGDCQCn
LULU LULU

Weight of sample (kg)

-m-i I4+4H

Dr-m — — CO — 'd-

— COCO'^d-LC) — CMlD Or^CM'd"(T> CM(T) CO IT) — — CO — 'd"

comuDLD^u^iouD <X)uOLOLDLT) LO lOLO uD (x> cM'^irmin

Oulodus murnndalensis

M

So

Sb

Sc

Polygnathus spp.
(see Tables 5 and 6
for differentiation
of Pa elements)

Pa

Pb

M

Sa

Sc

Ozarkodina buchanensis

Pa

4

2

5

7

1

10

8

Pb

1

3

2

1

3

1

Sa

Sb

Sc

1

1

1

0 excavata excavafa

Fh

Pb

Sc

0. linearis

Pa

Pb

0 prolata

Pa

1

Pb

M

Sa

Sb

Sc

Pandorinellina exigua
exigua

Pa

Pb

M

Sb

Sc

Belodella devonica
Belodella res/mo
Belodella triangularis
Drepanodus s.p
Panderodus unicostatus
(undifferentiated)
Panderodus recurvatus

2

M

3

S

8

4

21

7

12

18

29

9

1

1

7

2

1

2

1

1

2

1

1

1

1

1

1

1

II

23

5

6

1

1

1

t

1

18

1

1

1

3

4

1

4

19

1

4

7

1

II

37

7

II

4

2

1

3

1

1

1

15

3

2

62

iq

4

1

15

6

2

1

2

2

2

2

19

8

8

1

2

2

8

2

1

1

1

1

2

1

14

1

1

1

6

18

2

1

1

2

1

2

5

1

1

MAWSON: EARLY DEVONIAN CONODONTS

261

TABLE 3. Distribution of conodont elements in samples principally from the Taravale Formation at Buchan-

Murrindal and The Basin, eastern Victoria.

Meires obove bose of
section

1

o

C£>

sr

O

h-

o

o

o

U5

C

o

5s

1

c

o>

(S

-

o

o

r-

o

5s

o

o

o

lO

s

o

sr

o

o

o

o

o

o

o

o

o

o

o

?

o

o

Oolley's Ck ( Mo 2 ; _|

Junction Rocky 8 Oolley's 40 ( Mo 3) 1

2

2

i

2

6-a

2

2

2

E

o

o

2

2

O

q:

o

<n

2

o

£

Somple number

K

B

o

uc

H/i

N

C

AV

:s

E

o

N1

R/

h-

:e

S

:c

■ic

N

■T

r

5

1

o

o

o

O

o

K

S

o

LO

CO

o

o

ME

E'

o

5

SE

o

o

CT

o

10

o

s

o

o

o

o

o

o

2

E

O

2

CC

2

E

o

2

Weight of somple (kq)

6 1

58

55

59

62

52

5 1

49

54

56

59

61

62

65

69

52

5-6

53

22

6-3

62

68

82

55

65

66

69

87

52

5 1

49

62

68

59

49

43

6 1

64

68

51

46

41

84

6 1

58

59

&8

54

64

Ou/odus murnndo/ens/s

Po

Pt

M

Sr

2

t

Sb

Sc

3

Po/ygno/hus spp
(See Tobies 5 a 6
ditferenliotion of
Po elements 1

R3

5

4

3

1

5

2

9

4

1

18

2

5

10

4

?7

2

2

1

1

3

4

3

Pt

3

2

i

2

II

4

?

7

2

2

2

2

2

2

1

M

1

7

So

2

1

1

Sb

1

2

2

3

3

2

6

2

3

1

1

Sc

2

2

2

13

2

12

3

8

2

2

3

Ozotkodma buchonensis

Ro

2

2

4

2

12

6

2

10

09

15

15

2

10

13

9

2

4

1

1

3

12

n

4

5

14

Pb

1

1

1

2

4

2

1

1

2

2

3

M

1

1

2

1

r

1

So

1

1

1

Sb

3

3

3

2

1

1

Sc

9

2

2

2

2

3

2

3

3

8

Oencovofo encovofo

Pn

1

5

Pb

M

So

Sb

Sc

0 lineor/s

Po

1

1

1

5

2

1

2

2

1

1

Pb

2

1

1

M

1

1

Sc

1

Sb

Sc

2

1

1

0 prolofo

Fb

Pb

1

1

1

1

39

4

17

2

2

1

2

2

2

M

So

1

Sb

Sc

2

7

3

PandonneWma exigua
exiguo

Hj

10

4

i

?

4

1

2

7

1

1

2

41

38

II

50

1

7

3

3

M

Sn

-

.Sb

Sc

14

4

3

4

Belodello devomco
Belodello resimo
Belodello rnangulons

1

to

3

2

3

2

Dreponodus sp
Ponderodus L/mcosloius
(undif ferentioted)

2

4

2

7

M

1

2

b

1

2

1

8

4

1

9

1 !

6

3 1 j8l ; ! 1 1 . j h'<

Lineage 1 (text-fig. 6). From P. dehiscens dehiscens developed P. nothoperhonus sp. nov. {= P. aflf.
P. perbonus sensu Klapper and Johnson 1975; see below) with a small to medium-sized, fairly
shallow basal cavity that is variably inverted at the posterior end of the unit. P. nothoperhonus sp.
nov. gave way, via intermediate forms, to P. inversus in which the basal cavity is almost entirely
inverted except for a relatively large pit anterior to the inward deflection of the keel. Again via
intermediate forms, P. inversus can be seen to have developed into P. serotinus which is characterized
aborally by a basal cavity entirely inverted posterior of the pit (situated close to the sharp inward
deflection of the unit), by development of a lip on the outer margin of the pit, and, orally, by the
outer margin being higher than the carina and inner margin. The distinctive lip is supported by a
build-up of material aborally or a bulge in the platform as shown by the deflected striations (PI. 33,
fig. 9). The reduction in size of the basal cavity in this lineage results from progressive inversion of
the cavity (text-fig. 6).

Lineage 2 (text-fig. 6). P. dehiscens abyssus subsp. nov. developed into P. perbonus, both forms
having a basal cavity with a V-shaped profile. P. perbonus in turn gave way through intermediate
forms to a ‘large-lipped’ P. perbonus herein described as P. labiosus sp. nov. It is as if the flanges of
the V-shaped cavity of P. perbonus ‘seamed up’, leaving the ‘lips’ protruding. Transitional forms

262

PALAEONTOLOGY, VOLUME 30

TABLE 4. Distribution of conodont elements in samples principally from the Buchan Caves Lime-
stone, Shanahan Limestone, and the Taravale Formation at Bindi, eastern Victoria.

Metres above base
of section

0

6

CD

ro

(M

(D

rO

N

N

li)

N

CO

0)

N

(VJ

0

po

(M

CD

0)

CD

N

CD

N

N

CD

N

0)

ro

(VJ

N

N

CM

CD

CM

0)

ro

(M

m

1

ro

0

vt

0)

d)

CD

CD

CD

CM

rO

N

1

0)

0)

CD

0)

s

N

CD

ro

s

CD

0)

CD

CD

0

6

0)

1 96 6-10601

0

CM

0

CO

0

Shonarbon Ls.,600mNNE of SBI 1

0

0

0

CD

S

CD

CM

lO

to

0)

N

rO

CD

(M

CD

N

m

0)

s

to

0

in

CVJ

CD

cn

(M

CD

ro

CD

CD

00

1

fO

CM

N

CM

CD

N

1 86 7 -90 9

Sample number

0

£

3

0

V

z

0

m

B

If)

ro

0

If)

ff)

If)

N/

ro

(\J

t

w

\N

If)

N

W

in

(VJ

Z

m

ro

0)

(VJ

A

0)

ro

CD

ro

0

m

CD

CD

m

0

m

G

m

0

CD

CD

in

JL

in

CD

m

0

CD

L

0

N

in

CD

Y

0

0

0

1

m

in

S

0

N

ro

CD

E(

s

;t

m

0

CM

N

CO

0

CD

0

CM

?

N

1

0

CM

CM

0

s

CM

1

0

vf

(M

C

SE

CD

5

0

a>

n

0)

0

>

0

p

CT

S

L

E

S

Cs

CD

V)

0

H/

IIV

of

H

or

Bl

(V

m

(0

M

tN

e:

IN

th

ac

10

(D

m

FH

ftk

5T

It

At

ife

o

CO

RO

A

ON

A

lie

in

m

m

UC

E

N

w

CD

CD

(D

3H

")

N

CD

CD

h

0)

a

£

0

CO

s

IN

w

1-

=^0

FO

ro

h

T

Rf

in

h

SA

FA

AA

ID

1-

M

RA

Tl

0)

h

pt

V/

or

iti

h

_E

\L

to

lO

h

E

M-

to

h

CD

P

S

L

C

CD

U

J

<

CD

OL

.IN

<

s

CJ

J

<

CD

T

IE

C

CD

U

J

<

(D

H

ST

E

m

CJ

J

<

CO

A

or

:t

u

J

<

CO

Rt

JE

10

VI

N

X

0)

0

b

c

0

u

c

3

Weight of sample (kg)

f£)

in

CD

ih

If)

0)

't

CD

CD

0)

in

CD

N

CD

CVJ

d)

CD

CM

m

in

m

m

CD

0)

0

m

lO

CD

S

CD

m

0

CD

CD

in

vJ-

m

CD

m

K

in

CD

in

m

m

CD

0

CD

U)

CD

m

m

m

CD

m

CD

m

(M

CD

m

CD

M-

CD

CD

(M

CD

0

CD

N

CD

CD

CM

Oulodus murnndalensis

Pa

1

Pb

M

2

2

Sa

Sb

1

Sc

1

1

1

1

Polygnathus spp.

(See Table 6 for
differenfiQtion of
Pa elements)

Po

2

II

22

1

5

1

1

3

17

4

9

2

1

26

25

1

4

7

l£

6

Pb

m

3

3

?

1

2

3

1

1

2

3

2

1

2

1

8

2

1

1

2

1

1

1

1

1

1

Sa

1

Sb

1

3

1

3

1

Sc

1

3

7

1

1

1

1

1

4

4

?

P

4

3

1

1

4

1

2

Ozarkodino buchonensis

Pa

14

Pb

M

Sa

Sb

Sc

3

0 excavata excovola

Pa

1

Pb

1

1

M

Sa

1

Sb

Sc

i

1

1

1

0 linearis

Pa

5

1

3

1

Pb

M

Sa

1

1

Sb

Sc

0 prolata

Pa

5

8

12

7

31

39

22

18

16

21

21

2

4

20

1

5

2

8

20

33

1

39

18

5

6

1

Pb

1

8

10

7

1

4

6

2

4

1

1

2

2

13

4

1

M

6

1

1

3

1

1

2

1

So

1

1

1

1

Sb

4

2

2

2

5

Sc

1

2

3

7

4

1

?

z

1

1C

2

4

8

3

24

14

2

2

2

1

Pandonnellina exigua
exigua

Pa

6

49

13

13

9

23

8

7

4

7

1

4

1

5

1

3

5

t

II

6

4

1

IQ

4

T

2

1

Pb

4

1

2

4

1

1

1

2

1

2

M

2

1

2

2

1

1

Sa

Sb

Sc

51

2

4

3

3

3

1

2

1

Belodella devonico
Belodello resimo
Belodella triangularis
Drepanodus sp.
Panderodus unicostatus
(undifferentiated)

Panderodus valgus

1

1

P

1

-

1

1

2

1

1

1

1

2

2

2

2

1

1

M

7

S

1

28

2

3

1

1

1

1

2

2

1

2

2

1

5

3

2

3

15

r

5

6

2

1

2

1

1

4

MAWSON: EARLY DEVONIAN CONODONTS

263

TABLE 5. Distribution of Pa elements of Polygnathus spp. at Buchan, eastern Victoria.

P dehiscens dehiscens
Transitional to P nolhoperbonus
P nothoperbonus
Transitional to Pmversus
P inversus

Transitional to P serotinus
P serotinus
P dehiscens obyssus
Transitional to P perbonus
P perbonus
P tabiosus

Tronsitionol to P pseudoserotinus
P pseudoserolinus

TABLE 6. Distribution of Pa elements of Polygnathus spp. at The Basin and Bindi, eastern Victoria.

BONANZA

GULLY

SECTION

'b oO
m CTl o SD O ip ^ -rf

roro>^ir)C£><i>— CM

I I I I I 1 I I

O) LO O m roo
CMrO'=^’iOlO^(X>CM

■z. o —

O ^

CD

P dehiscens dehiscens
Transitional to P nothoperbonus
P nothoperbonus
Transitional to P inversus
Pmversus

Transitional to P serotinus
P serotinus
P dehiscens obyssus
Transitional to P perbonus
P perbonus
P tabiosus

Transitional to P pseudoserotinus
P pseudoserotinus

53

4

5

ojro

1-1-

<OH
— 1—0
I^COLU
triiJCD

81^
C/>— lo

•sf so t^
o

SLOCOMBES SECTION
Buchan Caves L/s
at the Basin

oocuoioooomooo

c\j'd'ir>(otx)c£>coioir)0 — lO
Q ^ CM CMCM

SHANAHAN

LIMSTONE

SECTION

Shanahan L/s E side of gully
Taravale Fm below SBI

LD

CD CD
(/) CD

are seen from F. labiosus to a form superficially similar to P. serotinus, herein described as P.
pseudoserotinus sp. nov., where the small, subcircular, shelf-like protuberance is suspended (rather
than supported by a bulge in the platform on the outer side of the pit as in P. serotinus), representing
the vestigial lips of P. labiosus. The intermediate forms show development of the higher outer
margin as the basal cavity ‘seams up’ asymmetrically leaving a lip on the outer margin of the pit.
The reduction in size of the basal cavity in this lineage has resulted from the progressive ‘seaming
up’ of the cavity (text-fig. 6).

Quo vadis polygnathidsl Although the Buchan and Bindi samples have yielded no polygnathids
younger than P. serotinus, faunas of the serotinus and younger zones are known from elsewhere in
eastern Australia, notably from the Broken River (Telford 1975) and Timor, New South Wales
(Pedder, Jackson and Ellenor 1970). The best sequences for scrutiny of serotinus and later zones
spanning the Early Devonian-Middle Devonian boundary are to be found in the Broken River

264

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 6. Proposed lineages of Early Devonian Polygnathus in Australia; the right-hand branch

is a northern hemisphere lineage.

MAWSON: EARLY DEVONIAN CONODONTS

265

area (Mawson et al. 1985), which includes the type locality for P. serotinus (Telford 1975). From
examination of Telford’s original material of P. serotinus and topotype material (Mawson et al.
1985), the lip adjacent to the basal pit is buttressed by shell material rather than being a free, ledge-
like lip. Localities yielding P. serotinus also contain specimens intermediate between P. labiosus and
P. pseudoserotinus.

Weddige and Ziegler (1979, fig. 1, p. 160), using material from the Rhenish Slate Mountains,
proposed a ‘tree’ with three main branches to depict the evolution of polygnathids. All three
branches, the linguiformis branch, the robusticostatus branch, and the costatus branch, are repre-
sented in Australian faunas, including the Broken River faunas of north Queensland (Telford 1975;
Mawson et al. 1985). As Australian faunas do not contain the dehiscens-gronbergi-laticostatus
lineage (see also Klapper and Johnson 1980), forms ancestral to the costatus main branch probably
arrived in Australia towards the end of the Early Devonian when seaways were becoming less
restricted (Charpentier 1984) and there was an increase in cosmopolitanism of conodont faunas
(Telford 1972, 1979; Klapper and Johnson 1980; Charpentier 1984).

DACRYOCONARIDS OF THE TARAVALE FORMATION

Initially it was envisaged that this study would include a parallel zonation of dacryoconarids and
conodonts of the Taravale Formation at Buchan. Although over two hundred horizons in the
Gelantipy Road section and a further forty in the entrance to the Buchan Caves Road section have
been sampled, washed, and picked and many dozens of thin sections prepared, the work has been
hindered by the lack of complete specimens; most of the thousands of specimens picked from the
samples lack the initial chamber, making identification, in many instances, impossible. Thin section-
ing of the specimens contained within limestone nodules has not gone far towards solving the
problem as the crystallinity of preservation of most material has, at high magnification, obscured
structural details.

Some species of dacryoconarids from the Taravale Formation are new, and some similar to, if
not the same as, undescribed species from La Grange, France, presently being studied by Professor
Hubert Lardeux (pers. comm.). Work on the identification and zonation of the dacryoconarids will
continue with a view to calibrating the conodont and dacryoconarid successions for the dehiscens
to serotinus zones.

AGE OF AMMONOIDS PREVIOUSEY DESCRIBED FROM THE
TARAVALE FORMATION

One of the objectives of the present investigation was to provide a firmer basis for dating the
Taravale Formation goniatites and bactritids. When these were first recorded from Buchan by
Teichert (1948), the association of Gyroceratites and Lobobactrites was taken to be indicative of an
early Middle Devonian age. At that time there were very few records of ammonoids from horizons
that were of indubitable late Early Devonian age; all of these, mostly poorly known and, in terms
of modern stratigraphic imperatives, imprecisely located stratigraphically, were from Germany or
Bohemia. The fauna was subsequently expanded by Erben (1964, 1965) who concluded that the
Buchan ammonoids must be of late Emsian age; this was the view advocated by Chlupac (1976) in
a review of the oldest goniatite faunas and their stratigraphic significance. Subsequently, House
(1979) placed the two genera based on type species from the Taravale Formation, Teicherticeras
and Talenticeras, doubtfully within the early but not earliest Emsian.

Conodont studies have vastly improved correlations of ammonoid-bearing units and time-ranges
for many early ammonoids, thus providing a framework within which their early evolution can be
more precisely understood. Among such studies is that of the Nandan facies of South China by
Xian et al. ( 1 980) who combined biostratigraphic studies of conodonts, dacryoconarids, ammonoids,
trilobites, and tabulate corals. Interestingly, they documented the co-occurrence of Anetoceras and

266

PALAEONTOLOGY, VOLUME 30

Erhenoceras with Polygnathus perbonus and Nowakia barrandei Boucek and Prantl, but not with P.
dehiscens, though with the range of Anetoceras possibly extending down into an interval wherein P.
dehiscens may overlap with P. perbonus. The situation seems to be similar in the Bohemian where
the earliest goniatites appear at the top of the Zlichov Limestone, again associated with N. barrandei
and with P. perbonus or P. Iperbonus according to Chlupac et al. (1977). In a review of the Devonian
biostratigraphy of Guangxi, Bai et al. (1980, p. 5, fig. 4) showed that the range of Anetoceras
coincides with that of P. dehiscens for approximately 8 m in the Chongzhou Formation. Bultynck
and Hollard (1980, pp. 13-14, figs. 2 and 3) showed A. advolvens occurring with P. gronbergi (their
Fauna III) and with Icriodus bilatericrescens (their Fauna II). The latter may be a pre-gronbergi-
perbonus Zone ammonoid occurrence, but in an assemblage lacking polygnathids it is difficult to
detennine if Fauna II belongs to the upper dehiscens Zone or to the lower gronbergi Zone because
I. bilatericrescens spans both zones. There can be little doubt that the type localities for the oldest
goniatites at Buchan occur within the dehiscens Zone (text-fig. 7), within the upper part of the zone,
in early rather than late Zlichovian.

X-5
X-4
X-3
X-2
- I
X I
2

X 3
X 4

5^

5-

3

xio— ^

-

— O

r

XI4^

^16-SJb-

18^'^'

I C

- ^
CD

19-

X2I-

- o
c

23

o

. Q_

^25

27-

29

31

Section terminated at 118 9m
(equivalent to 67 9m of
Taravale Fm in cutting between
localities 31 and -5-5 )

- Teicherhceras desiderotus
(TeichertKtype locality)

100 m Talenliceros lalenli
Erben (type locality)
Teicherticeras sp.Dand Isp.E
of Erben

' Talenliceros sp., Loboboctnfes
inopinafus, L. sp and
Bacfrifes sp.

■Lobobactriles sp

— lowest horizon sampled
50m

Lower part of Taravale Formafion
concealed by alluvium .colluvium
and soil. Lowest horizon sampled
(BCE 31) IS approximately 52m
above top of Buchan Coves Limestone

'BCE' Section (cutting along entrance to Buchan Caves Reserve)

■X = horizons that yielded conodonts

TEXT-FIG. 7. Detail of Buchan Caves Entrance (BCE) section showing strati-
graphic position of goniatites, bactritids, and polygnathid conodonts.

MAWSON: EARLY DEVONIAN CONODONTS

267

The conodont correlations for the various goniatites and bactritids from theTaravale Formation
can be summarized as follows:

1. The type locality for Talenticems talenti Erben and the horizons of Teicherticeras n. sp. D
and n. sp. E of Erben (1965) fall well within the dehiscesn Zone (text-fig. 7, Table 3).

2. The type locality of Teicherticeras desideratum (Teichert) occurs low in the perbouus Zone
(text-fig. 7, Tables 3 and 5).

3. The type locality for Lobobactrites inopinatus Teichert (his loc. C, 1948), from beds in the
centre of the anticline from which the Gelantipy Road stratigraphic section was commenced (text-
fig. 7, Tables 1 and 5), is from the dehiscens Zone. The paratype, from Teichert’s locality B, is not
precisely located (confirmatory conodont evidence is not available), but is surely within the perbouus
Zone. The best estimate of this horizon, taking into account the outcrops around the first hairpin
bend on the Gelantipy Road north of Buchan, is that it is approximately equivalent to the sampled
horizons 8. 8. 4. 2 or 8. 8. 4. 3 on the Gelantipy Road section, i.e. a few metres above the horizon of
the holotype of T. desideratum. Teichert (1948) believed these two horizons to be the same; this
may well be the case.

4. The localities for the holotype and paratypes of Bactrites howitti Teichert (Iocs. D and E of
Teichert 1948) were not sampled for conodonts but were considered by Teichert to be approximately
equivalent stratigraphically. Their position, about 90 m above the top of the Buchan Caves Lime-
stone, suggests a horizon about the same as that of the type locality for Talenticeras talenti in the
entrance cutting to the Buchan Caves Reserve, i.e. within the dehiscens Zone; but the suggestion of
decreasing thickness of section north-eastward towards The Basin may indicate that these horizons
are slightly younger, i.e. within the lower part of the perbouus Zone. Several as yet unidentified
specimens of Bactrites, seemingly belonging to this species, have been collected from dehiscens
horizons in the entrance cutting to the Buchan Caves Reserve.

In summary, the ammonoid faunas of the Taravale Formation are restricted to the dehiscens and
perbouus zones, with at least three species of goniatites and one (possibly two) species of bactritids
occurring in the dehiscens Zone, thus making them among the oldest firmly dated ammonoids in
the world.

CONCLUSIONS

Basic data for evaluation of four of the Australian Early Devonian {dehiscens to serotinus zones)
conodont faunas has been presented. There are, admittedly, a number of elements (some new) that
at this stage appear to be endemic. Some endemism in Australian faunas has been pointed out
previously (Telford 1972, 1979; Fahraeus 1976; Klapper and Johnson 1980; Charpentier 1984), but
its degree in Australian conodont faunas for the late Early Devonian is not sufficient to confound
international correlation. The apparent extensions of range of some previously described species
produce no profound discrepancies that would call into question significant parts of the present
zonal scheme for this part of the Early Devonian. The present scheme has evolved rapidly, particu-
larly over the past decade, and has been thoroughly reviewed and reworked recently (Klapper and
Ziegler 1979; Klapper and Johnson 1980; Ziegler and Klapper 1985). The increased data presented
here for at least four of the zones has reinforced the applicability of this zonal scheme to Australian
faunas.

Dehiscens Zone

The key species of this zone. Polygnathus dehiscens, is found at the base of the Taravale Formation
at Buchan (text-figs. 1 and 2; OTRC section) and high in the Buchan Caves Limestone at Bindi
(text-fig. 3, BON section), confirming the diachronous nature of the top of the Buchan Caves
Limestone between Buchan and Bindi. With P. dehiscens dehiscens occurs P. d. abyssus subsp. nov.
(whose range is similar to that of P. d. dehiscens)', P. d. dehiscens extends upwards to 105-6 m above
the top of the Buchan Caves Limestone in the Taravale Eormation at Buchan, and P. d. abyssus
extends to 116 m above the base of the same section (Table 5). The type locality for Talenticeras
talenti Erben lies between beds BCE 12 and BCEll, 100 m above the base of the Buchan Caves

268

PALAEONTOLOGY, VOLUME 30

Limestone in the section at the entrace to the Buchan Caves, within the dehiscens Zone. Sixteen
metres higher in the section, within bed BCE3, the type locality for Teicherticeras desideratus
(Teichert) falls low in the perbonus Zone.

Elsewhere in Australia, conodont faunas of the dehiscens Zone have been documented from New
South Wales at Ravine (Flood 1969) and Wee Jasper (Philip and Jackson 1967; Pedder, Jackson
and Philip 1970).

Perbonus Zone

As the differences in the ranges of P. perbonus (from 1 16 m to 271 m above the base of the Buchan
Caves Limestone) and P. nothoperbonus (from 1 19 m to 272 m) are not great in the Gelantipy Road
section at Buchan (Table 5), they are taken to be approximate stratigraphical equivalents. Apart
from at Buchan, Bindi, and The Basin, the perbonus Zone is known to occur in Australia at Wee
Jasper (Philip and Jackson 1967; Pedder, Jackson and Philip 1970). Pickett (1980) reported a single
specimen of P. perbonus from Cobar; the specimens referred to as ‘P. perbonus late form’ by Pickett
(1978) from Mount Frome, are probably P. perbonus sensu stricto.

Inversus Zone

P. inversus and P. labiosiis seem to have broadly the same stratigraphic range, but P. labiosus arises
29 m lower in the Taravale Formation along the Gelantipy Road. P. labiosus overlaps considerably
in range with P. perbonus, but in the Gelantipy Road section there is a gap of 3 m between the
highest occurrence of P. perbonus and the first occurrence of P. inversus. Elsewhere in Australia,
faunas of inversus zone age occur in Queensland at the Broken River (Telford 1975) and in the
Nogoa anticline, Springsure, Queensland (Fordham 1976).

Serotinus Zone

P. pseudoserotinus, assumed to be broadly equivalent stratigraphically to P. serotinus, first appears
in the Gelantipy Road section 419 m above the top of the Buchan Caves Limestone; the entry of P.
serotinus is some 34 m higher (Table 5). At Bindi, in the SALC (south arm of Limestone Creek)
section, both forms occur (Table 6). It is assumed that these occurrences represent very early forms
of P. serotinus as they occur in sequence with or after P. inversus. Older forms of P. serotinus occur
in the Jessey Springs section at the Broken River (Mawson et al. 1985); in this section, the upper
boundary of the serotinus Zone is placed at last occurrence of Pandorinellina exigua e.xigua (Mawson
et al. 1985). Key species of this zone have been reported from Wee Jasper, New South Wales (Philip
and Jackson 1967; Pedder, Jackson and Philip 1970), Mount Frome, New South Wales (Pickett
1978), and the Nogoa anticline, Springsure, Queensland (Fordham 1976).

SYSTEMATIC PALAEONTOLOGY

As the purpose of this study was primarily biostratigraphic, and because the Pa elements of multielement
assemblages show most obvious changes through time, only the Pa elements of the multi-element species of
Polygnathus, Oiilodus, Pandorinellina, and Ozarkodina found at Buchan and Bindi are described in detail.
Other elements are illustrated but not discussed at length. Non-platform elements have not been differentiated
for each of the polygnathid species. No multi-element reconstructions have been attempted for coniform
elements that have been illustrated (PI. 41) but not described. Type and figured specimens are housed in the
collections of the National Museum of Victoria, Melbourne (NMVP) and the Australian Museum, Sydney
(AMF). Precise horizon and locality data for each sample number can be obtained by reference to text-figs.

1 5 and tables 1 -6.

Family hibbardellidae Muller, 1956
Genus oulodus Branson and Mehl, 1933

1933 On/or/n.v Branson and Mehl, p. 116.

1935a GjT0g7zat/!u.? Stauffer, p. f44.

19356 Barbarodina Stauffer, pp. 602-603.

MAWSON: EARLY DEVONIAN CONODONTS

269

1969 Ligonodina Bassler; Jeppsson, pp. 20-21.

1971 Delotaxis Klapper and Philip, p. 446.

Type species. Oulodus serratus 1930).

Remarks. For discussion of the genus and variation within it, see Mawson (1986).

Oulodus nmrrindalensis (Philip, 1966)

Plate 31, figs. 1-9

1966 Lonchodina n. sp. Philip, p. 446, pi. 3, fig. 24 {non figs. 19 and 20) [Pa element].

1966 Lonchodina nmrrindalensis n. sp. Philip, p. 446, pi. 4, figs. 9- 14 [Pb element].

1966 Plectospathodus extensns laceriosus n. subsp. Philip, p. 448, pi. 1, figs. 25-28 [M element].

1966 Trichonodella sp. cf. T. inconstans Walliser; Philip, p. 451, pi. 4, figs. 24 and 25 [Sa element].

1966 Hindeodella equidentata Rhodes; Philip, p. 445, pi. 3, fig. 1 [Sc element].

Holotype. Specimen 8850/32 (Philip 1966, pi. 4, figs. 9, 10, 12) from the Murrindal Limestone (Loc. 6 of Philip
1966, text-fig. 1; see also text-fig. 1 herein) of Buchan, Victoria.

Diagnosis. A species of Oulodus in which all elements are characterized by the irregularity in size
and arrangement of denticles, and the deeply excavated basal cavity beneath each element.

Material. Six specimens of the Pa element from four localities; seventy-three non-platform elements from
twenty-two localities (Tables 1-4).

Discussion. Except for the Sb element, all elements of O. murrindalensis were figured by Philip
(1966) in his study of the conodont faunas from the Buchan Group. The Sb element is digyrate
(‘lonchodiniform’) and, like the Sa element (alate— ‘trichonodelliform’) and the Sc elements (bipen-
nate— ‘ligonodiniform’) of the symmetry transition series, has irregularly spaced denticles of varying
sizes, some discrete and some tending to be fused at their bases. Specimens of O. murrindalensis
collected by Philip (1966) were confined to the perbonus Zone. This study shows the highest
occurrence of O. murrindalensis as bed 8.12 in the Gelantipy Road section within the perbonus
Zone. The first occurrence is in bed OTRC 2, low in the Taravale Formation at Buchan, and in
sample BON 25-27.5, high in the Buchan Caves Limestone at Bindi, confirming its presence also in
the dehiscens Zone.

The irregularity of denticulation in O. murrindalensis clearly separates it from older species of
Oulodus described from Windellama, New South Wales (Mawson 1986).

Family POLYGNATHiDAE Bassler, 1925
Genus polygnathus Hinde, 1879

Type species. Polygnathus dubius Hinde, 1879.

Discussion. See Klapper in Ziegler (1973) for discussion of the genus. Pa elements of P. dehiscens
dehiscens, P. d. abyssus, P. perbonus, P. nothoperbonus, P. inversus, P. labiosus, P. serotinus, and P.
pseudoserotinus are described below in full. The non-platform elements (Klapper and Philip 1971;
Klapper in Ziegler 1973) have been described previously (see Philip 1966) and are illustrated herein
(PI. 36, figs. 11-18). As the majority of conodonts for this study were recovered from the nodular
limestone within the sequence of mudstones, shales, and impure limestones of the Taravale Forma-
tion at Buchan and Bindi, it is not surprising that the numbers of specimens, especially the less
robust non-platform elements, are relatively low. It is surprising, however, that only seven specimens
of the Sa (alate— ‘diplodellan’-type) element of the polygnathid apparatuses have been recovered
from the 325 samples that yielded conodonts for this study. Philip (1966) recorded the presence of
the Sa element, 'Roundya perbona' from the Murrindal Limestone but not from the Buchan Caves
Limestone at Buchan, negative evidence consistent with a pre-dehiscens Zone age for the faunas of
the Buchan Caves Limestone at Buchan. As discussed earlier, the uppermost Buchan Caves Lime-
stone at Bindi, containing both P. dehiscens and P. perbonus, is younger than the top of the Buchan

270

PALAEONTOLOGY, VOLUME 30

Caves Limestone at Buchan. From both high and low in the sections at Wee Jasper, New South
Wales, rare occurrences of the Sa element, ‘Hibbardella perbona have been reported (Pedder,
Jackson and Philip 1970). Altogether 711 polygnathid Pa elements that can be speciated with
certainty from ninety-eight localities (Tables 5 and 6), a further forty-three Pa elements that cannot
be assigned with certainty because of preservation, juvenile forms, etc., and 459 non-platform
elements from 105 localities (Tables 1-4) have been recovered from Buchan and Bindi localities.

Polygnathus dehiscens Philip and Jackson, 1967

Amended diagnosis. Pa elements have a large basal cavity occupying most of platform except for
crimp. Cavity is V-shaped or flat or in form of extremely shallow trough at posterior end.

Discussion. This study shows that the profile of the basal cavity of P. dehiscens varies from flat to
distinctly V-shaped, and on this basis two subspecies are distinguished: the nominate subspecies
and P. dehiscens abyssus n. subsp.

Polygnathus dehiscens dehiscens Philip and Jackson, 1967

Plate 32, figs. 1-10; Plate 36, fig. 6

1967 Polygnathus linguiformis linguiformis Hinde; Adrichem Boogaert, p. 184, pi. 3, fig. 1.

1967 Polygnathus linguiformis dehiscens n. subsp. Philip and Jackson, p. 1265, figs. 2i-k (listed
incorrectly as h-j in figure caption), 3n.

1969 Polygnathus lenzi n. sp. Klapper, pp. 14-15, pi. 6, figs. 9-18.

1969 Polygnathus webbi excavata n. subsp. Carls and Gandl, pp. 193-195, pi. 18, figs. 9-13.

1969 Pohgnathus linguiformis foveolata Philip and Jackson; Carls and Gandl, p. 196, pi. 18, figs. 14-
19,22.

1970 Polygnathus linguiformis dehiscens Philip and Jackson; Philip and Jackson in Pedder, Jackson
and Philip, p. 216, pi. 40, figs. 18, 20.

1971 Polygnathus dehiscens Philip and Jackson; Fahraeus, pp. 677-678, pi. 77, figs. 1-12.

1972 Polygnathus lenzi Klapper; Uyeno in McGregor and Uyeno, p. 14, pi. 5, figs. 10-12.

1974 Polygnathus dehiscens Philip and Jackson; Klapper in Perry et al., p. 1087.

1975 Polygnathus dehiscens Philip and Jackson; Klapper and Johnson, pp. 72, 73, pi. 1, figs. 1-8,
13-16.

1976 Polygnathus dehiscens Philip and Jackson; Lane and Ormiston, pi. 1, figs. 17-20.

1976 Polygnathus dehiscens Philip and Jackson; Bultynck, pi. 11, figs. 1-15.

1977 Polygnathus dehiscens Philip and Jackson; Klapper in Ziegler, pp. 447-448, Polygnathus pi. 8,
figs. 7 and 8.

1977 Polygnathus dehiscens Philip and Jackson; Savage, pi. 1, figs. 29-36.

1978 Polygnathus dehiscens Philip and Jackson; Apekina and Mashkova, pi. 74, figs. 1, 9; pi 75,
figs. 1 and 2.

1979 Polygnathus dehiscens Philip and Jackson; Lane and Ormiston, pi. 5, figs. 24-26, 35, 36.

EXPLANATION OF PLATE 31

Figs. 1-9. Oulodus murrindalensis (Philip). Pa elements: 1, NMVP 99075, sample SLO 200, x45. 2, NMVP
99076, sample SLO 60, x 45. Pb elements: 3, NMVP 99077, sample OTRC 2, x 60. 4, NMVP 99078,
sample OTRC 2, x 60. M element: 7, NMVP 99079, sample OTRC 2, x 45. Sa elements: 8, NMVP 99080,
sample SLO 130, x 45; 9, NMVP 99081, sample SLO 10, x 45. Sb elements: 5, NMVP 99082, sample SLO
60; 6, NMVP 99083, sample SLO 52, x 45.

Figs. 10-16. Ozarkodina excavata excavata (Branson and Mehl). Pa element: 1 1, NMVP 99084, sample 8.10,
x45. Pb element: 12, NMVP 99085, sample BON 25-27.5, x45. M element; 10, NMVP 99086, sample
OTRC 1, x45. Sb elements: 13, NMVP 99087, sample BCE-4, x45; 14, NMVP 99088, sample BON 25-
27.5, X 45. Sc elements: 15, NMVP 99089, sample Forrest Motel, x 45; 16, NMVP 99090, sample SLO 60,
x45.

PLATE 31

MAWSON, Oidodus, Ozarkodina

272

PALAEONTOLOGY, VOLUME 30

1980 Polygnathus dehiscens Philip and Jackson; Bultynck and Hollard, pi. 2, fig. 5.

1980 Polygnathus dehiscens Philip and Jackson; Uyeno and Klapper, pi. 8.1, figs. 1-4.

1980 Polygnathus dehiscens Philip and Jackson; Xiong in Xian et al., pi. 22, figs. 1-4, 9-12, 19, 20.

Diagnosis. ‘Representative Pa elements of Polygnathus dehiscens dehiscens have a large basal cavity
occupying most of platform except for crimp. Cavity is flat or in form of extremely shallow trough
at posterior end.’ (Klapper in Ziegler 1977, p. 447.)

Material. Twenty-seven specimens from eleven localities; twenty-eight specimens transitional between P.
dehiscens dehiscens and P. nothoperbonus from seventeen localities (Tables 5 and 6).

Discussion. In the material examined from Buchan and Bindi, two clusters of specimens can be
distinguished: one in which specimens have a flat basal cavity, and one in which specimens have a
deep V-shaped basal cavity. As a polygnathid lineage arises from each of these forms, the configura-
tion of the basal cavity is taken as a sufficient criterion on which to base two subspecies, forms with
the deeper cavity being discriminated as a new subspecies, P. dehiscens abyssus, described below.

Polygnathus dehiscens abyssus subsp. nov.

Plate 34, figs. 1-7; Plate 36, fig. 1

1969 Polygnathus linguiformis dehiscens Philip and Jackson; Flood, p. 9, pi. 2, figs. 1-6.

1970 Polygnathus linguiformis dehiscens Philip and Jackson; Philip and Jackson in Pedder, Jackson
and Philip, pi. 40, figs. 21 and 23.

1978 Polygnathus lenzi Klapper; Wang and Wang, pp. 340-341, pi. 41, figs. 1-3, 7-9, 24-26.

Derivation of name, abyssus (Lat.) = deep, in reference to the deep, V-shaped basal cavity.

Holotype. AMF 66031 (PI. 34, figs. 3 and 4) from the Buchan Caves Limestone, 1T4 to 12-3 m above the base
of the BON section (sample BON 36-39) at Bindi, Victoria.

Material. Seventy-eight specimens from twenty-four localities; thirty-three specimens transitional between P.
dehiscens abyssus and P. perbonus from nineteen localities (Tables 5 and 6).

Diagnosis. Pa elements have basal cavity occupying most of the platform except for the crimp with
lips of the basal cavity forming a V-shaped profile in cross section.

Discussion. Specimens previously designated P. dehiscens from the Lick Hole Limestone at Ravine,
New South Wales have a deeper basal cavity than specimens illustrated from Taemas, New South
Wales; compare specimens from Ravine (Flood 1969, pi. 11, figs. 5 and 6; Philip and Jackson in
Pedder, Jackson and Philip 1970, pi. 40, fig. 23) with material from the Cavan Limestone, Taemas
(Ibid. pi. 40, fig. 20— aboral view of holotype). Collections made by Flood and Philip and Jackson
from Ravine, and by Philip and Jackson from Taemas, were examined at Sydney University and
New England University, Armidale in order to confirm the above.

EXPLANATION OF PLATE 32

Figs. 1-10. Polygnathus dehiscens dehiscens Philip and Jackson. 1 and 2, lower and upper views of AMF
66007, sample 7.17, a specimen transitional between P. pireneae and P. d. dehiscens, x 60. 3 and 4, lower
and upper views of AMF 66008, sample OTRC 5, x 60. 5, lower view of AMF 66009, sample OTRC 5,
X 60. 6-10, specimens transitional between P. d. dehiscens and P. nothoperbonus. 6 and 7, lower and upper
views of AMF 66013, sample 8.10.2, x 60; 8, lower view of AMF 66010, sample 8.9, x 75; 9 and 10, upper
and lower views of AMF 6601 1, sample Ma7, x 60.

Figs. 11-15. P. nothoperbonus sp. nov. 1 1 and 12, upper and lower views of AMF 66016, sample BON 46-50,
X 60. 13, lower view of AMF 66015, sample OTRC 16, x 60. 14 and 15, lower and upper views of AMF
66018, sample 8.9.2, x 45.

PLATE 32

MAWSON, Polygnathus

274

PALAEONTOLOGY, VOLUME 30
Polygnathus inversus Klapper and Johnson, 1975
Plate 33, figs. 3-8; Plate 36, figs. 8 and 9
Synonymy. See Klapper and Johnson (1975, p. 73; 1980, p. 453).

Material. Seventy-eight specimens from thirteen localities; forty-one specimens transitional between P. inversus
and P. serotinus from nine localities (Tables 5 and 6).

Discussion. Specimens of P. inversus from Buchan and Bindi show the same range of variation as
the Nevada material (Klapper and Johnson 1975), from P. inversus s. s. to forms transitional to P.
serotinus in which an incipient bulge surmounted by a lip is noticeable on the outer side of the basal
pit, but the height of the inner and outer margin on the oral surface anterior of the angular deflection
of the platform is still the same. The pit, anterior groove, and posterior keel area are bounded by
narrow bands of lamellae edge (inverted area) indicative of the plane of attachment of the basal
body.

Polygnathus labiosus sp. nov.

Plate 35, figs. 1-9; Plate 36, figs. 3 and 4

Derivation of name, labiosus (Lat.) = large-lipped, in reference to the flared nature of the flanks of the basal
cavity.

Holotype. AMF 66043 (PI. 35, figs. 5 and 6) from the Taravale Formation, 340-5 m above the base of the
Gelantipy Road section (sample 15.2) at Buchan, Victoria.

Diagnosis. Pa elements have a basal cavity with large flaring lips in the centre of the unit anterior
to the inward deflection of platform. Cavity is joined posterior of lips and extends as groove
anterior of lips. Transverse ridges cross posterior third of platform and can be either complete or
interrupted.

Material. Forty-two specimens from twelve localities; twenty-eight specimens transitional between P. labiosus
and P. pseudoserotinus from nine localities (Tables 5 and 6).

Discussion. The strong inwardly deflected posterior platform of P. labiosus shows the same variation
in arrangement of transverse ridges as in P. perbonus. The parallel striations of lamellae bound-
ing the posterior keel and cavity in P. inversus are not developed in P. labiosus as the basal
body material is restricted to the interior of the lips of the basal cavity and does not extend to the
posterior as in P. inversus. The flaring lips of P. labiosus are accentuated in specimens where
basal body material is preserved (PI. 35, fig. 2) but is clearly differentiated from the extended lips
of P. labiosus by the manner of attachment, composition, and colour. There is a gradation from
P. perbonus to P. labiosus, but the latter can be easily distinguished by the extended lips of the
basal cavity.

EXPLANATION OF PLATE 33

Figs. 1 and 2. Polygnathus nothoperbonus sp. nov. Upper and lower views of AMF 66020, sample T3, a
specimen transitional between P. nothoperbonus and P. inversus, x 60.

Figs. 3-8. P. inversus Klapper and Johnson. 3 and 4, lower and upper views of AMF 66033, sample T3, x 75.
5, lower view of AMF 66021, sample T2, x75. 6-8, specimens transitional between P. inversus and P.
serotinus. 6 and 7, lower and upper views of AMF 66024, sample T3, x 90; 8, lower view of AMF 66026,
sample T3, x 90.

Figs. 9-12. P. serotinus ‘delta morphotype’ Telford. 9, lower view of AMF 66027, sample EB 3, x75. 10,
lower view of AMF 66028, sample T3, x 45. 11 and 12, lower and upper views of AMF 66029, sample T2,
x90.

PLATE 33

MAWSON, Polygnathus

276

PALAEONTOLOGY, VOLUME 30
Polygnathus nothoperbonus sp. nov.

Plate 32, figs. 11-15; Plate 33, figs. 1 and 2; Plate 36, fig. 7

1975 Polygnathus aff. P. perbonus (Philip); Klapper and Johnson, p. 74, pi. 2, figs. 1-10.

1979 Polygnathus aff. P. perbonus (Philip); Lane and Ormiston, p. 62, pi. 8, figs. 26 and 27.

1980 Polygnathus aff. P. perbonus (Philip); Uyeno and Klapper, pi. 8.1, figs. 5 and 6; pi. 8.3, figs. 11
and 12.

Derivation of name, notho (Gr.) = spurious, bastard, in reference to its superficial similarity to P. perbonus.

Holotype. University of Iowa 38018 (Klapper and Johnson 1975, pi. 2, figs. 7 and 8) from the Baratine Member
of McColley Canyon Formation at Lone Mountain, Nevada (J-30-73, 172 m above formation base at loc. 1
of Klapper and Johnson 1975, fig. 2).

Diagnosis. Representative Pa elements of P. nothoperbonus have a medium-sized basal cavity ex-
panded beneath the central part of the platform anterior of the sharp, inward deflection of the
platform, inverted posteriorly, and extended anteriorly as a narrow groove. The cavity is flat or
shallow. Transverse ridges crossing posterior third of oral platform usually interrupted. Anterior
outer platform margin at about same height as inner margin.

Material. Ninety-four specimens from thirty-two localities; thirty-one specimens transitional between P.
nothoperbonus and P. inversus from thirteen localities (Tables 5 and 6).

Discussion. P. aff. P. perbonus {sensu Klapper and Johnson 1975) is here given specific status for
the following reasons: it can be clearly separated from P. perbonus by its shallow basal cavity, the
lips (or lateral flanks) of which are only slightly V-shaped, and by the clear inversion of the posterior
portion of the cavity; and it completes the P. serotinus lineage. In their original description and
diagnosis of P. perbonus, Philip and Jackson (1967, p. 1265) made no special mention of the oral
ornamentation, although Klapper and Johnson (1975), in an amended diagnosis of the form, stated
that ‘Transverse ridges cross [the] posterior third of the platform’. They remarked that although P.
perbonus had not been found in Nevada, P. aff. P. perbonus had, the latter having a shallower basal
cavity and showing some variation in the posterior transverse ridges compared with P. perbonus ‘in
which the transverse ridges are not interrupted’.

From a study of polygnathid faunas from Buchan, and especially from collections of topotype
material from the Murrindal Limestone (Hyland and Mawson, in prep.), it appears that P. perbonus
exhibits variation in arrangement of the posterior transverse ridges on its oral surface. Text-fig. 8
compares oral views of the holotype of P. linguiformis foveolata Philip and Jackson (= P. perbonus)
and specimens of P. aff. P. perbonus from Nevada. It appears that the variation in the transverse
ridges in P. aff. P. perbonus falls within the range of variation shown by P. perbonus. The difference
then between P. perbonus and P. aff. P. perbonus (sensu Klapper and Johnson 1975) is the depth of
the basal cavity, and it is on the basis of this feature that P. nothoperbonus sp. nov. is herein erected
to include P. aff. P. perbonus.

From the Buchan and Bindi material it is clear that P. nothoperbonus is intermediate between P.
dehiscens dehiscens and P. inversus, the latter showing complete inversion of the basal cavity
posterior to the cavity pit. The oral ornamentation of P. nothoperbonus shows the same variation
as P. inversus. The occurrence of P. nothoperbonus elsewhere (e.g. Sor Fiord, Ellesmere Island; see
Uyeno and Klapper 1980, p. 85), entirely below the first occurrence of P. inversus and occurring
together with P. inversus (e.g. Blue Fiord, Ellesmere Island; Uyeno and Klapper 1980), lends
support to the reality of the dehiscens dehiscens serotinus lineage occurring both in the northern
hemisphere and in Australia.

Polygnathus perbonus (Philip, 1966)
Plate 34, figs. 8-13; Plate 36, fig. 2
Synonymy. See Klapper and Johnson (1975, p. 74; 1980, p. 454).

MAWSON: EARLY DEVONIAN CONODONTS

111

Material. 159 specimens from thirty-three localities (Tables 5 and 6).

Discussion. P. perbonus is characterized by a medium-sized, V-shaped basal cavity expanded beneath
the central part of the platform anterior of the sharp inward deflection of the platform and
continuing posteriorly as a keel, where the basal cavity has ‘seamed’ together. There appear to be
no parallel striations flanking the keel as basal body material is restricted to the inner portion of
the cavity. Although Philip (1966, p. 448) noted that ‘deflected posterior portion of platform bears
transverse ridges’, he illustrated (1966, pi. 2, figs. 35, 36, 39) specimens where these transverse ridges
are not continuous; the specimen he illustrated as pi. 2, fig. 36 was later chosen as the holotype for
P. foveolatus (= P. perbonus) by Philip and Jackson (1967). The posterior portion of the platform
of P. perbonus, therefore, is characterized by transverse ridges which are not always continuous
(see above, in discussion of oral ornamentation of P. nothoperbonus).

TEXT-FIG. 8. Line-drawing comparison of Polygnathus aff. P. perbonus of
Klapper and Johnson (1980) and the holotype of P. foveolata = P. perbonus
(Philip). A, P. aff. P. perbonus (Klapper and Johnson 1975, pi. 2, fig. 7). b,
P. aff. P. perbonus (Klapper and Johnson 1975, pi. 2, fig. 1). c, P. perbonus
(Philip 1966, pi. 2, fig. 39); note transverse lines not continuous across
posterior third of platform.

Polygnathus pseudoserotinus sp. nov.

Plate 35, figs. 10-12; Plate 36, fig. 5

1979 Polygnathus serotinus &\ph3Lmorp\\otypeTe\ior&, Lane and Ormiston, p. 63, pi. 7, figs. 13 and 37.

Derivation of name, pseudo (Gr. pseudes) = false, in reference to superficial similarity to P. serotinus Telford.

Holotype. AMF 66049 (PI. 35, figs. 11 and 12) from the Taravale Formation, 371 m above the base of the
East Buchan section (sample EB 3) at Buchan, Victoria.

Diagnosis. Pa elements have small pit located slightly anterior of inward deflection of platform. On
outer side of pit, a small, subcircular, shelf-like protuberance is suspended above platform. Cavity
is entirely joined posterior of pit and extends anteriorly of pit as shallow groove. Anterior outer
margin is higher and wider than inner margin.

Material. Thirteen specimens from seven localities (Tables 5 and 6).

Discussion. Although there is a superficial similarity between P. pseudoserotinus and P. serotinus,
the main difference lies in the nature and mode of formation of the tiny lip on the outer side of

278

PALAEONTOLOGY, VOLUME 30

the basal pit. In P. pseudoserotinus the lip is suspended above the platform rather than supported
by a solid buttress of platform material, as occurs in P. serotinus. Plate 35, fig. 1 1 shows the lack of
exposed edges of lamellae in the zone parallel to the keel in P. pseudoserotinus compared with the
clear development of this zone in P. serotinus (PI. 33, figs. 9 and 11). Lane {in Lane and Ormiston
1979) noted this difference when distinguishing two of his three morphotypes of P. serotinus in the
Salmontrout River area, Alaska. Unfortunately, his collections did not include sufficient material
to see how each evolved from P. dehiscens: one by inversion of the basal cavity {sensu Lindstrom
1964), the other by seaming-up of the basal cavity. It could be that specimens described by Wang
(1979) as P. declinatus, and later assigned by Uyeno and Klapper (1980) to P. inversus transitional
to P. serotinus, belong to the new species P. pseudoserotinus. Wang’s description (1979, p. 407) fits
that of P. pseudoserotinus but it is not clear from his pi. 1, figs. 12-22 whether the ‘ffange-like
anterior outer margin’ has resulted from inversion or ‘seaming-up’ of the basal cavity. His line
drawing (Wang 1979, fig. 3, p. 402) appears to indicate some inversion; if this is so, his specimens
should remain assigned to P. inversus transitional to P. serotinus, but it not, P. declinatus has
priority over P. pseudoserotinus.

Polygnathus serotinus Telford, 1975

Discussion. Occurring in most instances with P. inversus, the specimens of P. serotinus recovered
from Buchan and Hindi are the ‘early’ form and match very closely with Telford’s type material
from the Dip Creek Limestone Member of the Broken River Formation, north Queensland (Telford
1975, pi. 7, figs. 5-8). Lane {in Lane and Ormiston 1979, p. 63) recognized three morphotypes of P.
serotinus: ‘alpha morph’ in which ‘the protuberance on the lower side usually is formed as a shelf-
like extension of the basal pit . . .’; ‘delta morph’ in which ‘the protuberance is formed as a shelf-
like extension of the basal pit that is supported in its entire extent by a solid shaft . . .’; and ‘gamma
morph’ with its rounded or quadrate outline and a protuberance consisting of ‘both a bulge in the
platform proper and a remnant shelf-like extension of the basal pit’. Lane and Ormiston’s (1979)
‘alpha morph’ has been described herein as P. pseudoserotinus as it belongs to a different lineage to
P. serotinus s. s., having evolved from the dehiscens abyssus stock by the ‘seaming up’, rather than
inversion, of the basal cavity.

‘gamma morphotype’

1978 Polygnathus serotinus Telford; Klapper et al., pi. 1, figs. 9 and 10 [Pa element].

1978 Polygnathus serotinus Telford; Apekina and Mashkova, pi. 77, fig. 6 {non figs. 1 and 2).

1979 Polygnathus serotinus gamma morphotype Telford; Lane and Ormiston, p. 63, pi. 8, figs. 2, 6,
13-16, 19-22, 32, 33 [Pa elements].

1980 Polygnathus serotinus Telford; Xiong, pp. 97-98, pi. 25, figs. 17-20 [Pa element].

Diagnosis. A morphotype of P. serotinus characterized by ‘a rounded or quadrate outer margin at
the point of junction of the small posterior platform and the main platform. On the lower surface,
the protuberance consists of both a bulge in the platform proper and a remnant shelf-like extension
of the basal pit’ (Lane and Ormiston 1979).

EXPLANATION OF PLATE 34

Figs. 1-7. Polygnathus dehiscens abyssus subsp. nov. 1 and 2, lower and upper views of AMF 66030, sample
BON 36-39, X 75. 3 and 4, upper and lower views of AMF 66031, sample BON 36-39, holotype, x 60. 5,
lower view of AMF 66032, sample OTRC 5, x 60. 6 and 7, specimen transitional between P. d. abyssus
and P. perbonus, upper and lower views of AMF 66034, sample OTRC 5, x 60.

Figs. 8-13. P. perbonus (Philip). 8 and 9, upper and lower views of AMF 66035, sample BON 46-50, x 60. 10
and 1 1, lower and upper views of AMF 66036, sample BCE 20, x 60. 12 and 13, upper and lower views of
AMF 66037, sample 15.2, x90.

PLATE 34

MAWSON, Polygnathus

280

PALAEONTOLOGY, VOLUME 30

Discussion. As no specimens of P. serotinus ‘gamma morph’ have been found in association with P.
serotinus ‘delta morph’ in the Buchan and Bindi region, it may be that the former is found only in
the upper part of the serotinus Zone. If this can be proved, a subdivision of the serotinus Zone may
be possible. Until further sampling is carried out, both in Victoria and in areas such as the Broken
River region, the potential for this remains uncertain. Specimens illustrated by Xiong (1980) from
south China show the typical rounded posterior margin of this morphotype. In his discussion of P.
serotinus, Xiong suggested that specimens having a rounded posterior margin should perhaps be
separated from those with a straight posterior margin; he did not, however, indicate at what level
this separation should be regarded — morphotype, subspecies, or species.

‘delta morphotype’

Plate 33, figs. 9-12; Plate 36, fig. 10

1967 Polygnathus linguiformis linguiformis Hinde; Philip and Jackson, pp. 13-14, text-fig. 2a (non
2b, 2c).

1970 Polygnathus linguiformis linguiformis Hinde; Philip and Jackson in Pedder, Jackson and Philip,
pp. 216-217, pi. 40, figs. 6 and 8 (non figs. 9 and 10).

1974 Polygnathus perbonus n. subsp. D, Klapper in Perry et al., pp. 1089 and 1091, pi. 8, figs. 9-13,
15,16.

1975 Polygnathus sp. nov. D, Klapper and Johnson, pp. 74, 75, pi. 3, figs. 1, 2, 8-10.

1975 Polygnathus foveolatus serotinus subsp. nov. Telford, pp. 43 and 44, pi. 7, figs. 5-8 (non figs.
U4).

1975 Polygnathus totensis sp. nov. Snigireva, p. 27, pi. 4, figs. 3 and 4.

1976 Polygnathus serotinus Telford; Bultynck, pi. 10, fig. 23; pi. 1 1, fig. 21.

1976 Polygnathus foveolatus Telford; Fordham, pi. 5, figs. 5-8, 15, 16, 29, 30, 34 (non figs. 9, 10, 13,
14, 31, 33).

1977 Polygnathus serotinus Telford; Weddige, pp. 319-320, pi. 4, figs. 77-79.

1977 Polygnathus serotinus Telford; Klapper in Ziegler, pp. 495-496, Polygnathus pi. 9, figs. 4 and 5.

1978 Polygnathus serotinus Telford; Pickett, pi. 1, figs. 23-25; pi. 2, fig. 18.

1978 Polygnathus serotinus Telford; Klapper et al., pi. 1, figs. 30 and 31 (non figs. 9 and 10).

1978 Polygnathus serotinus Telford; Apekina and Mashkova, pi. 76, fig. 9; pi. 77, figs. 1 and 2 (non
fig. 6).

1979 Polygnathus serotinus delta morphotype Telford; Lane and Ormiston, p. 63, pi. 8, figs. 8-10,
34, 35.

1980 Polygnathus serotinus Telford; Xiong, pp. 97-98, pi. 23, figs. 1-8; pi. 25, figs. 5-16, 23, 24, 26,
27 [Pa elements].

Amended diagnosis (modified from Klapper in Ziegler 1977 and Lane and Ormiston 1979). Pa
elements have a small pit located just anterior to the sharp inward deflection of the keel. Small,
subcircular, shelf-like protuberance, supported in its entire extent by a solid shaft from the main
platform’s lower surface, occurs on the outer side of the pit. Cavity entirely inverted posterior of
pit. Flange-like anterior outer margin is distinctly higher than carina and inner margin, and separ-
ated from carina by wide, deep, adcarinal trough.

EXPLANATION OF PLATE 35

Figs. 1-9. Polygnathus labiosus sp. nov. 1 and 2, upper and lower views of AMF 66041, sample BON 46-50,
X 60. 3 and 4, upper and lower views of AMF 66044, sample 15.2, x 60. 5 and 6, lower and upper views of
AMF 66043, holotype, sample 15.2, x 90. 7-9, specimens transitional between P. labiosus and P. pseudoser-
otinus. 1, lower view of AMF 66045, sample 15.06, x 90; 8, lower view of AMF 66046, sample 16.01, x 90;
9, lower view of AMF 66047, sample EB 5, x 90.

Figs. 10-12. P. pseudoserotinus sp. nov. 10, lower view of AMF 66048, sample 15.01, x 75. 11 and 12, lower
and upper views of AMF 66049, sample EB 3, holotype, x 90.

PLATE 35

MAWSON, Polygnathus

282

PALAEONTOLOGY, VOLUME 30

Material. Fifteen specimens from five localities (Tables 5 and 6).

Discussion. Most specimens from Buchan and Bindi, Victoria, and the Broken River, north Queens-
land, belong to Lane’s ‘delta morphotype’; a small number of specimens fitting Lane’s ‘gamma
morphotype’ definition are found in Broken River. Forms transitional between P. inversus and P.
serotinus occur with P. serotinus delta morph low in the section at Jessey Springs; they are included
with the latter because they share more characteristics diagnostic of P. serotinus than P. inversus.
On their upper surface the outer margin is somewhat higher than the inner margin and the carina,
and a deep adcarinal trough is developed between the carina and the outer margin, reminiscent of
P. serotinus. Posterior to the pit on the lower surface, however, the cavity is not completely inverted,
and the distinctive lip of P. serotinus is not fully developed.

Genus ozarkodina Branson and Mehl, 1933
Type species. Ozarkodina typica Branson and Mehl, 1933.

Discussion. The multi-element composition of Ozarkodina is as follows: Pa element is carminate
(‘spathognathodontan’); Pb element is angulate (‘ozarkodinan’); M element is dolabrate (commonly
‘neoprioniodontan’ or less commonly ‘synprioniodontan’); Sa element is alate (‘trichonodellan’X
Sb element is bipennate (‘hindeodellan’); Sc element is digyrate (‘plectospathodontan’). The only
basis for differentiating Pandorinellina and Ozarkodina is the Sa element: in Ozarkodina the posterior
process is expressed only by a slight swelling on the posterior base of the cusp (‘trichonodellan’
element) whereas in Pandorinellina the posterior process is well developed (‘diplodellan’ element)
(Klapper and Philip 1971; Klapper in Ziegler 1973). Pandorinellina n. sp. O of Klapper (1977) is
referred to Ozarkodina because there is a dearth of ‘diplodellan’ elements in the conodont faunas at
Buchan and Bindi, and, moreover, a ‘trichonodellan’ form that conforms with the symmetry
transition series is available.

Ozarkodina buchanensis (Philip, 1966)

Plate 37

1966 Spathognathodus steinhornensis buchanensis n. subsp. Philip, pp. 450-451, pi. 2, figs. 1-15 {non
figs. 16-28) [Pa element].

1966 Ozarkodina denckmanni Ziegler; Philip, pp. 446-447, pi. 4, figs. 16 and 17 (non figs. 15, 18-20)
[Pb element].

1966 Neoprioniodus bicurvatus (Branson and Mehl); Philip, p. 446, pi. 3, figs. 14-16 (non figs. 12 and
13) [M element].

EXPLANATION OF PLATE 36

Fig. 1. Polygnathus dehiscens abyssus subsp. nov., lower view of NMVP 99001, sample SLO 60, x 45.

Fig. 2. P. perbonus (Philip), lower view of NMVP 99002, sample OTRC 5, x 60.

Figs. 3 and 4. P. labiosus sp. nov. 3, lower view of NMVP 99003, sample 15.2, x 45. 4, lower view of specimen

transitional between P. labiosus and P. pseudoserotinus, NMVP 99004, sample 15.2, x 60.

Fig. 5. P. pseudoserotinus sp. nov., lower view of NMVP 99005, sample SALC 6, x 60.

Fig. 6. P. dehiscens dehiscens Philip and Jackson, lower view of NMVP 99006, sample OTRC 5, x 45.

Fig. 7. P. nothoperbonus sp. nov., lower view of NMVP 99007, sample BON 60.5-65, x 60.

Figs. 8 and 9. P. inversus Klapper and Johnson. 8, lower view of NMVP 99008, sample T3, x 90. 9, lower
view of specimen transitional between P. inversus and P. serotinus, NMVP 99009, sample EB 3, x 45.

Fig. 10. P. serotinus ‘delta morphotype’ Telford, lower view of NMVP 99010, sample T3, x 45.

Figs. 1 1-18. P. spp. Pb elements: 1 1, NMVP 9901 1, sample OTRC 1, x45; 12, NMVP 99012, sample OTRC
1, x45; 13, NMVP 99013, sample OTRC 1, x45; 14, NMVP 99014, sample SLO 60, x 45. M elements: 15,
NMVP 99015, sample Ma 10, x45; 16, NMVP 99016, sample Ma 10, x 45. Sb element: 18, NMVP 99018,
sample SB40-60, x 45. Sc element: 17, NMVP 99017, sample Ma 2, x 45.

PLATE 36

MAWSON, Polygnathus

284

PALAEONTOLOGY, VOLUME 30

1966 Trichonodella symmetrica pinnula n. subsp. Philip, pp. 452-453, pi. 4, figs. 1-6 [Sa element].

1966 Plectospalhodus alternatus Walliser; Philip, p. 448, pi. 3, figs. 10, 17, 21, 25 [Sb elements].

1966 Hindeodella priscilla Stauffer; Philip, p. 445, pi. 3, figs. 2, 7, 9 (non figs. 6, 8, 1 1, 18) [Sc elements].

1970 Spathognathodus steinhornensis optinms Moskalenko; Pedder, Jackson and Philip, p. 218, pi. 38,
figs. 4-6 (non figs. 7, 10-12) [Pa element].

1971 Ozarkodina buchanensis (Philip); Klapper and Philip, p. 448, fig. 10 [Pa, Pb, M, Sa, Sb, Sc
elements].

1973 Ozarkodina buchanensis (Philip); Klapper in Ziegler, pp. 215-216, Ozarkodina pi. 1, figs. P, O,
N, Al, A2, A3 [Pa, Pb, M, Sa, Sb, Sc elements].

1973 Spathagnathodus steinhornensis cf. buchanensis Philip; Cooper, p. 80, pi. 2, figs. 8, 9, 12; pi. 3,
figs. 2-5 [Pa element].

1979 Ozarkodina buchanensis (Philip); Lane and Ormiston, pp. 54-55, pi. 2, figs. 32 and 35; pi. 3,
fig. 13 [M, Sa, Pa elements].

Material. 385 specimens of the Pa element from sixty localities; 183 non-platform elements from fifty-four
localities (Tables 1 and 4).

Discussion. This apparatus, a Type 1 apparatus of Klapper and Philip (1972), was first illustrated
from material collected from the top of the Buchan Caves Limestone (Philip 1966). Material from
a stratigraphically similar locality, both above and below the contact of the Buchan Caves Limestone
and the Taravale Formation at Murrindal (‘P’ section; text-figs. 1 and 5), has yielded a similar
fauna with Pa, Pb, and Sc elements represented. In faunas from the basal Taravale Formation at
Slocombes (SLO section; text-fig. 1) all elements are present. The highest occurrence of O. buchanen-
sis is in bed 7.14.4 in the Gelantipy Road section at Buchan (text-figs. 1 and 4) showing that, at
Buchan at least, it does not range higher than the dehiscens Zone.

The older species O. remscheidensis remscheidensis closely resembles O. buchanensis but they can
be differentiated by the more strongly recurved lateral processes, smaller pit, and more compressed
cusp of the Sa element of the latter (Klapper in Ziegler 1975, p. 215), and by the shorter, higher
blade and the more centrally positioned basal cavity of the Pa element. O. remscheidensis repetitor
also resembles O. buchanensis, but the lack of symmetry in the arrangement of the denticles of the
latter separates the two. O. buchanensis is younger than the two subspecies of O. remscheidensis,
the highest occurrence of O. r. remscheidensis being in the delta Zone (Klapper and Johnson 1980,
p. 416) and the highest occurrence of O. r. repetitor in the pesavis Zone (Klapper and Johnson 1980,
p. 417) (text-fig. 9).

Bed-by-bed sampling of the Buchan Caves Limestone at Buchan, Bindi, and The Basin is in
progress; this should provide more precision for the lower limit of the range of O. buchanensis.
The fauna from Loyola, Victoria (Cooper 1973), incidentally, contains thirty-six specimens of O.
buchanensis occurring with Polygnathus pireneae Boermsa (identified by Cooper as Spathognathodus
trilinearis sp. nov. and P. sp., respectively). The occurrence of these forms together is indicative of
the kindlei Zone (Klapper and Johnson 1980, p. 418).

EXPLANATION OF PLATE 37

Figs. 1 -20. Ozarkodina buchanensis (Philip). Pa elements: I -3, lateral, lower, and upper views of NMVP 99019,
sample P 6.1; 4, upper view of NMVP 99020, sample P 6.1; 5 and 6, lateral and lower views of NMVP
99021, sample P 6.1; 7, lateral view of NMVP 99022, sample SLO 40-60; 8, lateral view of NMVP 99023,
sample P 7.1; 9 and 10, lateral and lower views of NMVP 99024, sample P 4. 1; 11, lateral view of NMVP
99025, sample SLO 60. Pb elements: 13, NMVP 99027, sample Ma 10; 14, NMVP 99028, sample SLO 250; 15,
NMVP 99029, sample Ma 2. M elements: 16, NMVP 99030, sample Taravale Fm. below SB 1; 17,
NMVP 99031, sample SLO 60. Sa element: 12, NMVP 99026, sample SLO 130. Sb element: 19, NMVP
99033, sample OTRC 2. Sc elements: 18, NMVP 99032, sample 8.5.2; 20, NMVP 99034, sample SLO 60.

All x60.

PLATE 37

MAWSON, Ozarkodina

286 PALAEONTOLOGY. VOLUME 30

TEXT-FIG. 9. Range chart of a selection of spathogna-
thodontan (Pa) elements occurring in the Early De-
vonian. Conodont zones are after Klapper and
Johnson (1980) with the Early Devonian-Middle De-
vonian boundary between patulus and partitus zones
as determined by the International Subcommission
on Devonian Stratigraphy in 1980 (Ziegler and
Klapper 1985). Ranges shown as a solid bar are taken
from Klapper and Johnson (1980); those with diag-
onal shading are additional information derived from
this study (Tables 1-6).

Ozarkodina excavata excavata (Branson and Mehl, 1933)

Plate 31, figs. 10-16

Synonymy. See Jeppsson (1974, pp. 19-20), Klapper {in Ziegler 1975, pp. 225-226), Klapper and Murphy
(1975, pp. 34-37), Barrick and Klapper (1976, pp. 78-79), and Klapper and Johnson (1980, p. 450).

Material. Nineteen specimens of the Pa element from seven localities; nineteen non-Pa elements from thirteen
localities (Tables 1 and 4).

Discussion. O. e. excavata occurs low in the sections at Buchan and Bindi {dehiscens and low
perbonus zones). Apart from the Sa element, all elements (Pa, Pb, M, Sb, Sc) are represented. Philip
(1966) identified ‘5.’ inclinatus inclinatus (Rhodes) [Pa element], ‘O.’ media Walliser [Pb element],
and "Trichonodella' excavata (Branson and Mehl) [Sa element] from McLarty’s ridge (Loc. 6), about
the middle of the Murrindal Limestone (text-fig. 1). He also identified ^Hindeodella' equi-dentata
Rhodes [Sc element] from the same locality and from the top of the Buchan Caves Limestone (Loc.
3), as well as 'Neoprioniodus' bicurvatus (Branson and Mehl) [M element] from his Loc. 6, higher in

EXPLANATION OF PLATE 38

Figs. 1-17. Ozarkodina linearis (Philip). Pa elements: I, lateral view of NMVP 99035, sample Shanahan Ls.,
x45; 2, lateral view of NMVP 99036, sample BON 25-27.5, x45; 3, lateral view of NMVP 99037, sample
BON 60.5-65, x 45; 4, lateral view of NMVP 99038, sample SLO 60, x 45; 5, lower view of NMVP 99039,
sample BON 36-39, x 45; 6, lateral view of NMVP 99040, sample SLO 220, x 45; 7, upper view of NMVP
99041, sample SLO 60, x 45; 8, lateral view of NMVP 99042, sample BON 60.5-65, x 60; 9 and 10, lower
and lateral views of NMVP 99043, sample BON 60.5-65, x 60. Pb elements: 11, NMVP 99044, sample SLO
60, x45; 12, NMVP 99045, sample Ma 10, x45. M element: 14, NMVP 99046, sample Ma 10, x45. Sa
element: 13, NMVP 99047, sample Ma 10, x 45. Sb element: 15, NMVP 99048, sample Forrest Motel, x 45.
Sc elements: 16, NMVP 99049, sample Ma 10, x 45; 17, NMVP 99050, sample Ma 10, x 45.

PLATE 38

MAWSON, Ozarkodina

288

PALAEONTOLOGY, VOLUME 30

the Murrindal Limestone at his Locs. 7 and 8, and from the top of the Buchan Caves Limestone at
his Locs. 2 and 3. Klapper and Johnson (1980, table 5, p. 419) concluded that O. e. excavata persists
as high as the perbonus Zone at Wee Jasper, New South Wales. Its occurrence at Buchan in the
perhonus Zone is therefore not surprising.

Ozarkodina linearis (Philip, 1966)

Plate 38

1966 Eognathodus linearis n. sp. Philip, pp. 444-445, pi. 4, figs. 33-36; text-fig. 3 [Pa element].

1966 Ozarkodina sp. cf. O.jaegeri Walliser; Philip, p. 447, pi. 4, figs. 31 and 32 [Pb element].

1966 Trichonodella inconstans Walliser; Philip, p. 451, pi. 4, figs. 21, 23, 27 {non fig. 30) [Sa element].
1966 Lonchodina n. sp. Philip, p. 446, pi. 3, fig. 19 (non figs. 20 and 24) [Sb element].

1966 Lonchodina n. sp. Philip, p. 446, pi. 3, fig. 20 (non figs. 19 and 24) [M element].

71969 Spathognathodm linearis (Philip); Flood, pi. 2, fig. 12 [Pa element].

1969 Ozarkodina typica australis Philip and Jackson; Flood, pi. 1, fig. 4 [Pb element].

1969 Lonchodina n. sp. Philip; Flood, pi. 1, fig. 8 [M element].

1969 Ligonodina salopia Rhodes; Flood, pi. 1, fig. 9 [Sc element].

1970 Spathognathodus linearis (Philip); Philip and Jackson in Pedder, Jackson and Philip, p. 217,
pi. 38, figs. 16-21 [Pa element].

1970 Ozarkodina typica australis n. subsp. Philip and Jackson in Pedder, Jackson and Philip, p. 215,
pi. 39, figs. 1-4, 11-15 [Pb element].

1970 Ligonodina salopia Rhodes; Philip and Jackson in Pedder, Jackson and Philip, p. 213, pi. 40,
figs. 1, 3, 4 [Sc element].

1970 Lonchodina greilingi Walliser; Philip and Jackson in Pedder, Jackson and Philip, pi. 37, fig. 18
(non figs. 14 and 15) [M element].

1970 Lonchodina sp. B, Philip and Jackson in Pedder, Jackson and Philip, p. 213, pi. 37, fig. 27 [Sb
element].

1970 Trichonodella inconstans Walliser; Philip and Jackson in Pedder, Jackson and Philip, p. 218,
pi. 37, figs. 17, 19, 21 [Sa element].

1971 Spathognathodus linearis (Philip); Fahraeus, p. 679, pi. 77, fig. 39 [Pa element].

Amended diagnosis. Pa elements have a subrectangular blade, slightly convex anteriorly and con-
cave posteriorly in lateral view. Denticles short and stubby; basal cavity with widely flaring sub-
symmetrical lobes.

Material. Forty-three specimens of the Pa element from twenty-two localities; sixteen non-platform elements
from thriteen localities (Tables 1-4).

Discussion. Philip (1966) recovered elements of O. linearis from the top of the Buchan Caves
Limestone at Buchan {dehiscens Zone), but the present study has shown them to occur in small
numbers high in the perbonus Zone, e.g. in bed 12.1 in the Gelantipy Road section (Table 1). The
Pa element of O. linearis can be distinguished from the Pa element of O. druceana from Cobar,
New South Wales (Pickett 1980) by being more rectangular in lateral view, rather than decreasing
dramatically posteriorly. The basal cavity of O. eurekaensis from the Roberts Mountains, Nevada,

EXPLANATION OF PLATE 39

Figs. 1-16. Ozarkodina prolata sp. nov. Pa elements: 1-3, lateral, upper, and lower views of NMVP 99051,
holotype, sample 8.8.10, x 60; 4, lateral view of NMVP 99052, sample BON 50-60, x 90; 5 and 6, lower
and lateral views of NMVP 99053, sample 8.8.10, x 60; 7 and 8, lateral and lower views of NMVP 99054,
sample Forrest Motel, x90; 9 and 10, lateral and upper views of NMVP 99055, BON 65-70, x 90; 11,
lateral view of NMVP 99056, sample BON 15-19.5, x 60. Pb element: 15, NMVP 99057, sample SB 40-60,
X 90. M element: 14, NMVP 99058, sample SB 40-60, x 90. Sb element: 16, NMVP 99059, sample SB 40-
60, X 90. Sc elements: 12, NMVP 99060, sample Taravale Fm. below SB 1, x 90; 13, NMVP 99061, sample
SB 40-60, x90.

PLATE 39

MAWSON, Ozarkodina

290

PALAEONTOLOGY, VOLUME 30

is much narrower than that of O. linearis and does not expand rapidly to maximum width midway
along the blade. The Pb, M, and symmetry transition elements of O. linearis, unlike those of O.
druceana, are conventionally assigned to the genus Ozarkodina; those assigned to O. druceana by
Pickett (1980) belong to a species of Amydrotaxis, possibly a morphotype of A. johnsoni Klapper
and Murphy.

Ozarkodina prolata sp. nov.

Plate 39

1966 Spathognathodus steinhornensis buchanensis n. subsp. Philip, pp. 450-451, pi. 2, figs. 16-21, 24-
28 {non figs. 1-15, 22, 23), text-fig. 86, 8c [Pa element].

1971 Spathognathodus optimus Moskalenko; Fahraeus, pp. 679-680, pi. 77, figs. 15-18, 23, 24, 31
{non figs. 19-21) [Pa element].

1976 Ozarkodina remscheidensis (Ziegler); Savage, p. 1182, pi. 1, figs. 1-12 {non figs. 13-15) [Pa, Pb,
M, Sa, Sb, Sc elements].

1977 Pandorinellina cf. P. optima (Moskalenko); Savage et al., p. 2934, pi. 2, figs. 11-14 [Pa element].
1980 Pandorinellina n. sp. O, Klapper and Johnson, p. 451 .

Derivation of name, prolatus (Lat.) = extended, elongated, in reference to the long, relatively narrow blade of
the Pa element.

Holotype. NMV P99051 (PI. 39, figs. 1 -3) from the Taravale Formation, 51-2 m above the base of the Gelantipy
Road section (sample 8.8.10) at Buchan, Victoria.

Diagnosis. Pa elements have a long, relatively narrow, straight or slightly curved blade with
numerous, irregular, moderately small, triangular denticles increasing in size towards the anterior.
Widely flaring, rounded lobes of basal cavity are situated medially, taper anteriorly, and continue
posteriorly as a groove.

Material. 1,128 specimens of the Pa element from 104 localities; 460 non-platform elements from seventy-four
localities (Tables 1-4).

Discussion. Klapper and Johnson (1980, p. 451) assigned this species to Pandorinellina, a genus that
differs from Ozarkodina by the Sa element being ‘diplodellan’ rather than ‘trichonodellan’ in form
(Klapper in Ziegler 1975, p. 317). The prolata apparatus is placed in Ozarkodina on two grounds:
there is no ‘diplodellan’ element that will ‘fit’ the reconstruction; and, from an evolutionary view-
point, it seems that the prolata line belongs to the ozarkodinids (see below).

The Pa element of O. prolata was first illustrated from Buchan by Philip (1966) who assigned it
to Spathognathodus eosteinhornensis buchanensis, a form recovered from samples of older strata in
the Buchan Caves Limestone. The following features of O. prolata separate the two: the longer
blade; the slightly smaller denticles and their more uniform size (except anteriorly where two
or three are clearly larger and higher); and the more centrally situated basal cavity that tapers
anteriorly. Although these features separate the two species, it is clear that O. buchanensis is the

EXPLANATION OF PLATE 40

Figs. 1-17. Pandorinellina exigua exigiia (Philip). Pa elements: 1 and 2, lateral and lower views of NMVP
99062, sample BON 13.5-15, x 60; 3 and 4, lateral and lower views of NMVP 99063, sample BON 50-60,
X 60; 5 and 6, lateral and upper views of NMVP 99064, sample Ma 4, x 60; 7 and 8, lateral and lower
views of NMVP 99065, sample BON 220-240, x 60. Pb elements: 9, NMVP 99066, sample SB 40-60, x 45;
10, NMVP 99067, sample OTRC 1, x45; 1 1, NMVP 99068, sample OTRC 1, x45. M element: 14, NMVP
99069, sample SLO 60, x 45. Sa elements: 12, NMVP 99070, sample Buchan Caves Ls. at The Basin, x 45;
13, NMVP 99071, sample Taravale Fm. below SB 1, x45. Sb elements: 15, NMVP 99072, sample SLO 85,
x45; 17, NMVP 99074, sample SLO 250, x 45. Sc element: 16, NMVP 99073, sample Ma 10, x45.

PLATE 40

MAWSON, Pandorinellina

292

PALAEONTOLOGY, VOLUME 30

precursor of O. prolata and, in Australia at least, probably replaces the remscheidensis repetitor
steinlwrnensis miae transition of Bultynck (1971). The ranges so far established for the above (text-
fig. 9) support this conclusion.

Specimens of O. prolata have been recovered from a section at La Grange, France (P. Bultynck,
pers. comm.) where they occur in the same sample (A8/10) with P. steinhornensis steinlwrnensis.
The Pa element of the older P. optima (Moskalenko) illustrated by Moskalenko (1966) has a much
wider blade and a more anteriorly located basal cavity than O. prolata. An interesting study of the
Pa element of P. exigua midundenta Wang and Ziegler has been made by Bai (1985). Forms
referred to by Bai as ‘alpha’ and ‘beta’ morphotypes of P. midwidenta may be synonymous with O.
buclianensis and O. prolata respectively but, without information regarding the Sa element of the
apparatuses, it is unwise to synonymize them.

Genus pandorinellina Muller and Muller, 1957
Type species. Pandorinellina insita Stauffer, 1 940

Pandorinellina exigua exigua (Philip, 1966)

Plate 40

1966 Spathognathodus exigua n. sp. Philip, pp. 449-450, pi. 3, figs. 26-37, text-fig. 7 [Pa element].

1966 Neoprioniodus bicurvatus (Branson and Mehl); Philip, p. 446, pi. 3, fig. 13 {non figs. 12, 14-16)
[M element].

1966 Spathognathodus frankenwaldensis Bischoff and Sannemann; Clark and Ethington, pp. 685-686,
pi. 82, figs. 15 and 21 [Pa element].

1 970 Spathognathodus steinhornensis exiguus Philip; Philip and Jackson in Pedder, Jackson and Philip,
pp. 217-218, pi. 38, figs. 7, 8, 10, 11, 13 [Pa element].

1971 Spathognathodus optimus Moskalenko; Fahraeus, pp. 679-680, pi. 77, figs. 19 and 20 {non figs.
15-18, 21, 24, 3 1 ) [Pa element].

1971 Spathognathodus exiguus Philip; Fahareus, pp. 678-679, pi. 77, figs. 25-30, 32 [Pa element].

1972 Spathognathodus n. sp. A, Uyeno in McGregor and Uyeno, p. 13, pi. 5, figs. 19-21, 30-32
[Pa element].

1972 Ozarkodina n. sp. A, Uyeno in McGregor and Uyeno, p. 13, pi. 5, figs. 4 and 5 [Pb element].

1973 Pandorinellina exigua exigua (Philip); Klapper in Ziegler, p. 319, pi. 2, fig. 2 [Pa element].

1974 Pandorinellina exigua exigua (Philip); Klapper in Perry et al., pp. 1086-1087, pi. 6, figs. 12 and 13
[Pa element].

1975 Spathognathodus exiguus Philip; Weyant, pi. 1, figs. 1-8 [Pa element].

1975 Spathognathodus exiguus Philip; Telford, pp. 58 and 60, pi. 14, figs. 10-18 [Pa element].

1978 Ozarkodina sp. Pickett, p. 100, pi. 1, figs. 8-1 1 {non figs. 12-16, 26, 27) [Pa element].

1979 Pandorinellina exigua (Philip); Lane and Ormiston, pp. 58 and 59, pi. 6, figs. 15, 20, 24 {non figs.
25 and 26) [Pa element].

EXPLANATION OF PLATE 41

Figs. 1-4. Belodella devonica (Stauffer). 1, NMVP 99091, sample Ma 7, x 60. 2, NMVP 99092, sample Ma 7,
X 60. 3, NMVP 99093, sample OTRC top 1 m, x 60. 4, NMVP 99094, sample Ma 7, x 60.

Figs. 5-8. B. resima (Philip). 5, NMVP 99095, sample Ma 7, x 60. 6, NMVP 99096, sample Ma 7, x 60. 7,
NMVP 99097, sample Ma 8, x 60. 8, NMVP 99098, sample Ma 7, x 60.

Fig. 9. B. triangularis (Stauffer), NMVP 99099, sample Ma 7, x 60.

Fig. 10. Panderodus valgus (Philip), NMVP 99100, sample Ma 7, x 60.

Fig. 1 1. Drepanodus sp., NMVP 99101, sample Ma 7, x 60.

Figs. 12 and 13. P. unicostatus (Branson and Mehl). 12, NMVP 99102, sample Ma 7, x 60. 13, NMVP 99103,
sample Ma 7, x 60.

Figs. 14 and 15. P. recurvatus (Rhodes). 14, NMVP 99104, sample Ma 7, x 60. 15, NMVP 99105, sample Ma
7, x300.

PLATE 41

MAWSON, Belode/la, Panderodus, Drepanodus

294

PALAEONTOLOGY, VOLUME 30

1979 Ozarkodina denckmatmi Ziegler; Lane and Ormiston, pi. 6, fig. 23 [Pb element].

1980 Patidorinellina exigua exigua (Philip); Uyeno and Klapper, pi. 8.1, figs. 25-27 [Pa element].

Material. 608 specimens of the Pa element from eighty-six localities; 281 non-platform elements from fifty-
one localities (Tables 1 -4).

Discussion. The Pa element of P. e. exigua first occurs at Buchan and Bindi together with Polygnathus
dehiscens. Although Pandorinellina exigua philipi does not accompany P. e. exigua in any of these
samples, transitional forms between them are found low in the dehiscens Zone; these have a very
narrowly expanded basal cavity posterior to the lobes, so narrow as to be almost a groove like that
found in P. e. philipi. Similar transitional forms from high in the section at Royal Creek, Yukon
Territory, have been illustrated by Klapper (1969, pi. 5, figs. 1-5). Towards the top of the section
at Buchan, the basal cavity of specimens of P. e. exigua tends to taper more evenly, with less
constriction at the posterior of the basal cavity lobes than in specimens recovered from lower levels.
These forms appear to be transitional between P. e. exigua and P. expansa Uyeno and Mason, a
form that shares with P. e. exigua a characteristic blade with a high anterior third somewhat offset
from the remaining, gently arched two-thirds. Pickett (1978) discussed specimens he referred to
"Ozarkodina’’ sp. from the Mount Frome Limestone, New South Wales, recovered from beds below
the first occurrence of P. expansa s. s. It appears from the shape of the basal cavity that the Pa
elements figured by Pickett (1978, pi. 1, figs. 8-11) are similarly a form transitional between P. e.
exigua and P. expansa, rather than juveniles of P. palethorpei. The Pb element of P. e. exigua has a
short, narrow posterior blade (PI. 40, fig. 10), a characteristic developed to an even greater degree
in P. expansa', this further supports the phylogenetic relationship of the two. From the Ogilvie
Formation, Northern Yukon, Perry et al. (1974) documented the presence of P. e. exigua in
faunas containing P. perhonus, with P. expansa occurring higher in the sequence accompanied by
P. serotinus.

Acknowledgements. This work owes much to encouragement and advice given by Professor John Talent. His
previous work in the area and involvement in activities of the lUGS Subcommission on Devonian Stratigraphy
highlighted the potential significance of the project. He and other members of the lUGS Subcommission,
especially Professors Gilbert Klapper and Willi Ziegler, commented on the manuscript and were generous
with their wisdom. The figures and tables were drafted by Rod Bashford and Julie Browne.

REFERENCES

ADRiCHEM BOOGAERT, H. A. VAN. 1967. Devonian and Lower Carboniferous conodonts of the Cantabrian
Mountains (Spain) and their stratigraphic application. Leid. geol. Meded. 39, 129-192, 3 pis.

AL-RAWi, D. 1977. Biostratigraphische Gliederung der Tentaculiten-Schichten des Frankenwaldes mit Con-
odonten und Tentaculiten (Unter- und Mittel-Devon; Bayern, Deutschland). Senckenberg. leth. 58, 25-79,
8 pis.

APEKiNA, L. s. and MASHKOVA, T. V. 1978. Conodonts. Pis. 73-78. In sokolov, b. s. and garkovets, v. g. (eds.).

Field Sess. Int. Subcommission Devonian Stratigraphy, Samarkand, USSR, Atlas Paleontol. Plates.

BAi, s. 1985. Morphologic analysis of a Devonian conodont species Pandorinellina midundenta. Rec. geol. 1985,
3-10, 1 pi. Res. Beijing Univ.

ning, z., jin, s. and he, j. 1980. Devonian biostratigraphy of Guangxi, with remarks on the Lower/Middle

Devonian boundary, pp. 1-14. 26th Inter. Cong. Geol., Strat. & Paleont. Sect., Dept. Geol., Beijing, Univ.
BARRiCK, J. E. and KLAPPER, G. 1976. Multielement Silurian (late Llandoverian-Wenlockian) conodonts of the
Clarita Formation, Arbuckle Mountains, Oklahoma, and phylogeny of Kockelella. Geologica Palaeont. 10,
59-100, 4 pis.

BOERSMA, K. T. 1974. Description of certain Lower Devonian platform conodonts of the Spanish central
Pyrenees. Leid. geol. Meded. 49, 285-301, 4 pis [Imprint 1973.]

BRADLEY, K. 1969. Gcology of the Murrindal River-Yalmy River area east of Buchan. Proc. R. Soc. Viet. 82,
277-285.

BRANSON, E. B. and MEHL, M. G. 1933. Conodont studies No. 1: Conodonts from the Harding Sandstone of
Colorado; Bainbridge (Silurian) of Missouri; Jefferson City (Lower Ordovician) of Missouri. Univ. Mo.
5'tnu'. 8(1), 5-72, 9pls.

MAWSON: EARLY DEVONIAN CONODONTS

295

BULTYNCK, p. 1971. Le Silurien superieur et le Devonien inferieur de la Sierra de Guadarrama (Espagne
centrale). Deuxieme partie: assemblages de conodontes a Spathognathodus. Bull Inst. r. Sci. nat. Belg.
47 (3), 43 pp., 5 pis.

1976. Le Silurien superieur et le Devonien inferieur de la Sierra de Guadarrama (Espagne centrale).

Troisieme partie: elements icriodiformes, pelekysgnathiformes et polygnathiformes. Ibid. 49 (Sci. Terre 5),
74 pp., 1 1 pis.

and HOLLARD, H. 1980. Distribution comparee de Conodontes et Goniatites Devoniens des Plaines du

Dra, de Ma’der et du Tafilalt (Maroc). Leiivense geol. Meded. 1, 73 pp., 10 pis.

CARLS, p. and gandl, j. 1969. Stratigraphie und Conodonten des Unter-Devons der Ostlichen Iberischen
Ketten (NE-Spanien). Neues Jb. Geol. Paldont. Abb. 132, 155-218, 6 pis.

CHARPENTiER, R. R. 1984. Conodonts through time and space: Studies in conodont provincialism. Spec. Pap.
geol. Soc. Am. 196, 11-32.

CHATTERTON, B. D. E. 1979. Aspects of late Early and Middle Devonian conodont biostratigraphy of western
and north-western Canada. Spec. Pap. geol. ^455. Can. 18, 161-231, 9 pis. [Imprint 1978.]

chlupaC, I. 1976. The oldest goniatite faunas and their stratigraphical significance. Lethaia, 9, 303-315.

lukeS, p. and zikmundova, j. 1977. Barrandian field trip guidebook, 1977. 23 pp. Field Conf Inti.

Subcom. Devonian Stratigraphy. [Limited distribution.]

CLARK, D. L. and ETHiNGTON, R. L. 1966. Conodonts and biostratigraphy of the Lower and Middle Devonian
of Nevada and Utah. J. Paleont. 40, 659-689, pis. 82-84.

COOPER, B. J. 1973. Lower Devonian conodonts from Loyola, Victoria. Proc. R. Soc. Viet. 86, 77-84, pis. 2
and 3.

ERBEN, H. K. 1964. Die Evolution der altesten Ammonoidea (Lieferung I). Neues Jb. Geol. Paldont. Abb. 120,
107-212, pis. 7-10.

1965. Die Evolution der altesten Ammonoidea (Lieferung II). Ibid. 122, 275-312, pis. 25-27.

FAHRAEUS, L. E. 1971. Lower Devonian conodonts from the Michelle and Prongs Creek Formations, Yukon
Territory. J. Paleont. 45, 665-683, pis. 77 and 78.

1976. Possible Early Devonian conodontophorid provinces. Palaeogeogr. Palaeoclimat. Palaeoecol. 19,

201-217.

FLETCHER, K. 1963. The Snowy River Volcanics west of Buchan, Victoria. Proc. R. Soc. Viet. 76, 169-179.

FLOOD, p. G. 1969. Lower Devonian conodonts from the Lick Hole Limestone, southern New South Wales.
J. Proc. R. Soc. N.S. W. 102, 5-9, 2 pis.

FORDHAM, B. G. 1976. Geology and Lower-Middle Devonian coral conodont biostratigraphy of the Nogoa
anticline, Springsure district, central Queensland. Proc. R. Soc. Qd, 87, 63-76, pis. 2-5.

HILL, D. 1950. Middle Devonian corals from the Buchan district, Victoria. Proc. R. Soc. Viet. 62, 137-164.

HOUSE, M. R. 1979. Biostratigraphy of the early Ammonoidea. Spec. Pap. Palaeont. 23, 263-280.

JEPPSSON, L. 1974. Aspects of Late Silurian conodonts. Fossils Strata, 6, 1-54, 12 pis.

KLAPPER, G. 1969. Lower Devonian conodont sequences. Royal Creek, Yukon Territory, and Devon Island,
Canada; with a section on Devon Island stratigraphy by A. R. Ormiston. J. Paleont. 43, 1-27, 6 pis.

1977. Lower and Middle Devonian conodont sequence in central Nevada; with contributions by D. B.

Johnson. Univ. Calif. Riverside, Campus Mus. Contr. 4, 33-54.

and JOHNSON, d. b. 1975. Sequence in conodont genus Polygnathus in Lower Devonian at Lone Mountain,

Nevada. Geologica Palaeontol. 9, 65-83, 3 pis.

and JOHNSON, g. j. 1980. Endemism and dispersal of Devonian conodonts. J. Paleont. 54, 400-455,

4 pis.

and MURPHY, M. A. 1975. Silurian-Lower Devonian conodont sequence in the Roberts Mountains

Formation of central Nevada. Univ. Calif. Pubis geol. Sci. Ill, 62 pp., 12 pis.

and PHILIP, G. M. 1971. Devonian conodont apparatuses and their vicarious skeletal elements. Letbaia, 4,

429-452.

1972. Familial classification of reconstructed Devonian conodont apparatuses. Geologica Palae-
ontol. SBl, 197-1 13, 4 pis.

and ZIEGLER, w. 1979. Devonian conodont biostratigraphy. Spec. Pap. Palaeont. 23, 199-224.

KROMMELBEiN, K. 1954. Devonische Ostracoden aus der Gegend von Buchan und von der Kiiste der Waratah
Bay, Victoria, Australien. Senckenberg. letb. 35, 193-229.

LANE, H. R. and ORMISTON, A. R. 1976. The age of the Woodchopper Limestone (Lower Devonian), Alaska.
Geologica Palaeontol. 10, 101-107, 1 pi.

1979. Siluro-Devonian biostratigraphy of the Salmontrout River area, east-central Alaska. Ibid.

13, 39-96, 12 pis.

296

PALAEONTOLOGY, VOLUME 30

LiNDSTROM, M. 1964. Cotiodonts. Elsevier Publishing Co., Amsterdam, London and New York, 196 pp.

MCGREGOR, D. c. and UYENO, T. T. 1972. Devonian spores and conodonts of Melville and Bathurst Islands,
District of Franklin. Geol. Surv. Pap. Can. 71-13, 37 pp., 5 pis.

MAWSON, R. 1986. Early Devonian (Lochkovian) conodont faunas from Windellama, New South Wales.
Geologica Palaeontol. 20, 39-71, 9 pis.

JELL, J. s. and TALENT, J. A. 1985. Stage boundaries within the Devonian: implications for application to

Australian sequences. Coiirr. Forschinst. Senckenberg, 75, 1-15.

MOSKALENKO, T. A. 1966. Pervaya nakhodka Pozdnesiluriyskikh konodontov v Zeravshanskom Khrebte.
Palaeont. Z. 1966 (2), 81-92, 1 pi.

PEDDER, A. E. H., JACKSON, J. H. and ELLENOR, D. w. 1970. An interim account of the Middle Devonian Timor
Limestone of north-eastern New South Wales. Proc. Linn. Soc. N.S.W. 94, 242-272, pis. 14-24.

and PHILIP, G. M. 1970. Lower Devonian biostratigraphy in the Wee Jasper region of New South

Wales. J. Paleont. 44, 206-251, pis. 37-50.

PERRY, D. G., KLAPPER, G. and LENZ, A. c. 1974. Age of the Ogilvie Formation (Devonian), northern Yukon:
based primarily on the occurrence of brachiopods and conodonts. Can. J. Earth Sci. 11, 1055-1097, 8 pis.

PHILIP, G. M. 1966. Lower Devonian conodonts from the Buchan Group, eastern Victoria. Micropaleontology,
12, 441-460, 4 pis.

and JACKSON, j. h. 1967. Lower Devonian subspecies of the conodont Polygnathus linguiformis Hinde

from southeastern Australia. J. Paleont. 41, 1262-1266.

PICKETT, J. 1978. Conodont faunas from the Mount Frome Limestone (Emsian/Eifelian), New South Wales.
Bull. Bur. Miner. Resour. Geol. Geophys. Aust. 192, 97-107, 2 pis.

1980. Conodont assemblages from the Cobar Supergroup (Early Devonian), New South Wales. Alcher-

inga, 4, 67-88.

SAVAGE, N. M. 1976. Lower Devonian (Gedinnian) conodonts from the Grouse Creek area, Klamath Moun-
tains, northern California. J. Paleont. 50, 1180-1 190, 3 pis.

1977. Lower Devonian conodonts from the Gazelle Formation, Klamath Mountains, northern Califor-
nia. Ibid. 51, 57-62, 1 pi.

CHURKIN, M. and eberlein, g. d. 1977. Lower Devonian conodonts from Port St. Nicholas, southeastern

Alaska. Can. J. Earth Sci. 14, 2928-2936, 2 pis.

snigireva, m. p. 1975. Novye Konodonty iz srednedevonskikh otlozheniy severnogo Urala. Palaeont. Z. 1975
(4), 24-30, 4 pis. [In Russian.]

TALENT, J. A. 1956. Devonian brachiopods and pelecypods from the Buchan Caves Limestone, Victoria. Proc.
R. Soc. Viet. 68, 1-56, 5 pis.

1965. The stratigraphic and diastrophic evolution of central and eastern Victoria in Middle Palaeozoic

times. Ibid. 79, 179-195.

1967. Eastern Victoria. Australia-New Zealand Assoc. Adv. Sci., Excursion Handb., Sec. C, 69-85.

1969. The geology of east Gippsland. Proc. R. Soc. Viet. 82, 37-60.

TEiCHERT, c. 1948. Middle Devonian goniatites from the Buchan district, Victoria. J. Paleont. 22, 60-67,
pi. 16.

and TALENT, J. A. 1958. Geology of the Buchan area. East Gippsland. Mem. geol. Surv. Viet. 21, 56 pp.

TELFORD, p. G. 1972. Conodonts. Pp. 83-88. In talent, j. a. et al. Provincialism and Australian Early Devonian
faunas. J. geol. Soc. Aust. 19, 81-97.

1975. Lower and Middle Devonian conodonts from the Broken River Embayment, north Queensland,

Australia. Spec. Pap. Palaeont. 15, 96 pp., 16 pis.

1979. Devonian conodont distribution— provinces or communities? Pp. 201-213. In gray, j. and boucot,

A. J. (eds.). Historical biogeography, plate tectonics and the changing environment. Proc. 37th Ann. Bio.
Coll., Oregon State Univ. Press, 500 pp.

UYENO, T. T. and klapper, g. 1980. Summary of conodont biostratigraphy of the Blue Fiord and Bird
Fiord Formations (Lower-Middle Devonian) at type and adjacent areas, south-western Ellesmere Island,
Canadian Arctic Archipelago. Geol. Surv. Pap. Can. 80-lC, 81-93, 3 pis.

WANG, c. 1979. Some conodonts from Sipai Formation in Xiangzhou of Guangsu. Acta palaeont. sin. 18,
395-408, 1 pi.

and WANG, z. 1978. Early and Middle Devonian conodonts of Kwangsi and Yunnan. In Inst. Geol. Min.

Res., Chinese Acad. Geol. Sci. (eds.). Symposium on the Devonian System of South China, 1974.
396 pp., 53 pis. Geological Press, Peking.

and ZIEGLER, w. 1983. Devonian conodont biostratigraphy of Guangxi, South China, and the correlation

with Europe. Geologica Palaeontol. 17, 75-107, 8 pis.

MAWSON: EARLY DEVONIAN CONODONTS

297

WEDDIGE, K. 1977. Die Conodonten der Eifel-Stufe im Typusgebiet und in benachbarten Faziesgebieten.
Senckenberg. leth. 58, 271-419, 6 pis.

and ZIEGLER, w. 1977. Correlation of Lower/Middle Devonian boundary beds. Newsl. Stratigr. 6,

67-84.

1979. Evolutionary patterns in Middle Devonian conodont genera Polygnathus and Icriodus. Geo-

logica Palaeontol. 13, 157-164.

WEYANT, M. 1975. Sur I’age du membre inferieur de la Formation Blue Fiord dans le Sud-Ouest de File
Ellesmere (Archipel. Artique Canadien) d’apres les Conodontes. Newsl. Stratigr. 4, 87-95, 1 pi.

XIAN, s., WANG, s., ZHOU, X., XIONG, J. and ZHOU, T. (eds.). 1980. Nandan typical stratigraphy and palaeontology
of the Devonian in South China, 161 pp., 48 pis. Guizhou People’s Public. Press.

XIONG, J. 1980. Conodonts. Pp. 82-100. In xian, s. et al. (eds). Ibid.

ZIEGLER, w. (ed.). 1973. Catalogue of conodonts, /, 504 pp., 27 pis. E. Schweizerbart’sche Verlagsbuchhandlung,
Stuttgart.

(ed.). 1975. Ibid. II, 404 pp., 25 pis.

(ed.). 1977. Ibid. Ill, 574 pp., 39 pis.

and KLAPPER, G. 1985. Stages of the Devonian. Episodes, 8 (2), 104-109.

RUTH MAWSON

School of Earth Sciences
Macquarie University

T ypescnpt received 2 J anuary 1 986 New South Wales 2 1 09

Revised typescript received 26 June 1986 Australia

THE OLDEST AMMONOID ‘COLOUR’ PATTERNS:
DESCRIPTION, COMPARISON WITH NAUTILUS,

AND IMPLICATIONS

by ROYAL H. MAPES and debra a. sneck

Abstract. Twelve Lower Triassic ammonoid specimens that retain four different ‘colour’ patterns are de-
scribed. These ‘colour’ patterns are the oldest known for ammonoids; three genera (Dieneroceras, Prosphin-
gites, and Owenites) are represented. Four factors contribute to the conclusion that these ‘colour’ patterns
were deposited at the time of growth: (1) the transverse bands are bilaterally symmetrical; (2) the coloration is
confined to the outer layer of the test; (3) the pattern is disrupted by sublethal damage to the conch; and (4)
the ‘colour’ is observable through the Rimzelschicht and is present under the dorsal shell. The ammonoid
‘colour’ patterns differ from modern Nautilus by being present through the terminal growth stage and by
being concordant with the growth lines; Nautilus loses the colour banding at maturity and has a discordant
relationship of colour patterns with growth lines. The transverse ‘colour’ patterns on ammonoids appear to
be less sophisticated than those observed in Nautilus', this suggests that the colour patterns on ammonoids
may have served a different primary function than that of camouflage.

‘Colour’ pattern preservation on the conchs of fossil cephalopods is a rare phenomenon. Even
when the shell survives the processes of fossilization in excellent condition, including the retention
of aragonite, the original coloration is never preserved. Therefore, detection of ‘colour’ patterns on
fossil cephalopods is dependent on the recognition of consistent non-random patterns of pigmen-
tation expressed in shades of brown or grey that are interpreted to reflect the original colour pattern
(Teichert 1964). Reports of such patterns on ammonoids are rare.

‘Colour’ patterns on nautiloids (Gordon 1964; Teichert 1964; Windle 1973) and bactritoids
(Mapes 1979) are known from localities that have co-oceurring ammonoids; however, there are
eurrently no Palaeozoic occurrences of ammonoid ‘colour’ patterns reported. Mesozoic ammonoid
examples are limited to twelve reports from widely distributed localities in Europe, North America,
and Japan. Cretaceous ammonoid ‘colour’ patterns include Tetragonites (Tanabe and Kanie 1979),
Protexanites (Matsumoto and Hirano 1976; Tanabe and Kanie 1979), and Paratexanites (Matsu-
moto and Hirano 1976). Jurassic occurrences include Amaltheus (Wright 1881; Schindewolf 1928;
Spath 1935; Arkell 1957; Pinna 1972), Androgynoceras (Spath 1935; Arkell 1957), Arietites (Greppin
1898), Asteroceras (Arkell 1957; Manley 1977), Leioceras (Greppin 1898; Arkell 1957), Pleuroceras
(Arkell 1957; Pinna 1972), and Tragophylloceras (Arkell 1957). The oldest reported occurrence of
an ammonoid ‘colour’ pattern is on a Lower Triassie specimen of Owenites koeneni from Nevada
(Tozer 1972).

Recent examination of a collection of ammonoids from the same Nevada locality that yielded
the specimen described by Tozer, revealed additional specimens of Owenites with two distinctly
different types of ‘colour’ patterns. Additionally, distinctly different patterns of relict pigmentation
were detected on specimens of Prosphingites and Dieneroceras.

In this paper we analyse these oldest-known fossil ‘colour’ patterns on ammonoids and eompare
them with the true colour patterns on modern Nautilus.

LOCALITY AND REPOSITORY

Aeeording to Kummel and Steele (1962), the outcrop which yielded this collection of ammonoids is
in the Meekoceras gracilitus Zone (Lower Triassic), which is exposed in the SWl/4, SWl/4, sec. 34

(Palaeontology, Vol. 30, Part 2, 1987, pp. 299-309, pi. 42.|

© The Palaeontological Association

300

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1. Locality map showing the location of the Crittenden Springs Triassic ammonoid locality in

Nevada.

and the Wl/2, SWl/4, sec. 3; T.42N., R.69E. (Dairy Valley 15' Quadrangle). This locality
is approximately 1-2 km (10 mile) north-east of Crittenden Springs, Elko County, Nevada
(text-fig. 1).

Ammonoids are extremely common and typically well preserved at the Crittenden Springs
locality. Of the specimens collected, approximately one in 500 retains remnant ‘colour’ patterns
(E. Noble, pers. comm.).

All of the specimens described in this report are reposited at the Department of Geology,
University of Iowa, Iowa City, Iowa.

MAPES AND SNECK: AMMONOID COLOUR’ PATTERNS

301

MATERIAL AND DESCRIPTION

‘Colour’ patterns were observed on twelve specimens. Most specimens retain part of the body
chamber; but all save one are missing the aperture. Some have an incomplete test due to mechanical
exfoliation produced by cracking out the specimens from the enclosing limestone. Other specimens
have weathered surfaces; where this occurs, the alteration appears to have obliterated any trace of
remnant ‘colour’ patterns on the conch.

On the Triassic ammonoid with the ‘colour’ pattern described by Tozer (1972), pigmentation is
found in the ‘outer layer of test’, and as he pointed out, the ‘outer test’ also contains growth lines
and ornament. These three morphological features are located in the porcelaneous ostracum of
Nautilus that is secreted by the apertural edge of the mantle. Examination of natural breaks on the
twelve newly discovered specimens show that the ‘colour’ pattern is also confined to the outer layer
of the test.

TABLE 1. Measurements (in millimetres) of ‘colour’ pattern bearing specimens of Dieneroceras, Prosphingites,
and Oweuites from Lower Triassic sediments at Crittenden Springs, Nevada. An asterisk indicates an estimated

measurement.

Specimen

Diameter

Umax

Height

Width

Dieneroceras spathi

SUI 49171

27-9

9-3*

11-2

5-6*

Prosphingites slossi

SUI 52308

24-7

8-6

100

150

SUI 52316

19-6

7-7

905

11-5

SUI 52317

181

80

7-8

121

SUI 52318

16-8

5-6

71

10-8

Oweuites cf. koeneni

SUI 52309

340*

6-5

13-8

10-9

SUI 52310

19-6

4-2

8-7

91

SUI 52311

25-3

61

120

10-5

SUI 52312

37-8*

12-6*

18-4

151*

SUI 52313

35-5

7-7

14-5

12-9

Owenites sp.

SUI 52314

370

9-9

17-8

12-9

SUI 52315

340

6-3

161

12-2

The ‘colour’ pattern on Prosphingites slossi consists of continuous, radial bands which extend
symmetrically from the umbilical seam across the venter to the opposite umbilical seam (PI. 42,
figs. 1-3). This pattern is present on four specimens (SUI 52309, 52316-52318) (Table 1; PI. 42,
figs. 1-5). On specimen SUI 52308 the venter is enclosed in matrix, and approximately half of the
outer whorl can be interpreted as missing because remnants of test are present on the umbilical
seam on the preceding whorl. The ‘colour’ pattern conforms to shallow constrictions in the test
which have an average spacing of 1-3 mm. Rimzelschicht is present on part of the conch, and the
pigmentation is faintly discernible under this shell layer. The remaining three specimens have a
similar but fainter colour pattern.

The ‘colour’ pattern on Dieneroceras spat hi (SUI 49171) consists of a dark greyish-brown longi-
tudinal band on the venter and a dark-grey longitudinal band at the umbilical seam. Faint medium-
grey transverse bands are present on the ribs located on the umbilical and lateral region of the
conch (PI. 42, fig. 4). These transverse bands connect with the longitudinal bands at the umbilical
seam and on the venter. Only one side of the specimen is exposed; the other side is covered with
matrix. Because of this, the bilateral symmetry of the pattern cannot be confirmed (Table 1).

302

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 2. Disruption of ‘colour’ pattern and Runzelschicht on Owenites cf. koeneni. a, SUI 52313
showing a massive repaired break that interrupts and somewhat offsets the transverse ‘colour’ bands
on the lateral part of the conch (see arrow), x 3-2. b, SUI 5231 1 lightly coated with ammonium chloride
to enhance the Runzelschicht layer, x 4-6; the colour pattern on this specimen, which can be observed
faintly through the Runzelschicht, is more clearly shown on Plate 42, figs. 10 and 1 1.

Part of the body chamber is missing on this specimen. In this area, dorsal shell is exposed, and
where this shell layer is missing on the venter, the greyish-brown longitudinal band is present.

Two different patterns are present on Owenites. The most frequently occurring pattern in the
collection is that on Owenites cf. koeneni (SUI 52309-52313) where the shading consists of simple
transverse dark-grey bands that cross the venter and disappear on the umbilical lateral part of the
conch (PI. 42, figs. 8-15; text-fig. 2a, b). The ‘colour’ bands coincide with shallow constrictions and

EXPLANATION OF PLATE 42

Figs. 1-15. ‘Colour’ patterns exhibited on Prosphingites slossi, Dieneroceras spathi, and Owenites. All are
photographed under xylene to enhance the colour pattern. 1-3, P. slossi (SUI 52308), left, ventral, and right
sides respectively showing transverse ‘colour’ bands, x 1-2. 4, D. spathi (SUI 49171), left side showing
longitudinal bands on the venter and umbilical areas and faint transverse bands on ribs, x 1-5. 5 and 6,
Owenites sp. (SUI 52314) left and right lateral views showing longitudinal lateral band and transverse bands
that cross the venter, x 1 -4. 7, Owenites sp. (SUI 52315) left lateral view showing longitudinal and transverse
bands, x 1-5. 8-15, Owenites cf. koeneni (SUI 52309-523011, 52313) showing transverse ‘eolour’ bands,
magnifications are x 1-6, x 1-3, x 1-2, and x 1-5, respectively.

PLATE 42

12

13

14

15

MAPES and SNECK, ammonoid colour patterns

304

PALAEONTOLOGY, VOLUME 30

are concordant with the growth lines. Two of these specimens (SUI 52309, 52312) are essentially
complete and show indications of mature modifications. A third specimen (SUI 52313) may also be
a mature specimen based on ‘colour’ pattern and constriction approximation.

The other ‘colour’ pattern is on specimens of Owenites sp. (SUI 52314, 52315) (PI. 42, figs. 5-7).
The umbilical area has a brownish-grey tone that abruptly changes to a broad dark-grey band in
the mid-lateral region of the conch. Radiating ventrally from this dark-grey band are dark-grey
transverse stripes that cross the venter. These transverse stripes are confined to shallow constrictions
and follow the growth lines (Table 1). The pattern is best preserved on specimen SUI 52315; the
bilateral symmetry of the pattern can be observed best on specimen SUI 52314.

Because the specimens assigned to Owenites display two distinctly different ‘colour’ patterns,
assignment at the species level is made with reservation. Additional studies of conch morphology
and suture patterns are necessary to determine whether the two ‘colour’ patterns are species specific
in Owenites, sexual dimorphism within O. koeneni, or normal variation within a species of Owenites.
At this time, we are reluctant to try to resolve this problem because exposing the suture pattern on
the specimens would destroy the preserved ‘colour’ pattern.

DISCUSSION

The colour patterns on the various species of modern Nautilus have been illustrated and generally
described by numerous investigators (Willey 1902; Stenzel 1964; Cowen et al. 1973). At one time,
these were considered to be a species level character (see Saunders 1981 for summary). However,
more recent detailed observations by Ward et al. (1977) indicate that several colour polymorphs of
N. pompilius are present and these colour pattern variations have no correlation to sex or depth
and cannot be used by themselves to determine species in Nautilus.

As Cowen et al. (1973) indicate, there is little or no correspondence between the bands of reddish-
brown pigment and the shell growth lines. The growth lines on mature Nautilus outline a pronounced
ocular sinus on each side of the conch and a modest hyponomic sinus at the venter. Earlier stages
of the ontogeny show the growth line configuration to be similar but not as pronounced. The colour
patterns of reddish-brown bands also generally form a sinus on the venter, but on the lateral parts
of the conch, the bands commonly cross the growth lines, coalesce, diverge and/or die out. Often in
the umbilical region, the conch is a reddish-brown colour; sometimes stripes penetrate into this
region giving darker bands of pigmentation on the reddish-brown umbilical area. In all cases, when
Nautilus nears maturity, the reddish-brown bands fade on the venter and ventrolateral portions of
the conch. Thus, this part of the mature conch is a white to cream colour. Notably many specimens
retain reddish-brown stripes in the umbilical region during this stage of ontogeny. Some specimens
retain these umbilical stripes to the mature aperture; in others these stripes die out, and the terminal
part of the conch is without reddish bands but may retain the less pronounced reddish-brown
umbilical shading.

The fossil ammonoids from Nevada have colour patterns with differences from those observed
in Nautilus. Noteworthy in the ammonoids is the coincidence between the growth lines and pigmen-
tation. The growth lines of the Triassic ammonoid specimens are curved adorally outlining a shallow
ocular sinus; the radial ‘colour’ bands always conform to this configuration. Also, when ‘colour’ is
present, the growth lines are generally coarser.

In Nautilus, coloration is due to the presence of melanin, a pigment-causing protein. The chemistry
of the ‘colour’ in ammonoids has not been established. Emplacement of metabolic waste material
in Nautilus has been postulated by Pruvot-Fol (1935); this possibility cannot be ruled out for the
Triassic ammonoids.

COLOUR PATTERNS AND MATURE MODIFICATIONS

Many of the mature modifications of Nautilus were described by Willey (1902). A more recent
summary and some additional features were provided by Davis (1972) and Collins et al. (1978).
Some of the external mature modifications that occur in the last stages of growth in Nautilus include

MAPES AND SNECK: AMMONOID ‘COLOUR’ PATTERNS

305

the cessation of secretion of colour patterns on the ventral and ventrolateral portions of the
conch, deepening of the ocular sinuses, a change in coiling rate (i.e. body-chamber contraction and
straightening), and expansion of the aperture (males only; this can also be considered a sexually
dimorphic character).

Mature modifications have been recognized in fossil ammonoids for more than 100 years; perhaps
the most complete list dealing with the morphological changes that occur at or near maturity is
provided by Davis (1972). His list of eleven characteristics includes those known to occur in both
the fossil cephalopods and Recent Nautilus.

The external character that most strongly suggests specimens SUI 52309, 52312, and 52315 were
mature or nearing maturity is the change of coiling. Other external characteristics such as apertural
expansion and deepening of the ocular sinuses in Nautilus cannot be used with confidence in these
fossil forms. The development of constrictions can only be used on specimen SUI 52312 which has
a partly intact aperture. Lappets cannot be used since the terminal ends of the body chambers are
not present on any except specimen SUI 52312 which does not develop this morphological feature.

The ‘colour’ patterns expressed on the fossil ammonoids are somewhat different from those
patterns that are present on Nautilus. In the fossil forms the change in coiling suggests the approach
of sexual maturity and cessation of shell secretion. However, unlike Nautilus, which during terminal
growth and/or the approach of sexual maturity, ceases to produce colour patterns on the venter
and ventrolateral parts of the conch, the Triassic ammonoids maintain the ‘colour’ banding pattern
to the aperture. In the vicinity of the aperture the distance between the ‘colour’ bands generally
decreases and they become approximated. Although this ‘colour’ banding approximation cannot
be proved to indicate sexual maturity, it is important that all known specimens which exhibit other
mature modifications also had approximation of the ultimate constrictions and also of the ‘colour’
bands.

BREAKAGE AND REPAIR ON COLOUR PATTERNS

Sublethal breakage and subsequent shell regeneration in Nautilus is a phenomenon that has only
recently received study even though it was originally noted (but not analysed) by Willey (1902, fig.
1 5). In 1974 Nautilus shell regeneration was extensively analysed using scanning electron microscopy
by Meenakshi and others. Bond and Saunders (1984) and Bond (1984) discussed sublethal predation
in both fossil ammonoids and Nautilus. In the latter study, Bond used Nautilus as a generalized
model and focused on the interruption of growth lines to determine sublethal events. However, the
only study to mention injury and associated disruption of colour patterns on Nautilus was by
Arnold (1 985, p. 388).

A brief analysis of about fifty available specimens (mostly juvenile) of Nautilus suggests that four
types of sublethal damage will alter the colour patterns on repaired shell. The actual reddish-brown
colour on the conch is not changed by sublethal damage; in all four types, the change is a product
of interruption and/or regeneration of missing shell in the body chamber region of the conch.

In the terminology developed by Bond (1984), the four major types of breaks that interrupt the
colour patterns on Nautilus are as follows: (1) Massive— removal of relatively large pieces of test;
damage may extend from umbilicus to umbilicus; the breaks include V’s, crescents, and scallops.
Colour interruptions include sudden termination of reddish-brown colour at the break followed by
post-predation shell that typically is white or cream colour. In some cases nacreous shell without
colour is deposited (or at least not repaired) at the break. Mantle damage may be associated with
this type of break; repairs with this type of damage are described under the third type of break (see
below). (2) Moderate— damage is less severe than massive but breaks are still in V’s, crescents, and
scallops; breaks may extend laterally from venter to umbilicus. Colour pattern interruption is the
same as previously described under massive breaks; however, the interruptions are usually less
pronounced, and many breaks show no appreciable colour interruption. (3) Narrow-piercing— the
breaks, when present, were narrow and deep and affected the mantle; lateral test damage is limited
in extent. Colour patterns are sometimes profoundly interrupted and the interruption can continue

306

PALAEONTOLOGY, VOLUME 30

to terminal growth stages (more than one whorl). Initial post breakage shell deposition is either
black organic material or white coloured shell that forms a narrow band. On many specimens the
white coloured shell gradually narrows and becomes a lighter reddish-brown shade than the typical
transverse colour bands. In some cases the transverse colour bands secreted after the predatory
event are offset by this type of trauma. (4) Minor— damage is less pronounced than moderate with
only several growth lamellae missing and the breakage extends for only short distances on the
conch; breaks can form small scallops, V’s or be irregular. Colour pattern interruptions are minor
and only take place at the edges of established reddish-brown bands. Frequently, because of the
somewhat irregular nature of the reddish-brown patterns laid down on Nautilus, the interruption is
difficult to detect even though the minor breakage is clearly expressed by the growth-line interrup-
tion.

In the Triassic ammonoids from Nevada, four specimens (SUI 52310, 52312, 52313, 52317) show
evidence of sublethal breakage with evidence of colour pattern disruption; several of the remaining
specimens show sublethal predation but colour pattern disruption is not evident. The disruptions
in ‘colour’ patterns on the Triassic specimens fall into three of the four colour interruption categories
observed in Nautilus. The only category not represented is the narrow-piercing break.

On specimen SUI 52310, the sublethal damage is confined to a single place on the venter of the
conch and is interpreted as moderate. The colour pattern in the vicinity of the break is a typical
transverse band that normally would continue across the venter of the conch (text-fig. 2a). However,
the breakage spans the width of the ‘colour’ band, and the colour band terminates abruptly at the
break on both sides of the venter. The shell in the repaired break is a distinctly lighter colour tone
that is similar to the grey ‘colour’ found between the darker ‘colour’ bands.

Two minor sublethal events interrupted normal ‘colour’ pattern development on specimen SUI
52312. One break is a shallow indentation of the test that involves several growth lines on the lateral
position of the conch. The ‘colour’ pattern interruption is a decrease in width of the dark transverse
band at the position of the break (text-fig. 3b). The adjacent dark bands apicad and orad of the
break do not show this decrease in width. The other ‘colour’ pattern interruption is located more
orad on the conch and is a V-shaped notch on the venter. The notch occurs at the orad edge of the
dark colour band, and at this place the dark band has been removed. The post-trauma shell repair
is of the lighter-grey tone typical of that observed between the darker ‘colour’ bands.

The third specimen (SUI 52313) has a massive sublethal break that extends across the venter to
the umbilicus. ‘Colour’ pattern interruption is restricted to the lateral part of the conch where two
dark transverse bands are offset (text-fig. 2c).

The fourth specimen (SUI 52317) also has massive sublethal damage. The ‘colour’ pattern
interruption is minimal since most of the breakage is located between two constrictions that are a
uniform shade of grey.

THE RELATIONSHIP OF DORSAL SHELL AND COLOUR PATTERNS

In Nautilus, the dorsal shell is composed of the black layer, which is made up of organic material,
and a nacreous aragonite layer that covers the black layer and is usually confined to the posterior
end of the body chamber. These two shell layers are known to completely cover the colour patterns
of the preceding whorl of the conch. Also, these layers are considered to be homologous to the
wrinkle layer (or Runzelschicht) found in Paleozoic and Mesozoic ammonoids (Stenzel 1964; House
1971;Tozer 1972).

Ten of the Triassic ammonoids from Nevada (SUI 52308-52311, 52314-52318, 49171) are suffi-
ciently broken back (i.e. missing part or all of the body chamber) to expose the wrinkle layer. Of
these, five (SUI 52308, 52309, 52314, 52315, 52318) do not have sufficient exposure or quality of
preservation to allow evaluation of the underlying ‘colour’ pattern and wrinkle layer relationship,
and one (SUI 52310) is a phragmocone with only a small patch of wrinkle layer preserved— this
specimen has a somewhat frosted appearance and the ‘colour’ pattern is subdued. On specimen
SUI 49171, no wrinkle layer is exposed although dorsal shell is present. This shell layer is covered

MAPES AND SNECK: AMMONOID ‘COLOUR’ PATTERNS

TEXT-FIG. 3. Transverse ‘colour’ patterns on Owenites cf. koeneni (SUI 52312). a, Minor sublethal damage
on the lateral part of the conch involving several growth lines that causes the ‘colour’ band to thin (see
arrow), x3T. b, c, Right and left sides respectively showing the bilateral symmetry of the transverse

‘colour’ bands, x T5.

by a dark organic-looking material that may be equivalent to the black layer in Nautilus. Where
the dorsal shell has been chipped away, the dark ‘colour’ band on the venter is preserved. Of the
remaining three specimens, one (SUI 5231 1) is about one-third body chamber which lacks the test
and two-thirds phragmocone which is covered with well-developed wrinkle layer (text-fig. 2b).
Despite presence of the wrinkle layer, the ‘colour’ pattern that underlies the wrinkle layer is relatively
clearly defined (PI. 42, figs. 10 and 1 1). Specimen SUI 52317 is a phragmocone that is also essentially
completely covered with wrinkle layer. As with specimen SUI 52310, the ‘colour’ pattern is clearly
exhibited, although the pattern is less pronounced than specimen SUI 52308 which has essentially
no exposure of this extra layer of test. Specimen SUI 52316 has a well-exposed wrinkle layer in
which the ‘colour’ pattern does not show through. On this specimen about one-fifth of a whorl has
wrinkle layer with two faint constrictions.

308

PALAEONTOLOGY, VOLUME 30

CONCLUSIONS AND SPECULATIONS

Crittenden Springs is an important locality because exceptional conditions of preservation have
made possible the documentation of a variety of the oldest-known ammonoid ‘colour’ patterns. Prior
comparisons of fossilized cephalopod coloration with that of Nautilus may have caused transverse
bands conforming to growth lines to be discounted as a relict colour pattern and attributed to a
phenomenon of preservation, possibly related to shell density and matrix lithology. Based on these
Triassic specimens, we are convinced the colour present is primary, being incorporated at the time of
growth. Four factors indicate this: (1) the transverse bands are bilaterally symmetrical; (2) the color-
ation is confined to the outer layer of test; (3) the pattern is disrupted by sublethal damage to the
conch; and (4) the colour is observable through Runzelschicht and/or dorsal shell.

Based on examination of the four ‘colour’ patterns described herein, differences in pigmentation
between Triassic ammonoids and modern Nautilus become apparent. While the ammonoid speci-
mens retain an approximately regular pattern through maturity, adult Nautilus loses pigmentation
at a genetically predetermined time of growth so that the terminal ventral surface of the conch is
white. Also, the transverse ‘colour’ banding of the ammonoids conforms to growth lines and
constrictions of the shell, whereas in Nautilus, the pattern and growth lines are discordant.

These two factors may indicate a profound difference in function of the colour patterns. For
Nautilus to construct such an arrangement of colour, secretion of pigmentation must occur at
different positions on the aperture with growth. Continuity of colour only exists as a function of a
specific preprogramming which also controls the cessation of colour banding with the onset of
maturity. The result is that the adult Nautilus displays irregular, disruptive coloration when viewed
from above or laterally and is without pigment when seen from below. Thus, the animal can be
considered as camouflaged in its environment (Cowen et al. 1973). The function of ammonoid
‘colour’ patterns and the method of emplacement are not known. Conformity between coloration,
constrictions, and growth lines and their presence throughout life may indicate that either this
combination of morphological characteristics was not as sophisticated as in Nautilus, or the pigmen-
tation does not have a specific purpose such as camouflage. Previously, when ‘colour’ patterns have
been described on fossilized cephalopods, the patterns have been compared to the pigmentation of
Nautilus. Although ammonoids and Nautilus have biologic similarities, it may not be appropriate
to try to force interpretations of ammonoid palaeobiology based on comparisons with living
Nautilus.

Acknowledgements. We wish to give our sincere thanks to Mr Ed Noble, El Cajon, California, and Mr Jim
Jenks of Salt Lake City, Utah, for providing the specimens on which this report is based. These specimens
were graciously and generously donated for research. T. P. P. Parsons is thanked for her aid in the photographic
processing of some of these specimens. The Cartographic Center at Ohio University is acknowledged for its
help in preparing the locality map. Additionally, we wish to thank the anonymous reviewer for his helpful
comments and suggestions. Acknowledgement is made to the donors of the Petroleum Research Fund,
administered by the American Chemical Society, for support of this research (PRF No. 15821-AC2).

REFERENCES

ARKELL, w. j. 1957. Introduction to Mesozoic Ammonoidea. In moore, r. c. (ed.). Treatise on invertebrate
paleontology. Part L, Mollusca, L81-L129, Geological Society of America and University of Kansas Press,
Boulder, Colorado and Lawrence, Kansas.

ARNOLD, j. M. 1985. Shell growth, trauma and repair as an indicator of life history for Nautilus. Veliger, 27,
386-396.

BOND, p. N. 1984. Sublethal predation of upper Mississippian (Chesterian) ammonoids. M.Sc. thesis (unpub-
lished), Bryn Mawr College.

BOND, p. N. and SAUNDERS, w. B. 1984. Evidence of predation in Mississippian ammonoids. Abstr. Progms,
geol. Soc. Amer. 16, 126.

COLLINS, D., WESTERMANN, G. E. G. and WARD, p. D. 1978. The mature Nautilus: its shell and buoyancy. Ibid.
10, 382.

MAPES AND SNECK: AMMONOID ‘COLOUR’ PATTERNS

309

COWEN, R., GERTMAN, R. and wiGGETT, G. 1973. Camouflage patterns in Nautilus, and their implications for
cephalopod paleobiology. Lethaia, 6, 201-213.

DAVIS, R. A. 1972. Mature modification and dimorphism in selected Late Paleozoic ammonoids. Bull. Am.
Paleont. 62, 130 pp.

GORDON, M. 1964 [1965]. Carboniferous cephalopods of Arkansas. Prof. Pap. U.S. geol. Surv. 460, 322 pp.
GREPPiN, E. 1898. Description des fossiles du Bajocien superieur des environs de Bale. Mem. Soc. paleont.
Suisse, 25, 1-52.

HOUSE, M. R. 1971. The goniatite wrinkle-layer. Smithson. Contrib. Paleohiol. 3, 23-32.

KUMMEL, B. and STEELE, G. 1962. Ammonites from the Meekoceras gracilitus Zone at Crittenden Spring, Elko
County, Nevada. J. Paleont. 36, 638-703.

MANLEY, c. E. 1977. Unusual pattern preservation in a Liassic ammonite from Dorset. Palaeontology, 20, 913-
916.

MAPES, R. H. 1979. Carboniferous and Permian Bactritoidea (Cephalopoda) in North America. Paleont. Contr.
Univ. Kans. 64, 75 pp.

MATSUMOTO, T. and HiRANO, H. 1976. Colour patterns in some Cretaceous ammonites from Hokkaido. Trans.
Proc. palaeont. Soc. Japan, (n.s.) 102, 334-342.

MEENAKSHI, V. R., MARTIN, A. w. and WILBUR, K. M. 1974. Shell repair in Nautilus macromphalus. Marine Biol.
27, 27-35.

PINNA, G. 1972. Prezenza di trace di coloure sul guiscio di alcune ammoniti della famiglia Amalthidea Hyatt
1877. Atti. Soc. ital. Sci. nat. 113, 193-200.

PRUVOT-FOL, A. 1935. Remarques sur le Nautile. Int. Congr. Zool. 3, 1652-1663. Twelfth session, Lisboa.
SAUNDERS, w. B. 1981. The species of living Nautilus and their distribution. Veliger, 24, 8-17.

SCHINDEWOLF, o. H. 1928. Uber Farbstreifen bei Amaltheus (Paltopleuroceras) spinatus (Brug.). Palaont. Z. 10,
136-143.

SPATH, L. F. 1935. On colour markings in ammonites. Ann. Mag. nat. Hist., ser. 10, 15, 395-398.

STENZEL, H. B. 1964. Living Nautilus. In MOORE, R. c. (ed.). Treatise on invertebrate paleontology. Part K,
Mollusca, K59-K93. Geological Society of America and University of Kansas Press, Boulder, Colorado
and Lawrence, Kansas.

TANABE, K. and KANiE, Y. 1979. Colour markings in two species of tetragonitid ammonites from the Upper
Cretaceous of Hokkaido, Japan. Sci. Rep. Yokosuka Cy Mus. 25, 1-6.

TEICHERT, c. 1964. Morphology of hard parts. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part
K, Mollusca, K13-K53. Geological Society of America and University of Kansas Press, Boulder, Colorado
and Lawrence, Kansas.

TOZER, E. T. 1972. Observations on the shell structure of Triassic ammonoids. Palaeontology, 15, 637-654.
WARD, p. D., STONE, R., WESTERMANN, G. and MARTIN, A. 1977. Notes ou animal weight, cameral fluids,
swimming speed, and colour polymorphism of the cephalopod Nautilus pompilius in the Fiji Islands. Ibid.
3, 377-388.

WILLEY, A. 1902. Contribution to the natural history of the pearly nautilus. Zoological results based on material
from New Britain, New Guinea, Loyalty Islands and elsewhere, collected during the years 1895, 1896 and 1897.
Part 6, 691-803. Cambridge University Press, Cambridge, England.
wiNDLE, D. L. 1973. Studies in Carboniferous nautiloids: cyrtocones and annulate orthocones. Ph.D. thesis
(unpublished). University of Iowa.

WRIGHT, T. 1881. Monograph on the Lias ammonites of the British Islands. Palaeontogr. Soc. [Monogr.], 4,
205-328.

ROYAL H. MAPES
DEBRA A. SNECK

Department of Geological Sciences
Ohio University
Athens, Ohio 45701
USA

Typescript received 7 January 1986
Revised typescript received 7 May 1986

^'1

•J

‘ S5 <«'

I;

i!'j-

V

Vk

EARLY CRETACEOUS BELEMNITES FROM
SOUTHERN MOZAMBIQUE

by PETER DOYLE

Abstract. Neohibolites ewaldi (v. Strombeck, 1861) and Per at obelus foersteri sp. nov. are described from the
Upper Aptian sediments of the Rio Maputo river, near Catuane, southern Mozambique. Peratobelus has a
wider distribution than was first thought, occurring on both Tethyan (Mozambique) and Pacific (Antarctica,
Australia) coasts of Gondwana. The presence of Neohibolites in Mozambique strengthens the hypothesis of a
widespread migration of this genus in the Aptian-Albian.

The following is a study of fifty-two complete and fragmentary belemnite rostra collected by Drs
H. Wachendorf and R. Forster from one stratigraphic section on the Rio Maputo river, near
Catuane, southern Mozambique (26° 49' 50" S, 32° 13' 15" E). The stratigraphy of this section has
been discussed in detail by Henriques Da Silva (1962: locality B), Wachendorf (1967), and more
recently by Forster (1975). The succession exposed there consists of more than 10 m of glauconitic
sandstones, marls, and occasional limestones, containing a rich ammonite fauna (including species
of Tropaeum, Ammonitoceras, and Acanthohoplites) indicative of an Upper Aptian age (Henriques
Da Silva 1962; Wachendorf 1967; Forster 1975). This succession forms part of the transgressive
Maputo Formation (Barremian-Turonian) which overlies the Triassic-Jurassic Karroo Volcanics
in this region.

The belemnites belong to the genera Neohibolites and Peratobelus, and were first recognized as
such by Wachendorf (1967: discussion by Schmid, p. 279). Previously, there have been only scant
references to Cretaceous belemnites from Mozambique, although the rich belemnite fauna of
adjacent Madagascar is well known (e.g. Besairie 1930). Newton (1924) referred to a few specimens
from Port Amelia, northern Mozambique, and these belong to the early Cretaceous genus Duvalia
(or Pseudoduvalia) (BM.C. 25957-25958). Two other belemnite fragments from nearby may be
juvenile Hibolithes, or even Neohibolites (BM.C. 25955-25956). Spath (1939) described a new species
of diplobelinid belemnite from the Upper Aptian of Chalala, southern Mozambique {Cono-
teuthis rennieri) which was later redescribed by Jeletzky (1981) as type of his new diplobelinid
genus Chalalabelus. Bowen ( 1 963) determined an Upper Aptian palaeotemperature of approximately
17° C using belemnites from Manyola Drift, southern Mozambique.

The purpose of this paper is to describe in detail the Rio Maputo belemnites, and discuss
their biogeographical significance. Unless otherwise stated, all specimens are housed in the
Bayerische Staatssammlung fiir Palaontologie und historische Geologie, Munich and are pre-
fixed by: 1967 xxvii. The terminology used below is discussed in a separate paper on the Dimito-
belidae (Doyle 1987u), and it may also be useful to consult Stevens (1965). Abbreviations are as
follows: BM, British Museum (Natural History); L, total preserved length of rostrum; X, length
from apex to rostrum to position of maximum inflation; Dvmax, maximum dorsoventral diameter;
Dlmax, maximum lateral diameter; Dvl; dorsoventral diameter at position of Dlmax. All measure-
ments in mm.

[Palaeontology, Vol. 30, Part 2, 1987, pp. 311-317, pi. 43.|

© The Palaeontological Association

312

PALAEONTOLOGY, VOLUME 30

SYSTEMATIC PALAEONTOLOGY

Order belemnitida Zittel, 1895
Suborder belemnopseina Jeletzky, 1965
Family belemnopseidae Naef, 1922
Genus neohibolites Stolley, 1911

Type species. Belenmites ewatdi v. Strombeck, 1861, by subsequent designation of Gorn (1968, p. 383).
Diagnosis. See Stolley (191 lo, p. 174), Swinnerton (1955, p. xxxix), Spaeth (1971, p. 56).

Neohibolites ewaldi {\. Strombeck, 1861)

Plate 43, figs. 1-5

1847 Belenmites semicanaliculatus Blainville; d’Orbigny, p. 23, pi. ix, figs. 7 and 8.

*1861 Belenmites Ewaldi v. Strombeck, p. 34.

191 \a Neohibolites Ewaldi (v. Strombeck); Stolley, p. 31, pi. I, figs. 1-20.

191 la Neohibolites clava Stolley, p. 37, pi. I, figs. 21-29; pi. II, figs. 1-12.

191 la Neohibolites inflexus Stolley, p. 42, pi. I, fig- 30; pi. II, figs. 13-26.

1955 Neohibolites ewaldi (v. Strombeck); Swinnerton, p. 64, pi. xvi, figs. 8-26; pi. xvii, figs. 1-14.

Type specimen. Lectotype (designated by Swinnerton 1955, p. 65), the original of d’Orbigny (1847, pi. ix,
fig. 7), Aptian, southern Prance.

Material. Three rostra from bed 2 (235, 236, 238); nineteen rostra from bed 7 (185-203); nine rostra from
beds 10-11 (208-217); and two rostra from bed 14 (223, 224). Upper Aptian, Rio Maputo section (Forster
1975, p. 24, fig. 6), southern Mozambique.

Dimensions L

X

Dvmax

Dlmax

Groove Length

208

50-8

17-5

7-0

7-7

8-4

209

51-9*

—

7-9

8-3

8-7

210

43-9

140

7-4

71

np

211

34-8f

16-2

7-3

6-8

np

185

50-6

23-7

7-2

7-5

np

186

44-3t

23-4

7-3

7-3

np

187

30-3

11-3

4-7

4-9

np

188

35-6

—

7-1

7-5

np

189

60-2

20-2

8-5

8-8

7-6

190

49-4

18-3

7-5

7-7

np

191

50-5

20-4

8-4

8-5

np

192

44-4t

19-5

7-4

7-4

np

193

40-0t

19-3

6-5

6-7

np

196

35-3*

—

5-9

61

np

235

460

16-9

7-1

7-1

np

*; apex missing, f: alveolar region missing, np: groove not preserved.

Description. A group of medium sized, hastate to subhastate Neohibolites, with a total length approximately
6-5 times the maximum dorsoventral diameter (Dvmax). The outline and profile are similar and usually hastate
or subhastate and symmetrical. Most are hastate, resembling N. clava Stolley and N. inflexus Stolley (see
Stolley 191 la; Swinnerton 1936-1955), while the larger subhastate forms are closer to N. ewaldi sensii stricto.
The apex of all these forms is moderately acute. Transverse sections are circular in the stem and apical regions,
becoming compressed and subquadrate in the alveolar region.

A short, deep, well-defined alveolar groove is seen in some of the specimens, despite the destruction of the
alveolar region. Doppellinien (double lateral lines) in these specimens are not clearly observed due to poor
preservation. Decay of the alveolar region is common, and more hastate individuals superficially resemble the
late Cretaceous genus Actinocamax Miller. Owing to the loss of the alveolus, no information is available
concerning the phragmocone of this species. The apical line is central and ortholineate.

DOYLE: CRETACEOUS BELEMNITES FROM MOZAMBIQUE

313

Remarks. Swinnerton (1936-56), p. 64) recognized that the species N. clava and N. inflexus described
by Stolley (1911a) from the Aptian of Germany were in fact varieties of N. ewaldi (v. Strombeck),
differentiated mainly by size and regularity of destruction of their alveolar regions (Stolley \9\\b,
p. 183). A similar range of variation in N. ewaldi was seen in single horizons (e.g. beds 7, 10-1 1) of
the Rio Maputo section. Consequently, Swinnerton’s interpretation is followed here in contrast to
the continuous N. ewaldi-N. clava- N. inflexus lineage envisaged by Stolley (19116) and Mutterlose
et al. (1983). N. ewaldi differs from N. minimus (Miller) and its allied forms, which are smaller, less
hastate with slightly flattened venters.

The specimens of N. ewaldi from southern Mozambique were previously referred to as N. cf.
inflexus Stolley by Wachendorf (1967, p. 279) and Forster (1975, p. 25).

Family dimitobelidae Whitehouse, 1924
Genus peratobelus Whitehouse, 1924

Type species. Belemnites oxys Tenison-Woods, 1884, by original designation.

Diagnosis. See Whitehouse (1924, p. 410), Stevens (1965, p. 61), Doyle (1985, p. 27, fig. 4, 1987a).

Peratobelus foersteri sp. nov.

Plate 43, figs. 6-12

Type specimens. Holotype, 229, bed 14. Paratypes, 204, bed 7; 219, 220, beds 10-11; 111, 228, bed 14. Upper
Aptian, Rio Maputo section (Forster 1975, p. 24, fig. 6), southern Mozambique.

Other material. One rostrum from bed 2 (237); two rostra from bed 7 (205, 207); two rostra from beds 10-1 1,
(221, 222); and five rostra from bed 14 (230-234). Upper Aptian, Rio Maputo section (Forster 1975,
p. 24, fig. 6), southern Mozambique.

Derivation of name. In recognition of the work of Dr R. Forster.

Diagnosis. Small, conical Peratobelus. Outline symmetrical, conical, cylindriconical to subhastate.
Profile symmetrical, conical to cylindriconical. Transverse sections subquadrate to pyriform.
Ventrolateral alveolar grooves slightly sinuous.

Dimensions L

X

Dvl

Dlmax

Groove Length

204

27-6

10-5

3-6

4-1

17-3

207

19-5*

7-7

3-9

4-0

np

219

34-2

10-8

5-1

5-6

22-4

220

29-3

—

—

—

20-7

221

24- 1

10-6

3-7

4-0

130

221

311

9-3

4-5

4-9

21-7

228

35-4

11-2

5-5

6-1

26-2

229

33-9

IM

4-5

51

23-8

230

30-5

11-4

50

61

24-0

231

26-7*

—

—

—

13-4

232

22-7

7-9

4-2

4-5

12-5

233

19-5*

9-8

3-6

4-3

7-6

237

28-9

11-7

5-9

6-4

15-8

*: apex missing, np: groove not preserved.

Description. Small, conical Peratobelus. Total length approximately 6 times the dorsoventral diameter at the
position of the greatest lateral diameter (Dvl). The outline is symmetrical and generally cylindriconical. While
some forms are conical, most are cylindriconical to subhastate with a slightly inflated stem region. The apex
is acute, and often attenuate in juvenile specimens. The profile is symmetrical or almost symmetrical and
usually conical or cylindriconical, depending on the flatness of the venter. The stem region is not inflated in
profile. Transverse sections are subquadrate to rounded subquadrate, roundness increasing adapically. The

314 PALAEONTOLOGY, VOLUME 30

venter is generally flattened, the stem and apical regions being slightly depressed in contrast to the slightly
compressed alveolar region.

Two long, deep ventrolateral alveolar grooves run parallel to the venter for two thirds to three quarters of
the length of the rostrum. Where the outline is inflated, the otherwise straight grooves may develop a ‘kink’.
Adapically the grooves may be slightly dorsally deflected, but there is no evidence to suggest they are prolonged
as lateral lines, seen in Dimitobelus. As in all Peratobelus species, lateral lines are poorly preserved, but some
dorsolateral alveolar flattening is present. The alveolar region is preserved entire, with no development of a
pseudalveolus. No phragmocones were present in any of the specimens, but the alveolus penetrates approxi-
mately one third of the rostrum.

Remarks. P. foersteri sp. nov. is typical of its genus. It is cylindriconical with a normal alveolar
region, a robust transverse section and extremely long ventrolateral alveolar grooves. In form it
resembles the type species P. oxys (Tenison-Woods), but differs in its smaller size, its conical profile,
and its less hastate outline. P. bauhinianus Skwarko also approaches P. foersteri sp. nov., but is
distinguished by its more regular subhastate form (Skwarko 1966, p. 124, pi. 15, figs. 7-11). Finally,
P. foersteri sp. nov. resembles Tetrabelus willeyi Doyle (Doyle 1987a) because of its conical form,
but the latter is distinguished by its marked, ventrally curving alveolar grooves.

P. foersteri sp. nov. was previously referred to as Peratobelus sp. nov. by Wachendorf (1967,
p. 279) and by Forster (1975, p. 25).

PALAEOBIOGEOGRAPHICAL IMPLICATIONS

The discovery of Neohibolites and Peratobelus in southern Mozambique is an important addition
to our understanding of the distribution of these genera. Stevens (1963, 1965, 1973) and Doyle
(1985, 19876) have discussed the distribution of these belemnites in the Aptian and Albian. They
suggested that at this time Neohibolites and the related genus Parahibolites migrated widely away
from their Tethyan origins, penetrating deep into both hemispheres. Thus in the Aptian of the
Southern Hemisphere, N. minimus and its allies are recorded from South America and Antarctica
(Stevens 1965; Willey 1973) (text-fig. 1), and by the Albian/Cenomanian they are recorded in
addition from Madagascar, southern India, and Japan (Stevens 1965). Doyle (1985) has suggested
that Neohibolites was also present in Australia and New Guinea at this time. Forms related to N.
ewaldi are known from South America in the Aptian (Fiddle 1946; Stevens 1965), as shown in text-
fig. 1. Both N. minimus and N. ewaldi are extremely common in the Aptian and Albian of the
Northern Hemisphere, and are especially well known in north-west Germany and England (Stolley
191 Ifl, 6; Swinnerton 1936-1955; Spaeth 1971).

By contrast, Peratobelus and the Dimitobelidae had a much more restricted distribution, being
found only within the 30 °S Cretaceous palaeolatitude (Stevens 1973; Doyle 1985, 19876; see text-
fig. 1). The Aptian genus Peratobelus was previously thought to have been restricted to the Pacific
coast of Gondwana, similar to Dimitobelus in the Albian, and in contrast to the trans-Gond-
wanian Tetrabelus (Doyle 19876). The discovery of Peratobelus in southern Mozambique, still within
the 30 °S palaeolatitude (see Smith et al. 1981, p. 34) (text-fig. 1) indicates a wider distribution
than its successors Dimitobelus and Tetrabelus, occurring on the Pacific and Tethyan coasts of

EXPLANATION OF PLATE 43

Figs. 1-5. Neohibolites ewaldi (v. Strombeck), ventral outlines and right profiles, x 1. Aptian, Rio Maputo,
southern Mozambique. 1, large individual without apex, 209. 2, large individual with ?postmortal borings,
189. 3, hastate individual, 191. 4,208. 5,185.

Figs. 6-12. Peratobelus foersteri sp. nov., ventral outlines and right profiles, x 1. Aptian, Rio Maputo,
southern Mozambique. 6, holotype, 229. 7, paratype, robust individual, 228. 8, paratype, 219. 9, paratype,
227. 10, paratype, squat individual with flat venter, 220. 11, paratype, juvenile, 204. 12, hastate individual,
230.

PLATE 43

DOYLE, Neohiholites, Peratobelus

316

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1 . Palaeogeographic reconstruction of southern Gondwana in the
Aptian. Numbered localities are as follows: 1, Patagonia; 2, Alexander
Island, Antarctica; 3, New South Wales, Australia; 4, South Australia;
5, Queensland, Australia; 6, Northern Territories, Australia; 7, southern

Mozambique.

Gondwana in the Antarctic Peninsula, Australia, and southern Africa. The discovery of Neohibolites
in association with Peratobelus in Mozambique is similar to the occurrence of Paraliibolites with
Tetrabelus in the Antarctic Peninsula and India (Doyle 1985), and strengthens the hypothesis of a
widespread Neohibolites! Parahibolites migration in the Aptian and Albian.

Acknowledgements. I thank Dr R. Forster for the opportunity to work on these specimens, and Dr M. K.
Howarth and D. Phillips for critical reading of the manuscript. This work was carried out while I was in
receipt of a NERC Research Fellowship at the British Museum (Natural History). The palaeogeographic map
was drawn with the aid of the ATLAS map making program, courtesy of Dr A. G. Smith, University of
Cambridge.

REFERENCES

BESAiRiE, H. 1930. Recherches geologiques a Madagascar. Bull. Soc. Hist. not. Toulouse, 60, 1-264.

BOWEN, R. 1963. Measurement of paleotemperatures of the Upper Aptian of Mozambique, Africa, and Middle
Cretaceous paleoclimatology. Am. J. Sci. 261, 566-570.

DOYLE, p. 1985. ‘Indian’ belemnites from the Albian (Lower Cretaceous) of James Ross Island, Antarctica.
Bull. Br. Antarct. Surv. 69, 23-34.

1987a. The Cretaceous Dimitobelidae (Belemnitida) of the Antarctic Peninsula. Palaeontology, 30, 147-

177.

DOYLE: CRETACEOUS BELEMNITES FROM MOZAMBIQUE

317

1987^. The belemnite family Dimitobelidae in the Cretaceous of Gondwana. Int. Un. geol. Sci., Series A

(in press).

FORSTER, R. 1975. Die geologische Entwicklung von Siid-Mozambique seit der Unterkreide und die Ammo-
niten-Fauna von Unterkreide und Cenoman. Geol. Jb. B, 12, 3-324.

CORN, N. K. 1968. Systematics of Early Cretaceous Belemnopsinae. Paleont. J. 2, 383-384.

HENRiQUES DA SILVA, G. 1962. Amonites do Cretacico inferior do rio Maputo (Catuane-Mogambique). Bolm
Servs Ind. Minas Geol. Lourenqo Marques, 29, 7-32.

JELETZKY, J. A. 1981. Lower Cretaceous diplobelinid belemnites from the Anglo-Paris Basin. Palaeontology,
24,115-145.

LiDDLE, R. A. 1946. The Geology of Venezuala and Trinidad, 890 pp. Paleontological Research Institution,
Ithaca, New York.

MUTTERLOSE, J., SCHMID, F. and SPAETH, c. 1983. Zur Palaobiogeographie von Belemniten der Unter-Kreide in
NW-Europa. Zitteliana, 10, 293-307.

NEWTON, R. B. 1924. A Contribution to the palaeontology of Portugese East Africa. Trans geol. Soc. S. Afr. 26,
141-159.

ORBIGNY, A. DE. 1847. Paleontologie franqaise. Terrains cretaces. Supplement, 28 pp. Paris.

SKWARKO, s. K. 1966. Cretaceous stratigraphy and paleontology of the Northern Territory. Bull. Bur. Miner.
Resour. Geol. Geophys. Aust. 73, 1-135.

SMITH, A. G., HURLEY, A. M. and BRIDEN, J. c. 1981. Plianerozoic paleocontinental world maps, 98 pp. Cambridge
University Press, Cambridge.

SPAETH, c. 1971. Untersuchungen an Belemniten des Formenkreises um Neohibolites minimus (Miller 1826)
aus dem Mittel und Uber Alb, Nordwestdeutschlands. Beib. geol. Jb. 100, 1-127.

SPATH, L. F. 1939. On a new belemnoid {Conoteuthis rennieri) from the Aptian of the Colony of Mozambique.
Bolm Servs Ind. Minas Geol. Lourenqo Marques, 2, 1-16.

STEVENS, G. R. 1963. Faunal realms in Jurassic and Cretaceous belemnites. Geol. Mag. 100, 481-497.

1965. The Jurassic and Cretaceous belemnites of New Zealand, and a review of the Jurassic and

Cretaceous belemnites of the Indo-Pacific region. Pal. Bull. geol. Surv. N.Z. 36, 1-283.

1973. Cretaceous belemnites. In hallam, a. (ed.). Atlas of Palaeobiogeography, 259-274. Elsevier, Amster-
dam.

STOLLEY, E. 191 lu. Beitrage zur Kenntnis der Cephalopoden der norddeutschen unteren Kreide. 1: Die belemni-
tiden der norddeutschen unteren Kreide (Aptiens und Albiens). Geol. paldont. Abh. n.f. 10, 201-272.

1911^. Studien an den Belemniten der unteren Kreide Norddeutschlands. Jber. niedersdchs. geol. ver. 4,

174-191.

STROMBECK, A. VON. 1861. Ucber den Gault und insbesondere die Gargas-Mergel im nordwestlichen Deutsch-
land. Z. dt. geol. Ges. 13, 20-60.

swinnerton, h. h. 1936-1956. A Monograph of British Eower Cretaceous belemnites. Palaeontogr. Soc.
[Monogr.], 1-86.

WACHENDORF, H. 1967. Zur Unterkreide-Stratigraphie von Siid-Mo9ambique. Neues Jb. geol. Paldont., Abh.
129-303.

WHiTEHOUSE, F. w. 1924. Dimitobelidae— a new family of Cretaceous belemnites. Geol. Mag. 61, 410-416.

WILLEY, L. E. 1973. Belemnites from south-eastern Alexander Island. II. The occurrence of the family Belemno-
pseidae in the Upper Jurassic and Lower Cretaceous. Bull. Br. Antarct. Surv. 36, 33-59.

PETER DOYLE

Department of Palaeontology
British Museum (Natural History)

Manuscript received 24 January 1986 Cromwell Road

Revised manuscript received 15 April 1986 London SW7 5BD

S-; *

••II

TAXONOMY, EVOLUTION, AND FUNCTIONAL
MORPHOLOGY OF SOUTHERN AUSTRALIAN
TERTIARY HEMIASTERID ECHINOIDS

by KENNETH J. MCNAMARA

Abstract. Eight species of hemiasterid echinoids, assigned to the genera Hemiaster (Bolbaster) and Psepho-
aster gen. nov., are described from Late Eocene to Middle Miocene strata of southern Australia. Seven of the
species are new: the Late Eocene H. (B.) subidus, the Early Oligocene H. {B.) dolosus, the Early Miocene H.
(B.) verecundus, the Middle Miocene H. (B.) callidus, the Late Eocene P. lissos, the Late Oligocene P.
apokryphos and the Early Miocene P. klydonos. A neotype is selected for H. (B.) planedeclivis Gregory, 1890.
The five species of H. {Bolbaster) are considered to have formed a single evolutionary lineage, as are the three
species of Psephoaster. Directional trends in many morphological features in these two lineages are interpreted
as reflecting both paedomorphosis and peramorphosis. The sediments in which the echinoids lived became
progressively finer-grained from the Eocene to the Miocene and the morphological changes in the two lineages
are considered to reflect adaptations by descendant morphotypes to the occupation of finer-grained sediments.
Many of the morphological changes are common to both lineages. Species of Psephoaster are considered to
have been shallow burrowers, while species of H. {Bolbaster) are interpreted as having burrowed more deeply
in the sediment.

Hemiasterid echinoids form a relatively minor part of an otherwise rich spatangoid echinoid
fauna which occurs in parts of the Tertiary sedimentary sequence of southern Australia. The only
hemiasterid species to have been described previously from these rocks is Hemiaster planedeclivis
Gregory, 1890, from the Early/Middle Miocene Morgan Limestone exposed in the banks of the
Murray River in South Australia. Since Gregory’s description there has been a tendency to call any
specimen of Hemiaster found in Australian Tertiary deposits H. planedeclivis, specimens attributable
to Hemiaster having subsequently been found in Late Eocene to Early/Middle Miocene strata (Tate
1891; Clark 1946).

Extensive collecting in recent years, by R. and F. Foster, of Tertiary sequences in the St Vincent
and Murray Basins, South Australia, and in the Torquay and parts of the Otway Basins, Victoria,
has yielded a number of forms superficially resembling H. planedeclivis. These specimens have come
from the Late Eocene Tortachilla Limestone and Early Oligocene Port Willunga Formation, in
coastal cliff sections in the St Vincent Basin, south of Adelaide, South Australia; from coastal cliff
sections in the Late Oligocene Jan Juc Formation and Early Miocene Puebla Formation in the
Torquay Basin, south-west of Melbourne, Victoria; from the Early Miocene Mannum Formation
and the Early/Middle Miocene Morgan Limestone in cliff's exposed by the Murray River in the
Murray Basin, South Australia; and from a coastal section exposing the Middle Miocene Port
Campbell Limestone, in the south-eastern Otway Basin, Victoria (text-fig. 1). Specimens collected
from these units are described below and assigned to eight species, seven of which are new. Five of
the eight species are placed in Hemiaster, all being members of the subgenus Bolbaster. The other
three species are considered to belong in a separate, previously unrecognized genus, Psephoaster
gen. nov.

The five species of H. {Bolbaster) and the three species of Psephoaster are considered to form
two evolutionary lineages. Most of the diagnostic morphological features which separate members
of the lineages show temporal, directional, morphological transformations. These morphological
changes are examined and compared with the ontogenetic development of species of Hemiaster in

[Palaeontology, Vol. 30, Part 2, 1987, pp. 319-352, pis. 44-48.] © The Palaeontological Association

320

PALAEONTOLOGY, VOLUME 30

Middle

Miocene

Early

Miocene

Late

Oligocene

Early

Oligocene

Late

Eocene

Port

Willunga

Formation

3

Chinaman Gully &

Blanclic Point l-ormations

Tortachilla
Limestone 1,2

A. St Vincent Basin (L)

Morgan Limestone ^

Mannum ^

Buccleuch

Group

Port

Campbell 5
Limestone

Gellibrand

Marl

Nirranda

Sub-group

B. Murray Basin (\V) C. Otway Basin (SL) U. Torquay Basin

TEXT-FIG. 1. Simplified stratigraphy of Tertiary strata of the eastern St Vincent Basin, western Murray Basin,
south-eastern Otway Basin, and the Torquay Basin in southern Australia (after Abele et al. 1976 and Lindsay
1985) and locality map showing distribution of hemiasterid species.

McNAMARA: TERTIARY ECHINOIDS

321

order to assess the importance of heterochrony in the evolution of the two lineages. The functional
significance of these morphological changes is assessed, particularly in relation to the character
of the sediment in which the echinoids are preserved. It is considered that the specimens are
autochthonous, the frequent presence of fine spines still adhering to some tests indicating that
the specimens both lived and died within the sediment. Discussion of the relationship between
morphology and enclosing sediment is therefore considered to be valid.

The relationship of morphology to sediment type has been assessed for a number of spatangoid
groups in recent times (McNamara 1982a, 1985; McNamara and Philip 1980, 1984; Smith 1980Z),
1984). The present study of evolutionary trends and their functional significance in two closely
related spatangoid genera provides further evidence to support the view that morphological evolu-
tion in many spatangoids was strongly affected by the character of the sediment in which the
eehinoid lived and upon which it fed.

Materials and Methods. The collections upon which this study is based are housed in the Museum of Victoria,
Melbourne (NMV), the South Australian Museum, Adelaide (SAM), the Western Australian Museum, Perth
(WAM), and the private collection of E. and F. Holmes. Measurements were made to an accuracy of 01 mm
with a vernier calliper or an ocular graticule fitted to a Wild binocular microscope. A number of parameters
are expressed as percentages of maximum test length (% TL).

EVOLUTIONARY TRENDS AND THEIR FUNCTIONAL
SIGNIFICANCE IN HEMI ASTER (BOEBASTER) AND PSEPHO ASTER

Morphological evolution o/H. (Bolbaster)

Five species of H. {Bolbaster) occur in Eocene to Miocene strata in southern Australia. The oldest
is the Late Eocene (Aldingan) H. {B.) suhidus sp. nov., which is succeeded by the Early Oligocene
(Willungan) H. {B.) dolosus sp. nov., the Early Mioeene (Longfordian) H. (B.) vereciindiis sp. nov.,
the late Early/early Middle Mioeene (Batesfordian/Balcombian) H. (B.) planedeclivis Gregory, and
finally the late Middle Mioeene H. (B.) callidus. sp. nov. (text-fig. 2). It is considered that these five
species form part of a single evolutionary lineage. They oecur in relatively close geographical
proximity to one another. The two earliest speeies occur in the Tortachilla Limestone and Port
Willunga Formation, respectively, in the cliffs south of Adelaide. H. (B.) verecundus oecurs 650 km
to the south-east in the Puebla Formation on the southern Victoria coast, H. (B.) ccdlidus 200 km
to the west in the Port Campbell Limestone, and H. {B.) planedeclivis 500 km north-west in the
Morgan Limestone, which outcrops in the Murray River Cliffs. The wide ranges of some modern
Australian taxa, such as more than 4,000 km for both Breynia desorii (McNamara 1982a) and
Protenaster australis (McNamara 1985), argue, in addition to the morphologieal criteria discussed
below, for the five species of H. {Bolbaster) all forming part of a single evolutionary lineage.

Some ten morphological characters of the test undergo directional morphological change along
the H. {Bolbaster) lineage over a period of twenty-eight million years from the Late Eoeene to
Middle Miocene. The charaeters which show these directional morphological transformations are:
test height, width, and profile; length and depth of petals; length and depth of ambulacrum III, and
extent of tuberculation; shape and disposition of the pore pairs in the petals; position of the apical
system; thickness of the peripetalous fasciole; shape of the aboral interambulaera; density of ambital
and aboral tuberculation; and form of the plastronal plating.

Test shape. During the eourse of the evolution of the H. {Bolbaster) lineage in southern Australia
the test shape underwent a marked change, from ovoid in the earlier species, to near circular in
outline in the later species (text-fig. 3b). In the Late Eocene H. {B.) subidus the width of the test is
90-96 % TL. The test widened in the three succeeding species (93-98 % TL in H. {B.) dolosus, 97 %
TL in H. {B.) verecundus and 91-101 % TL in H. {B.) planedeclivis, apart from one specimen which
is 88 % TL). In the youngest species, the Middle Miocene H. {B.) callidus, the width is even greater
at 97-103 %TL.

322

PALAEONTOLOGY, VOLUME 30

Middle

Miocene

Ma

15-

PsephoasteiA \ | Hemiaster (Bolbaster)

\ callidus ^

\ \ H. (B.) planedeclivis

Early

1 \ \

■

Miocene

20-

P. klydonos \ \

■

H. (B.) verecundus

Late

25-

1 \\

Oligocene

30-

P. apokryphos \ \

\ >

Early

\

1

Oligocene

35-

\

I

H. (B.) dolosus

Late

Eocene

■

1

P. lissos

H. (B.) subidus

sediment
grain size

r

TEXT-FIG. 2. Stratigraphic and sedimentological distribution of species of Hemiaster (Bolbaster) and Psepho-

aster in the Tertiary of southern Australia.

Concomitant with this increase in test width there was an increase in test height and a change in
the profile of the posterior surface of the test (text-figs. 3a, 8). In the Late Eocene H. {B.) subidus
the posterior surface of the test is inclined such that interambulacrum 5 aborally overhangs the
posterior of the plastron. With the evolution of the lineage the posterior face changed its orientation
to become vertical in H. {B.) dolosus and H. {B.) verecundus, then slightly inclined forward in H. {B.)
planedeclivis, and strongly inclined forward in the youngest species, H. (B.) callidus, such that both
the periproct and posterior of the plastron are visible from above.

The general result of the increase in test width and height along the lineage was for the test to
become more spherical in shape; the increase in relative test height occurred both by a foreshortening
of the test and by increased swelling of interambulacrum 5.

Petals and ambulacrum III. These structures underwent a number of morphological changes, princi-
pally involving: progressive deepening; modest relative shortening of the anterior petals; change in
the structure of the pore pairs in the petals, from each pore being elongate and proximal to the
other pore of the pair in the early species, to nearly circular and widely separated in the later species;
and decrease in primary tuberculation in the interporiferous zone.

In the two oldest species, the Late Eocene H. {B.) subidus and the Early Oligocene H. (5.) dolosus,
the anterior petals are 28-32 % TL and 29-30 % TL in length, respectively. The Early Miocene H.
{B.) verecundus is characterized by its particularly short anterior petals, 20-22 % TL in length. The
two youngest species, H.{B.) planedeclivis and H. (B.) callidus, have slightly longer petals, but shorter
than those of the two oldest species, being 24-28 % TL and 25-26 % TL in length, respectively (text-
fig. 4a). This reduction in petal length corresponds to a reduction in the number of pore pairs. For
instance, the anterior petals of the oldest species, H. (5.) subidus, have eighteen to twenty-seven
pore pairs, whilst the youngest species, H. {B.) planedeclivis and H. {B.) callidus, have thirteen to

. Mio. M. Mio. L. Mio.

McNAMARA; TERTIARY ECHINOIDS

323

A

H. (B.) callidus

H. (B.) planedeclivis

o

J

o

ui

u

o

w

J

H. (B.) dolosus

H. (B.) suhidus
70

Test height (% TL)

TEXT-FIG. 3. Ranges (horizontal lines) and means (vertical lines), for species of Hemiaster (Bolbaster)
in the Tertiary of southern Australia of test height, test width, width of peripetalous fasciole, and
distance between apical system and anterior ambitus; all expressed as percentages of test length.

324

PALAEONTOLOGY, VOLUME 30

B

■V li. (B.) callidus

H. (B.) planedeclivis

H. (B.) verecundus

% -V

H, (B.) dolosus

H. (B.) subidus

TEXT-FIG. 4. Ranges (horizontal lines) and means (vertical lines), for species of Hemiaster {Bolhaster) in the
Tertiary of southern Australia, of the length of the anterior petals, expressed as percentages of test length
(left); and illustrations of the posterior row of pore pairs of anterior petals of the five species of H. (Bolhaster),
illustrating the progressive temporal reduction in size of the pores and their increasing separation (right).

twenty-one and sixteen to twenty, respectively (text-fig. 4b). Similarly, the posterior petals show a
reduction, with thirteen to nineteen pore pairs in H. (B.) subidus, eight to fourteen in H. (B.)
plamleclivis, and eleven or twelve in H. (B.) callidus.

As the anterior petals became shorter, from the Late Eocene to the Middle Miocene, so both
they and the posterior petals also became progressively deeper. A similar deepening also occurred
in ambulacrum III. Furthermore, as the petals and ambulacrum III deepened, so the aboral interam-
bulacra became increasingly swollen. This effectively increased the depth of the petals and ambu-
lacrum III even further. Consequently, the smooth aboral surface of the early species of H.
(Bolhaster) progressively changed through the Tertiary to give way to a coarsely corrugated surface
in the youngest species, H. (B.) callidus (PI. 47, fig. 4).

Along with these changes in petal topography, the shape and disposition of the pore pairs in the
petals underwent a unidirectional morphological change. The pairs within each pore pair in the
Late Eocene H. (B.) subidus are elongate and situated close together, the distance between the pores
in each pair being less than the length of the pore (text-fig. 4b). In the descendant Early Oligocene
H. (B.) dolosus the pores are still as elongate, but are set a little further apart within each pair. The

McNAMARA: TERTIARY ECHINOIDS

325

pores became less elongate in the Early Miocene H. {B.) verecundus and set even wider apart, the
distance between the pores in each pair being greater than the length of the pore. In the Early/Middle
Miocene H. {B.) planedeclivis the pores became almost circular and separated by a distance greater
than their diameter. The final species in the lineage, the Middle Miocene H. (B.) callidus, also
possesses nearly circular pore pairs, but they are more widely separated than those of its ancestor,
H. (B.) planedeclivis, being separated by a distance up to twice the diameter of the pores.

Although the pores within each pair became progressively more widely separated, the interpori-
ferous zone between the anterior and posterior rows of pore pairs in each petal did not become
narrower. Instead, the petals became slightly wider to accommodate the expansion. The widening
rows were also accommodated by the increase in depth of the petals. However, although the
interporiferous zone did not change in width as the lineage evolved, it underwent a change in
character. In early species, such as H. (B.) subidus and H. (B.) dolosiis, the interporiferous zone is
covered by a mixture of primary and miliary tubercles (PI. 44, fig. 4). As the lineage evolved, so the
density of primary tuberculation decreased, with the result that the youngest species, H. {B.) callidus,
possesses only miliary tubercles. The same degree of reduction in primary tubercles is also evident
in ambulacrum III. H. (5.) subidus possesses a maximum of seventeen tubercles in ambulacrum III;
H. {B.) dolosus has up to nine; H. {B.) verecundus five; H. (B.) planedeclivis six; H. (B.) callidus
generally none, though rarely one or two. Tubercles were preferentially lost from the adapical part
of the ambulacrum. Thus, in H. (B.) subidus primary tubercles are spread over the entire length of
ambulacrum III, but are confined to the adambital half in H. (B.) dolosus and the adambital third
in H. (B.) verecundus and H. [B.) planedeclivis. Where primary tubercles do occur in H. {B.) callidus,
they are found close to the peripetalous fasciole.

Apical system. A consequence of the foreshortening of the test and the inclination of the posterior
surface of the test was a relative anterior movement of the apical system. It is posterior of centre in
the oldest species, H. (B.) subidus and H. {B.) dolosus, almost central in H. {B.) verecundus and H.
(B.) planedeclivis, and slightly anterior of centre in H. {B.) callidus (text-fig. 3d). With the anterior
migration of the apical system through the lineage, a change in the orientation of the anterior petals
might be expected; but this was not the case. There was, however, a change in the outline of the
peripetalous fasciole, from oval in the earlier species, with the long axis sagittal, to nearly circular
in H. (B.) callidus. Ambulacrum III also shortened as the apical system migrated anteriorly, from
45 % TL in the Late Eocene H. (B.) subidus to 40 % TL in the Middle Miocene H. (B.) callidus.

Peripetalous fasciole. One of the more dramatic morphological changes along the lineage was the
increase in width of the peripetalous fasciole (text-fig. 3c). This increase in width was not over the
entire fasciole, but was confined to the plate margins. In the earliest species, the Late Eocene H.
{B.) subidus, the fasciole is narrow (2-4-3-3 % TL) and of nearly even width throughout its course
(PI. 48, fig. 8). Where it crosses the ambulacra it covers no more than two plates in each column
(text-fig. 5a). As the species evolved along the lineage, so the fasciole progressively widened from
3-8 % TL in the Early Oligocene H. {B.) dolosus, to 5-7 % TL in H. (B.) verecundus, 5-7 % TL in H.
(B.) plandeclivis, and 7-3-9-5 % TL in H. {B.) callidus. In the later species the fasciole broadened at
all plate boundaries; at the centre of each ambulacral plate the fasciole is a little broader than in H.
(B.) subidus. Whereas the fasciole covers only two plates in each ambulacral column in H. {B.)
subidus, it covers five or six plates in each column in the youngest species, H. (B.) callidus (text-fig.
5b). The reduction in petal length and pore pair number in the petals along the lineage, and the
increase in the number of small, fasciole-bearing ambulacral plates, implies that the fasciole in-
creased in width at the expense of the pore pairs in the petals. In H. {B.) callidus the fasciole covers
almost the entire lateral margin of the interambulacral plates (PI. 48, fig. 7), whereas in H. (B.)
subidus it covers only one-sixth of the lateral plate margin.

Aboral tuberculation. In the Late Eocene H. {B.) subidus the aboral and ambital tuberculation
consists of relatively widely spaced primary tubercles set in a matrix of miliary tubercles (PI. 48,
fig. 8). As the lineage evolved, so the proportion of primary to miliary tubercles increased to such

326

PALAEONTOLOGY, VOLUME 30

an extent that in the Middle Miocene H. (B.) callidus the miliary tubercles are almost absent, the
primary tubercles being densely concentrated (PI. 48, fig. 7). The primary tubercle concentration
(measured on the aboral surface between the peripetalous fasciole and the ambitus) increases from
3-2 mm^^ in the oldest species, H. (B.) subidus, to 7-8 mm^^ in the youngest species, H. (B.) callidus
(text-fig. 9a).

Imm

Imm

TEXT-FIG. 5. Camera lucida drawings of the distal portion of the posterior ambulacrum and the peripetalous
fasciole (dotted), a, H. (Bolbaster) subidus sp. nov., SAM P26554, holotype. b, H. (B.) callidus sp. nov., NMV
PI 00503, holotype, illustrating its wider peripetalous fasciole, covering a greater number of ambulacral plates

than in H. [B.) subidus.

Adoral surface. This underwent less morphological evolution than the aboral surface in the H.
{Bolbaster) lineage. Changes were confined to the peristome and the plastron. The peristome became
positioned further from the anterior ambitus in later species. The peristomial margin became
increasingly raised and notched as the lineage evolved. The labrum also elongated and came to
project more strongly anteriorly in later species. Furthermore, the adoral interambulacra surround-
ing the peristome became a little more swollen.

The only other morphological changes on the adoral surface of the test involved the two plastronal
plates, interambulacra 5a and 5b. In H. {B.) subidus the plates are markedly asymmetric, being in a
primitive amphisternous condition. In the youngest species, H. {B.) callidus, the plates are almost
symmetrical (text-fig. 6).

Morphological evolution of Psephoaster

Three species of Psephoaster are known from Eocene to Miocene strata of southern Australia. The
oldest species is the Late Eocene (Aldingan) P. lissos sp. nov., which is succeeded by the Late
Oligocene (Janjukian) P. apokryphos sp. nov. and the Early Miocene (Longfordian) P. klydonos sp.
nov. (text-fig. 2). These species probably form part of a single evolutionary lineage. Like the H.

McNAMARA: TERTIARY ECHINOIDS

327

2mm

2mm

2 mm

TEXT-FIG. 6. Camera lucida drawings of the labrum and anterior plastron, a, Hemiaster (Bolbasler)
subidus sp. nov., SAM P26554, holotype. b, H. {B.) planedeclivis Gregory 1890, NMV P78460. c,
H. (B.) callidus sp. nov., NMV P100503, holotype. They illustrate the adaxial temporal migration

of the interambulacrum 5 suture.

(Bolbaster) lineage, the species in this lineage occur over a geographical range of some 650 km, well
within the extent of range of many modern spatangoid species.

Seven morphological characters of the test undergo directional morphological change along the
Psephoaster lineage over a period of twenty million years from the Late Eocene to Early Miocene.
Directional morphological transformations occur in the test height and profile, width and depth of
the petals, position of the apical system, shape of the aboral interambulacra, size and position of
the pore pairs in the petals, width of the peripetalous fasciole, and density of ambital tuberculation.
Apart from the change in pore pair shape, all other changes parallel those which occur in the H.
(Bolbaster) lineage.

Test shape. As the Psephoaster lineage evolved, so the test became relatively shorter and higher.
This occurred by a preferential increase in height of the posterior of the test in interambulacrum 5
(text-fig. 7). The height of the test relative to the length of the test increased from 66 % in the Late
Eocene P. lissos, to 67-71 % in the Late Oligocene P. apokryphos, and 79-87 % (apart from one
specimen at 72 %) in the Early Miocene P. klydonos. A concomitant change was the development
of a prominent rostrum in interambulacrum 5, both adorally and aborally. With this increase in test
height, the aboral surface, particularly in the central part of the test, became more steeply inclined.
The increase in height also resulted in the anterior movement of the apical system, from being 55 %
TL from the anterior ambitus in P. lissos, to 49 % TL in P. apokryphos, and 44-49 % in P. klydonos.

TEXT-FIG. 7. Camera lucida drawings of the lateral profiles of the three species of Psephoaster. a, P. lissos sp.
nov., SAM P26560, holotype. b, P. apokryphos sp. nov., NMV P100506, holotype. c, P. klydonos sp. nov.,
SAM P24631, holotype. They illustrate the temporal relative increase in test height.

328

PALAEONTOLOGY, VOLUME 30

Petals. The petals of the Late Eocene P. lissos are relatively narrow (3 % TL), very shallow, and
bear very small, widely spaced, circular pore pairs. As the lineage evolved, so the petals widened
(4-5 % TL in the Late Oligocene P. apokryphos, 5-6 % in the Early Miocene P. klydonos), became
a little deeper, and the pore pairs enlarged, the two rows of pore pairs in each petal becoming
situated closer together (text-fig. 13). The depth of the petals was effectively further increased by
the development of gently swollen adapical interambulacra in P. klydonos, in a similar manner to
those of H. {Bolhaster) callidus.

Peripetalous fasciole. The Psephoaster lineage parallels the H. {Bolhaster) lineage in its development
of an increasingly broader peripetalous fasciole in younger species. In the Late Eocene P. lissos the
fasciole is very thin (L5% TL), thread-like, and of even width throughout (PI. 47, fig. 3). The
fasciole is almost twice as wide in the Late Oligocene P. apokryphos, reaching a maximum width of
2-5 % TL where it crosses the ambulacra. It is slightly narrower across the interambulacra. This
variation in fasciole width is accentuated in the Early Miocene P. klydonos (PI. 48, fig. 1), which
reaches up to 6 % TL in width where it crosses the ambulacra and the interadial suture (text-fig.
13). Here it is about three times wider than where it crosses the middle of the interambulacral
plates.

Tuberculation. Aboral and ambital primary tuberculation increased in density from the Late Eocene
P. lissos, with a density of 3 mm“^, through the Late Oligocene P. apokryphos with 5 mm“^, to the
Early Miocene P. klydonos with 7-5 mm^^ (text-fig. 9b). As the primary tuberculation increased in
density, so the intervening miliary tubercles decreased in density. The rate of change in density of
primary tuberculation in Psephoaster, from 3 mm“^ in the Late Eocene to 7-5 mm~^ in the Early
Miocene, was a little greater than the increase in H. {Bolhaster), which over the same period
increased from 3-2 to 4-8 mm~^. However, by the end of the Middle Miocene the tuberculation
density in H. {Bolhaster) reached a similar concentration to that of Psephoaster, being 7-8 mm“^ in
H. {B.) callidus.

Role of heterochrony in the evolution of the H. (Bolhaster) and Psephoaster lineages.

Nearly all of the morphological changes which occurred along the H. {Bolhaster) and Psephoaster
lineages may be interpreted as being products of variation in the rate of ontogenetic development
of various structures (heterochrony). Certain morphological features are considered to have under-
gone an increased rate of development (acceleration), resulting in peramorphosis {sensu Alberch et
al. 1979; McNamara 1986), while others experienced a decreased rate of development (neoteny),
resulting in paedomorphosis.

A number of examples of heterochrony in echinoid evolution have recently been documented.
McNamara and Philip (1980) and McNamara (1982a) interpreted many of the morphological
changes which occurred with the evolution of S. {Schizaster) from S. {Paraster) in the Australian
Tertiary as having been produced by the action of both acceleration and hypermorphosis (extention
of ontogenetic allometries) resulting in peramorphosis. Conversely, the living species of Breynia are
thought to have evolved by paedomorphosis (McNamara 19826). The changes in test architecture
in Australian Tertiary species of Pericosmus are considered to have occurred by hypermorphosis,
resulting in peramorphosis (McNamara and Philip 1984). Similarly, changes in the form of phyllodal
plates in the Australian spatangoid Protenaster have been interpreted (McNamara 1985) as a
product of acceleration, resulting in peramorphosis. Hypermorphosis has been invoked by Smith
and Paul (1985) for the evolution of Discoides favrina (Desor) from D. suhucula (Leske) in the
Cenomanian of Devon. In all of these examples the morphological evolution within each lineage is
either paedomorphosis or peramorphosis. The species in each lineage lie along discontinuous
morphological gradients, termed (McNamara 1982a) either paedomorphoclines (as in Breynia) or
peramorphoclines (as in Schizaster, Pericosmus, and Protenaster).

Recently, McKinney (1984) has demonstrated that some structures in an Oligopygus lineage are
paedomorphic, while others are peramorphic in the same species. A similar situation occurs in
species in the H. {Bolhaster) and Psephoaster lineages where some of the morphological characters

McNAMARA: TERTIARY ECHINOIDS

329

of descendant species may be interpreted as paedomorphic, whilst others may be peramorphic. As
McKinney (1984, p. 415) noted, such dissociated heterochronic events have been documented in a
number of living organisms. Any particular developmental pattern for any structure may be dis-
rupted to varying degrees, potentially producing an almost endless variety of descendant hetero-
chronic morphotypes. The selection pressures which act on these descendant morphotypes in H.
{Bolbaster) and Psephoaster are discussed below.

In order to assess which particular structures in the species of H. {Bolbaster) and Psephoaster are
paedomorphic and which are peramorphic, it is necessary to be able to document the species’
ontogeny. The only described ontogenies of species of Hemiaster are those of the living H. expergitus
Loven, 1874 (Mortensen 1907) and the Paleocene H. (Leymeriaster) targari McNamara and Philip,
1987. These two species both follow similar ontogenetic pathways. Comparison of adult morpho-
logies of the five species of H. {Bolbaster) and the three species of Psephoaster with the ontogeny of
H. {L.) targari allows heterochronic patterns to be assessed.

During the ontogeny of H. {L.) targari the test lengthened, narrowed, became less inflated, and
the posterior face changed from being anteriorly inclined to near vertical; the petals lengthened and
deepened, and the pore pairs changed from being circular to elongate; ambulacrum III also deepened
and a weak anterior notch developed; the apical system underwent a relative posterior migration,
moving from anterior of centre to posterior of centre; the density of the tuberculation increased; on
the adoral surface the interambulacra became relatively less inflated close to the peristome; the
peristome itself migrated anteriorly and the labrum developed an anterior lip; and the plastron,
which was very convex in juveniles, became flatter during ontogeny (McNamara and Philip 1987).

Of the morphological trends which have been documented in the H. {Bolbaster) lineage, ten
follow the opposite trend to the ontogenetic development, with the result that later species became
increasingly juvenile in appearance in these particular characters. This increasing degree of paedo-
morphosis resulted in the establishment of paedomorphoclines for many of the structural changes.
Paedomorphoclines can be observed in the change in test outline, from oval to circular; the relative
increase in test height (text-fig. 8); the increasing inclination of the posterior surface of the test; the
shortening of the petals and ambulacrum III, and reduction in number of pore pairs in the petals;
the development of rounded pores from elongate pores in the petals; the reduction in primary
tuberculation in the petals; the reduction in primary tuberculation in the petals and ambulacrum
III; the relative anterior movement of the apical system; the broadening of the peripetalous fasciole;
the posterior migration of the peristome; and the swelling of the interambulacra close to the
peristome. Three evolutionary changes in the Psephoaster lineage which also occur in the H.
{Bolbaster) lineage, namely increase in test height, anterior displacement of the apical system, and
broadening of the fasciole, are also interpreted as being paedomorphic changes.

With structures such as the width of the peripetalous fasciole, in which the ancestral morphotypes
underwent negative allometry during growth, the degree of negative allometry decreased along the
lineage and approached isometry. Conversely, the petals grew with positive allometry, but the
paedomorphocline developed because the degree of positive allometry was reduced along the lineage
such that it approached closer to isometry. Reducing either positive or negative allometries closer
to isometry results in paedomorphosis. Increasing positive or negative allometries away from
isometry results in peramorphosis. Three structures form peramorphoclines in both the H. {Bolbas-
ter) and Psephoaster lineages: the deepening petals, the swelling of the interambulacra adapically,
and the increase in aboral primary tuberculation. Furthermore, the broadening petals and increase
in pore pair size are peramorphic features in Psephoaster.

Psephoaster itself may be interpreted as a paedomorphic hemiasterid. Juvenile hemiasterid charac-
ters which it possesses as an adult include: the petals, which are shallow to flush with the test
surface; the very small, simple pore pairs in the petals; and the lack of a sunken ambulacrum III,
which also has very small pore pairs.

This obvious structural plasticity in the two lineages is comparable with that observed in other
spatangoid and holasteroid lineages (McNamara and Philip 1980, 1984; McNamara 1982a, b, 1985;
Gale and Smith 1982; Smith 1984). Such plasticity has been interpreted (McNamara in press) as

330

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 8. Camera lucida drawings of the lateral profiles of four of the Australian Tertiary species
of Hemiaster {Bolbaster) and the juvenile and adult profiles of an Australian Paleocene species of
Hemiaster. They illustrate the paedomorphic change in lateral test profile from the Paleocene to Middle
Miocene and its relationship to the reduction in sediment grain size.

being caused by ‘plate translocation’. This involves the dissociation of growth allometries of each
coronal plate and the capacity of plates to resorb stereom, as well as grow, at their margins at
different rates, thereby providing a great range of morphotypes available for selection and sub-
sequent genetic stabilization.

Functional significance of the morphological evolution of the H. (Bolbaster) anr/ Psephoaster lineages.
Spatangoid echinoids are a highly successful group of echinoderms which are adapted to living on
and within a wide range of sediment types. Not only have some evolved a number of morphological

McNAMARA: TERTIARY ECHINOIDS

331

structures which allow them to live buried in coarse to fine-grained sediments, but they are also
adapted to feeding directly from the sediment. Consequently, any unidirectional morphological
trends occurred in response to selection for test morphologies which are better adapted either to
inhabiting, or feeding from, a different sediment grain size.

A number of spatangoid evolutionary lineages have been described. The morphological evolution
in all these examples was directed toward the occupation of finer-grained sediments. Thus, morpho-
logical changes along the S. (Paraster)-S. (Schizaster) lineage have been interpreted (McNamara
and Philip 1980) as adaptations to burrowing in progressively finer-grained sediment along a
peramorphocline. The morphological changes involved: swelling of interambulacrum V aborally; a
posterior migration of the apical system; decreased petal depth; increased depth of ambulacrum III
and increase in the number of pore pairs; slightly anterior migration of the peristome; and increased
development of an anterior rim on the labrum. These changes are thought to have allowed descen-
dant species of 5. {Schizaster) to occupy finer-grained sediments than their ancestors. This necessi-
tated the development of structures which were adapted to optimum utilization of the water for
respiration, such water being restricted to fiowing down the funnel connected to the surface by the
funnel-building tube feet of ambulacra III. Earlier species of the S. (Paraster) morphotype inhabited
coarser-grained sediments which allowed water to bathe the entire test surface. The more impervious
nature of finer-grained sediments resulted in only those morphotypes with the ability to construct
mucus-lined funnels being able to inhabit such sediments.

Smith (1984) recently re-interpreted the evolution of the Cretaceous spatangoid Micraster in a
similar light. Nichols (1959a, b) had previously interpreted the morphological changes in Micraster
as adaptations to burrowing deeper in the sediment. However, Smith noted that changes such as
increased test height, posterior movement of the apical system, development of a keel in interambu-
lacrum 5, broadening of the test, deepening and increasingly tuberculate ambulacrum III, anterior
movement of the peristome, labrum enlargement, and increase in petal length (many of which occur
in the Schizaster and, as discussed below, in the H. (Bolbaster) and Psephoaster lineages) are all
adaptations by descendant morphotypes to inhabiting finer-grained sediments. Similar morphologi-
cal changes in the evolution of species of Pericosnms from the Australian Tertiary (McNamara and
Philip 1984) are also considered to be adaptations to inhabiting finer-grained sediments. Changes
in the structure of the phyllodal plates in Protenaster are similarly interpreted (McNamara 1985)
but, in this case, the adaptation was concerned largely with the organism’s ability to feed from a
finer-grained sediment.

Many of the evolutionary trends which occurred in the H. {Bolbaster) and Psephoaster lineages
are the same as those in the Schizaster, Micraster, and Pericosnms lineages, and for the same
reason— the occupation by descendant morphotypes of progressively finer-grained sediments. In
H. {Bolbaster) these adaptations included: broadening of the test, increase in test height, and
development of a keel in interambulacrum III (all of which improved water fiow over the test);
elongation of the labrum; and reduction in primary tuberculation in the petals and ambulacrum III
(allowing an increase in miliary currents and so improving ciliary current fiow). Water fiow over
the test was likewise improved in descendant species of Psephoaster by the increase in test height,
development of a keel in interambulacrum 5, deepening of the petals, and swelling of the aboral
interambulacra adapically.

Smith and Paul (1985) recently demonstrated how the ecophenotypic increase in test height in a
lineage of Discoides subucula corresponds to a decrease in sediment grain size. They interpreted the
higher test shape, resulting in the attainment of a more conical profile, as an adaptation for assisting
in preventing finer-grained sediment from falling through the shield of fine spines. Discoides, being
a holectypoid, lacked mucus-producing fascicles. The increase in test height in spatangoids is more
likely to have helped water fiow over the test and on to the respiratory tube feet-bearing petals at a
faster rate.

Analysis of the calcarenites which the five Australian Tertiary species of H. {Bolbaster) inhabited
reveals a progressive decrease in sediment grain size (text-fig. 9a), 98 % of sediment grains ranging
in diameter from 0-5-2-0 mm (1 % up to 6 mm) in the Late Eocene, to 0-05-0-2 mm (up to 0-5 mm)

332

PALAEONTOLOGY, VOLUME 30

in the Middle Miocene. The Eocene and Oligocene sediments are predominantly coarse-grained
bryozoal calcarenites. The finer-grained calcarenites of the Miocene deposits are composed largely
of foraminifers, with a minor component of fragments of echinoid tests and spines, and of molluscan
shells.

In studies of the relationship between grain size and permeability of both artificial (Fraser 1935;
Beard and Weyl 1973) and natural (Pryor 1973) sediments, it was shown that with decreasing grain
size there is a corresponding decrease in permeability. Although these authors also showed that
porosity may sometimes increase with decreasing grain size, other factors, such as sphericity and
angularity of the grains, play an important role in affecting sediment permeability. Sphericity of
grains is much lower in the Eocene and Oligocene sediments, which are dominated by fragments of
stick bryozoans. Fraser (1935) showed how permeability of sediments increases as particles depart
in shape from a perfect sphere. Permeability is higher because such irregular-shaped grains act as
bridges between other grains, resulting in looser packing of the sediment. Von Engelhardt and Pitter
(1951) demonstrated that looser packing results in higher sediment permeability.

The overall decrease in grain size from the Late Eocene to Middle Miocene of sediments inhabited
by species of H. (Bolbaster), combined with an increase in grain sphericity, is likely to have had a
greater impact on reducing overall sediment permeability than the small increase in sorting which
accompanied the decrease in grain size. Consequently, only descendant morphotypes with morpho-
logical adaptations suitable for inhabiting sediments of lower permeability were selected. In the
case of the H. {Bolbaster) lineage, the combination of paedomorphic and peramorphic features
which evolved were the most successful adaptations.

A parallel situation occurred in the Psephoaster lineage (text-fig. 9b), the sediment grains ranging
in diameter from 0-5 2 0 mm in the Late Eocene Tortachilla Limestone to OT-0-6 mm in the Late
Oligocene Jan Juc Formation and to 01-0-33mm in the Early Miocene Mannum Formation;
the sediments inhabited by the three species therefore diminished in grain size and increased in
sphericity.

In addition to the morphological trends common to H. {Bolbaster), Psephoaster, Schizaster,
Micraster, and Pericosmus, some have only been recognized in the H. {Bolbaster) and Psephoaster
lineages. Three significant trends which provide strong evidence for the morphological changes
reflecting adaptations to inhabiting finer-grained sediments are those involving the increase in
peripetalous fasciole width, change in tubercle density, and change in size and shape of the pore
pairs in the petals. All three character changes occur in both of these hemiasterid lineages.

It is likely that the reduction in density of aboral miliary tubercles, at the expense of the primary
tubercles, and the increase in width of the peripetalous fasciole are closely related. The increase in
fasciole width occurred by an increase in the density of tiny miliary tubercles which bear clavules.
As Chesher (1963, p. 560) noted for the spatangoid Moira, the clavules have a division of labour,
some parts producing mucus, others generating water currents from cilia arranged along the clavule
shaft. Increase in the density of mucus and current producing clavules will have benefited an echinoid
inhabiting fine-grained sediments. Smith (1984) observed a similar change in the distribution of
miliary tubercles on some later species of Micraster which inhabited finer-grained sediments than
their ancestors. Early species possess no aboral fasciole, only a dense array of miliary tubercles.
In some later species a peripetalous fasciole began to develop at the expense of the aboral miliaries.
Although Smith (19806) questioned how fasciole arrangement could be used to infer preferred
substrata in fossil echinoids, the positive relationship between increasing peripetalous fasciole width
and decreasing sediment grain size in H. {Bolbaster) and Psephoaster indicates that fasciole width
may prove to be of some general significance in assessing substratum preference in spatangoids.
The reduction in the number of primary tubercles in the interporiferous zones of the petals and
ambulacrum III in later species of H. {Bolbaster), at the expense of miliary tubercles, also suggests
a further increase in current-generating spines to improve water flow over the respiratory tube feet.

The relationship between the concentration of aboral primary tubercles and decreasing sediment
grain size has been examined by Smith (19806); in spatangoids, he noted how the tubercle concen-
tration increased as the sediment grain size decreased. Both H. {Bolbaster) and Psephoaster follow

Aboral primary tubercle concentration Aboral primary tubercle concentration

McNAMARA: TERTIARY ECHINOIDS

333

A

H. (B.) callidus (M. Micrcene)

b-

9_

/■/. (B.) planedeclivis (Bb/M. Miocene)

H. (B.) verecundiis (Ib Miocene)
H. (B.) dolosus (fb Oligocene)

//. (B.) siibidus (L. Eiocene)

I I I

0.5 1.0 1.5

Sediment grain size

A

2.0

8-

6-

9_

B

P. klydonos (Ib Miocene)

P. ap ^^ypfu^ (L. Oligocene)

P. lissos (L. Eocene)

— I —

0.5

1.0 1.5

Sediment grain size

A

— I —

2.0

TEXT-FIG. 9. Relationship of aboral primary tubercle concentration to sediment grain
size. A, in five species of Hemiaster (Bolbaster). b, in three species of Psephoaster.

334

PALAEONTOLOGY, VOLUME 30

this relationship (text-fig. 9). The increase in tubercle concentration is particularly evident between
the peripetalous fasciole and the ambitus, reflecting an increase in aboral spine concentration.
Aboral spines are used by burrowing spatangoids to transport sand over the aboral surface of the
test (Smith 1980(?). An increase in concentration of aboral spines is a natural corollary of inhabiting
a finer grained sediment, as a denser concentration is necessary to effectively transport smaller
sediment grains. A similar relationship of high aboral spine density with fine-grained sediments has
been observed in archiaciid cassiduloids by Smith and Zaghbib-Turki (1985).

In the petals of H. (Bolbaster), the paedomorphic evolution of widely spaced round pores from
closely spaced elongate pores is yet another morphological feature which may be interpreted as an
adaptation to inhabiting finer-grained sediments. Smith (1980a) noted how pores which are widely
separated have a more efficient gas exchange system in the tube feet than those which are closely
spaced. The wider spacing reflects the presence of a larger ampulla with many septa. It may be
argued that the increased spacing of the pores along the H. {Bolbaster) lineage reflects the possession
of more efficient respiratory tube feet in those species which inhabited finer-grained sediments. In
Psephoaster, improvements in the respiratory capabilities of the aboral ambulacral tube feet oc-
curred by an increase in size, thus allowing a greater degree of oxygen absorption.

The morphological adaptations in all eight species of the H. {Bolbaster) and the Psephoaster
lineages indicate that they lived buried in the sediment. However, there is no direct evidence to
indicate to what depth they burrowed. The co-existence of these two, probably closely related,
hemiasterid genera suggests that niche partitioning between them may have occurred by differences
in depth of burial within the sediment. Although five living species of Hemiaster are known (Mor-
tensen 1950), including some such as H. expergitus which are morphologically very similar to some
species within the H. {Bolbaster) lineage, little of their ecology was known until recently, beyond
the fact that they generally occupy fine-grained sediments in relatively deep water (140-3,000 m).

Smith (1980a, p. 53) recognized the presence of funnel-building tube feet in H. expergitus. These
are poorly specialized tube feet bearing a broad, circular disc, with scalloped margins, supported
by a rosette of ten or eleven rods. Each tube foot is associated with a partitioned isopore. The
presence of funnel-building tube feet in ambulacrum III, combined with the possession of a peri-
petalous fasciole, suggest that H. expergitus burrowed to some depth within the sediment. This
prediction has recently been confirmed by Gage et al. (1985) who reported the recovery of a
specimen of H. expergitus from a box core sample in which it was lying in its burrow 12 cm
from the sediment surface. A long, slightly oblique funnel linked it to a shallow depression and
opening in the sediment. The specimen was found in a soft, fine-grained sediment. Gage et al. (1985)
suggested that the wide peripetalous fasciole of H. expergitus was capable of generating a strong
current within its burrow.

The principal difference between Hemiaster and Psephoaster lies in the character of aboral
ambulacrum III, which provides the possible key to the niche partitioning of the H. {Bolbaster) and
Psephoaster lineages. H. {Bolbaster), like all other members of the genus, possesses a sunken
ambulacrum III aborally. Within this depressed ambulacrum the pore pairs occur as isopores which
are divided by a prominent interporal partition and are very similar in morphology to the equivalent
structure in H. expergitus. This, combined with the presence of a peripetalous fasciole which is very
broad in later species, and other morphological adaptations consistent with optimizing water current
flow over the test in a relatively fine-grained sediment, points to H. {Bolbaster) having burrowed
quite deeply in the sediment (text-fig. 10), perhaps in the region of 10-12 cm below the sediment/
water interface, similar to H. expergitus.

Psephoaster, on the contrary, has neither a sunken aboral ambulacrum III nor isopores with a
swollen interporal partition. The pore pairs in ambulacrum III are extremely small and widely
spaced, with an interporal partition flush with the surface of the test; they were probably sensory,
not mucus-generating. The presence of a peripetalous fasciole which is relatively narrower in species
of Psephoaster than in co-existent species of H. {Bolbaster), and the probable absence of funnel-
building tube feet, suggests that species of Psephoaster were shallow burrowers, perhaps burrowing
only to sufficient depth to cover the aboral surface of the test (text-fig. 10).

McNAMARA: TERTIARY ECHINOIDS

335

EARLY

MIOCENE

LATE

EOCENE

A

intermediate morphotypes and
intermediate sediment grain size

II. (bolbaster)

A

B

n'^ ':6' T ^ ‘^8,0 (3 ^.ry ' '* “Q n. ^ O'

.UV-y-y-Cl.n :0°-^ryr\f^^ nv->_ V

o-OA
' AO'

P o’n'^0 Pc

o Vt?^

PS^O'O

',V br\-'-_- : Q )

'' '

M^g:Sft.pAo«<.r^WSJS^4

AoSAA- "

) r^'

:0Pol

P;P'

^A6Av5:

H. (Bolbaster)

A :'°;

Q'p'^p;
P”rr <0
■Qo^'O'C
.•py^ 0

■009-O0

,(PM&

^■JiV.^^Docoy

TEXT-FIG. 10. Suggested niche partitioning, a. Early Miocene species inhabiting fine-grained sediment, b. Late
Eocene species of Psephoaster (shallow burrower) and Hemiaster {Bolbaster) (deeper burrower) inhabiting

coarse-grained sediment.

336

PALAEONTOLOGY, VOLUME 30
SYSTEMATIC PALAEONTOLOGY

Order SPATANGOIDA Claus, 1876
Family HEMiASTERiDAE Clark, 1917
Genus hemiaster Agassiz in Agassiz and Desor, 1847

Type species. Spatangus btifo Brongniart, 1822, p. 84.

Discussion. Lambert and Thiery (1924) subdivided Hemiaster {s.l.) into seven sections: Hemiaster
{s.s.), Leymeriaster, Mecaster, Gregoryaster, Integraster, Bolhaster, and Holanthus. Lambert (1931)
later abandoned Integraster, regarding it as being synonymous with Hemiaster (5.5.), but introduced
another, Catoproctus. Fischer (1966) regarded the sections as subgenera, but relegated Catoproctus
to the class of ‘doubtful nominal genera’. The many described species of Hemiaster and their wide
range of morphologies makes the use of the seven subgenera reasonably effective. The only problems
arise in the assignment of some species to particular subgenera, on account of the artificial nature
of the generic subdivision. The Australian Tertiary species of Hemiaster {s.l.) described herein are
assigned to Bolhaster.

Subgenus bolbaster Pomel, 1869
Type species. Spatangus prunella Lamarck, 1816, p. 33.

Emended diagnosis. Test spherical to subspherical, with anterior notch very faint or absent.

Discussion. The type species of Bolhaster, H. {B.) prunella from the Maastrichtian of Maastricht,
southern Netherlands, was described by Lamarck in 1816, but figured earlier in his Tableau encyclo-
pedicpie et methodique des trois regnes de la nature (Lamarck 1798, pi. 158, figs. 3 and 4). It was later
figured by d’Orbigny (1856, pi. 881, figs. 2-4); d’Orbigny’s specimen (see Mortensen 1950, fig. 280)
has a faint anterior notch, this feature being used as the principal diagnostic character of the
subgenus, along with its subspherical form (Lambert and Thiery 1924, p. 505; Mortensen 1950,
p. 384; Fischer 1966, pp. U558-559). However, Lamarck’s (1798, pi. 158, figs. 3 and 4) illustrations,
and specimens of H. {B.) prunella which I have examined (collected by Dr J. Geys from the
Maastricht Chalk at Limburg, Belgium), have no anterior notch present at all; these specimens are
about one-third of the size of d’Orbigny’s specimen. As shown below, in species of Hemiaster (s.l.)
the anterior notch develops and deepens during ontogeny; thus its absence in juveniles and small
adults of H. (B.) prunella, but presence in larger adults are both diagnostic characters of the
subgenus. Mesozoic species of Hemiaster (s.l.) tend to be characterized by the presence of a relatively
deep anterior notch which is likely to have begun development at an early stage of ontogeny. The
delay in onset of development of the anterior notch in some Late Cretaceous and Tertiary species
of Hemiaster (s.l.) (post-displacement of Alberch et al. 1979; see also McNamara 1986) has resulted
in this character being paedomorphic in H. {Bolhaster).

The five species of Hemiaster (s.l.) described below likewise do not possess an anterior notch and
have a subspherical shape similar to the type species; they are therefore considered to belong in the
subgenus Bolhaster. The few species of Bolhaster which have previously been described (see Lambert
and Thiery 1924, p. 505) range from the Maastrichtian to the Paleocene. Inclusion of the five Eocene
to Miocene Australian species within Bolhaster therefore greatly extends the range of the subgenus.

Furthermore, it is considered that some of the living species referred to Hemiaster (s.s.), namely
H. expergitus Loven, 1874 and H. gihhosus Agassiz, 1879, which lack an anterior notch and are
very similar in overall morphology to the Australian Tertiary species, also belong in Bolhaster.

Hemiaster {Bolhaster) suhidus sp. nov.

Plate 44; Plate 48, fig. 8; text-figs. 3, 4, 5a, 6a, 8, 10, 11a, 12a

Diagnosis. Test ovoid; apical system well posterior of centre; petals and ambulacrum III very
shallow; pores narrow, elongate, and closely positioned within each row; peripetalous fasciole

McNAMARA: TERTIARY ECHINOIDS

337

narrow, 2-3 % TL in width. Labrum barely projected anteriorly; plastron relatively long. Posterior
surface of test vertical.

Material. Holotype SAM P26554 and paratypes SAM P26555, P26556 and NMV P20484, P5321 1, from Late
Eocene (Aldingan) Tortachilla Limestone, Maslin Beach-Port Willunga district, south of Adelaide, South
Australia.

5mm

TEXT-FIG. 11. Camera lucida drawings of adoral plating, a, Hemiaster (Bolbaster) subidiis sp. nov., SAM
P26554, holotype. b, H. (B.) planedeclivis Gregory, 1890, NMV P78460.

Description. Test reaching maximum known length of 38 mm; ovoid, maximum width anterior of centre; width
90-96 % TL; highest posterior of apical system close to posterior ambitus; height 71-75 % TL; posterior face
vertical (text-fig. 8); aboral surface gently declined anteriorly. Interambulacra weakly raised adapically. Apical
system slightly sunken and set 55-60 % TL from anterior ambitus. Ambital tubercle density 3-2 mm~^.
Ambulacrum III weakly depressed; parallel sided, width 6-7 % TL; bears up to eighteen isopores, pores within
each pair being aligned at about 45° and being separated by a prominent interporal partition. Petals very
shallow. Anterior pair diverge at c. 105°; steadily increase in width distally to be 9 % TL wide; slightly flexed
distally; length of each petal 28-32 % TL; bear up to twenty-seven pore pairs in each row, fewer in smaller
specimens; pores slit-like (PI. 44, fig. 4), pairs not conjugate; distance between pores in each row less than
length of pore (text-fig. 4); pores in anterior row slightly smaller than those in posterior row adapically.
Posterior petals also up to 9 % TL in width; diverge at c. 80°; length 15-17 % TL; bear up to nineteen pore
pairs, fewer in smaller specimens. Peripetalous fasciole not indented between petals; relatively narrow (PI. 48,
fig. 8), 2-3 % TL; shows little appreciable widening opposite tips of petals (text-fig. 5a).

Adoral surface gently convex. Peristome semicircular, entirely bordered by slightly raised rim; width 15-
1 6 % TL; moderately sunken; posterior situated 28-29 % TL from anterior. Labrum long; posteriorly it narrows
slightly before flaring toward plastron; relatively long, 12 % TL. Phyllode with seven isopores in ambulacra II
and IV, four in ambulacrum III, and five in ambulacra I and V; pores separated by prominent interporal
partition. Plastron relatively long, 46 % TL; width 33-36 % TL. Periproct oval, small, slightly sunken; length
10%TL.

338

PALAEONTOLOGY, VOLUME 30

A

B

lOmm lOmm

TEXT-FIG. 12. Camera liicida drawings of aboral surfaces, a, Hemiaster (Bolhaster) suhidus sp. nov., SAM
P26555, paratype. b, H. (B.) callidus sp. nov., NMV P100503, holotype.

Discussion. H. {Bolhaster) suhidus is the oldest of the five known species of this subgenus. It is a
rare component of the rich Late Eocene Tortachilla Limestone fauna of South Australia. Compari-
son with other species of H. {Bolhaster) is made under those species. H. {B.) suhidus can be
distinguished from the type species, H. {B.) prunella, on the basis of its more posterior apical system,
relatively shorter petals, narrower test, and larger peristome. H. {B.) suhidus is similar to H. integer
Lambert (1933, pi. 3, figs. 5 and 6) from the early Turonian of D’Antantiloky, Madagascar, but
can be distinguished by its narrower ambulacrum III, more posterior apical system, and more
evenly sloping aboral surface. H. integer is herein considered to belong in Bolhaster. H. {B.) suhidus
also compares with H. madagascariensis Cottreau from the Late Maastrichtian of Madagascar
(Besairie 1930, pi. 26, fig. 12), but can be distinguished by its more posterior apical system and
longer, broader petals. H. madagascariensis also belongs in Bolhaster.

Hemiaster {Bolhaster) dolosus sp. nov.

Plate 45, figs. 1-3, 6; text-figs. 4 and 8

Diagnosis. Species of Bolhaster with nearly vertical posterior surface; nearly circular outline; moder-
ately impressed, distally flared anterior petals; relatively deeply impressed, broad ambulacrum III;
and short plastron.

EXPLANATION OF PLATE 44

Figs. 1 -5. Hemiaster {Bolhaster) suhidus sp. nov. 1 -3, SAM P26554, holotype. 4 and 5, SAM P26555, paratype;
both from Late Eocene (Aldingan) Tortachilla Limestone, Maslin Beach-Port Willunga district, south of
Adelaide, South Australia, All x 2 except fig. 1 ( x 2- 1 ) and fig. 4 ( x 5).

PLATE 44

McNAMARA, Hemiaster (Bolbaster)

340

PALAEONTOLOGY, VOLUME 30

Material. Holotype NMV P53172, from the Early Oligocene (Willungan) Ruwarung Member of the Port
Willunga Formation, Maslin’s Beach, Aldinga, South Australia. Paratypes SAM P26557-26559 and WAM
86.1206-1209 from the Port Willunga Formation in the sea cliffs at Seaford, South Australia. J. Murray
Lindsay (South Australian Department of Mines and Energy) has sampled Foraminifera from the section
and reports that it is Early Oligocene (planktic foraminiferal zones PI 9/20) in age. One further specimen,
SAM P221 1 1 from the Gambier Limestone, Mt Gambier, South Australia, is probably also a member of this
species.

Description. Test circular, reaching maximum known test length of 37 mm; maximum width at mid-test length,
being 93-98 % TL; highest mid-way between apical system and posterior ambitus; height 77-83 % TL; posterior
surface very slightly inclined; aboral surface moderately declined anteriorly. Interambulacra 2 and 3 moder-
ately raised adapically; other interambulacra only weakly raised adapically. Apical system slightly sunken, set
53-54 % TL from anterior ambitus. Ambulacrum III relatively strongly impressed, particularly adapically;
with slight adambital taper; width 7-8 % TL; bears up to seventeen isopores in each row; pores elongate and
within each pair separated by a distance equal to length of pore. Petals moderately impressed. Anterior pair
diverge at c. 115°; increase in width distally to be more than 10% wide; length of each petal 29-30% TL,
bearing up to twenty-four pore pairs in each row. Posterior petals slightly narrower than anterior pair; diverge
at c. 70°; length 13-18 % TL; bear up to fourteen pore pairs. Peripetalous fasciole follows similar path to that
in H. (B.) suhidm\ width 4 % TL, widening slightly opposite tips of petals.

Adoral surface moderately convex. Peristome semicircular, bordered by entire, raised rim; width 17 %TL;
slightly sunken; posterior situated 35 % TL from anterior. Labrum long; constricts to half width at about
mid-length, then widens slightly toward plastron. Phyllode with eight isopores in ambulacra I and IV, four in
ambulacrum III, and six in ambulacra I and V. Plastron relatively short, 43 % TL; width 36 % TL. Periproct
subcircular, small, situated high on posterior face; barely sunken; maximum diameter 9 % TL.

Discussion. H. (Bolbaster) dolosus occurs in the lower part of the Port Willunga Formation, the
base of which occurs 33 m above the top of the Late Eocene Tortachilla Limestone, within which
H. {B.) siibidus occurs. H. {B.) dolosus can be distinguished from the older species by its more
circular test, which is relatively higher and has a slightly more anteriorly inclined posterior face
(text-fig. 8). It is further distinguished by its deeper ambulacrum III and deeper petals, straighter
anterior petals, less divergent posterior petals, possession of keeled anterior interambulacra adapi-
cally, slightly wider peripetalous fasciole, more posteriorly positioned peristome, less constricted
labrum, and more swollen plastron.

H. (B.) dolosus can be distinguished from the similar H. {B.) integer from the early Turonian of
Madagascar (Lambert 1933) by its more posteriorly situated apical system, more evenly declined
aboral surface, and more swollen plastron. H. {B.) dolosus differs from H. (5.) madagascariensis
from the Late Maastrichtian of Madagascar (Besairie 1930) in its longer, more distally flared petals
and wider ambulacrum III. In its circular test, H. {B.) dolosus compares with H. (B.) prunella
Lamarck, 1816, from the Maastrichtian at Maastricht, but is distinguished by the complete absence
of a frontal notch in large adult specimens, its broader petals, less divergent posterior petals, and
possession of a peristome further from the anterior ambitus.

Hernias ter {Bolbaster) verecundus sp. nov.

Plate 45, figs. 4 and 5; Plate 46, figs. 1 and 2; text-figs. 4 and 8

Diagnosis. Species of H. {Bolbaster) with short, broad petals; anterior pair diverge at 100°; pores in

EXPLANATION OF PLATE 45

Figs. 1-3, 6. Hemiaster {Bolbaster) dolosus sp. nov. 1, 3, 6, NMV P53172, holotype, from the Early Oligocene
(Willungan) Ruwarung Member of the Port Willunga Formation, Maslin Beach, Aldinga, South Australia.
2, SAM P26557, paratype, from same horizon as holotype, in the sea cliffs at Seaford, South Australia.

Figs. 4 and 5. H. {B.) verecundus sp. nov. 4, NMV PI 8578, holotype; 5, NMV P78458, paratype; both from
the Early Miocene (Longfordian) Puebla Formation, Fisherman’s Steps, Torquay, Victoria.

All x2.

PLATE 45

McNAMARA, Hemiaster (Bolbaster)

342 PALAEONTOLOGY, VOLUME 30

petals slightly elongate; short poriferous ambulacrum III; and nearly centrally positioned apical
system.

Material. Holotype NMV PI 8578, from the Early Miocene (Longfordian) Puebla Formation, Fisherman’s
Steps, Torquay, Victoria. Paratypes NMV P18761, 20145, and 78458, from essentially the same locality and
horizon as the holotype.

Description. Test subcircular, reaching a maximum known test length of 30 mm; maximum width near mid-
test length, 97 % TF; apex mid-way between apical system and posterior ambitus; posterior face almost
vertical; aboral surface evenly and gently declined anteriorly. Apical system slightly sunken, set 51-54% TL
from anterior ambitus. Poriferous zone of ambulacrum III relatively short, 30-31 % TL; width 5% TL;
moderately incised, bearing up to fourteen isopores; interporal region swollen. Petals short; anterior pair 20-
22 % TL; relatively broad, 1 1 % TL; bearing up to seventeen pore pairs in each row; diverge at c. 100°; pores
slightly elongate; within each pair, pores separated from one another by a distance slightly greater than length
of pore; pores not conjugate. Posterior petals short, 13-14 % TL, bearing up to ten pore pairs in each row;
diverge at c. 70°. Peripetalous fasciole not indented between petals; 5-5 % TL in width; narrows slightly
between petals.

Adoral surface poorly known. Moderately convex. Plastron short and wide. Labrum projects quite strongly
across peristome; bordered by raised rim; posteriorly labrum widens steadily toward plastron. Nature of
phyllode unknown.

Discussion. H. (Bolhaster) verecundus compares with the Early Oligocene H. (B.) dolosus in its nearly
circular test outline, almost vertical posterior surface of the test, and relatively broad petals.
However, it can be distinguished by its much shorter petals, shorter poriferous zone of ambulacrum
III aborally, slightly more centrally placed apical system, marginally deeper petals, wider peripeta-
lous fasciole, and more widely separated, less elongate pores in each row of pore pairs in the petals.

H. (B.) verecundus is easily distinguished from the Late Eocene H. {B.) subidus by its wider,
shorter, deeper petals, more centrally placed apical system, less elongate test, wider peripetalous
fasciole, more widely spaced and less elongate pores in each pore pair in the petals, and deeper,
shorter poriferous tract of ambulacrum III aborally.

The Paleocene species H. {B.) hawkinsi Lambert, 1933, from Madagascar, is similar to H. (B.)
verecundus, but differs in its longer petals and poriferous tract of ambulacrum III. H. {B.) prunella
is also similar to H. (B.) verecundus, but the Australian species has less divergent, shorter, and wider
petals, as well as lacking the anterior notch in large adults.

Hemiaster (Bolbaster) planedeclivis Gregory, 1890

Plate 46, figs. 3-6; Plate 47, figs. 1 and 2; text-figs. 4, 6b, 8, 1 1b

1890 Hemiaster planedeclivis Gregory, pp. 488-489, pi. 14, figs. 6 and 7.

1891 Hemiaster planedeclivis Gregory; Tate, p. 277.

1892 Hemiaster planedeclivis Gregory; Bittner, pp. 366-367, pi. 2, fig. 4.

1914 Hemiaster planedeclivis Gregory; Chapman, p. 147, fig. 81a.

1924 Hemiaster {Integraster) planedeclivis Gregory; Lambert and Thiery, p. 504.

1946 Hemiaster planadeclivis [i/c] Gregory; H. L. Clark, p. 364.

Diagnosis. Posterior surface of test slightly inclined anteriorly. Petals moderately impressed, bearing
nearly circular pores; ambulacrum III narrow. Labrum projects quite strongly forward. Peristomial
margin with strongly indented raised rim.

EXPLANATION OF PLATE 46

Figs. 1 and 2. Hemiaster (Bolbaster) verecundus sp. nov., NMV P18761, paratype, from the Early Miocene
(Longfordian) Puebla Formation, Fisherman’s Steps, Torquay, Victoria, x 2.

Figs. 3-6. H. (B.) planedeclivis Gregory, 1890. 3-5, NMV P78461, neotype, from the late Early to early Middle
Miocene (Batesfordian/Balcombian) Morgan Limestone, Morgan, South Australia, x 2. 6, NMV P18016,
from the same horizon and locality as the neotype, showing detail of the peristome, x 10.

PLATE 46

McNAMARA, Hemiaster (Bolbaster)

344

PALAEONTOLOGY, VOLUME 30

Material. Gregory’s (1890) type specimen, said to be from Morgan, South Australia, was lodged in the Ipswich
Museum. However, it can not now be located. In order to provide taxonomic stability, a neotype NMV
P78461 (PI. 46, figs. 3-5), from the late Early to early Middle Miocene (Batesfordian/Balcombian) Morgan
Limestone at Morgan, South Australia, is chosen. This species is relatively common in the fine-grained
calcarenites of the Morgan Limestone in the region of Morgan on the Murray River, South Australia. NMV
PI 80 16 is from the same horizon and locality as the neotype. Six specimens (SAM P22111) are also known
from the upper siliceous horizon in the Gambier Limestone, Mt Schank, South Australia, preserved as flint
moulds.

Description. Test reaching a maximum length of 31 mm; ovoid to circular, maximum width slightly posterior
of centre, 88-100 % TL; highest mid-way between apical system and posterior surface, forming a keel; height
74-84 % TL; posterior face inclined forward at c. 10° to the vertical. Interambulacra 2 and 3 moderately raised
adapically. Interambulacra 1 and 4 slightly swollen in region of peripetalous fasciole. Apical system slightly
sunken, set 51-55 % TL from anterior. Ambulacrum III moderately depressed; width 5-7 % TL; bearing up
to twenty-one isopores. Petals moderately impressed. Anterior pair diverge at c. 1 15°; of even width for much
of that length; slightly flexed distally; length of each petal 24-28 % TL; bearing up to twenty-one sub-circular
pore pairs in each row; pores within each pair separated by a distance greater than width of pores; pores in
anterior row very much smaller than those in posterior row. Posterior petal slightly curved; diverge at c. 70°;
length 12-16 % TL; bearing up to fourteen pore pairs in each row, fewer in smaller specimens. Peripetalous
fasciole variable in width, being broadest at distal extremities of petals in middle of interambulacra, in other
words widest at suture lines between columns of plates; maximum width 5-7 % TL.

Adoral surface moderately convex. Peristome (PI. 46, fig. 6) lunate; entirely bounded by raised rim which is
often strongly indented; indentations consist of a narrow slit, corresponding to an intercolumnal suture line,
and a deeper, broader pit situated equidistantly between sutures. Labrum projects anteriorly to variable degree
and also ventrally; posteriorly constricted at one-third labral length from anterior, flaring toward plastron.
Phyllode with eight isopores in ambulacra II and IV, five in ambulacrum III, and six in ambulacra I and V.
Second plate of interambulacrum lb not generally bisecting first plate of interambulacrum 1 and second plate
of la; however in one specimen (NMV P78460) plate translocation has occurred resulting in bisection of these
plates (text-fig. 11b). Plastron nearly flat; length 43-51 % TL; width 34-38 % TL. Periproct small, circular;
diameter 8-1 1 % TL; not sunken.

Discussion. H. (Bolbaster) planedeclivis is most easily distinguished from all other species of H.
(Bolhaster) by the strongly indented nature of its peristomial margin. This is a most unusual feature
and is only found elsewhere, less well developed, in H. {B.) callidus. The notches in the margin
probably reflect attachment sites for the peristomial plates. Normally, in spatangoids, these have
only a membranous attachment. It seems that H. {B.) planedeclivis and, to a lesser degree, H. (B.)
callidus developed a notched peristomial margin, into which the distal margin of the peristomial
plates could fit. This would necessitate the presence of an enlarged process on the marginal peri-
stomial plates. Such a notched peristomial margin seems to be unique in spatangoids.

H. (B.) planedeclivis can further be distinguished from the older species H. {B.) subidus, H. {B.)
dolosus, and H. (B.) verecundus by its more anteriorly situated apical system, deeper petals, more

EXPLANATION OF PLATE 47

Figs. 1 and 2. Hemiaster (Bolbaster) planedeclivis Gregory, 1890, NMV P18016, from the late Early to early
Middle Miocene (Batesfordian/Balcombian) Morgan Limestone, Morgan, South Australia.

Figs. 3 and 10. Psephoaster lissos sp. nov., SAM P26560, holotype, from the Late Eocene (Aldingan) Kingscote
Limestone, Kingscote, Kangaroo Island, South Australia.

Figs. 4-8. H. (B.) callidus sp. nov. 4-6, NMV P100503, holotype, from the late Middle Miocene (Bairnsdalian)
Port Campbell Limestone, near the mouth of the Sherbrooke River, near Port Campbell, Victoria. 7, NMV
P100504, paratype, from same horizon and locality as holotype. 8, NMV P100505, paratype, from same
horizon and locality as holotype.

Figs. 9, 1 1, 12. P. apokryphos sp. nov., NMV P100507, paratype, from the Late Oligocene (Janjukian) Jan Juc
Formation between Bird Rock and Fisherman’s Steps, Torquay, Victoria.

All x2.

PLATE 47

McNAMARA, Heniiaster (Bolhasfer), Psephoaster

346

PALAEONTOLOGY, VOLUME 30

circular pores in the petals, more anteriorly inclined posterior surface of the test, broader peripetal-
ous fasciole, and more strongly projecting labrum. Furthermore, it has longer petals than H. {B.)
verecundus, a narrower ambulacrum III and more anteriorly positioned peristome than H. {B.)
dolosus, and much denser ambital tuberculation than H. (B.) subidus.

Of all the Australian species of H. {Bolbaster), H. (B.) planedeclivis is the most similar to H.
integer from the early Turonian of Madagascar (Lambert 1933). The Australian species can be
distinguished by its narrower ambulacrum III and slightly shorter petals. H. (5.) planedeclivis can
be distinguished from H. (B.) prunella by its shorter, less divergent posterior petals, more posterior
peristome, and more swollen aboral interambulacrum 5. H. (B.) planedeclivis is also quite similar
to H. (B.) madagascariensis (Besairie 1930), but has a more circular test outline and broader anterior
petals.

Hemiaster {Bolbaster) callidus sp. nov.

Plate 47, figs. 4-8; Plate 48, fig. 7; text-figs. 3, 4, 5b, 6c, 8, 10, 12b

Diagnosis. Test broad, with strongly anteriorly inclined posterior surface and swollen aboral inter-
ambulacra. Apical system central or slightly anterior of centre. Petals depressed. Peripetalous
fasciole very broad.

Material. Holotype NMV PI 00503 and paratypes NMV PI 00504, 100505, 100508, from the late Middle
Miocene (Bairnsdalian) Port Campbell Limestone near the mouth of the Sherbrooke River, near Port
Campbell, Victoria. Paratype NMV PI 00509 from the same horizon, 22 m above sea level. Amphitheatre,
east of Port Campbell, Victoria.

Description. Test reaching maximum test length of 35 mm; subcircular, often wider than long, width 97-103
% TL; highest point at prominent keel in interambulacrum 5; height 81-89% TL; posterior face inclined
anteriorly at about 15° to the vertical. Interambulacra all strongly raised aborally, reaching their apex close
to course of peripetalous fasciole; intercolumnal sutures run in depression. Ambital tubercle density 7-8 mm“^.
Ambulacrum III depressed; narrow, width 6-7 % TL; maximum number of pore pairs unknown; floor covered
by dense concentration of fine tubercles. Petals deep. Anterior pair diverge at 115°; steadily increase in width
distally; length of each petal 24-26 % TL; bearing up to twenty pore pairs in each row; pores almost circular
and, within each row, widely separated, interporal distance being about twice diameter of pores. Posterior
petals diverge at c. 80°; length 14-15 % TL; bearing up to twelve pore pairs in each row. Peripetalous fasciole
(PI. 48, fig. 7) very broad at intercolumnal sutures, particularly at distal extremities of petals (text-fig. 5),
ranging from 7 to 10 % TL in width. Ambital spines 2-3 mm long, with flattened distal tips. Subanal spines
5 mm long, with broad, spatulate tips.

Adoral surface moderately convex. Peristome slightly sunken; with raised rim, weakly notched, marginal
rim width 14-17 % TL; situated 31-40 % TL from anterior ambitus. Labrum anteriorly accuminate; narrows
slightly posteriorly, then flares broadly to plastron. Plastron gently convex; relatively broad, length 39-46 %
TL, width 37-42 % TL. Periproct small, circular, with diameter 9-10 % TL; not sunken.

Discussion. H. (Bolbaster) callidus can be distinguished from the other species in this subgenus by
its high, wide test with anteriorly inclined posterior surface, swollen aboral interambulacra, very
broad peripetalous fasciole, deeper petals, more widely separated pores within each petaliferous
row, and more anteriorly positioned apical system. In having relatively short petals, it resembles
the Early Miocene (Longfordian) H. (B.) verecundus. It can be distinguished, however, by its deeper
petals and deeper ambulacrum III, more anteriorly positioned apical system, and swollen aboral
interambulacra.

The anteriorly projecting labrum of H. (B.) callidus resembles that of H. (B.) planedeclivis, but
the more posteriorly situated peristome of H. (B.) callidus distinguishes the two species. In this
respect H. (B.) callidus resembles the Early Oligocene H. (B.) dolosus. However, the nature of the
aboral surface is sufficient to distinguish the two taxa. The near spherical test of H. (B.) callidus
resembles that of H. (B.) prunella. The two species can be distinguished by the relatively shorter,
deeper petals of H. (B.) callidus, its deeper ambulacrum III, more swollen adapical interambulacra,
and less swollen, non-tuberculate labrum.

McNAMARA: TERTIARY ECHINOIDS

347

Of all the Australian Tertiary species of H. (Bolhaster), H. (B.) callidus most closely resembles
the living Indo-West Pacific species H. (B.) gihbosus (Agassiz 1879, pi. 20, figs. 5-16) and the
Atlantic species H. {B.) expergitus (Mortensen 1907, pi. 2, figs. 1,4, 18, 20). H. (B.) callidus can be
distinguished from these species by its more central apical system, broader petals, and narrower
peristome.

Genus psephoaster gen. nov.

Type species. Psephoaster klydonos sp. nov.

Diagnosis. Test subspherical without anterior notch. Ambulacrum III not sunken at all aborally,
with much reduced pore pairs. Petals narrow, flush with test surface or slightly depressed; pores
circular, pairs not conjugate. Apical system ethmophract with four genital pores.

Discussion. Psephoaster can be distinguished from Hemiaster (s.l.) by the absence of a sunken
ambulacrum III and anterior notch, and by the single, parallel sided petals which are flush with the
test, or almost so. The only other hemiasterid with such simple aboral ambulacra is the Late
Cretaceous Vomeraster (see Mortensen 1950, p. 407). However, Psephoaster can be distinguished
by its more truncate posterior margin, shallower, more parallel-sided petals, and lack of intersutural
depressions (a distinctive character of Vomeraster).

Psephoaster lissos sp. nov.

Plate 47, figs. 3 and 10; text-figs 7a, 10, 13a

Diagnosis. Species of Psephoaster with relatively low test; apical system set well posterior of centre;
very narrow, slightly sunken petals with extremely small pore pairs.

Material. Holotype SAM P26560, from the Late Eocene (Aldingan) Kingscote Limestone, Kingscote, Kanga-
roo Island, South Australia. Paratype SAM P26561, from the Late Eocene (Aldingan) Tortachilla Limestone,
Maslin Beach, south of Adelaide, South Australia.

Description. Test reaching maximum known length of 24 mm; ovoid, maximum width central; width 84-90 %
TL; highest point midway between apical system and posterior; height 66 % TL; posterior face inclined slightly

5

mm

5mm

5mm

TEXT-FIG. 13. Camera lucida drawings of aboral surfaces, a, Psephoaster lissos sp. nov., SAM P26560, holo-
type. B, P. apokryphos sp. nov., NMV PI 00506, holotype. c, P. klydonos sp. nov., SAM P24631, holotype.

348

PALAEONTOLOGY, VOLUME 30

anteriorly; aboral surface gently inclined anteriorly in posterior half progressively steepening anteriorly. Apical
system slightly sunken and set 55 % TL from anterior ambitus. Ambital tubercle density 3 mm“^. Ambulacrum
III flush with test surface; narrow, width 5 % TL; bears about twelve minute pore pairs, pores within each
pair being aligned almost exsagitally. Petals slightly depressed, very narrow, 2-3 % TL. Anterior pair diverge
at about 120°; straight, parallel-sided; length 35 % TL; bearing up to twenty minute pore pairs, the distal four
or five of which are extremely reduced in size; pores circular, pairs not conjugate; rows separated by a distance
equal to about three times pore pair widths. Posterior petals diverge at about 75°; length 24 % TL; bearing up
to fourteen pore pairs, similar in size to those of anterior pair, and similarly degenerating distally before
reaching peripetalous fasciole. Peripetalous fasciole not indented between petals; narrow, 1-5 % TL.

Adoral surface poorly known. Peristome situated 23 % TL from anterior ambitus. Plastron strongly convex,
transversely. Periproct nearly circular, 14% TL in diameter; situated high on posterior surface. This surface
slightly depressed below periproct.

Discussion. Of the two known specimens of P. lissos only one, the holotype, has its aboral and
posterior surfaces preserved, while the paratype is largely covered by growths of bryozoans. How-
ever, sufficient details are preserved to indicate their conspecificity and close morphological simi-
larity. P. lissos occurs in the Late Eocene Kingscote and Tortachilla Limestones with the other
hemiasterid, PI. (B.) subidus. P. lissos can be distinguished by its finer aboral tuberculation, narrower
and longer petals which are hardly depressed, and much smaller pore pairs.

Pseplioaster apokryphos sp. nov.

Plate 47, figs. 9, 11, 12; Plate 48, figs. 4-6; text-figs. 7b and 13b

Diagnosis. Species of Pseplioaster with a test which has a strongly inclined, though flat, aboral
surface; petals and ambulacrum III not sunken; relatively long posterior petals; central apical
system; narrow plastron; prominent subanal rostrum.

Material. Holotype NMV P100506 and paratype NMV P100507 from the Late Oligocene (Janjukian) Jan Juc
Formation between Bird Rock and Fisherman’s Steps, Torquay, Victoria.

Description. Test reaching maximum known length of 20 mm; ovoid, maximum width slightly anterior of
centre, 83-87 % TL; highest point close to posterior ambitus in interambulacrum 4; height 67-71 % TL;
posterior face slightly overhung by aboral interambulacrum 5; aboral surface steeply inclined, but nearly flat,
until it plunges vertically to anterior ambitus. Apical system central at 49 % TL; not sunken. Ambital tubercle
density 5 mm~^. Ambulacrum III flush with surface of test, with indeterminate number of minute pore pairs.
Petals not depressed and poorly defined, aboral tuberculation on ambulacra of the same size and density
as on interambulacra. Anterior pair diverge at c. 110°; straight, parallel-sided, and relatively long; bearing
up to sixteen small single pore pairs which fail to reach peripetalous fasciole. Pairs separated by a distance
equal to one and a half to two times width of pore pairs. Posterior petals diverge at about 65°; length 28 % TL;

EXPLANATION OF PLATE 48

Figs. 1-3. Pseplioaster klydonos sp. nov. 1, SAM P24631, holotype, from the Early Miocene (Longfordian)
Mannum Formation, Murray River, South Australia, x 2. 2, WAM 86.296, paratype, from same horizon
and locality as holotype, x 2. 3, SAM P22017, paratype, from the same horizon as the holotype at

Punyelroo, South Australia, x 2.

Figs. 4-6, P. apokryphos sp. nov., NMV P100506, holotype, from the Late Oligocene (Janjukian) Jan Juc
Formation between Bird Rock and Fisherman’s Steps, Torquay, Victoria, x 2.

Fig. 7. Hemiaster (Bolbaster) callidus sp. nov., NMV PI 00503, holotype, from the late Middle Miocene
(Bairnsdalian) Port Campbell Limestone, near the mouth of the Sherbrooke River, near Port Campbell,
Victoria, showing detail of dense aboral tuberculation and broad peripetalous fasciole, x 8.

Fig. 8. H. (B.) subidus sp. nov., SAM P26555, paratype, from Late Eocene (Aldingan) Tortachilla Limestone,
Maslin Beach-Port Willunga district, south of Adelaide, South Australia, showing detail of sparse aboral
tuberculation and narrow peripetalous fasciole, x 8.

PLATE 48

McNAMARA, Psephoaster, Hemiaster (Bolbaster)

350 PALAEONTOLOGY, VOLUME 30

bearing up to twelve small pore pairs in each row. Peripetalous fasciole narrow, 2-5 % TL; situated quite close
to ambitus.

Adoral surface strongly convex transversely, with prominent subanal rostrum. Peristome small, width 16-
17 % TL, situated 22-23 % TL from anterior ambitus; slightly sunken. Labrum slightly projecting anteriorly;
long, parallel-sided. Nature of phyllode not known. Plastron narrow, width 43 % TL. Posterior face of test
depressed below periproct, depressed area narrowing adorally. Periproct situated high on posterior face,
slightly wider than long, 10-12 % TL.

Discussion. The Late Oligocene P. apokryphos can be distinguished from the Late Eocene P. lissos
by its flatter and more steeply inclined aboral surface, narrower test, less conspicuous petals with
slightly larger pore pairs, more central apical system, longer posterior petals, wider peripetalous
fasciole, and small periproct.

In the same cliff section near Torquay, Victoria, the hemiasterid H. {B.) verecundus occurs in the
Early Miocene Puebla Formation, which conformably overlies the Jan Juc Formation. P. apokry-
phos differs in its narrower test and inconspicuous, though longer, petals and ambulacrum III.

Psephoaster klydonos sp. nov.

Plate 48, figs. 1-3; text-figs. 7c, 10, 13c

Diagnosis. Species of Psephoaster with high test, slightly inflated interambulacra adapically; slightly
sunken, relatively broad petals; apical system anterior of centre; relatively broad peripetalous
fasciole and short posterior petals.

Material. Holotype SAM P24631 and paratypes SAM P565, P8933, and WAM 86.296, from the Early Miocene
(Longfordian) Mannum Formation, Murray River, South Australia. Paratypes SAM P22017 from the same
horizon at Punyelroo, South Australia and NMV P13167 from the same horizon near Morgan, South
Australia. Other material: three specimens from the Mannum Formation north of Younghusband, and from
Mannum, South Australia, in the private collection of E. and F. Holmes.

Description. Test reaching a maximum known length of 23 mm; ovoid, maximum width central; width 85-
96 % TL; highest point midway between apical system and posterior; height 79-87 % TL (although holotype
is only 72 % TL); posterior face vertical and slightly overhung by keel in interambulacrum 5; aboral surface
steeply inclined. Apical system slightly sunken and set 44-49 % TL from anterior ambitus. Ambital tubercle
density 7-5 mm~^. Ambulacrum III flush with test surface; narrow, width 4-5 % TL; bears up to eleven very
small pore pairs, pores within each pair being aligned 30° to exsagittal line. Petals slightly depressed, narrow,
5 % TL. Anterior pair diverge at c. 120°; straight parallel-sided, open distally; length 31-38 % TL, bearing up
to nineteen small pore pairs; pores circular, not conjugate; rows separated by a distance equal to width of
pore pairs. Posterior petals diverge at about 75°; length 17-19 % TL; bearing up to thirteen pore pairs, similar
in size to those of anterior pair. Peripetalous fasciole relatively broad; wider at coronal sutures than in centre
of plates, reaching up to nearly 6 % TL in width.

Adoral surface moderately arched transversely. Peristome situated 24-29 % TL from anterior ambitus;
lunate, width 15-20% TL; slightly sunken; has raised marginal rim. Phyllode with isopores, pores being
separated by prominent interporal partition; nine in ambulacra II and IV, seven in ambulacra I and V, and
five in ambulacrum III. Labrum moderately constricted, expanding posteriorly to plastron; projects strongly
anteriorly, more than halfway across peristome; marginal rim well developed. Plastron length 39-47 % TL,
width 34-36 % TL. Periproct small, 1 1 % TL in diameter; subanal area between periproct and projecting
subanal rostrum slightly depressed.

Discussion. P. klydonos differs from the Late Eocene P. lissos in its higher test, more anteriorly
eccentric apical system, broader petals, larger and more closely spaced pore pairs, more densely
tuberculate ambitus, broader peripetalous fasciole, and relatively smaller periproct. P. klydonos
differs from the Late Oligocene P. apokryphos in its higher test with more pronounced keel in
interambulacrum 5, more sunken and shorter petals, longer pore pairs, slightly more eccentric
apical system, broader fasciole, wider plastron, less strongly developed subanal rostrum, and more
prominent labrum.

McNAMARA: TERTIARY ECHINOIDS

351

REFERENCES

ABELE, C., GLOE, C. S., HOCKING, J. B., HOLDGATE, G., KENLEY, P. R., LAWRENCE, C. R., RIPPER, D. and THRELFALL,

w. F. 1976. Tertiary. In douglas, j. g. and ferguson, j. a. (eds.). Geology of Victoria. Spec. Pubis geol.
Soc. Aust. 5, 177-274.

AGASSIZ, A. 1879. Preliminary report on the ‘Challenger’ Echini. Proc. Am. Acad. Arts Sci. 14, 190-212.
AGASSIZ, L. and desor, e. 1847. Catalogue raisonne des echinides. Annis Sci. nai. Zool. (ser. 3), 7, 129-168; 8,
1-35, 355-380.

alberch, p., GOULD, s. J., OSTER, G. F. and WAKE, D. B. 1979. Size and shape in ontogeny and phylogeny.
Paleobiology^ 5, 296-317.

beard, d. c. and weyl, p. k. 1973. The influence of texture on porosity and permeability of unconsolidated
sand. Bull. Am. Ass. Petrol. Geol. 57, 349-369.

BESAIRIE, H. 1930. Recherches geologiques a Madagascar. Bidl. Soc. Hist. nat. Toulouse. 60, 345-612.

BITTNER, A. 1892. Uber Echiniden des Tertiars von Australian. Sber. Akad. Wiss. Wien {Abt. I), 101 (1),
331-371.

BRONGNiART, A. 1 822. Description geologique des environs de Paris, 428 pp. Paris.

CHAPMAN, F. 1914. Australasian fossils, 341 pp. Robertson, Melbourne.

CHESHER, R. H. 1963. The morphology and function of the frontal ambulacrum of Moira atropos (Spatangoida).
Bull. mar. Sci. Gulf Caribb. 13, 549-573.

CLARK, H. L. 1917. Hawaiian and other Pacific Echini. Spatangina. Mem. Mus. comp. Zool. Harv. 46 (2),
81-283.

1946. The echinoderm fauna of Australia. Its composition and its origin. Pubis Carnegie Instn, 566,

1-567.

CLAUS, c. 1876. Griidzuge der zoologie (3rd edn.) 1. Marburg and Leipzig.

d’orbigny, a. 1856. Echinodermes. In Paleontologie Francaise, series 1 (Terrains Cretaces), 6, 1-598.

FISCHER, A. G. 1966. Spatangoida. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part U. Echino-
dermata 3, Asterozoa-Echinozoa, pp. U543-U633. Geological Society of America and University of Kansas
Press, New York and Lawrence, Kansas.

FRASER, H. J. 1935. Experimental studies of the porosity and permeability of clastic sediments. J. Geol. 43,
910-1010.

GAGE, J. D., BiLLETT, D. s. M., JENSEN, M. and TYLER, p. A. 1985. Echinodcrms of the Rockall Trough and
adjacent areas. 2. Echinoidea and Holothurioidea. Bull. Br. Mus. nat. Hist. [Zool), 48, 173-213.

GALE, A. s. and SMITH, A. B. 1982. The palaeobiology of the Cretaceous echinoids Infulaster and Hagenowia.
Palaeontology, 25, 1 1-42.

GREGORY, J. w. 1890. Some additions to the Australian Tertiary Echinoidea. Geol. Mag. 27, 481-492.
LAMARCK, J. B. p. 1798. Tableau encyclopedique et metliodique des trois regnes de la nature. 2, Vers, coquilles,
mollusques et polypiers. Agasse, Paris.

1816. Histoire naturelle des animaux sans vertebres. HI (Echinides). Paris.

LAMBERT, J. 1931. Etude sur les echinides fossiles du Nord de L’Afrique. Mem. Soc. geol. Fr. 7, 1-228.

1933. Echinides de Madagascar communiques par M. H. Besairie. Annis geol. Serv. Mines Madagascar,

3, 1-49.

and THIERY, p. 1909-1925. Essai de nomenclature raisonnee des Echinides. Ferriere, Chaumont.

LINDSAY, J. M. 1985. Aspects of South Australian Tertiary foraminiferal biostratigraphy, with emphasis on
studies of Massilina and Subbotina. Spec. Pubis S. Aust. Dept. Mines Energy, 5, 187-231.

MCKINNEY, M. L. 1984. Allometry and heterochrony in an Eocene echinoid lineage: morphological change as a
by-product of size selection. Paleobiology, 10, 407-419.

MCNAMARA, K. J. 1982u. Taxonomy and evolution of living species of Breynia (Echinoidea: Spatangoida) from
Australia. Rec. W. Aust. Mus. 10, 167-197.

19826. Heterochrony and phylogenetic trends. Paleobiology, 8, 130-142.

1985. Taxonomy and evolution of the Cainozoic spatangoid echinoid Protenaster. Palaeontology, 28,

311-330.

1986. A guide to the nomenclature of heterochrony. J. Paleont. 60, 4-13.

In press. Plate translocation in spatangoid echinoids: its morphological, functional and phylogenetic

significance. Paleobiology.

and PHILIP, G. M. 1980. Australian Tertiary schizasterid echinoids. Alcheringa, 4, 47-65.

1984. A revision of the spatangoid echinoid Pericosmus from the Tertiary of Australia. Rec. W.

Aust. Mus. 11, 319-356.

352

PALAEONTOLOGY, VOLUME 30

MCNAMARA, K. j. and PHILIP, G. M. 1987. Paleoccnc spatangoid echinoids from the Giralia Range, Western
Australia. Rec. W. Aust. Mus. 13.

MORTENSEN, T. 1907. Echinoidea II. Dan. Ingolf Exped. 4 (2), 1-200.

1950. A monograph of the Echinoidea, 5 (7), Spatangoida II. Reitzel, Copenhagen.

NICHOLS, D. 1959n. Changes in the Chalk heart-urchin Micraster interpreted in relation to living forms. Phil.
Trans. R. Soc. B242, 347-437.

19597?. Mode of life and taxonomy in irregular sea-urchins. Spec. Pubis Syst. ^^5. 3, 61-80.

POMEL, A. 1869. Revue des Echinodermes et de leur classification pour servir d' introduction a I etude des fossiles,
67 pp. Deyrolle, Paris.

PRYOR, w. A. 1973. Permeability-porosity patterns and variations in some Holocene sand bodies. Bull. Am
Ass. Petrol. Geol. 57, 162-189.

SMITH, A. B. 1980n. The structure, function, and evolution of tube feet and ambulacral pores in irregular
echinoids. Palaeontology, 23, 39-83.

19807). The structure and arrangement of echinoid tubercles. Phil. Trans. R. Soc. B289, 1-54.

1984. Echinoid palaeobiology. Allen & Unwin, London.

and PAUL, c. R. c. 1985. Variation in the irregular echinoid Discoides during the early Cenomanian. Spec.

Pap. Palaeont. 33, 29-37.

and ZAGHBiB-TURKi, D. 1985. Les Archiaciidae (Cassiduloida Echinoidea) du Cretace Superieur de Tunisie

et leur mode de vie. Annls Paleont. 71, 1-33.

TATE, R. 1891. A bibliography and revised list of the described echinoids of the Australian Eocene, with
descriptions of some new species. Trans. Proc. R. Soc. S. Aust. 14, 270-282.

VON ENGELHARDT, w. and PiTTER, H. 1951. Uber die Zusammenhange zwischen Porositat, und Korngrosse bei
Sanden und Sandsternen. Heildelb. Beitr. Miner. Petrogr. 2, 477-491.

KENNETH J. MCNAMARA
Western Australian Museum

Typescript received 18 February 1986 Francis Street Perth

Revised typescript received 1 July 1986 Western Australia 6000

EVOLUTION AND PHYLOGENETIC
CLASSIFICATION OF THE DIPLOGR APT ACE A

by CHARLES E. MITCHELL

Abstract. The complex astogenetic patterns produced by the specialized, first few thecae (primordial thecae)
of graptoloid rhabdosomes were conserved during evolution and provide a reliable guide to 'propinquity of
descent’ among diplograptids. Using this principle one can redefine their taxonomy, establish phylogenetically
meaningful higher taxa, and obtain an improved understanding of diplograptacean phylogeny. The Diplograp-
tacea comprises four major subclades. 1, Orthograptidae: archaic orthograptids (species of the "Glyptograptus'
teretiusciiliis species group), Orthograptus and Amplexograplus, together with archiretiolitids, lasiograptids,
and "Climacograptus' typicalis-G.' lorratnensis groups; 2, Dicranograptidae: dicranograptids plus nemagrap-
tids; 3, Diplograptidae; pseudoclimacograptids and Climacograptus s.s., together with Diplograptus s.s. and
offshoots; 4, Monograptidae: ‘G.’ dentatus and descendants including Undulograpfus paradoxus, G. euglyphus,
‘C.’ uonnalis, and all the Silurian diplograptids including the uniserial monograptines. The first three families
dominated Ordovician faunas. Taxa with complex proximal end structures were succeeded during the Llandeilo
and Caradoc by taxa with simpler astogenies. Following extinction of the dominant Ordovician taxa, mono-
graptids (sensii lato) underwent explosive evolution in the Llandovery. New generic group taxa; Arclii-
climacograptus, Arulieimograptus, Diplacanthograptus, Eoglyptograptus, Genicidograptus, Hustedograptus,
Oelaudograptus, Pseudamplexograptus, and Urbanekograptus. New family group taxa: Eoglyptograptinae and
Orthograptidae.

The first-formed few thecae of graptoloid colonies exhibit specialized ontogenies. These specialized
features are largely prothecal in origin and are associated with the formation of the primary stipes.
Elies (1922), and later Bulman (especially 1933u and 1936), grouped graptoloid astogenies into a
series of ‘developmental types’ distinguished from one another by budding pattern, direction of thecal
growth, and position of the dicalycal theca in the budding sequence. These general developmental
types were defined following the recognition that graptoloid astogeny displayed a limited
range of basic patterns and that individual patterns characterized large segments of the Grapto-
loidea.

The early astogeny of members of the Suborder Diplograptina Bulman, 1 970 is especially complex.
Each of the first several thecae exhibit unique features and ontogenetic patterns that are orchestrated
to establish the foundation of the rhabdosome. This complex orchestration is, in turn, repeated with
great precision among all members of a given species— a regularity not unlike that of the metathecal
morphoclines exhibited by monograptids (Bulman 1968, p. 1353;Urbanek 1973). Despite these seem-
ingly useful features, graptoloid (and particularly diplograptinid) astogeny has been largely ignored
in both systematic and phylogenetic studies of these organisms. Astogenetic pattern has been seen as
simply another of the many features of graptoloid colonies that underwent extensive parallel change
(Bulman 1933a, p. 2) and so bears no consistent relationship to taxonomy: '\J]hc Diplograptus type of
development . . . exhibits considerable modifications which occur indiscriminately in the various gen-
era and sub-genera [of the Diplograptidae] ’ (1 933a, p. 3). This treatment of astogeny has remained

the standard approach, e.g. Urbanek’s (1959, p. 326) discussion of Gymnograptus diSioggny and Rick-
ards et al.'s (1977, p. 23) discussion of the appearance of the monograptid condition.

Cooper and Fortey (1982, 1983), Kearsley (1982, 1985), and Mitchell (1981, 1986) have argued
for a different interpretation of astogenetic similarity. We have each independently concluded
that graptoloid astogeny shows a striking parallelism with von Baer’s Law, i.e. that primordial

(Palaeontology, Vol. 30, Part 2, 1987, pp. 353-405.)

© The Palaeontological Association

354

PALAEONTOLOGY, VOLUME 30

astogenetic features were not altered with great ease or frequency but rather were highly conserved
during the evolution of graptoloid colonial design. The features of early astogenetic stages and the
sicula were more refractory to change than were later stages in astogeny. In most cases, detailed
structural and developmental similarities in early astogeny among graptoloids are homologies. Ac-
cordingly, these can and should be used to determine evolutionary relationships among graptolites
and to establish a phylogenetic classification. I have presented a theoretical basis for this view, together
with detailed supporting evidence, elsewhere ( 1 986; see also Gould 1 977 for a discussion of von Baer’s
law).

There are now sufficient data available on the astogeny of the Diplograptina to permit an accurate
survey of the range of their developmental patterns. I believe that the distribution of these data
across the group is also sufficient to trace the outlines of the phylogenetic history of this complex
and interesting group and to begin the reorganization of the traditional diplograptinid form taxa
into more meaningful units.

DIPLOGRAPTINID ASTOGENETIC PATTERNS

Primitively, the Diplograptina exhibit an early astogenetic pattern in which each of the first four
thecae have specialized ontogenies. For convenience, we may refer to these specialized, first few
thecae of graptolite colonies as primordial thecae (adapting somewhat a term employed early in the
study of graptolites: see Holm 1895), and to that part of astogeny that encompasses the growth of
these thecae as the primordial astogeny. Among primitive diplograptinids th2^ is dicalycal and thG
to th2^ include crossing canals. This is essentially the definition of the Diplograptus developmental
type of Elies (1922). Bulman (1936) subdivided this pattern into a number of ‘stages’ (again seen
solely as grades of organization) as part of his study of graptoloid orthogenesis. While retaining
the general definition for the Diplograptus developmental type, with its emphasis on three crossing
canals, Bulman (1963a, 1970) later abandoned these ‘stages’ and simply recognized two grades of
organization among the range of diplograptinid developmental patterns: 1, the primitive strepto-
blastic condition in which thF is S-shaped and initially grows upwards from its origin; and 2, the
derived prosoblastic condition in which thF is J-shaped and the initial upward growth is lost. This
simple structural distinction does indeed appear to have been erossed repeatedly during the evolu-

TEXT-FiG. 1. Thecal diagrams of the diplograptid astogenetic Patterns A-I (letter designations are those used

to refer to these patterns throughout the text).

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSI LIGATION

355

lion of the Diplograptina. Consequently, I do not place any great emphasis on it in the definition
of the diplograptinid primordial astogenetic patterns, except to note that the streptoblastic condi-
tion is restricted to the more primitive members of the group (i.e. those with astogenetic Patterns
A, B, and C, see discussion of these patterns below). I believe these astogenetic patterns have
strict phylogenetic significance.

The Diplograptina and the Dicranograptidae (including Leptograptus: see Finney 1985) are
characterized by the unique left-handed origin of thP from thlF Fuselli from the obverse side of

TABLE I. Generic assignments of Ordovician ‘diplograptid’ species with known primordial astogeny (* = type

species).

Pattern A

HUSTEDOGRAPTUS gen. nov.: Diplograptus notahilis Hadding, D. propinqims Hadding, D. uplandicus Wiman*,
Glyptograptus teretiusculus sensu Jaanusson, G. vikarbyensis Jaanusson. oelandograptus gen. nov.: Glypto-
graptus americanus Bulman, G. austrodentatus Harris and Keble, G. oelandicus Bulman*, G. sinodentalus
Mu and Lee.

Pattern B

eoglyptograptus gen. nov.: Glyptograptus cermius Jaanusson, G. deutatus (Brongniart)*, PseudocUmaco-
graptus jaroslovi Boucek. undulograptus Boucek, 1973: Climacograptiis paradoxus Boucek*.

Pattern C

DiCAULOGRAPTUS Rickards and Bulman, 1965: D. hystrix (Bulman)*, D. cumdiscus Finney, diplograptus
M’Coy, 1850: D. foliaceus (Murchison), D. molestus Thorslund, D. pristis (Hisinger)*. prolasiograptus
Lee, 1963: Lasiograptus haplus Jaanusson. pseudamplexograptus gen. nov.: Amplexograptus coelatus (Lap-
worth), A. munimentus Berry, A. maxwelli Ekstrom, Climacograptus distichus (Eichwald)*, C. meridionalis
Ruedemann, Pseudoclimacograptus vestergothicus Jaanusson and Skoglund. pseudoclimacograptus Pfibyl,
1947: p. (archiclimacograptus) subgen. nov.: P. angulatus augulatus (Bulman), P. angulatus sehyeiisis
Jaanusson*, P. cumbrensis Bulman, P. eurystoma Jaanusson, P. klabavensis Boucek, P. luperus Jaanusson,
P. marathonensis Clarkson, P. modestus Ruedemann, P. oliveri Boucek. urbanekograptus gen. nov.:
Gymnograptus retioloides (Wiman)*.

Pattern D

p. (pseudoclimacograptus) Pfibyl, 1947: P. clevensis Skoglund, P. scharenbergi (Lapworth)*.
CLIMACOGRAPTUS Hall, 1865: c. (climacograptus): C. bicornis (Hall)*, C. caudatus Lapworth, C. putdius
(Hall), C. styloideus Lapworth.

Pattern E

CLIMACOGRAPTUS (diplacanthograptus) subgcn. nov.: C. dorotheus Riva, C. longispinus T. S. Hall, C.
spiniferus Ruedemann*, C. venustus Hsu.

Pattern F

ARNHEiMOGRAPTUS gen. nov.: Glyptograptus anacanthus Mitchell and Bergstrom*, G. hudsoni Jackson, G.
lorrainemis Parks, geniculograptus gen. nov.: Climacograptus inuiti Cox, C. pygmaeus Ruedemann, C.
typicalis Hall*, gymnograptus Bulman, 1953: G. lumarssoni (Moberg)*.

Pattern G

amplexograptus Lapworth, Elies and Wood, 1907: A. bekkeri (Opik), A. elongatus Barrass, A. fallax
Bulman, A. cf. fallax Jaanusson and Skoglund, A. leptotheca (Bulman), A. maxwelli Decker, hallograptus
Lapworth, 1876: H. bimucronatus (Nicholson)*, lasiograptus Lapworth, 1873: L. harknessi (Nicholson).
NEUROGRAPTUS Elles and Wood, 1908: N.l bulmani Strachan, N. margaritatus Lapworth*. orthograptus
Lapworth, 1873: O. amplexicaulis Hall, O. apiculatus (Elles and Wood), O. gracilis Roemer, O. quadrimu-
cronatus (Hall)*, O. ruedemanni Gurley, O. truncatus (Lapworth). orthoretiolites Whittington, 1954: O.
hami Whittington*, paraorthograptus Mu et al., 1974: Climacograptus pacificus Ruedemann*.
PEiRAGRAPTUS Strachan, 1954: P. fallax Strachan*. pipiograptus Whittington, 1955: P. liesperus
Whittington*.

Pattern H

GLYPTOGRAPTUS Lapworth, 1873: Climacograptus angustus (Perner), C. brevis Elles and Wood, C. brevis
mutabilis Strachan, C. kuckersianus Wiman, C. normalis Lapworth, C. rotuudatus Jaanusson and Skoglund,
Diplograptus toernquisti Hadding, Glyptograptus euglyplms Lapworth.

356

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 2. Camera lucida drawings of species exhibiting Pattern A and Pattern C primordial astogenies. All
are reverse views unless noted otherwise.

A, E-G, K, Hustedograptus (gen. nov.) uplandicus (Wiman), Kukruse Stage (Cae), Estonia. A, Cn 59915, note
origins of metatheca of th2' and protheca of th2^ from paired foramina near thP, x 20. e, Cn 59913, note
prosoblastic form of crossing canal of thF, x 20. f, Cn 59916, bleached and cleared specimen; note small
upward growing flange present in ontogeny of th2‘, and prominent paired antivirgellar spines (cf. Pattern G,
text-fig. 9), X 20. g, Cn 59914, proximal end showing well-defined patch in region of prothecae of th2*-th2^
that corresponds to exposed descending portion of crossing canal of th2‘, x 10-5. k, Cn 59917, obverse view;
note gradient in thecal form from glyptograptid with cuspate apertures to orthograptid with paired apertural
spines, x 5.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSILICATION

357

thl ^ swing around the theca and form a hood over the foramen in the reverse side of thF (see, for
instance, text-figs. 2e, 6b, 9i, j). The foramen of thP is completed, as is the contribution of thP to
the hood of thl^, when the next fusellus passes around thl * to meet with the reverse wall of thP or
the sicula, rather than continuing around the leading edge of the hood (e.g. text-figs. 2e and 9a, d).
Following the completion of the hood of thP, the crossing canal of thP begins to grow downward,
across the reverse side of the sicula. I employ the term ‘crossing canal’ to refer only to the sicula-
crossing, tubular prothecae of primordial thecae. Diplograptinid structures force this restriction of
the term because species with a comparatively simple primordial astogeny and a delayed dicalycal
theca (such as Amplexograptus bekkeri, in which th3* is dicalycal) could be said to have four or five
crossing canals, several of which differ in no significant structural way from later thecae. Hence,
the term ‘crossing canal’ retains greater meaning if restricted in its application to the prothecae of
the specialized early thecae that cross the sicula.

Beyond its initial origin the growth of thP includes several major variants among the Diplograp-
tina and, together with the ontogenetic variations exhibited by the crossing canals of th2* and th2^,
these variants define nine basic primordial astogenetic patterns. The exact position of the dicalycal
theca contributes very little to the distinctiveness of the patterns, however. In almost any given
pattern, some species possess a delayed dicalycal theca. However, the level of the dicalycal theca’s
earliest occurrence within each astogenetic pattern is of significance. Consequently, I do not use
Cooper and Fortey’s (1983, p. 171) two diplograptinid ‘developmental types’, which they recognized
on the basis of the position of the dicalycal theca.

The nine diplograptinid primordial astogenetic patterns are illustrated diagrammatically in text-
fig. 1 and have been designated ‘A’-‘F, roughly in stratigraphic order of first appearance, elsewhere
(Mitchell 1986). I have chosen not to name them after seemingly typical species or genera. None
of these patterns are invariant. To name them after a particular taxon promotes a stereotypic and.

B-D, j, Oelandograptus (gen. nov.) austrodentatus oelandicus (Bulman in Skevington 1965) exhibiting Pattern
A primordial astogeny; Holen Limestone, Kunda Stage (Hunderumian Substage, D. hirundo Zone), Halludden,
Oland. B, c, Cn 59911 (from horizon— 120D), obverse and reverse views; note quasi-symmetrical disposition
of primordial thecae, x 20. d, Cn 59892 (Holm Collection), showing streptoblastic crossing canal of thF and
left-handed origin of th2\ x 20. j, Cn 59891 (Holm Collection), note visible th2' descending crossing canal
and undulating median septum formed by successive prothecae, x 9.

H, R, Dicranograptus nicholsoni longibasalis Ruedemann and Decker, Viola Springs Formation (0-3 m above
base of section D; Alberstadt 1973), Rocklandian Stage (upper C. bicornis Zone), Arbuckle Mtns., Oklahoma.
H, MCZ 9461/1, young growth stage equivalent to d; note dorsal notch, lateral lappets, and paired notches
adjacent to virgella, x 14. r, MCZ 9461/2, obverse view showing dicranograptid sicula and prominent nema,
X 9.

I, N, Q, Pseudoclimacograptus (Archiclimacograptus subgen. nov.) angidatus sebyensis Jaanusson exhibiting
Pattern C primordial astogeny; Holm Collection, Folkeslunda Limestone, Lasnamagi Stage CG.'teretiuscidiis
Zone). I, Cn 59885, Gardslosa, Oland, specimen showing streptoblastic thF with origin of th2‘ from its right
side (specimen damaged subsequent to sketching), x 20. n, q, Cn 59803, Sjostorp, Oland, obverse and reverse
views, X 20.

L, M, Hustedograptus (gen. nov.) teretiusculus sensu Jaanusson, 1960, Cn 59886, Folkeslunda Limestone,
Lasnamagi Stage (H. teretiusculus Zone), Sjostorp, Oland (Holm Collection), obverse and reverse views; note
paired lappets on dorsal margin of sicula, prominent crossing canal of th2‘, and dicalycal th2^, x 20.

O, P, Pseudamplexograptus (gen. nov.) distichus (Eichwald), exhibiting Pattern C primordial astogeny;
Folkeslunda Limestone, Lasnamagi Stage (‘G.’ teretiusculus Zone), Lerkaka, Oland (Holm Collection), o, Cn
59922, showing formation of th2*-th2^ and foramen from which th3’ arises, p, Cn 59921, note right-handed
origin of th2’ from thF, both x 20.

Abbreviations: an, ancora; av, antivirgellar spines; cc, crossing canal; dt, dicalycal theca; fl, upward growing
flange; fo, foramen; la, lappets; Is, list scar; m, mesial spine; ms, median septum; s, sicula; p, protheca; pr,
prothecal rods; v, virgella. Repositories: BMNH, British Museum (Natural History), London; MCZ, Museum
of Comparative Zoology, Harvard University; Cn, Naturhistoriska Riksmuseet, Stockholm; SM, Sedgwick
Museum, Cambridge University; 01 and Vg, Paleontological Institute, Uppsala University; USNM, United
States National Museum, Smithsonian Institution, Washington.

358

PALAEONTOLOGY, VOLUME 30

at times, seriously distorted view of the astogenetic patterns. Finally, these patterns do not apply
to the Silurian retiolitids (although most of the archiretiolitids are encompassed by the scheme).
Table 1 indicates the astogenetic pattern of more than eighty diplograptinid species known in
relief or from isolated preparations, and for which I have been able to obtain data. The stratigraphic
range of the genera exhibiting these patterns is indicated in text-fig. 17.

Pattern A (text-fig. 2a-h, j-m, r)

The sicula is straight. Its aperture is commonly plain except for a prominent virgella. In a few
species the aperture is elaborated in the form of a pair of antivirgellar lappets or antivirgellar spines.
The crossing canal of thF is usually streptoblastic. ThP gives rise to the crossing canal of th2*
from its left side by bifurcation of a broad hood formed early in the ontogeny of thF (text-fig. 2d,
h). The crossing canal of th2‘ also grows downward, toward the virgella and along the reverse wall
of thl k An isolated fusellar flange forms near the sicular aperture in advance of the approaching
crossing canal of th2’, and grows upwards. They fuse and form a symmetrical pair of foramina
from which the metatheca of th2' arises on the biological left side, and of th2^ on the right (text-
fig. 2a). From this point, both thecae grow upward and surround the crossing canal of th2\ which
remains visible for a large part of its length. There are, thus, four primordial thecae and three
crossing canals (thF-th2^). Either th2^ or some later theca may be dicalycal.

In obverse view, both thP and thF diverge widely from the sicula, forming a blunt to broadly
rounded proximal end that is usually sub-symmetrical. Th2’ and th2^ enclose the sicula together
with subsequent thecae. In species exhibiting a Pattern A astogeny, the median septum may be
undulating to straight, and their post-primordial thecae range in shape from glyptograptid to
orthograptid. Species exhibiting Pattern A include the earliest known diplograptinids. This develop-
mental pattern is also found throughout the Dicranograptidae, as noted by Bulman (1970, pp.
V76-V78).

Pattern B (text-flg. 3a-e, i, j)

The sicula is straight to slightly deflected. The sicular aperture bears only a short stout virgella.
ThP may be prosoblastic or streptoblastic. The crossing canal of thF grows downward obliquely
across the sicula and away from thl k The third theca arises from thl^ on its left side, as in Pattern
A. Th2^ grows downward at first and then turns upward before giving rise to th2^. Th2^ arises by a
pattern of differentiation like that of distal thecae. The dicalycal theca may be or later. The
median septum may be undulating to straight. There are three primordial thecae and two crossing
canals (thF and th2*). The proximal end is asymmetric with the first two thecal apertures at
markedly different levels. In obverse view the sicula is exposed only to the level of the aperture of
thl* or thF. Thecal shapes among species exhibiting Pattern B range from glyptograptid to
climacograptid. This pattern is relatively poorly known: only the primordial astogeny of Glypto-
graptus dentatus (Brongniart) and Undulograptus paradoxus (Boucek) (= Climacograptus
pauperatus Bulman) are known in any detail (Bulman 1963a).

Pattern C (text-figs. 2i, n-q, 4a-o)

The sicula is straight and generally slender. With rare exceptions (e.g. Dicaidograptus hystrix
(Bulman), text-fig. 4), the sicular aperture bears only a virgella. The growth of thl^-th2^ is like that
seen in Pattern A, except that the crossing canal of th2* arises from the right side of thF. Th2* (or
rarely, th2^) is dicalycal. Among a large number of early species, th3* originates from a foramen in
the metatheca of th2' (text-figs. 2o, Q, 4e). This feature commonly produces what appears to be a
continuous arch connecting th2^ and th3' (see text-fig. 4l). Because the crossing canal of th2* must
swing out and away from the sicula a considerable distance to grow around thF, it is commonly
exposed as a diamond-shaped patch in the central region of the rhabdosome above the th2^-th3'
arch. This feature is exhibited clearly by Pseudoclimacograptiis oliveri Boucek and P. angulatus
(Bulman 1953, text-figs. \h and 2b, respectively, but note that the origin of th2* in his text-figs. \c

TEXT-FIG. 3. Camera lucida drawings of species exhibiting Pattern B and Pattern H primordial astogenies. See
text-fig. 2 for explanation of abbreviations and specimen repositories.

A-E, Undulograptus paradoxus (Boucek), Seby Limestone, Lasnamagi Stage {D. murchisoni Zone), Seby,
Oland. A, B, OI unnumbered, reverse and obverse views of early growth stage showing streptoblastic thC and
delayed origin of th2*, x 32. c, D, Ol 983, obverse and reverse views; note budding sequence with dicalycal
th2\ X 16. E, Ol unnumbered, reverse view; note paired thL and th2*, x 20.

F-H, K, L, Glyptograptus brevis (Elies and Wood), ‘'Climacograptus band’, Balclatchie beds, Caradoc Series
CD.' midtidens Zone), Laggan Burn, Ayrshire, Scotland, f, MCZ 9462/1, oblique reverse view showing list
that links free reverse wall of thC with sicula, x 32. g, MCZ 9462/2, reverse view; note origin of th2‘, x 26.
H, MCZ 9462/3, obverse view; note origin of th2^ by simple distal differentiation, x 32. k, l, MCZ 9462/4,
reverse and obverse views; specimen partly flattened distally to present sub-scalariform view, x 1 5.

I, J, Eoglyptograptus (gen. nov.) dentatus (Brongniart) sensu Bulman, 1963a, Holen Limestone, Kunda Stage
(D. hifidus Zone), -1-04- 15D, Hagudden, Oland. i, Cn 59937, note shape of proximal end and cuspate thecal
apertures, x 10. j, Ol 1228, early th2' stage; see text and Skevington (1965) for further discussion, x 34.

360

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 4. Camera lucida drawings of species exhibiting Pattern C primordial astogeny. See text-fig. 2 for
explanation of abbreviations and specimen repositories.

A-F, Pseudoclimacograptus {Archiclimacograptus subgen. nov.) eiirystoma Jaanusson, Folkeslunda Lime-
stone, Lasnamagi Stage (‘G.’ teretiusculus Zone?), Gardslosa, Oland. a-d, Cn 59921, oblique right-lateral,
reverse, oblique reverse, and left-lateral views, respectively, showing construction of right-handed crossing
canal of th2‘ and its origin from thU, x 27. e, f, Cn 59922, oblique reverse and reverse views; note final hood-
like form of crossing canal of th2‘ and origin of th2^ and th3‘, x 27.

G-i, Prolasiograptus haplus {Jaanusson), Folkeslunda Limestone, Lasnamagi Stage (‘G.’ teretiusculus Zone?),
Gardslosa, Oland. g, h, Cn 59925, obverse and reverse views; note exposed patch of crossing canal of th2^ in

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

361

and Ic is shown, incorrectly, as left-handed). This morphology appears to be distinctive of species
with Pattern C primordial astogeny.

Species exhibiting Pattern C commonly possess a zigzag median septum, but it may become
straight (or nearly so) after the first few thecae, as in ‘C’ distichus (Eichwald) (Bulman 1932, pi. 4,
figs. 24 and 25) or 'A.' munimentus Berry (1964, pi. 14, text-figs. 1-4). The proximal end of
rhabdosomes exhibiting Pattern C is evenly rounded to blunt and generally broad. In obverse view
the sicula is exposed only to the level of the aperture of thl^ or thF. Sharply geniculate thecae
predominate in this species group.

Pattern D (text-figs. 5, 6, 7a-c)

The sicula in this group is rather broad for its length, and its axis is usually strongly deflected
toward the dorsal side of the sicula. The aperture bears only a virgella. The metasicula commonly
exhibits a series of dense, raised bands comprising two or three condensed fuselli (text-fig. 6o, n).
The prosicula is usually absent and is replaced by one or two stout rods that merge with the virgula.
ThP is small and possesses a tightly upturned metatheca that grows closely adpressed to its
protheca. ThF is prosoblastic and bifurcates shortly after crossing thl \ giving rise to th2‘ from its
right side, as in Pattern C (text-fig. 5c-e). The crossing canals of both thF and th2* grow across
the sicula in a nearly horizontal direction. The crossing canal of th2* ceases growth near the sicular
axis and exhibits a hood-like form. The flange, that in Patterns A and C had grown upward from
near the sicular aperture to fuse with the approaching edge of the crossing canal of th2\ appears in
Pattern D on the dorsal side of the crossing canal of thF (text-figs. 5m and 6g). As the flange grows
upward it is linked to the hood of th2’ by a list (text-figs. 5j, 6l, 7b). This event marks the
differentiation of the prothecae of th2^ and th2^. There are four primordial thecae and two crossing
canals (thF and th2’). Th2^ or some later theca is dicalycal. The median septum may be zigzagged,
may become straight distally, or may be straight throughout (as in C. bicornis) among species
known to possess this astogenetic pattern. The proximal end of these species is narrow and evenly
rounded. In obverse view the sicula is exposed only to the level of the aperture of thP or thF.
Post-primordial thecae range in shape from pseudoclimacograptid to climacograptid.

Pattern E (text-fig. 7d-l)

The sicula is like that described for Pattern D. The long virgella is angled across the sicular aperture.
The metatheca of thl' is tightly upturned and adpressed against its protheca. The hood over the
foramen of thF is completely enclosed by the metatheca of thlb The crossing canal of thF
originates as an isolated flange located on thl' below the foramen of thP (text-fig. 7h, j). From
this origin, it grows upward diagonally across the sicula with no downward component of growth.
Th2' arises from thF above a prominent growth-line unconformity by a pattern of differentiation
like that seen in the budding of distal thecae (text-fig. 7i, k). Accordingly, there are two primordial
thecae and one crossing canal (thF).

Rhabdosomes of species with this astogenetic pattern may be septate, with a straight median
septum and th2^ or some later theca dicalycal, or they may be aseptate. The proximal end is narrow

region between thl ' and thF in reverse aspect, x 14. i, Cn 59962, reverse view, proximal end broken to reveal
internal right-handed origin of crossing canal of th2', x 14.

j, K, M, N, Dicaulograptus hystrix (Bulman), Folkeslunda Limestone, Lasnamagi Stage ("G.' teretiuscidus
Zone?), Gardslosa, Oland, x 27. j, k, Cn 59928, reverse and left-lateral views; note disconformable crossing
of left-lateral wall of thC by newly differentiated, right-handed crossing canal of th2'. m, n, Cn 59927, reverse
and obverse views.

L, o, Pseudoclimacograptus (Archiclimacograptus subgen. nov.) luperus Jaanusson, Folkeslunda Limestone,
Lasnamagi Stage (‘G.’ teretiusculus Zone?), l, Cn 54587, Gardslosa, Oland, reverse view of holotype; note
early origin of th3' and extensive exposure of crossing canal of thF. o, Cn 59929, Lerkaka, Oland, reverse
view, early th2' growth stage with right-handed origin of th2' and broken metatheca of thC. Both x 27.

362

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 5. Camera lucida drawings illustrating Pattern D primordial astogeny in Climacograptus {Climaco-
graptus) sp. cf. C. (C.) caudatus Lapworth; Viola Springs Formation (101m above base of section H; Alberstadt
1973), Maysvillian Stage (‘C’ pygmaeus Zone), Arbuckle Mtns., Oklahoma. All illustrations are reverse views
unless noted otherwise. See text-fig. 2 for explanation of abbreviations and specimen repositories.

A, c, MCZ 9463/1, obverse and reverse views; note prosoblastic form of thF. b, MCZ 9463/2. d, MCZ
9463/3; note list linking crossing canal of thF with sicula and marking differentiation of right-handed th2L e,
F, MCZ 9463/4, reverse and left-lateral view of later stage in growth of crossing canal of th2L G, J, MCZ
9463/5, reverse and left-lateral views of early th2^ stage; list links upward-growing flange of protheca of th2*
with hood-like crossing canal, h, i, MCZ 9463/6, reverse and oblique left-lateral views, k, l, MCZ 9463/7,

MITCHELL; GRAPTOLOID EVOLUTION AND CLASSIFICATION

363

and unevenly rounded to somewhat rectangular. The markedly offset sicula lies almost entirely to
the right of the plane of symmetry in obverse view and is exposed to the level of the aperture of
thP (text-fig. 7l). Post-primordial thecae are sharply geniculate, climacograptid thecae with nearly
vertical supragenicular walls.

Pattern /^(text-fig. 8)

The sicula is slender and bears a prominent pair of antivirgellar spines in addition to the usual
virgella. The metatheca of thP is tightly recurved and grows upward along its protheca. ThP is
prosoblastic and the hood which covers the foramen of thF is generally short and largely or entirely
enclosed by thl ’ (text-fig. 8d, j, m). The crossing canal of thP arises as an isolated flange on thP
below the foramen (text-fig. 8e, j), as in Pattern E. It subsequently grows diagonally upward across
the sicula. Near the mid-line of the sicula, this upward-growing flange bifurcates to produce both
the metatheca of thfo and the protheca of th2* (text-fig. 8h, n). Th2^ arises by a pattern of
differentiation that is like that of all subsequent thecae (text-fig. 8h). There are three primordial
thecae and one crossing canal (th 1 ^).

Pattern F rhabdosomes are aseptate. The proximal end is strongly asymmetrical and generally
quite narrow compared to the distal width of the colony. In obverse view the sicula is exposed for a
large part of its length, nearly to the level of the aperture of th2k Post-primordial thecae range
from climacograptid with prominent genicular flanges to glyptograptid or nearly orthograptid.

Pattern G (text-figs. 9 and 10)

The sicula is long and slender with a virgella and a pair of antivirgellar spines on the sicular aperture.
Thl ^ gives rise to a prosoblastic thF. The crossing canal of thF grows diagonally downward across
the sicula in the form of a hood that is free on its ventral (proximal-ward) side (text-figs. 9a and
10a). It fuses with an upward-growing, wedge-shaped flange that arises partly on the sicula and
partly on thl ' near the sicular aperture. The hood of thfo continues to grow downward but now as
a complete tube (text-figs. 9d and 10c, e). The edge of the open ventral margin of the earlier hood
of thF bears a thickened rim and, together with one edge of the upward-growing flange, forms a
foramen from which th2' develops (text-fig. 9b). Thus, thfo and th2' form a somewhat asymmetrical
pair with a smooth arch connecting their prothecae across the reverse side of the sicula. The
protheca of th2' expands rapidly, growing upward along the dorsal wall of the crossing canal of th fo
(text-fig. 9e, e). Th2^ differentiates from the prothecae of th2' above a growth line unconformity, in
a fashion like that by which all subsequent thecae arise (text-figs. 9g and IOd). There are three
primordial thecae and one crossing canal (thF).

The first several thecal pairs alternate, and the dicalycal theca may be th3 ‘ or, more commonly,
some later theca. Many species exhibiting Pattern G are aseptate. Septate species have a straight
median septum. The proximal end is tapered bluntly and markedly asymmetric. On the obverse
side of the colony the sicula is visible for most of its length, up to a level near the aperture of th2*,
as in Pattern F. Post-primordial thecae range from orthograptid to amplexograptid or lasiograptid
in shape.

right-lateral and reverse views, showing relation of mature hood-like crossing canal of th2‘ to sicula and
thP. M, N, MCZ 9463/8, reverse and obverse views; note upward-growing flange along dorsal side of
thF. o, T, MCZ 9463/9, reverse and obverse views; note intercalation of short fuselli into reverse lateral wall
of protheca of th2’ in the region from which th2^ will soon appear (cf. text-fig. 5p, s). p, q, MCZ 9463/10,
reverse and obverse views, illustrating origin of th2^ by simple distal differentiation from protheca
of th2T R, MCZ 9463/11, sicula with slight secondary elongation; note proximally zig-zag septum and
dimple-like list scar (Is) corresponding to list shown in text-fig. 5j (cf. text-figs. 6t, l and 7b), x 14. s, MCZ
9463/12.

All except r x 25.

364

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 6. Camera lucida drawings of species exhibiting Pattern D primordial astogeny. All illustrations are
reverse views unless noted otherwise. See text-fig. 2 for explanation of abbreviations and specimen repositories.

A-i, Pseiidoclimacograptus (Pseudoclimacograptus) scharenhergi (Lapworth), ^Climacograptus band’, Balclat-
chie beds, Caradoc Series (’/).’ multidens Zone), Laggan Burn, Ayrshire, Scotland, a, MCZ 9464/1, ventral
view; note twisted form and raised, thickened fuselli. b, MCZ 9464/2, ventral view, showing left-handed origin
of crossing canal of thF. c, MCZ 9464/3, oblique reverse view; note prosoblastic form of thC. d, MCZ
9464/4. E, G, MCZ 9464/6, obverse and reverse views; note linkage of upward-growing flange to crossing canal

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIEICATION

365

Pattern //(text-figs. 3f-h, k, l, 11, 12k-o)

The sicula is commonly relatively short and broad (about 1 mm or less in length among many
Ordovician representatives), but varies greatly (e.g. Cystograptiis penna). The sicular aperture bears
only a virgella. However, the dorsal margin is often prolonged and somewhat incurvate. The
metatheca of thH is generally sharply upturned, as in Patterns E and F, and may partially enclose
its protheca on the obverse side. It also encloses the foramen and small hood of thH on its reverse
side. In many species with this pattern (but not '’CUmacograptus' kuckersianus Wiman), the reverse
wall of the metatheca of thP is free and is not anchored against cither the sicula or the protheca of
thP (text-fig. 3f). The crossing canal of thF begins as an isolated flange located on thP near the
sicular aperture. As it grows diagonally upward, across the sicula, its ventral edge grows along the
edge of the reverse wall of thl*. Early in the growth of the protheca of thF, th2^ begins to
differentiate— a process marked by the interfingering of wedge-shaped fuselli in the region adjacent
to thl ^ (text-figs. 1 1h, i and 12k). Th2^ differentiates in a fashion like that of later thecae (text-figs.
3h, 11m, 12k). There are three primordial thecae and one crossing canal (thF). Rhabdosomes are
usually septate with the dicalycal theca at the primitive location (th2*) or later. The majority
of Pattern H species exhibit a straight median septum, but among some Silurian species (e.g.
ClinocHmacograptus retroversus and Metaclimacograptus imdulatus) the median septum is undulate
to zigzagged.

The proximal end is quite narrow and sharply rounded or fusiform and strongly asymmetrical.
In obverse view, thF partly encloses the sicula, which is exposed only up to the level of the aperture
of thl* or, at most, nearly up to the level of the aperture of thH (text-figs. 3h, l, and lit, l).
Among Ordovician Pattern H species, post-primordial thecae are restricted to glyptograptid and
climacograptid in shape. The rhabdosomes tend to be narrow and parallel sided with little distal
widening. However, Silurian species encompass a broad range of diplograptinid thecal shapes and
colony forms.

Pattern / (text-fig. 12a-j)

The long sicula bears only a virgella projecting from its simple aperture. Thl ’ arises relatively close
to the sicular aperture and has an exceptionally short descending segment. The foramen of thH in
the right lateral wall of the protheca of thl* bears only a minute hood, or no hood at all. Thl*
turns upward very sharply, and grows upward with its left lateral wall partly or completely enclosing
its descending protheca in obverse aspect (text-fig. 12c, h). As thl* grows, its right lateral wall
begins to sweep out on to the metasicula and to enclose the latter’s reverse side (text-fig. 12a, b, d).
After this wall crosses the sicula’s mid-line and the theca approaches its mature length, an interthecal
septum appears that divides the right lateral wall into a metatheca of thl* and a protheca of thH
(text-fig. 12a, d). a growth line unconformity may also mark the separation of the protheca
of thF.

of th2*. H, I, MCZ 9464/7, obverse and reverse views; dicalycal th2^ arises by simple differentiation from distal
portion of protheca of th2‘.

J-P, CUmacograptus {CUmacograptus) hicornis (Hall), Viola Springs Formation (0-3 m above base of section
D; Alberstadt 1973), Rocklandian Stage (upper C. hicornis Zone), Arbuckle Mtns., Oklahoma, j, MCZ 9465/1,
showing paired prosicular rods, raised fusellar ridges on metasicula, and origin of th2^ near aperture of hood-
like crossing canal of th2* (cf. text-figs. 6l and 7a). k, MCZ 9465/2. l, MCZ 9465/4, proximal end of
rhabdosome with portion of reverse side broken away, revealing hood-like crossing canal of th2* and list that
linked crossing canal with reverse wall of protheca of th2* (cf. text-figs. 5j and 6o). m, n, MCZ 9465/5, ventral
and reverse views of immature metasicula, showing dorsal deflection and concave ventral side; note regularly
spaced fusellar rings and paired prosicular rods, o, MCZ 9465/3. p, MCZ 9465/6, early growth stage showing
prosoblastic thP.

All x35.

366

PALAEONTOLOGY, VOLUME 30

th|2fo

TEXT-FIG. 7. Camera lucida drawings of species exhibiting Pattern D and Pattern E primordial astogenies. All
illustrations are reverse views unless noted otherwise. See text-fig. 2 for explanation of abbreviations and
specimen repositories.

A, Pseudoclimacograptiis (Pseudoclimacograptus) scharenbergi (Lapworth), BMNH GS 74247, Balclatchie
beds, Caradoc Series (‘Z).’ miiltidens Zone), Laggan Burn, Ayrshire, Scotland; note origin of th2^ at point of
fusion between upward-growing flange of reverse prothecal wall and crossing canal of th2^; sicular apex sealed
but lacking normal prosicula.

B, c, P. (P.) clevensis Skoglund, Fjacka Shale, Jonstorp, Vastergotland. b, Vg 757, semi-relief specimen with
prominent list scar (Is) where list that links crossing canal of th2* to reverse prothecal wall has been pressed
through, c, Vg 758, obverse view; th2^ rapidly encloses strongly deflected sicula.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

367

Following its origin, thF continues the trend of the right lateral wall of thl ' and grows strongly
upward, often reaching the dorsal side of the sicula well above the latter’s aperture. This origin of
thF is quite late (in some species it does not appear until after the protheca of thl ^ has reached the
dorsal side of the sicula) compared to other diplograptinid astogenetic patterns. The differentiation
of thF follows a developmental pattern like that of all later thecae. Thus, Pattern I exhibits only
one primordial theca and no crossing canals. Although generally aseptate, some Pattern I species
have a straight median septum and a dicalycal th2* or some later theca. In at least some dimorpho-
graptids (e.g. Rhaphidograptus toernquisti) the metatheca of thF appears to be suppressed (text-fig.
12d) but th2* develops from the protheca of thF in the normal fashion.

The proximal end is sharply fusiform to acicular. In reverse aspect, a substantial portion of the
dorsal side of the sicula is visible. The combined right lateral wall of thl* and ventral wall of thl^
often have a concave curvature (especially in species with orthograptid or petalograptid colonies).
In obverse view the sicula is commonly exposed for most or all of its length. Nevertheless, thecae
are sufficiently elongate that it is usually enclosed by the level of the aperture of thl' or, more
rarely, by a level slightly below that of the aperture of thP. Also in obverse view, thl ' often appears
to grow directly upward from near the sicular aperture because of its enclosure of the short
descending portion of the protheca of thl'. Pattern I species (all are Silurian in age) possess post-
primordial thecae ranging in shape from climacograptid to orthograptid and petalograptid.

PHYTOGENY AND CLASSIFICATION OF THE DIPLOGRAPTACEA
Suborder virgellina Fortey and Cooper, 1986

Diagnosis. Graptoloids with a virgella. Primordial astogeny is of isograptid type or modified to
‘diplograptid’, ‘nemagraptid’, or ‘monograptid’ type. Rhabdosomes extensiform to reclined or
platycalycal scandent.

Superfamily diplograptacea Lapworth, 1873, emend.

Diagnosis. Horizontal to reclined and partly to wholly scandent, dipleural biserial and monoserial
virgellinids with a single dicalycal theca, delayed to th2' or some later theca, and three crossing
canals; thl' with metasicular origin and thF arising right-handedly from thl'.

Discussion. Text-fig. 13 depicts the cladistic relationships among the major diplograptacean taxa
(typified in the diagram by their astogenetic patterns), the stem group Oelandograptus gen. nov.
(‘Or,— see below), and the other taxa of the Virgellina Fortey and Cooper, 1986. The Phyllo-
graptidae, which possess an isograptid primordial astogeny (denoted by Ts’ in text-fig. 13),
form a convenient outgroup for comparisons among the Diplograptacea (see Cooper and Fortey
1983 for a discussion of the primitive status of this astogenetic pattern).

D-L, Climacograptus (Diplacanthograplus subgen. nov.) spiniferus Ruedemann, Viola Springs Formation,
Maysvillian Stage CC.' pygmaeus Zone), Arbuckle Mtns., Oklahoma; MCZ 9466 from 51 m above base of section
along Interstate Highway 35, adjacent to section H; MCZ 9467 from 104-5 m above base of section H; MCZ
9468 from 76-5 m above base of section J; MCZ 9469 from 30-5 m above base of section H (Alberstadt 1973).
D, MCZ 9466/1, early thl ' stage; note set of prosicular rods in place of normal prosicula and dorsal deflection
of metasicula. e, MCZ 9467/1, ventral view; note stirrup-like form of prosicular rods and prosoblastic crossing
canal of thC. f, MCZ 9466/2, ventral view; note characteristic twisted, asymmetric form of sicula, x 40. G,
MCZ 9468/1, thP formed by upward-growing flange, h, j, MCZ 9466/3, oblique reverse and left-lateral views;
flange of thH originates on reverse lateral wall of thl', not in continuity with earlier segment of its crossing
canal, i, MCZ 9469/1; note origin of th2' by simple differentiation from distal portion of protheca of thH,
X 40. K, L, MCZ 9467/2, reverse and obverse views; note dicalycal th2^ and strongly deflected sicula visible
for its entire length on obverse side.

All except f and i x 32.

368

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 8. Camera lucida drawings of species exhibiting Pattern F primordial astogeny. All illustrations are
reverse views unless noted otherwise. See text-fig. 2 for explanation of abbreviations and specimen repositories.

A-D, H-j, o, p, Geniculograptus (gen. nov.) pygmaeus (Ruedemann), Viola Springs Formation, Maysvillian
Stage (‘C.’ pygmaeus Zone), Arbuckle Mtns., Oklahoma; MCZ 9427 from 51 m above base of section along
Interstate Highway 35, adjacent to section H and MCZ 9470 from 104-5 m above base of section H (Alberstadt
1973). A, B, MCZ 9470/1, obverse and reverse views of specimen in early phase of formation of foramen of
the. c, MCZ 9470/2. d, MCZ 9427/4, completed foramen of thU and metatheca of thl * nearing maturity. H,
MCZ 9427/2, early th2* stage showing division of crossing canal of thC to form metatheca of thC and
protheca of th2L i, MCZ 9427/1. r, MCZ 9427/3, stage showing construction of protheca of thF by upward-

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

369

The cladistic relationships depicted in text-fig. 13 provide an outline of the branching history of
the group. It includes three alternative interpretations, differing in their treatment of the Dicrano-
graptinae and the Nemagraptinae. The basis for these alternatives is discussed below. As is readily
apparent from the cladograms (and as Bulman had suspected in 1963u, b), the Diplograptacea
comprises three large primary divisions: the Diplograptidae, the Monograptidae, and the Ortho-
graptidae. The roots of these families lie deep within the early history of the clade. Each family
exhibits a striking degree of parallelism during its evolutionary history and each includes a subgroup
that is characterized by a highly derived and highly simplified astogenetic pattern that becomes
dominant among the family’s representatives in Upper Ordovician or Silurian faunas. The highly
derived uniserial monograptines also belong taxonomically among the Diplograptacea, as do the
Dicranograptidae (see below).

Previously, when most graptolithologists considered the Monograptina to be a grade of organiza-
tion that stood above that of the Diplograptina (one which was achieved repeatedly by several
lineages), the ranking of these two taxa at the same level was appropriate (see Rigby 1986). However,
the monograptid condition is a synapomorphy that characterizes a subclade of the Diplograptacea
and is not a grade of organization. The close structural and cladistic relationships between the
traditional Monograptina and their antecedents among the dimorphograptines and glyptograptines
(discussed below in the section ‘Monograptidae’) indicate that there is no longer sufficient justifica-
tion for the retention of the high taxonomic rank usually accorded to this subclade. Instead,
I propose here a classification that emphasizes their evolutionary derivation from among the
diplograptaceans. In as much as the ICZN has extended the principle of priority to encompass
family group taxa, the entire family to which the monograptines belong must take the most senior
available name the Monograptidae. The Monograptidae, thus, includes Arenig biserial species as
well as the traditional Silurian and Devonian monograptine and cyrtograptine species.

It is very difiicult to establish a well-corroborated hierarchical branching among the three original
families. Archaic monograptids (e.g. 'Glyptograptus' dentatus, with a Pattern B astogeny) and
archaic diplograptids (in the form of Psendoclimacograptus cumhrensis and related forms, with a
Pattern C astogeny) each differ from the primitive diplograptaceans (Pattern A representatives of
the ‘G.’ ausirodentatus species group, herein recognized as Oelandograptus gen. nov.) in a unique
suite of derived characters. All astogenetic similarities that they exhibit with Pattern A are sym-
plesiomorphic, regardless of the cladistic position of the Dicranograptidae relative to other diplo-
graptaceans. However, data on the character of the thecal apertures among these graptolites appear
to be helpful. Very early in the history of the Orthograptidae, this group acquired a cuspate thecal
aperture (text-fig. 2l, m). An identical thecal form is present in ‘G.’ dentatus (text-fig. 3i) and ‘G.’
cernus and suggests that the Orthograptidae (excluding the stem group Oelandograptus gen. nov.)
and the Monograptidae are sister groups, and that the Diplograptidae branched from the stem
group lineage prior to the appearance of this thecal form and the separation of the Monograptidae
and Orthograptidae. Given the small size of the group and their paraphyletic status, the members

growing flange; internal portion of crossing canal enclosed by thU. o, MCZ 9470/3, obverse view showing
extensive exposure of sicula and alternating thecae, p, MCZ 9470/4.

E, F, G, N, G. (gen. nov.) typicalis (Hall), Viola Springs Formation, 51 m above base of section along
Interstate Highway 35, adjacent to section H (Alberstadt 1973), Maysvillian Stage (‘C.’ pygmaeus Zone),
Arbuckle Mtns., Oklahoma. E, MCZ 9426/4, incipient flange of thU. f, MCZ 9426/3. G, MCZ 9426/1, early
th2* stage showing paired growth of protheca of th2^ and metatheca of thU. n, MCZ 9426/2; note division of
protheca of thU to form its metatheca and descendant protheca of th2'.

K, L, M, G. (gen. nov.) inuiti (Cox), Maysvillian Stage (‘C.’ manitoulinensis Zone), Ungava Bay, Akpatok
Island, Canada, k, m, SM unnumbered (Cox Collection), reverse and left-lateral views showing internal portion
of crossing canal of thU enclosed by reverse lateral wall of metatheca of thP, x 32. l, SM A. 102372, showing
reverse lateral wall of protheca of thU formed by simple upward-growing flange.

All except k and m x 26.

TEXT-FIG. 9. Camera lucida drawings of species exhibiting Pattern G primordial astogeny. All illustrations are
reverse views unless noted otherwise. See text-fig. 2 for explanation of abbreviations and specimen repositories.

A-D, F-H, Orthograptus quadrimucronatus (Hall), ssp. Viola Springs Formation (51 m above base of section
along Interstate Highway 35, adjacent to section H; Alberstadt 1973), Maysvillian Stage (‘C’ pygmaeus Zone),
Arbuckle Mtns., Oklahoma, a, c, MCZ 9471/1, reverse and obverse views; note fusion of hood with upward-
growing flange of prosoblastic crossing canal of thH. b, MCZ 9471/2, proximal view illustrating origin of
flange of thF. d, MCZ 9471/3; note paired growth of thH and th2‘ following fusion of hood and flange, f,
MCZ 9471/4, th2^ arises from distal portion of protheca of th2* by simple differentiation, g, h, MCZ 9471/5,
reverse and obverse views; note extensively exposed sicula and delayed dicalycal theca, x 12.

E, Amplexograptus leptotbeca (Bulman), SM A. 723040, Balclatchie beds, Caradoc Series (‘Z).’ midtidens
Zone), Laggan Burn, Ayrshire, Scotland; showing presence of upward-growing flange in ontogeny of thH;
also note growth of protheca of th2’ back upon crossing canal of thP.

i, j, A. hekkeri (Opik), Cn 59938, Kukruse Stage, (N. gracilis Zone), Estonia; formation of prosoblastic
thP.

All except G and h x 26.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

371

TEXT-FIG. 10. Camera lucida drawings illustrating Pattern G primordial astogeny in Amplexograptus bekkeri
(Opik); Kukruse Stage {N. gracilis Zone), Estonia (Holm Collection). All illustrations are reverse views unless
noted otherwise. See text-fig. 2 for explanation of abbreviations and specimen repository.

A, Cn 59909; note upward growing flange, b, c, Cn 59939, obverse and reverse views, showing fusion of
hood-like crossing canal with small flange and early growth of protheca of th2L d, Cn 59942. e, Cn 59943;
note origin of th2* still in early prothecal stage, while thU nearly complete. F, Cn 2474c, obverse view; note
strong gradient in thecal form from orthograptid through lasiograptid to amplexograptid in distal thecae,
X 12. G, Cn 59941; note post-primordial pattern of differentiation in th2^ and th3L H, Cn 59940, broken
specimen showing course of internal canals.

All except f x 26.

372

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 1 1 . Camera lucida drawings illustrating Pattern H primordial astogeny in Glyptograptus kuckersi-
amis (Wiman); Kukruse Stage (N. gracilis Zone), Estonia (Holm Collection). All illustrations are reverse views
unless noted otherwise. See text-fig. 2 for explanation of abbreviations and specimen repository.

A, B, Cn 59933, ventral and reverse views; note form of sicular aperture, c-e, Cn 59930, obverse, reverse,
and oblique dorsal views, illustrating enclosure of foramen of thH by the metatheca of thlL f, Cn 59931,
crossing canal of thF formed by upward growing flange. G, h, Cn 59932, obverse and reverse views; note
origin of protheca of th2‘ by zigzag suturing of fuselli below interthecal septum, i, Cn 59934. j, Cn 54606,
obverse view; note rapid enclosure of sicula by level of aperture of thlL k, l, Cn 59935, reverse and obverse
views, X 12. M, Cn 59936, bleached specimen showing details of dicalycal budding and distal thecal ontogenies.

All except k and l x 26.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

373

of the Oelandograptus gen. nov. austrodentatus species group are treated as Incertae familiae within
the classification presented here.

Family dicranograptidae Lapworth, 1873, emend. Fortey and Cooper, 1986

Diagnosis. Diplograptaceans with uniserial and uni-biserial, partly scandent to reclined or extensi-
form rhabdosomes with three crossing canals and dicalycal th2*; sicula with aperture notched on
dorsal side and with paired clefts adjacent to the lanceolate virgella; thecae exhibit strong morpho-
logical gradient along stipes, with proximal-most thecae at least having isolate, introtorted apertural
regions; all thecae with sharp fusellar disconformity within the prothecal segment (see Finney 1985).

Discussion. Among the traditional Diplograptina, the Dicranograptidae, and in Nemagraptus (see
Finney 1985), the first dicalycal theca is delayed from its primitive position at thF to a position
at th2' or later, producing three crossing canals. Unlike Phyllograptus and the multi-stiped dicho-
graptids, no further dicalycal thecae occur in any of these taxa. As Finney (1985) has shown, the
lateral branches of Nemagraptus are cladia and are produced in a quite different way from the
branch dichotomies of the anisograptids and dichograptids. The possession of a delayed single
dicalycal theca distinguishes the advanced Virgellina from all previous Graptoloidea. An additional
synapomorphy is the metasicular origin of thU.

Nemagraptus shares a unique configuration of the sicular aperture with the Dicranograptidae
Lapworth, 1873, emend. Fortey and Cooper, 1986: the sicular aperture bears two broad lappets
that are separated from one another on the antivirgellar side of the sicula by a broad notch. These
lappets are also separated from the virgella by a pair of clefts that possess a flared lip (see text-fig.
2h, r). This synapomorphic condition is unknown in other Virgellina, including the traditional
Diplograptina, and unites the Dicranograptinae and the Nemagraptinae as the Dicranograptidae,
emend. However, the relationship of the Dicranograptidae to the rest of the Diplograptacea and its
status as a clade or paraclade are problematic. In addition to the dicalycal th2* mentioned above,
the traditional Diplograptina (or ‘diplograptids’ in the following discussion) and the Dicranograpti-
nae share additional unique astogenetic features— yet these are not shared with Nemagraptus: 1, in
contrast to the right-handed origin of thU from thl ' in the isograptid pattern, this theca arises left-
handedly in ‘diplograptids’ and dicranograptines; and 2, thU then follows a convoluted S-shaped
course that Bulman referred to as streptoblastic. Despite vague similarities in thecal form or more
intriguing similarities in growth direction, no dichograptid or other virgellinid is known to possess
these features. The dicranograptines possess a primordial astogeny that, apart from the form of the
sicular aperture, is indistinguishable from that of the primitive diplograptaceans — members of the
‘G.’ austrodentatus group (see Bulman 1945, 1947). These similarities (the left-handed origin of thU
from thU, and the streptoblastic form of this second theca), together with the dipleural scandent
rhabdosome architecture of Dicranograptus and the ‘diplograptids’, may be interpreted in several
ways (cf. text-fig. 13a-c).

First, they may be synapomorphies (text-fig. 13b). If so, this indicates that the ‘diplograptids’, as
traditionally construed, and the Dicranograptinae of the Dicranograptidae are sister groups. It
further implies that the ‘diplograptids’ have lost the unique form of the sicular aperture characteristic
of the paraclade Dicranograptidae. The clade ‘diplograptids’ -I- Dicranograptinae, in turn, shares
with the Nemagraptinae the synapomorphies of three crossing canals and a dicalycal th2 ' . However,
as Fortey and Cooper (1986) noted, this cladistic sequence conflicts sharply with the stratigraphic
order of appearance of these taxa. The diplograptids precede the earliest dicranograptines or
nemagraptines by at least the entire duration of the Llanvirn (Bulman 1960; Finney 1985). Further-
more, Finney has found that the thecae of Nemagraptus, like those of the Dicranograptinae, are
highly complex with a number of unique features that make them unlikely to be ancestral to the
‘diplograptid’ thecal structure.

Secondly, the Dicranograptidae as a whole may be a sister group to the ‘diplograptids’ (text-fig.
1 3c), but this does not alleviate the problems posed by the stratigraphic record of these graptolites.
Furthermore, it requires either that the similarities in proximal end structure and colony architecture

374

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 12. Camera lucida drawings of species exhibiting Pattern H and Pattern I primordial astogenies.
See text-fig. 2 for explanation of abbreviations and specimen repositories.

A, B, Petalograptus sp. 1, Cn 54917 and Cn 54916, lower Klubbudden Stage (M. turriculatus Zone), Dalarna,
Sweden; reverse views (after Hutt et al. 1970); note metatheca of thP in contact with reverse side of sicula;
also note pattern of simple differentiation of thP from protheca of thU; dorsal side of sicula extensively
exposed, x 32.

c, D, Rhaphidograptus toernquisti Elies and Wood, Cn 54910 and Cn 54915, upper Bollerup Stage (M.
gregarius Zone), Dalarna, Sweden; obverse and reverse views (after Hutt et al. 1970); note form of thU and
absence of metatheca of thU, x 12 and x 18.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

375

between the dicranograptinae and the ‘diplograptids’ are parallelisms, or that the ostensibly primi-
tive astogenetic and other features of Nemagraptus are not primitive but derived.

Thirdly, the Dicranograptidae may be a sister group of the Orthograptidae (text-fig. 13a). In this
case the unique dicranograptid sicular form must have been derived from the simple sicula of the
‘diplograptids’. In addition to the better fit of this hypothesis with the stratigraphic data, there are
two further lines of morphological evidence in its support: 1, several Llanvirn-Llandeilo species
among the archaic orthograptids, such as ‘G.’ vikarbyensis Jaanusson and ‘G.’ teretiusculus sensu
Jaanusson, which exhibit a Pattern A astogeny, have a flared sicular aperture (text-fig. 2l); this flair
is strongest on the dorsal side of the sicula and consists of two lobes separated by a broad,
shallow notch; this structure may be homologous with the unique lappets and dorsal notch of the
dicranograptid sicula; and 2, in N. gracilis the first two thecae have isolated, introtorted apertures
quite like those of Dicellograptus (see Finney 1985), while later thecae have simpler apertural regions
that are neither isolated nor introtorted. If this condition, like the astogenetic pattern, is primitive
relative to the condition in Dicellograptus (in which all thecae have isolated and introtorted aper-
tures), then this complex thecal form must have arisen at the proximal end and spread distally
during the course of evolution. This, however, conflicts sharply with the general conservatism of
the primordial thecae in virgellinid colonies. Alternatively, if the thecal characters in N. gracilis are
derived relative to those of early dicellograptids like D. vagus Hadding, then the complexity of thP
and thP may be the result of conservatism in the proximal end while distal thecae became simplified.

Subfamily dicranograptinae Lapworth, 1873, emend. Finney, 1985

Diagnosis. Dicranograptids with Pattern A astogeny; complex introtorted, introverted thecae with
isolated apertures and lateral apertural processes; reclined uniserial unbranched stipes to uni-
biserial, partly scandent rhabdosomes, at least some of which possess a virgula.

Generic group taxa. Dicranograptus Hall, Dicellograptus Hopkinson, and Leptograptus Lapworth.

E-G, Petalograptus sp. 2 (after Hutt et al. 1970). e, Trinity College Dublin, TCD 8272a, b, Balbriggan Co.,
Dublin, Eire (M. turriculatus Zone); obverse view; note simple form of sicular aperture, x 9. f, g, Cn 54920
and Cn 54921, lower Klubbudden Stage {M. turriculatus Zone), Dalarna, Sweden; reverse views; note narrow,
acicular proximal end and short interthecal septum between thU and thU (cf. text-fig. 12a), x 1 1.

H, P. insectiformis (Nicholson), Cn 54913, Bollerup Stage (M. gregarius Zone), Dalarna, Sweden; obverse
view (after Hutt et al. 1970); note short downward-growing portion of protheca of thU and strongly upward-
growing metatheca of thH; virgella with ancora, x 30.

I, P. obuti (Rickards and Koren'), Tchernyshev Central Geol. Mus., Leningrad (topotype collection),
Sakmara Formation, Llandovery Series, Mugodjary Range, South Urals, USSR; reverse view (after Rickards
and Koren' 1974); note form of thU, which arises well above sicular aperture, and growth lines that suggest
late differentiation of thU, x 10.

j, P. palmaeus (Barrande), USNM 161811, Descon Formation, Llandovery Series {M. gregarius Zone),
south-eastern Alaska; reverse view (after Churkin and Carter 1970), x 6.

K, o, Glyptograptus sp. cf. G. scalaris, Birmingham University unnumbered, Jupiter Formation, Llandovery
Series, south shore Anticosti Island, Quebec. K, (after Barrass 1954) reverse view of proximal end fragment
showing origin of thU and probable dicalycal th2L x 36. o, ventral view of sicular fragment with protheca of
thU; note absence of distinct crossing canal of thU, x 36.

L, Paraclimacograptus innotatus obesus (Churkin and Carter), USNM 161611, Descon Formation, Llan-
dovery Series (M. cyplius Zone), south-eastern Alaska; obverse view (after Churkin and Carter 1970); note
strongly acicular proximal end and absence of antivirgellar spines, x 7.

M, "Diplograptus' modestus diminutus Churkin and Carter, USNM 161701a, Descon Formation, Llandovery
Series (P. acuminatus and C. vesiculosus zones), south-eastern Alaska; obverse view (after Churkin and Carter
1970), x7.

N, Glyptograptus gnomus Churkin and Carter, USNM 161644, Descon Formation, Llandovery Series (M.
cyphus Zone), south-eastern Alaska; reverse view (after Churkin and Carter 1970), x 7.

376

SUBORDER

PALAEONTOLOGY, VOLUME 30

VIRGELLINA

SUPER-

FAMILY

I

Diplograptacea

I 1

Stem-group
Phyllograptidae
I 1

FAMILY

Orthograptidae Dicranograptidae

I 1 I 1

Monograptidae

r

T

Diplograptidae Phyllograptidae

I 1 I 1

Diplograptacea

1

Dicrano- Phyllo-
graptidae graptidae

I 1 I 1

Diplograptacea

I 1

Phyllo- Dicrano-
graptidae graptidae

I 1 I 1

"diplograptids”

8, 9, ±

62, 63

A' A” N Is

23-- ho

6--

4, 20, 26E:

lO

--27

B

19, 21, 22 = -
1-3

Hi»0|

,5,7:|=

Is N A” A'

27--

<) I

-1-6
1>h f23

19, 21, 22, EE

“diplograptids”

EE 4, 20, 26

4, 6, 8, 9

EE 1-3, 5, 7

TEXT-FIG. 13. Cladograms illustrating the possible general relationships among the major virgellinid clades,
symbolized by the letter designation of their primordial astogenetic pattern (see text-fig. 1; Is, isograptid; M,
monograptid; Ol, Oelandograptus (gen. nov.) austrodentatus and related species) and their classification.

A, preferred relationships with the Dicranograptidae as sister group to the Orthograptidae. Synapomorphies;
I, virgella present; II, dichotomy d3 suppressed (see Fortey and Cooper 1986); 1, metasicular origin of thl*; 2,
dichotomy dl delayed to th2'; 3, dichotomy d2 suppressed; 4, thU with left-handed origin from thl‘; 5,
metatheca of th2‘ and protheca of th2^ arise from paired foramina formed by fusion of downward-growing
crossing canal of th2* with upward-growing flange; 6, scandent dipleural rhabdosome architecture; 7, capacity
for cladia generation; 8, sigmoidal thecae; 9, cuspate thecal apertures; 10, right-handed origin of th2‘; 11,
sharply geniculate thecae; 12, introverted thecal apertures; 13, asymmetrical proximal end; 14, origin of th2^
delayed to distal portion of protheca of th2^ and removed from set of primordial thecae; 15, elaborated
sicular aperture; 16, asymmetrical proximal end among advanced species; 17, paired antivirgellar spines among
advanced species; 18, delay of dicalycal theca to th3^ among advanced species; 19, complex notched and
lappet-bearing sicular aperture; 20, colony achieves capacity for wholly or partially uniserial stipes; 21, isolated,

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSILICATION

377

Subfamily nemagraptinae Lapworth, 1873, emend. Finney, 1985

Diagnosis. Dicranograptids with horizontal, cladia bearing, uniserial stipes; primordial astogeny
nemagraptid (see Finney 1985) with right-handed origin of thF from thP.

Generic group taxa. Nemagraptus Emmons. Nemagraptine status of Amphigraptus Lapworth, Pleurograptus
Nicholson, and Syndyograptus Ruedemann is probable but remains to be demonstrated.

Family orthograptidae fam. nov.

Diagnosis. Diplograptaceans with Pattern A astogeny or the derived Patterns G or F; sicular
aperture simple or with antivirgellar spines; thecae sigmoidal primitively but becoming orthograptid
or amplexograptid among forms with fully sclerotized periderm, and lasiograptid to retiolitid among
others; generally with cuspate apertures or with paired apertural spines.

Discussion. Text-fig. 14 presents a detailed cladogram that includes representatives of the broad
range of orthograptids. Text-fig. 17 illustrates the approximate geochronological range of the
redefined and new orthograptid genera. Among species with Pattern A astogeny, several (e.g.
' Diplograptus' uplandicus Wiman; see text-fig. 2) exhibit a markedly asymmetrical proximal end;
thF is prosoblastic or nearly so; the crossing canal of th2' is quite short and the sicula bears a

introverted thecal apertures; 22, geniculate dicranograptid thecae with prominent growth line unconformity
in protheca; 23, uni-biserial condition; 24, metathecae of thl ' and thF horizontal, forming uniserial, reclined
to horizontal stipes; 25, thF with right-handed origin from thl'; 26, introversion and apertural isolation of
post-thP thecae reduced or lost; 27, cladia present; 28, sicula extensively exposed on obverse side; 29, thF-
th2' form as a pair and with th2^ non-primordial; 30, second and third crossing canals lost; 31, metatheca of
tliF and protheca of th2' formed by fusion of upward-growing flange and hood-like crossing canal of thF;
32, protheca of thF is simple upward-growing flange that is not connected to its reduced hood-like crossing
canal; 33, metatheca of thF and protheca of th2‘ formed by division of protheca of thF; 34, thl' with
metatheca closely pressed against its protheca; 35, crossing canal of thF suppressed; 36, protheca of thl'
partly surrounded by its metatheca; 37, th2* no longer a primordial theca; 38, protheca of thF forms from
isolated, upward-growing flange; 39, reverse wall of metatheca of thl ' commonly free prior to growth of thF;
40, reverse lateral wall of metatheca of thl' in contact with reverse side of sicula; 41, thF no longer a
primordial theca; 42, capacity to produce proximally uniserial, scandent colonies; 43, sicula extensively exposed
along its obverse side; 44, capacity to produce ancora from virgella; 45, uniserial rhabdosome; 46, thl no
longer a primordial theca; 47, sicula with porus and lacuna stages in formation of foramen of thl; 48, crossing
canal of th2' visible as oval or diamond-shaped patch positioned on or near median plane of rhabdosome;
49, th2^ no longer primordial; 50, dicalycal theca delayed to th2^ or later; 51, asymmetrical sicula strongly
deflected to dorsal side; 52, metasicula with bands comprising condensed fuselli; 53, prosicula commonly
absent; 54, crossing canal of th2' reduced to hood-like form; 55, protheca of th2' formed by upward-growing
flange that does not fuse with crossing canal; 56, crossing canal of thF suppressed; 57, protheca of thP
formed from isolated, upward-growing flange; 58, th2' no longer primordial; 59, virgella reflected across
sicular aperture; 60, undulating median septum; 61, sigmoidal thecae with short, nearly vertical supragenicular
wall and gentle geniculum. Retained primitive characters: a, quasi-symmetrical proximal end; b, left-handed
origin of th2'; c, paired origins of metatheca of th2' and protheca of th2' (see 5, above); d, left-handed origin
of thF from thl'; e, broad proximal end with th2' occupying region between protheca and metatheca of
thl '; f, th2' with descending crossing canal; g, retained characters 10-12 (see above).

B, alternative cladogram with Dicranograptidae as paraphyletic stem-group occupying intermediate position
between ‘diplograptids’ and Phyllograptidae. Synapomorphies 1-20, 22-25, 27-61 as in a; 21, thl' and thF
introtorted with isolated, introverted apertures; 26, thecal characteristics of thl' and thF extended through-
out rhabdosome; 62, loss of complex features of sicular aperture; 63, loss of prothecal folds and fusellar uncon-
formity in protheca. Retained primitive characters a-g as in a; h, horizontal to reclined stipes (24 of a,
above); i, right-handed origin of thF (25 of a, above).

c, second alternative cladogram with Dicranograptidae as sister-group to traditional ‘diplograptids’, but
requiring several parallelisms. Synapomorphies and retained primitive characters as in b.

378

PALAEONTOLOGY, VOLUME 30

prominent pair of antivirgellar spines. These species are combined in the new genus Hustedograptus
defined below. The overall configuration of the proximal end of Hustedograptus gen. nov. is
remarkably similar to that of species of Orthograptus (particularly species of the O. calcaratus
group) and Amplexograptus with their Pattern G astogeny (compare text-fig. 2f, k with text-figs.
9g and IOf). While not exactly identical in these advanced Pattern A species and in Orthograptus
etc., I believe that this complex form produces a gestalt {sensu Fortey and Jefferies 1982) which,
together with the presence of the paired antivirgellar spines, constitutes a high burden synapo-
morphy between these groups. This gestalt synapomorphy is maintained largely intact through-
out the group of orthograptid species. Such a relationship is also supported by similarities in the form
of the thecal apertures among these taxa. The orthograptid cuspate thecal aperture, which is promi-
nently displayed in species of Hustedograptus gen. nov., is similarly developed in species of Amplexo-
graptus and most of the other Peiragraptinae. In the distal thecae of H. uplandicus, these cusps
develop into prominent lateral apertural spines of the same form and construction as those on the
thecae of species in the O. calcaratus and O. quadrimucronatus groups. I see no justification for
segregating those species such as O. gracilis, which lack apertural spines, into Rectograptus Pfibyl.
This taxon is defined on the basis of the loss of a single, relatively simple character, and there is no
reason to assume that such a loss should be unique.

Pattern G species are linked, in turn, to those with Pattern F ('Climacograptus' typicalis Hall and
relatives; see Table 1 and text-fig. 8) by the shared presence of paired antivirgellar spines, extensive
exposure of the sicula for most of its length in obverse view, the possession of three primordial
thecae, and only one crossing canal. These synapomorphies are further strengthened by the great
thecal similarities between these ‘climacograptids’ and peiragraptines such as A. leptotheca Bulman.
Indeed, the generic placement of such species as ‘C.’ inuiti Cox has been a persistent problem that
reflects their close relationship.

Several of the genera included within this family require significant revision and restriction in
their scope. Orthograptus does not properly include the Silurian species that have been referred to
it. Orthograptus Lapworth, with type species O. quadrimucronatus (Hall), is a well-known genus
comprising a coherent set of similar species, all of which possess a Pattern G astogeny. The Silurian
species that have been referred to Orthograptus (e.g. "O' obuti Rickards and Koren') exhibit a
Pattern I astogeny and consequently have a narrow, acicular proximal end. These forms may be
subsumed within Petalograptus Suess, among the Monograptidae.

Amplexograptus has been a source of persistent confusion. The type species A. perexcavatus
Lapworth has no holotype and Bulman (1962) selected a neotype (Birmingham University specimen
BU 1297) from among the specimens referred to this species and figured by Elies and Wood (1907).
As Bulman noted, two distinct species have been confused under this name. Bulman chose a neotype
that matches well with the age and morphology of Lapworth’s original taxon (Lapworth 1876,
1877). However, it is clearly the other species, A. fallax Bulman, that Lapworth (1880) had before
him when he made the distinction between typical Diplograptus species and members of the species
group that Elies and Wood (following Lapworth’s lead) later named Amplexograptus. Thus, al-
though it has been referred to as T. perexcavatus Lapworth, the type species of Amplexograptus is
in fact the biological entity A. fallax Bulman. The species group associated with A. fallax is by far
the better known. Consequently, the interests of taxonomic stability will be best served by retaining
the name Amplexograptus for these species, rather than for the "A.' perexcavatus group. According
to ICZN guidelines, I am preparing an application requesting that the Commission exercise its
plenary powers to suppress A. perexcavatus and establish A. fallax as the neotype of the genus. The
biological entity originally described by Lapworth (1877) as D. perexcavatus is probably not an
orthograptid. Rather, both it and D. pristis (Hisinger) (the type species of Diplograptus M‘Coy)
appear to have a Pattern C astogeny similar to that of ‘C.’ distichus (see the discussion for the
Family Diplograptidae).

The genera Glyptograptus and Climacograptus, as presently used, are extremely heterogeneous
and include species from each of the three superfamilies. Among the orthograptids, at least two
separate species groups with glyptograptid thecae exist: 1, the ancestral diplograptaceans of the

MITCHELL; GRAPTOLOID EVOLUTION AND CLASSIFICATION

379

FAMILY
SUB-

ORTHOGRAPTIDAE

I > r

0) 0)

S <0

FAMILY ^ I Orthograptinae

8 E I 1

c to

Peiragraptinae

Lasiograptinae

Dicranograptidae

Monograptidae

TEXT-FIG. 14. Cladogram and classification of the Orthograptidae, including its relationship to the stem group
Oelandograptus gen. nov. and the other diplograptacean families. Synapomorphies 1-61 as in text-fig. 13a;
64, orthograptid thecae; 65, sharply geniculate thecae with long straight infragenicular wall, short supragenic-
ular wall, and prominent genicular spines; 66, amplexograptid thecae with genicular flanges present in distal
thecae; 67, amplexograptid thecae throughout; 68, aseptate; 69, thP sharply upturned at dorsal side of sicula;
70, genicular spines; 71, uni-biserial; 72, gymnograptid thecae with lists; 73, elongate climacograptid thecae in
proximal end; 74, glyptograptid thecae; 75, lasiograptid thecae; 76, thecae with reduced fusellar periderm and
clathria; 77, lateral (septal) spines or scopulae; 78, bifid genicular spines on post-thP thecae; 79, lacinia
developed from genicular spines; 80, fusellar periderm of all thecae reduced to clathria except for sicula and
initial bud; 81, archiretiolitid-like clathrial astogeny; 82, loss of septal spines and scopulae; 83, reduction of
flange in ontogeny of crossing canal of thF; 84, fusellum of all post-thF thecae reduced to clathria; 85, loss
of reverse wall of th P in region of foramen of th F; 86, pipiograptid thecae; 87, loss of lacinia; 88, simplification
of clathria to produce ‘orthograptid’ thecae; 89, nema in obverse wall; 90, loss of septal spines. Retained
primitive characters: j, sigmoidal glyptograptid thecae; k, amplexograptid thecae.

380

PALAEONTOLOGY, VOLUME 30

‘G.’ austrodentatus group and descendants in the ‘G.’ teretiusculus species group (here recognized as
Oelandograptus gen. nov. and Hustedograptus gen. nov., respectively); and 2, the minor ‘G.’ anacan-
thus-G.' hudsoni species cluster (grouped here as Arnheimograptus gen. nov.). Climacograplus-\\kQ
species exist in the form of ‘C.’ typicalis, ‘C.’ inuiti, and similar species. These species have a thecal
form and astogeny different from that of the type species of Hall’s genus, C. bicornis. C. bicornis is
fully septate with a dicalycal th2^ and a Pattern D astogeny. Its thecae have vertical supragenicular
walls, semicircular apertures without apertural cusps, and no genicular flanges. These features
indicate a relationship between Climacograptus s.s. and the advanced pseudoclimacograptids among
the diplograptidae (see below). I propose to combine the species of the ‘C.’ typicalis group in a new
taxon, Genicidograptus gen. nov. From text-fig. 14 we can extract the following classification:

Subfamily orthograptinae subfam. nov.

Diagnosis. Aseptate to septate species with straight median septum; Pattern A or Pattern G astogeny;
strongly asymmetrical proximal end with sicula extensively exposed on obverse side; sicula com-
monly bearing paired antivirgellar spines.

Generic group taxa. Hustedograptus gen. nov.; Orthograptus Lapworth (= Rectograptus Pfibyl, non Ditto-
graptus Obut and Sobolevskaya).

Genus hustedograptus gen. nov.

Type species. Diplograptus uplandicus Wiman, 1895, from erratic boulders of Chasmops ( = Dalby) Limestone,
Upland, Sweden; Viruan Series (N. gracilis Zone ?). The reported occurrence of this species in boulders of
Centaurus Limestone ( = Folkeslunda Limestone, in part) (Wiman 1 895) has not been confirmed by subsequent
studies of this unit.

Diagnosis. Thecae smoothly sigmoidal glyptograptid in the proximal end, becoming orthograptid
distally; thecal apertures normal to rhabdosome axis or slightly introverted, with prominent paired
cusps or spines on lateral margin and with concave ventral apertural margin; median septum
straight with dicalycal theca th2* or substantially delayed; primordial astogeny follows Pattern A
but with short descending portion in crossing canal of th2^; proximal end broad and weakly
to markedly asymmetric; sicula with simple aperture, or aperture bearing paired lappets, or anti-
virgellar spines flanking concave dorsal margin.

Species included. D. notabilis Hadding, D. propinquus Hadding, G. teretiusculus (sensu Jaanusson 1960),
D. uplandicus Wiman, and G. vikarbyensis Jaanusson.

Discussion. Based on a restudy of the type specimens, the proximal end structure of ‘Z>.’ uplandicus
given by Wiman (1895) appears to be inaccurate. The structure is correctly illustrated in text-fig. 2,
based on abundant and excellently preserved material isolated from a limestone sample in the Holm
collections of the Naturhistoriska Riksmuseet, Stockholm. This sample is labelled "Diplograpsus,
Kuckers C2b’ and is lithologically identical to the distinctive Kukruse Limestone of Estonia. It is
my intention that the type species of the genus Hustedograptus gen. nov. be the biological entity
embodied by the Estonian material, an entity that I believe to be synonymous with Wiman’s D.
uplandicus.

Hustedograptus gen. nov. differs from Orthograptus Lapworth in its primordial astogeny (Pattern
A as opposed to Pattern G), the exposure of the sicula only to the level of the aperture of thF on
the obverse side of the rhabdosome, and by the presence of glyptograptid proximal thecae. Addition-
ally, in Orthograptus the dicalycal theca is never as early as the second thecal pair (in the type species,
O. quadrimucronatus Hall, th5^ or a later theca is dicalycal) and the sicula always bears paired anti-
virgellar spines, while in many species of Hustedograptus gen. nov. antivirgellar spines are absent.
Eor comparisons with Oelandograptus gen. nov. and Eoglyptograptus gen. nov. see discussion of
these taxa below.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIEICATION

381

Subfamily peiragraptinae Jaanusson, 1960, emend.

Diagnosis. Orthograptids with Pattern G or Pattern F primordial astogeny; strongly geniculate
thecae bearing genicular spines or flanges or both; rhabdosome generally aseptate and with fully
sclerotized periderm.

Generic group taxa. Group 1: Amplexograptus Elies and Wood, emend., Paraorthograptus Mu et al., 1974,
and Peiragraptus Strachan, 1954. Group 2: Geniculograptus gen. nov., Arnheimograptus gen. nov., and Gymno-
graptus Bulman.

Discussion. The Peiragraptinae comprises two informal generic groups. Within Group 1 (peiragrap-
tids), Paraorthograptus and Peiragraptus are little modified from Amplexograptus and both retain
the basic amplexograptid theca. Within Group 2 (geniculograptids), all three genera share the
unique Pattern F astogeny and narrow, gradually widening proximal end, but exhibit variously
modified thecal forms.

Genus amplexograptus Elies and Wood, 1907

Proposed Neotype species. Amplexograptus fatlax Bulman, 1962 (subject to approval by the ICZN), Hartfell
Shales, Scotland; Caradoc Series (principally C. wilsoni Zone).

Emended diagnosis. Thecae amplexograptid with short, slightly outwardly inclined supragenicular
walls, sharp geniculum bearing a genicular flange, and with cuspate thecal apertures horizontal to
slightly everted. Rhabdosomes may be partly septate, with th3* or some later theca dicalycal, but
are more commonly aseptate. Primordial astogeny follows Pattern G. Paired antivirgellar spines
generally present on dorsal margin of sicular aperture. ThF invariably bears a subapertural or
mesial spine but thG commonly does not.

Species included. A.fallax Bulman, D. leptotheca Bulman, A. maxwelli Decker, Climacograptus maniloulinensis
Parks, A. prominens Barrass, and C. bekkeri Opik.

Discussion. This name is here applied only to forms with amplexograptid thecae and a Pattern G
astogeny (see discussion of the Family Orthograptidae above). Amplexograptus is most similar to
Geniculograptus gen. nov. but differs in the form of its proximal end. Geniculograptus species such
as G. inuiti (Cox), possess a narrow proximal end based on a Pattern F astogeny in which the first
theca is tightly upturned and closely pressed against its protheca. Its metatheca extends distally to
the level of the bud of thF or beyond. In contrast, in Pattern G proximal ends thF is always
separated from its protheca by a gap through which th2' develops, and its aperture seldom reaches
the level of its primary bud (cf. text-figs. 8f, l, o, 9d, e, h, IOe, e, h). Homeomorphic members of
the Diplograptidae (e.g. M.’ munimentus Berry or "A.' confertus Lapworth) differ in their possession
of simple, semicircular to introverted thecal apertures that lack the lateral lappets of amplexograptid
thecae, and a Pattern C astogeny with dicalycal th2F as well as by the rapid enclosure on the
obverse side of the colony of the sicula, which lacks antivirgellar spines.

Genus geniculograptus gen. nov.

Type species. Climacograptus typicalis Hall, 1865, Lexington Limestone and Kope Formation, Cincinnati
Region, USA; Blackriveran to Maysvillian Stages (C. americanus to A. manitoulinensis zones).

Diagnosis. Aseptate, gradually widening rhabdosomes with narrow proximal end; slightly outwardly
inclined amplexograptid thecae bearing a variably prominent genicular flange; primordial astogeny
follows Pattern F; sicula is extensively exposed on the obverse side of the rhabdosome and bears
paired antivirgellar spines in addition to the virgella. ThF may or may not bear a mesial spine.
ThU bears no spines.

Species included. C. inuiti Cox, C. latus Elies and Wood, C. typicalis magnificus Twenhofel, and C. pygmaeus
Ruedemann.

382

PALAEONTOLOGY, VOLUME 30

Discussion. Species of Geniculograptus gen. nov. most resemble those of Amplexograptus in the form
of their thecae, but they differ astogenetically (see discussion above for Amplexograptus). They also
resemble certain members of the Monograptidae, such as Paraclimacograptus innotatus and P.
nevadensis, which also possess amplexograptid-like thecae with prominent genicular flanges, but
again these taxa differ in the form of their proximal end and rhabdosome architecture. Among
Geniculograptus species, the sicula bears prominent antivirgellar spines and is extensively exposed
on the obverse side of the rhabdosome, which is aseptate. Those of Paraclimacograptus, with their
Pattern H astogeny, are generally septate, the sicula is rapidly enclosed by the early thecae, and
there are no antivirgellar spines (see Table 2).

Genus arnheimograptus gen. nov.

Type species. Glyptograptus lorrainensis anacanlhus Mitchell and Bergstrom, 1977, Arnheim Formation, Cin-
cinnati Region, USA; Richmondian Stage {A. manitoulinensis Zone).

Diagnosis. Minute aseptate species with glyptograptid thecae; thecal apertures undulating to cuspate;
thl ' may or may not possess a mesial spine; primordial astogeny follows Pattern F.

Species included. G. anacantlius Mitchell and Bergstrom, G. hudsoni Jackson, and G. lorrainensis Parks.

Discussion. The rhabdosomes of these species exhibit a proximal end which is nearly identical to
that of Geniculograptus gen. nov. species. The common ancestor of the Arnheimograptus species
probably arose from one of these by the loss of the distinctive genicular flanges, converting the
amplexograptid thecae of Geniculograptus gen. nov. to the glyptograptid thecae of Arnheimograptus.
Also like Geniculograptus gen nov., this taxon differs from similar looking species of Glyptograptus
in being aseptate, by exhibiting a long slender sicula that is extensively exposed on the obverse side
of the colony, and in possessing antivirgellar spines on the dorsal margin of the sicula.

Genus gymnograptus Bulman, 1953, emend.

Type species. Gymnograptus linnarssoni (Moberg, 1896), Ogygiocaris Shale, Crassicauda (= Furudal) Lime-
stone, Baltoscandia; Uhakuan Stage (H. teretiusculus Zone).

Discussion. Taxon remains as described by Bulman except that it is here restricted to species which,
like the type species, possess a Pattern F primordial astogeny. Thus, species such as ‘G.’ retioloides
(Wiman), with its Pattern C astogeny, belong among the Diplograptidae (see below). G. linnarssoni
differs from typical Pattern F proximal form in that the sicula possesses only a single dorsal
antivirgellar spine.

Subfamily lasiograptinae Lapworth, 1879, emend.

Diagnosis. Species with Pattern G primordial astogeny; thecal periderm reduced to absent; thecae
with prominent lists and commonly bearing lacinia derived from genicular and lateral spines; thecae
lasiograptid to highly stylized, polygonal clathrium.

Generic group taxa. Group 1: Lasiograptus Lapworth, Hallograptus Lapworth, Orthoretiolites Whittington,
Neurograptus Elies and Wood, Pipiograptus Whittington, and INymphograptus Elies and Wood. Group
2; Plegmatograptus Elies and Wood, Arachniograptus Ross and Berry, Phormograptus Whittington, and
Archiretiolites Eisenack.

Discussion. This subfamily includes two generic groups that are clearly related but none the less
distinct in their colonial architecture. Group 1 (lasiograptids) consists of forms exhibiting lasiograp-
tid to Pipiograptus-Wkt thecae with prominent genicular spines; sicula and at least thU partly
sclerotized; typical Pattern G astogeny; clathria, lateral (septal) spines, and at least partial lacinia
common. Hallograptus has the appearance of a primitive stem group with respect to the core of
this assemblage because of its lack of a lacinia and somewhat better sclerotized thecae (which
resemble the proximal thecae of A. bekkeri and possess weaker clathria than the main group of

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

383

lasiograptids). [I consider the absence of a lacinia and the generally simple architecture of Ortho-
retiolites to be a derived condition based on considerations discussed more fully elsewhere.] This
group includes several retiolitid-like taxa formerly classed in the polyphyletic Archiretiolitinae
Bulman, 1955. Nymphograptus is included here with some reservation because its structure is very
incompletely known.

Group 2 (archiretiolitids) comprises retiolitid rhabdosomes with a sclerotized sicula bearing
antivirgellar spines; thecal fusellum represented only by clathria; clathrium generally irregularly
organized and complexly connected with well developed lacinia; primordial astogeny like that of
Archiretiolites. Although the phytogeny of these highly derived and structurally reduced archiretiol-
itid colonies is somewhat difficult to evaluate, their orthograptid sicula with antivirgellar spines and
a lacinia derived from paired genicular or subapertural spines allies the archiretiolitids with the
lasiograptids among the Orthograptidae, and clearly separates them from any close phylogenetic
relationship with the Silurian Retiolitinae.

The Lasiograptinae include an unusually large number of monotypic genera and it is quite likely
that, as these taxa become better known, several may prove to be synonymous (e.g. Pipiograptus with
Neurograptus). As discussed by Finney (1980) the abrograptid Reteograptus of the Phyllograptidae
Lapworth (emend. Cooper and Fortey, 1982) is distinguished from the archiretiolitids by its pos-
session of an isograptid primordial astogeny and a small, simple sicula that lacks antivirgellar
spines.

Incertae familiae

Genus oelandograptus gen. nov.

[= Undulograptus partim imkm?,, 1980; non UnduhgraptusBoucs^k, 1973]

Type species. Glyptograptus austrodentatus oelandicus Bulman, 1963a, Helen Limestone (= Glaukonithaltig
gra Vaginatumkalk of Holm and Bulman), Oland; Ontikan Series, Kunda Stage {D. hinmdo and D. hifidiis
zones).

Diagnosis. Median septum undulatory, weakly sigmoidal thecae with long, outwardly inclined
infragenicular wall, sharply rounded geniculum and short, nearly vertical supragenicular wall.
Thecal apertures slightly everted and undulatory with a concave ventral margin. Sicular aperture
simple. Primordial astogeny follows Pattern A and the proximal end is evenly rounded to somewhat
blunt and nearly symmetrical.

Species included. G. a. americanus Bulman, G. austrodentatus Harris and Keble, G. a. oelandicus Bulman, and
G. sinodentatus Mu and Lee; G. curvithecatus Mu and Lee is imperfectly known but may also belong here.

Discussion. Oelandograptus gen. nov. differs from its contemporaries Hustedograptus gen. nov.,
Undulograptus Boucek, and Pseudoclimacograptus (Arcliiclimacograptus) subgen. nov. in several
respects. Species of Oelandograptus gen. nov. are most similar to the archaic Pseudoclimacograptus
species of P. {Arcliiclimacograptus) subgen. nov. but differ in possessing a Pattern A astogeny, an
undulatory median septum, and weakly geniculate thecae with apertures normal to the rhabdosome
axis; the latter exhibit a Pattern C astogeny (see Table 2), generally have a sharply zigzag median
septum, and pseudoclimacograptid thecae with introverted apertures. Undulograptus Boucek, as
redefined here, is a monotypic taxon with a narrow proximal end based on primordial astogenetic
Pattern B and a climacograptid thecal form exhibiting a nearly vertical supragenicular wall. Hustedo-
graptus gen. nov. has more strongly glyptograptid proximal thecae, thecae with prominently cuspate
apertures, and a more asymmetrical proximal end (still based on Pattern A), together with a straight
median septum.

Orthograptid history. The Orthograptidae were an important constituent of graptoloid faunas from
the early Llanvirn to the late Ashgill. Their initial diversity was eclipsed by that of the Diplograptidae
but advanced orthograptids dominated the faunas of the Upper Ordovician. This situation changed

384

PALAEONTOLOGY, VOLUME 30

radically at the end of the Ordovician when the orthograptids appear to have become extinct. All
of the early Silurian diplograptaceans, based on their possession of astogenetic Patterns H and 1,
were monograptids. [I am indebted to Anton Kearsley who pointed out to me, in 1981, the
magnitude of these late Ordovician and early Silurian extinctions and their effects on the taxonomic
composition of the Silurian diplograptacean radiation.] Thus, the Silurian species currently referred
to the genera ' Amplexograptus' and ‘‘Orthograptus' are not particularly closely related to the species
of these Ordovician taxa.

One outstanding problem for the systematics of Ordovician diplograptaceans is the fact that our
knowledge of the astogeny of many of the Arenig to Llandeilo species previously included in
Diplograptus, Glyptograptus, and Amplexograptus is quite limited. Some of these species may possess
either a Pattern A or G astogeny, while others will certainly be found to exhibit a Pattern C or
perhaps even a Pattern B proximal end structure. Additional study of material from this interval,
preserved in relief or in isolated preparations, is needed to resolve the cladistic relationships and
systematic associations of the early diplograptaceans. Such study is of particular importance because
it is in just this interval that several of the fundamental steps in diplograptacean evolution occurred,
including the establishment of all four of the diplograptacean families.

Family diplograptidae Lapworth, 1873, emend.

Diagnosis. Rhabdosomes generally septate with pseudoclimacograptid to climacograptid and am-
plexograptid to orthograptid (rarely gymnograptid or lasiograptid) thecae, and with primordial
astogenetic Pattern C or its derivatives D and E. Sicula commonly deflected toward its dorsal side
and rapidly enclosed by the second thecal pair. Sicular aperture simple, lacking paired antivirgellar
spines. Many species with a three-vaned nematularium formed from an intact nema.

Discussion. The second major subclade among the Diplograptacea is the pseudoclimacograptines
and their descendants, the diplograptines and true climacograptines. Text-fig. 1 5 presents a dado-
gram of the branching history of the taxon. The primitive diplograptid astogenetic pattern. Pattern
C, differs from the ancestral diplograptacean Pattern A in the right-handed rather than left-handed
origin of th2‘ from thF. The crossing canal of th2*, on its way toward the virgella, grows out from
the sicula and arcs around the crossing canal of thP (text-figs. 2l, o, p and 4). This peculiar origin
of th2^ (best known in P. eurystoma Jaanusson, P. angulatus, ‘C.’ distichus Eichwald, and ‘G.’
retioloides) is shared with the advanced diplograptid pattern. Pattern D (e.g. P. scharenbergi; cf.
text-fig. 6), clearly indicating that these two groups share a close common ancestry. This relationship
is also fully corroborated by the correspondence in thecal form and rhabdosome architecture among
these graptolites. Indeed, the form of the thecae and shape of the proximal end indicate that the
advanced pseudoclimacograptines, such as P. scharenbergi, appear to be directly descended from
an archaic form with a Pattern C astogeny (e.g. P. eurystoma). The species ‘L.’ haplus Jaanusson,
with its compact proximal end and Pattern C astogeny, also appears to be very closely allied to
pseudoclimacograptines like P. eurystoma.

Primordial astogenetic Pattern D is not confined to species with pseudoclimacograptid rhabdo-
somes. It is also present in C. bicornis (Hall), the type species of Climacograptus (text-fig. 6j-p), and
several other related species (see text-fig. 15). This high burden synapomorphy unambiguously allies
Climacograptus, sensu stricto, with the advanced pseudoclimacograptines and removes them from
the possibility of any close phylogenetic relationship to species of either the Geniculograptus typicalis
group (see above) or the ‘C.’ brevisj^C.' normalis group, which possess a Pattern H development
and are allied to the Monograptidae (see below). Finally, several species such as ‘C.’ distichus, ‘C.’
meridionalis Ruedemann, and 'A.' munimentus Berry possess a Pattern C astogeny. Given the
relationships of C. bicornis to advanced pseudoclimacograptids with a Pattern D astogeny, these
species must have acquired their climacograptid thecae independently of C. bicornis and related
species (in contrast to Riva’s 1976 suggestions). Hence, they too must be classed separately from
the true climacograptines. However, it should not be difficult to recognize these as a separate group,
given their proximally zigzag median septum and broad, blunt proximal end.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

385

> >

10 12

11

92

100

48-

91+

51

101

102
— h-

49,50,52-55

104

99--

3 < <
C/) LL LL

103

Archiclimacograptus ?
camptochilus

Archiclimacograptus

angulatus

■ Archiclimacograptus luperus

. Archiclimacograptus
eurystoma

. Prolasiograptus haplus

■ Pseudoclimacograptus

scharenbergi

■ Pseudoclimacograptus

clevensis

Incertae

Subfamiliae

Climacograptus (C.) bicornis

93-

94-

105,106 *—m-Climacograptus (D.) spiniferus
56-59

Dicaulograptus hystrix

Pseudamplexograptus distichus
Diplograptus pristis
“Lasiograptus" armatus
Urbanekograptus retioloides

O

TEXT-FIG. 15. Cladogram and classification of the Diplograptidae. Synapomorphies 1-61 as in text-fig. 13a;
91, thU and thP with open curvature producing a broad, blunt proximal end; 92, metathecae of thl' and
thl^ with pronounced convex ventral walls and introverted apertures; 93, thl‘ and thP with subhorizontal
metathecae and generally with horizontal to everted apertures throughout; 94, 'amplexograptid' thecae with
simple semicircular to introverted apertures and short, nearly vertical supragenicular walls; 95, straight median
septum; 96, distal thecae orthograptid to glyptograptid; 97, genicular spines; 98, ‘gymnograptid’ thecae with
well-defined clathria; 99, unique spinose dorsal margin of sicular aperture and with isolated, strongly intro-
verted and spinose thecal apertures bearing fenestrate lateral processes; 100, flange of th2^ formed on dorsal
side of thF; 101, narrow, evenly rounded proximal end with tightly upturned thU; 102, open, semicircular
apertures; 103, ‘lasiograptid’ thecae lacking genicular spines or lacinia but with clathria; 104, climacograptid
thecae with relatively long, straight, and nearly vertical supragenicular wall; 105, straight median septum; 106,
narrow, nearly parallel-sided rhabdosome. Retained primitive characters: I, archaic geometry of descending
crossing canal of th2' along mid-line of colony in narrow proximal end; m, broad proximal end.

Species of the C. spiniferus Ruedemann species group, with their Pattern E astogeny, share with
C. bicornis and C. cf. caudatus (Strachan 1974, cf. pi. 6, figs. 1, 7, 9 with figs. 2 and 3, and text-fig. 7
herein) an unusual sicula; the prosicula is absent and replaced by a stirrup-like set of rods; the
metasicula is strongly deflected to the dorsal side and bears a series of regularly spaced dark bands.
These synapomorphies link the C. spiniferus and C. bicornis species groups. Based on the overall
morphology, proximal end structure, and stratigraphic occurrence of these species, Riva (1976)
arrived at the conclusion that C. spiniferus is descended from C. bicornis. A similarly close relation-
ship is shown in text-fig. 15. Despite the highly simplified astogeny of the distinctive Pattern E
species, they are best retained within Hall’s genus, although they can easily and usefully be recogni-
zed as a subgenus: C. {Diplacanthograptus) subgen. nov.

386

PALAEONTOLOGY, VOLUME 30

The taxa Metaclimacograptus and Clinoclimacograptus, recognized by Bulman and Rickards
(1968) and included as subgenera of Pseudoclimacograptus, appear to possess a Pattern H astogeny.
Michael Melchin (pers. comm.) has acquired numerous isolated growth stages of several metaclima-
cograptid and clinoclimacograptid species during his studies of Llandoverian biserial diplograptace-
ans from arctic Canada. These specimens confirm that even the most pseudoclimacograptid-like of
these species, M. orientalis Obut and Sobolevskaya, possess a Pattern H astogeny. They are, thus,
homeomorphic with Ordovician pseudoclimacograptines and represent a Silurian ‘re-invention’ of
this rhabdosome architecture among the Monograptidae, following the extinction of pseudoclima-
cograptine diplograptids in the latest Ordovician.

The Diplograptidae appear to have undergone a significant dichotomy early in their history. In
the second branch of the family, thP and thF acquired a more horizontal growth form with only
the apertural regions sharply upturned. This configuration gave these diplograptids a broad and
rather blunt proximal end that commonly attains widths nearly as great as the maximum colony
width. This group, the subfamily Diplograptinae, includes species with a much broader range of
thecal types than exist among the Climacograptinae. Within the subfamily occur several species-
groups in which rhabdosomes exhibit thecae homeomorphic with those of such orthograptid taxa
as Amplexograptus, Orthograptus, Hallograptus, and Gymnograptus. The most important of these
taxa are Diplograptus M‘Coy and Pseudamplexograptus gen. nov. The latter comprises the archaic
amplexograptus-like species, such as "A.' confertiis (Lapworth), "A.' munimentus Berry, and ‘C.’
distichus (Eichwald), and it includes most of the species from the Arenig to the Llandeilo that have
formerly been assigned to Amplexograptus— 'whtre. they have resided uncomfortably (see Bulman
1962).

The status of Diplograptus remains particularly problematic. D. pristis (Hisinger) is known only
from flattened material, but Skoglund’s (1963) preparations of isolated, flattened specimens from
the type area in Sweden provide useful information. Its broad, blunt proximal end that rapidly
encloses a short, stout sicula lacking antivirgellar spines suggests a Pattern C astogeny. Thecal
characters of the proximal end are also like those of Pseudamplexograptus gen. nov. This suite of
characters exhibited by D. pristis appears to be shared with many other species assigned to this
genus. Additionally, I have isolated specimens of a speeies similar to D. decoratus Harris and
Thomas from the Table Head Formation, Western Newfoundland, and its growth stages exhibit a
Pattern C astogeny. Finally, it is noteworthy that, like many pseudoclimacograptines, species of
Diplograptus (sensu stricto) often possess a retuse, three-vaned nematularium derived from the
distal extremity of their intact nema (see Ruedemann 1904; Mitehell and Carle 1986). To my
knowledge, nematularia of the conulare-iypQ (Muller and Schauer 1969) do not occur in other
diplograptacean families during the Ordovician.

Not all of the Arenig to Caradoc species that have been classed with Diplograptus exhibit these
features, however. ‘Z).’ propinquus Hadding and ‘£).’ notabilis Hadding belong among the Pattern
A-bearing Hustedograptus gen. nov., in which weak thecal gradients encompass glyptograptid to
orthograptid shapes. Still others (e.g. ‘Z>.’ toernquisti Hadding) possess a Pattern H astogeny that
places them among the Monograptidae. Considerable additional work needs to be done on this
group.

Subfamily climacograptinae Freeh, 1897, emend.

Diagnosis. Diplograptids with zigzag to straight median septum and pseudoclimacograptid to
climacograptid thecae; thH and thF grow distally in a gentle arc producing a rounded proximal
end. Primordial astogeny is Pattern C or D.

Generic group taxa. Pseudoclimacograptus Pfibyi, s.s. (non Metaclimacograptus Bulman and Rickards; non
Clinoclimacograptus Bulman and Rickards), comprising the subgenera P. (Pseudoclimacograptus) Pfibyi and
P. (Archiclimacograptus) subgen. nov.; Prolasiograptus Lee; Climacograptus Hall, 1865, emend., comprising
the subgenera C. (Climacograptus) Hall and C. (Diplacanthograptus) subgen. nov.; Dicaulograptus Rickards
and Bulman.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

387

Subgenus pseudoclimacograptus (pseudoclimacograptus) Pfibyl, 1947, emend.

Type species. Climacograptus scharenbergi Lapworth, 1876, Balclatchie beds and Lower Hartfell Shale, Scot-
land; Caradoc Series (‘D.’ multidens and ‘C’ wilsoni zones).

Diagnosis. Climacograptines with narrow, rounded proximal end and Pattern D astogeny, including
a dicalycal th2^. Region between the metatheca of thP and thF on the reverse side occupied by
the protheca of th2L

Species included. Those few species known to possess this suite of characters include P. scharenbergi (Lapworth)
and P. clevensis Skoglund.

Subgenus pseudoclimacograptus (archiclimacograptus) subgen. nov.

Type species. Pseudoclimacograptus angulatus sebyensis Jaanusson, 1960, Grey Seby Limestone, Seby and
Folkeslunda Limestone, Sjostorp, Oland; Viruan Series, Lasnamagi Stage (D. murcliisoni Zone).

Diagnosis. Taxon with broadly rounded proximal end and Pattern C astogeny; median septum
sharply zigzag to undulatory. Th2* generally dicalycal and with the region between the metatheca
of thP and thF on the reverse side occupied by an exposed patch of the right-handed crossing
canal of th2* flanked or enclosed by the prothecae of both th2^ and th2^.

Discussion. P. {Archiclimacograptus) subgen. nov. differs from the nominate subgenus by its pos-
session of a relatively broad and blunt proximal end, based on a Pattern C astogeny, and a
dicalycal th2\ in contrast to the rather more narrow and evenly rounded proximal end of P.
{Pseudoclimacograptus), with its Pattern D astogeny and dicalycal th2^ (see Table 2). It differs from
Pseudamplexograptus gen. nov. in the form of the thecae and median septum (see below). Species
of Metaclimacograptus and Clinoclimacograptus exhibit a Pattern H astogeny that produces a
slender, nearly parallel-sided rhabdosome and allies them unambiguously with the Monograptidae.
They also possess a complexly folded, rather than strictly zigzag, median septum.

Although the species P.? camptochilus Skevington and P.? formosus Mu and Lee are known from
isolated and well-preserved relief material, respectively, and possess the thecal characters of the
genus, they are doubtfully included in P. {Archiclimacograptus) subgen. nov. The symmetrical form
of their proximal end, in which the descending crossing canal of th2^ lies along the mid-line of the
colony and is extensively exposed, suggests a Pattern A rather than Pattern C astogeny. This
remains to be confirmed from the study of early growth stages, however. In any event, their primitive
geometry places this species group as a paraphyletic stem group in text-fig. 15.

Species included. Structurally well-known members of this taxon include P. angulatus angulatus (Bulman), P.
a. sebyensis Jaanusson, P. luperus Jaanusson, P. marathonensis Clarkson, P. modestus (Ruedemann), and P.
oliveri Boucek. The form of the proximal end of P. eurystoma Jaanusson closely resembles that of typical
members of P. {Pseudoclimacograptus), differing only in its retention of a Pattern C astogeny with an exposed
patch of the crossing canal of th2* between the apertures of thP and thC on the reverse side of the
rhabdosome.

Genus prolasiograptus Lee, 1963, emend.

Type species. Lasiograptus retusus Lapworth, 1880, ‘upper Llandeilo shales of the neighbourhood of Llandrin-
dod Wells, Radnorshire’, Wales; Llandeilo Series {N. gracilis Zone ?).

Diagnosis. Taxon restricted to Climacograptinae with Pattern C astogeny, lasiograptid thecae, and
without lacinia.

Discussion. Distinguished from similar looking species of Lasiograptus by its proximal end structure,
absence of antivirgellar spines, and simple thecal apertures.

Species included. Known to include L. haplus Jaanusson, 1960, in addition to the type species.

388

PALAEONTOLOGY, VOLUME 30

Genus climacograptus Hall, 1865, emend.

Type species. Graptolithus hicornis Hall, 1848, Austin Glen Greywacke, Norman’s Kill, New York (but also
common in equivalent units world wide); Mohawkian Series, Blackriveran to Shermanian stages {N. gracilis
and midtidens zones).

Emended diagnosis. Climacograptines with climacograptid thecae, bearing semicircular thecal exca-
vations that lack apertural cusps, nearly vertical supragenicular walls, and a sharp geniculum
without genicular flanges. Proximal end narrow, evenly rounded to blunt, based on a Pattern D or
Pattern E astogeny. Sicula strongly deflected toward its dorsal side, generally lacking a normal
prosicula, and with an aperture bearing only a prominent virgella. Rhabdosome septate with
proximally zigzag to straight median septum. Th2^ generally dicalycal.

Discussion. Species of Climacograptus Hall differing from similar looking taxa in Geniculograptus
gen. nov., Glyptograptus, Pseudamplexograptus gen. nov., and Undulograptus in the form of the
thecae, or the proximal end (including the configuration of the sicula), or both. Genus consists
of two subgenera.

Subgenus climacograptus (climacograptus) Hall, 1865, emend.

Type species. Graptolithus hicornis Hall, 1848, see above.

Diagnosis. Characters of genus, but restricted to species with Pattern D astogeny.

Species included. Species well enough known to be assigned with confidence include C. caudatus Lapworth,
C. hastatus T. S. Hall, C. raricaudatus Ross and Berry, C. tubuliferous Lapworth, Diplograpsus minimus
Carruthers, and G. putillus Hall.

Discussion. Among species of the nominate subgenus the crossing canal of thF exhibits an evenly
rounded curvature and constant diameter as it sweeps across the reverse side of the sicula at the
level of the sicular aperture. The globular protheca of th2^ occupies the region encircled by thH
and commonly bears a distinct dimple that corresponds to the end of a list which links the enclosing
protheca with the hood-like crossing canal below (see text-figs. 5 and 6). The sicula commonly bears
a long stiff virgella that projects downward. The first thecal pair may be without spines, or thl'
alone, or both it and thF may possess prominent mesial spines.

C. {Climacograptus) is most similar to Pseudamplexograptus gen. nov. but differs in that the
latter, despite their climacograptid thecae, retain a Pattern C astogeny, which produces a broad,
blunt proximal end and a rather wide, tabular rhabdosome (see also Table 2). The sicular form of
Climacograptus is highly distinctive but difficult to observe. Some species of C. {Climacograptus),
such as C. (C.) caudatus (see text-fig. 6r), exhibit a proximally zigzag median septum, but in all
species the median septum is straight after the first few thecal pairs, thus distinguishing them from
species of Pseudoclimacograptus.

Subgenus climacograptus (diplacanthograptus) subgen. nov.

Type species. Climacograptus spiniferus Ruedemann, 1908, lower Utica Shale, Hudson and Mohawk River
valleys; Mohawkian and Cincinnatian Series, Blackriveran to Edenian Stages (C. americanus to C. pygmaeus
zones).

Diagnosis. Species of Climacograptus with Pattern E primordial astogeny and a narrow, asymmetri-
cal proximal end with the sicula lying almost entirely to the right of rhabdosome mid-line in obverse
view; sicular aperture oriented at about 70° from the rhabdosome axis and bearing a stout (but not
necessarily long) virgella deflected across the sicular aperture. Virgella commonly matched by a
long mesial spine on thU such that they form a pair which, in an undeformed state, is symmetrical
about the rhabdosome axis. More rarely thU bears a small mesial spine as well.

Species included. C. dorotheus Riva and C. spiniferus Ruedemann. At present, the astogenetic details of C.
venustus Hsu, C. longispinus T. S. Hall, and related species are unknown, but their possession of a reflected

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

389

virgella that is symmetrical with a mesial spine on thP, plus a strongly asymmetrical proximal end, both
suggest that they belong here rather than in C. {Climacograptus).

Discussion. Differs from C. {Climacograptus) in its possession of a Pattern E astogeny and in the
unique configuration of its sicula.

Subfamily diplograptinae Lapworth, 1873, emend.

Diagnosis. Diplograptids with a broad and blunt proximal end that arises from a strong lateral
component in the growth of the first thecal pair and a Pattern C primordial astogeny; thecae
amplexograptid to orthograptid, or rarely gymnograptid.

Generic group taxa. Diplograptus M'Coy, Pseudamplexograptus gen. nov., and Urhanekograptus gen. nov. The
status of Diplograptus remains uncertain due to our poor understanding of the type species, D. pristis (Hisinger)
(see discussion of Diplograptidae). The species of the ‘L.’ armatus-porrectus Bulman group may also belong
in this family. These species appear to combine amplexograptid thecae bearing prominent genicular spines
with a diplograptid-like proximal end, but the details of their development are unknown. Similarly, many of
the other species referred to ^ Lasiograptus' and " Hallograptus' among the Llanvirn and Llandeilo faunas will
probably prove to belong here.

Genus pseudamplexograptus gen. nov.

Type species. Lomatoceras distichus Eichwald, 1840, Lasnamiigi strata, Estonia, but best known from the Seby
and Folkeslunda Limestones, Oland, Sweden; Lasnamiigi Stage (upper D. murchisoni Zone).

Diagnosis. Species having broad, tabular, and nearly parallel-sided rhabdosomes with amplexo-
graptid thecae throughout. Thecal excavations deep and semicircular to somewhat restricted, com-
monly with strong apertural selvage; apertures horizontal to introverted and lacking lateral cusps.
The supragenicular wall is short and commonly of similar height to that of the thecal excavation;
geniculum sharp and frequently with genicular flange. Proximal end blunt and nearly as wide as
distal colony width. Primordial astogeny follows Pattern C; th2* generally dicalycal with proximally
zigzag to straight, complete median septum. Sicula exposed on obverse side only to level of thG
aperture, or slightly above, and bearing only a stout virgella.

Species included. C. confertus Lapworth, C. distichus Eichwald, C. meridionalis Ruedemann, A. latus Bulman,
A. maxwelli Ekstrom {non Decker), and A. munimentus Berry.

Discussion. Pseudamplexograptus differs from Amplexograptus in its proximal end structures (Pat-
tern C rather than Pattern G astogeny and morphological correlates: see Table 2), absence of
antivirgellar spines, extent of exposure of the sicula, and in the simple thecal excavations. For
comparison with Climacograptus and Pseudoclimacograptus see discussion of these taxa above.

Genus urbanekograptus gen. nov.

Type species. Climacograptus retioloides Wiman, 1895, from erratic boulder of Scandinavian origin but
probably from the Crassicauda { = Furudal) or Ludibundus (= Dalby) Limestones, Sweden; Uhaku or
Kukruse Stages {H. teretiusculus or N. gracilis zones).

Diagnosis. Diplograptines with gymnograptid thecae bearing complex spinose genicular processes.
Pattern C primordial astogeny, and nearly symmetrical, blunt proximal end. ThD and thG with
orthograptid shape and prominent apertural spines.

Discussion. Taxon distinguished from homeomorphic Gymnograptus by its astogeny, the broad
shape of the proximal end with its subhorizontal first two thecae and by the absence of antivirgellar
spines. Taxon presently monotypic.

Summary history of the Diplograptidae

Compared to the other diplograptacean superfamilies the Diplograptidae form a relatively small
and close-knit assemblage. They appear to have achieved their maximum diversity and peak abun-

390

PALAEONTOLOGY, VOLUME 30

dance early in the history of the diplograptacean radiation— during the Llanvirn and Llandeilo
(text-fig. 17), when members acquired a range of thecal shapes and rhabdosome designs that are
strikingly similar to those evolved later among the Orthograptidae. But by the mid-Caradoc, the
diplograptids had begun to wane in importance, losing their status as common and numerous
components of the diplograptid fauna. Species of Climacograptus (particularly in the subgenus
Diplacanthograptus), however, did remain as highly distinctive elements and continued to evolve
rapidly (hence their common use in zonation and chronostratigraphic correlation). Yet, they too
were extinguished in the Ordovician-Silurian mass extinction. The Diplograptidae, like the Ortho-
graptidae, apparently made no contribution to the great Silurian diplograptacean renaissance.

Family monograptidae Lapworth, 1873, emend.

Diagnosis. Rhabdosomes with narrow, asymmetrical proximal end and simple sicula; colonies may
be biserial, uni-biserial, or fully uniserial. In biserial taxa the first two thecae are closely pressed to
the sicula and lack mesial spines. Primitively, thecae are glyptograptid to climacograptid, but are
modified to petalograptid, pseudoclimacograptid, or variously isolate, lobate, hooked, or triangu-
lar—particularly among the Monograptinae. Primordial astogeny is Pattern B, modified to Patterns
H, I, or the monograptid pattern. Silurian representatives develop virgellar meshworks and ancora-
based retiolitid colonies while others develop thecal and sicular cladia to re-establish multi-stiped,
spreading colony forms.

Discussion. The fourth major division of the Diplograptacea encompasses ‘G.’ dentatus Brongniart,
its congeners, and their descendants. Bulman (1963fi) considered the prosoblastic form of thF in
the astogeny of ‘G.’ dentatus to be distinctive of that species and all its descendants. This has proved
not to be true, however. Derived members of both the Orthograptidae and Diplograptidae also
develop a prosoblastic thD, i.e. species with Patterns E, F, and G. The distinctive elements of a
Pattern B astogeny, of which ‘G.’ dentatus is the prime example, comprise the suppression of the
upward-growing flange present in the ontogeny of th2* in the primitive diplograptid pattern (Pattern
A), the consequent J -shaped growth of the crossing canal of th2', and the late origin of th2^ by
a pattern of differentiation like that of distal thecae. Thus, this pattern has only two crossing
canals and three primordial thecae, compared to the primitive pattern of three and four respec-
tively. The Llanvirn-Llandeilo species that possess this pattern accordingly exhibit a relatively
advanced proximal structure compared to their contemporaries among the Orthograptidae and
Diplograptidae.

Text-fig. 13 indicates that the species group with Patterns H and I shares a common ancestry
with Pattern B species. Since Pattern I is restricted to Silurian species, I have based the inference of
common ancestry on characters shared between Patterns B and H. Pattern H astogeny is highly
simplified and possesses few unique characters apart from the loss of the more complex features of
other diplograptacean astogenies. Thus, establishing the sister group relations of Pattern H species
poses a difficulty. The overall shape of the proximal end, as well as the ontogeny of thF, are like
those in species with astogenetic Patterns E and F— both of which are also highly derived and
simplified patterns. However, species with Patterns E and F primordial astogenies exhibit distinctive,
derived features of the sicula or of its relationship to the colony that unambiguously ally them with
the orthograptids or diplograptids and exclude any close relationship with Pattern H species. For in-
stance, in the geniculograptids (Peiragraptinae, generic group 2), which possess a Pattern F astogeny,
the sicula is extensively exposed on the obverse side of the colony and its aperture bears a pair of
antivirgellar spines— all synapomorphies shared among the Peiragraptinae as a whole. In contrast,
species with Pattern H retain the primitive conditions; the sicula is rapidly enclosed by thF and
th2‘ on the obverse side of the rhabdosome and its aperture does not bear antivirgellar spines.
These are also still present in C. {Diplacanthograptus) spiniferus (the only species with a Pattern E
astogeny in which the details of its course are known), but, once again, the sicula exhibits a suite of
unique features: the metasicula is strongly defieeted toward its dorsal side and exhibits a series of
regularly spaced bands comprising condensed fuselli, while the prosicula is replaced by a set of rods

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

391

that unite to form the nema. These characters are shared with C. {Climacograptus) bicornis and
other species with a Pattern D astogeny. Since this suite of synapomorphies appeared among the
Climacograptinae prior to the derivation of the Pattern E astogeny, and since the species with a
Pattern H astogeny lack all of these characteristic sicular features, the similarity between these
highly simplified primordial astogenetic patterns must be analogous and must have been derived in
parallel. Furthermore, th2' is commonly the dicalycal theca in Pattern H species, but all known
Pattern F species are aseptate and no species of Climacograptus s.s. (whether with a Pattern D or E
astogeny) is known to have a dicalycal theca earlier than th2^.

The combination of a prosoblastic th2* that gives rise to th2^ from its upward-growing segment,
and which is also dicalycal, is found only in species with a Pattern H or a Pattern B astogeny.
Additionally, a thecal form strikingly like that of the early glyptograptines of the G. euglyphus
group is present in ‘G.’ jaroslovi (Boucek). Contrary to the inexplicable thecal diagram given by
Boucek (1973, text-fig. 36d), "G.' jaroslovi appears to exhibit a Pattern B proximal end structure.
Given these resemblances, I conclude that Glyptograptus s.s. (but including members of the ‘C.’
brevis- C.' normalis and related lineages) is the sister group of the primitive, Glyptograptus-Wko., and
Pattern B-bearing species (e.g. ‘G.’ dentatus and ‘’G.' jaroslovi) grouped below as Eoglyptograptus
gen. nov.

It is among species with a Pattern H astogeny that we at last encounter the diplograptaceans
with glyptograptid thecae that are congeneric with Diplograpsus tamariscus Nicholson, the type
species of Glyptograptus Lapworth. The Climacograptus-hke species of this group, however, are
only homeomorphic with C. bicornis and not closely related to it, as discussed above. Pfibyl’s (1947)
taxon Paraclimacograptus, with ‘C.’ innotatus as its type, is available to accommodate those Silurian
(and possibly Ordovician) glyptograptines with climacograptid thecae and prominent genicular
flanges. However, it is unclear to what extent the other Silurian and Ordovician ‘climacograptines’
such as ‘C.’ rectangularis, ‘G.’ normalis, ‘C.’ mohawkensis, and ‘C.’ brevis constitute a true clade
separate from Glyptograptus (see Bulman \963b, p. 413; 1970, p. VI 03). Among the many Llando-
verian glyptograptine species the distinction between Glyptograptus and ^Climacograptus' is entirely
arbitrary (see Rickards et al. 1977, p. 19). At the present time, it seems preferable to group all these
species together under the genus Glyptograptus Lapworth. Detailed morphometric studies may help
to delineate some useful and recognizable subclades within this complex array of structurally simple
taxa.

Pattern I-bearing species, which comprise the Retiolitinae (including Petalograptus and Cephalo-
graptus, as well as the ‘retiolitids’ themselves: see text-fig. 16) and the Dimorphograptinae, share
with Pattern H-bearing species, the Glyptograptinae: 1 , the propensity of the right lateral wall of
the metatheca of thH to be free of its protheca on the reverse side, and thus to form a free-standing
edge, as in ‘C.’ brevis, or a broad reverse wall that extends on to the sicula, as in Petalograptus (see
text-fig. 12a, b); 2, the tendency for the left lateral wall to enclose much or all of the descending
portion of its protheca on its obverse side; and 3, the continued presence of the plesiomorphic
dicalycal th2L The primordial astogenetic Patterns H and I characterize all of the biserial Silurian
diplograptaceans with the exception of the retiolitines. Silurian species currently identified as Diplo-
graptus, Amplexograptus, or Orthograptus will have to be either subsumed by Glyptograptus or
Petalograptus or renamed if their heritage is to be properly reflected and justice done to the true
magnitude of the Late Ordovician extinction. None possess the characteristic astogenies of their
Ordovician homeomorphs. Obut (1949), Obut and Sobolevskaya (1968), and others have erected a
number of genera based on these unique Llandoverian glyptograptines and their phylogenetic
significance needs to be established.

The retiolitines have an even more highly derived proximal end structure than the other mono-
graptids. Since the early thecae and even the metasicula are wholly unsclerotized, apart from the
stylized clathrium, it is hardly possible to compare their primordial astogeny with that of the non-
retiolitid diplograptaceans. None the less, they too possess proximal end structures that ally them
with a sister group— in this case Petalograptus. In their recent work on the retiolitines. Bates and
Kirk (1984) demonstrated that several Petalograptus species, such as ‘O.’ obuti, possess an ancora.

392

PALAEONTOLOGY, VOLUME 30

FAMILY

SUB-
FAMILY r

MONOGRAPTIDAE

Glyptograptinae

Dimorphograptinae Retiolitinae

O)

o

LLI

3

2

c

■§

a

O)

o

UJ

X

%

CO

CO

CL

O)

O

%

c

O CD

o

2

Q.

§;

CD

■§

Cj

0)

3

a

CO

V.

O)

o

c

CO

-c

QJ

O)

3

O)

o

0

CO

1
o
o
.c

O

S

o

CO

<0

a

CO

O)

■§

5

CO

CO

CL

13

o-

c

O)

:§

CL

CO

ct

Oi

o

•c

9-

o

9

Q

3

Q.

CO

O)

■§

O)

o

'co

2

”5

CD

C

o

o

CD

O

s

<x>

CL

C

O)

3

c

o

CO

E

CO

3

a

CO

k.

CD

O

o

3

Q.

2

Q.

CO

b)

o

tr

i>s

O

TEXT-FIG. 16. Cladogram and classification of the Monograptidae. Synapomorphies 1-61 as in text-fig. 13a;
107, th2* J-shaped and without upward growing flange in its prothecal ontogeny; 108, thecal elongation
to produce double sigmoidal curvature; 109, undulatory median septum; 110, thecal apertures hooded by
overhanging genicular flanges; 111, orthograptid thecae; 112, strongly acicular proximal end with dorsal side
of sicula free for nearly its entire length; 113, metatheca of thP becomes upturned well above sicular aperture;
114, metatheca of thP reduced; 115, metatheca of thF absent but protheca retained; 116, thP absent; 117,
thecae inclined at high angle to axis of rhabdosome and with everted aperture; 118, thP and thF with concave
ventral walls; 119, rhabdosome aseptate; 120, thecae elongate and with great overlap; 121, ancora incorporated
into thecal clathria; 122, fusellar periderm reduced to clathria; 123, metasicula suppressed; 124, clathrial
elements corresponding to edges of interthecal septa present; 125, ‘reticulum’ that forms separate, lacinia-like
mesh which encloses, but is free of, clathria along median region of rhabdosome; 126, clathria lacks elements
showing any clear correspondence with interthecal septa; 127, reticulum entirely dependent on clathria; 128,
cladia present. Retained primitive character f as in text-fig. 13a; n, dicalycal th2^; o, fusellar periderm present.

The conformity between the complex structures of the Petalograptus ancora and the ancora of the
Silurian retiolitids suggests that these structures are homologous. This relationship raises questions
about the nature of the so-called clathria of these graptolites. If their ancora is derived phylogenetic-
ally not from the fusellum of the thecae, but rather from a lacinia-like set of rods that arise from
the virgella independently of the thecae, then the retiolitine skeletal framework can hardly be
considered a clathria in the same sense that it is among the Lasiograptinae or the Abrograptinae.
This appears to be the case at least in obuti, where the ancora produces a lacinia-like structure
that grows upwards to enclose an otherwise non-retiolitid-like and fully sclerotized rhabdosome. In

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

393

any case, the morphogenesis of these retiolitine colonies must be quite different from that of the
Lasiograptinae (including the arehiretiolitids). Not only is the proximal end strueture of the Ordovi-
cian retiolite-type diplograptaceans derived from that of Orthograptiis and related taxa, but here
the lacinia is developed primarily from bifurcated thecal spines and lateral spines located along the
dorsal clathria, and is anehored to an otherwise normal set of virgellar and antivirgellar spines.

Text-fig. 16 presents the Monograptinae as a sister group to the Retiolitinae plus the Dimorpho-
graptinae, with their Pattern I astogeny, because in all these taxa thP is no longer a primordial
theca. As discussed below, this relationship explains many of the troubling similarities between
the Monograptinae and the Dimorphograptinae— similarities that are not accounted for in the
phylogenetic hypotheses advanced by Rickards and Hutt (1970) and Rickards, Hutt and Berry
(1977). It also provides insights into the significance of some of the unique features of the Mono-
graptinae that previously did not appear to be directly related to their origins.

As Bulman (1970, pp. V108-V109) noted, the elimination of thF from the astogenetic sequence
has been a stumbling block to theories concerning the origin of the monograptine design. He
pointed out that the suppression of the dicalycal theca is not sufficient to produce a monograptid
rhabdosome and cited the partly monoserial species Peiragraptus fallax as an example. Indeed, the
suppression of the dicalycal theca is not sufficient, but neither is its location or suppression of critieal
significance to this problem. Nevertheless, the structure of P. fallax is helpful in understanding the
evolution of a fully monograptid rhabdosome.

In P. fallax the proximal end consists of three alternate thecae, thH, thF, and th2', after which
the rhabdosome is uniserial. This condition, in which thH is retained while th2^ is lost, is not the
consequence of the suppression of the second bud of the dicalycal theca. In no species with a Pattern
G astogeny is th2^ dicalycal; A. bekkeri has a dicalycal th3* and this is the earliest dicalyeal theca
of any Pattern G species known to me. Furthermore, all of the advanced peiragraptine species to
which P. fallax is most closely related are aseptate, e.g. A. prominens Barrass and Paraorthograptus
pacificus (Ruedemann). The proximal end configuration of Peiragraptus fallax suggests that it is
not the location or presence of a dicalycal thecae that is problematic. Rather, the cause of this
proximally biserial and distally uniserial rhabdosome form appears to be the configuration of the
highly conservative primordial thecae. Both this species and all others with a Pattern G astogeny
possess three primordial thecae, and it is precisely these three that are retained in their primitive
alternating form in P. fallax. Th2^ may be the first theca to be suppressed because it is the first
non-primordial theca. ThF cannot be reoriented or eliminated until it is liberated from its role in
the primordial astogeny. Following such liberation in the Pattern I astogeny, the metatheca of thH
is reduced and eventually eliminated from the astogeny of both the dimorphograptines and the
monograptines. The dimorphograptines remain fundamentally biserial diplograptaceans, however,
perhaps because of the diplograptid ontogeny of the sole primordial theca, thl *, which they retained.
The first theca includes a downward-growing prothecal segment in which a foramen for thF
develops. The dimorphograptines, with their uni-biserial architecture, appear not to have been
involved in the ancestry of the monograptines. Indeed, Li (1985) has recently demonstrated that
the early dimorphograptines (e.g. D. elongatus of the P. acuminatus and C. vesiculosus zones) have
longer uniserial sections than do those of succeeding zones, suggesting that evolution in this group
favoured the accumulation of more fully biserial rather than monoserial species.

The fully monograptid eondition of the Monograptinae arose through the loss of the charaeteristic
primordial features of thl. The morphogenesis of the sicula and the mode of origin of thl reflect
the loss of the primordial status of thl. In all of the Graptoloidea except the Monograptinae, the
first theca arises through a resorption foramen. Among the Monograptinae a sinus forms in the
aperture as the metatheca grows. The protheca of thl crosses the virgella and then grows directly
upward without any noteworthy ontogenetie specializations. This coincidence between the occur-
rence of a wholly unprecedented structural change in the sicula and the early ontogeny of thl on
the one hand, and the possession of a radically new rhabdosomal form on the other, could be
unrelated to the acquisition of the monograptid condition but this is unlikely. The sicula is the most
structurally and morphologically conservative portion of graptolite colonies. The resorption

394

PALAEONTOLOGY, VOLUME 30

foramen is plesiomorphic with respect to the entire Graptolithina and the nematophorous sicula,
with a primordial thl, is plesiomorphic to the Graptoloidea. The shift from a resorption foramen
did not involve the simple loss of that feature but, rather, required an alteration of metasicular
ontogeny.

The fusellar morphology of the metasicula provides some information about the origins of the
monograptine sicular ontogeny: the configuration of growth lines during the sinus and lacuna stages
of porus formation are remarkably like the corresponding stages in the formation of the foramen
for thF in the descending protheca of thl * among other diplograptaceans. The metasicular fuselli
arc out from the sicular profile and loop back distalward. They make contact with previous fuselli
before reaching the virgella (e.g. Bulman 1970, fig. 48.9; Walker 1953, text-fig. 2). In this way the
siculozooid formed a hooded foramen. The metasicula’s contribution to the hood ceased when
the next fusellus reached directly around to the virgella and so closed the open proximal end of
the sinus. The thl protheca arises unconformably from this hooded foramen. The configuration of
its protheca and metatheca are like those of all subsequent thecae.

I propose that this similarity between the sicular structures and mode of origin of thl in the
Monograptinae and the structure of the protheca of thP and mode of origin of thF in glyptograp-
tines is more than analogy— it reflects a common origin. The monograptid condition, like that of
the other fundamentally distinct primordial astogenetic patterns among graptolites, appears to have
arisen by an abrupt shift in the timing or coordination of a crucial event in the astogenetic sequence:
by a kind of colonial heterochrony. Beginning with a primordial astogeny in which only thP
retained its specialized role in astogeny, as in Pattern I, the essential primordial feature of the
protheca of thP (the hooded foramen for thF) was accelerated (displaced to an earlier stage in
astogeny) into the ontogeny of the sicula, thl was liberated from its role as a primordial theca and
the previously fully biserial colony acquired a fully uniserial architecture.

Rickards et al. (1977, pp. 36-39) advanced the theory that the monograptines arose in a saltatory
fashion within a dithyrial population of a glyptograptine species similar to G. persculptus in the G.
persculptus Zone. Although a plausible suggestion, their theory lacked both a convincing mechanism
and predictions by which it could be tested. Derived along a different route and based on a different
logic, the theory I have outlined above postulates a similar mode of origin for the monograptines.
It provides a mechanism for their proposed origin and a means of testing its explanatory power.
It differs only in suggesting an ancestor with a Pattern I primordial astogeny like that of a Petalo-
graptus or Parakidograptus species rather than the Pattern H-bearing glyptograptine favoured by
Rickards et al.

Rickards and Hutt (1970) were unable to determine whether Atavograptus ceryx exhibited a
descending portion in the ontogeny of thl and whether this theca developed from a primary notch
rather than a resorption foramen. If the theory presented above is correct, then suitable material
should reveal the astogeny and sicula of A. ceryx to be fully monograptid. If it proves not to have a
primary notch and a non-primordial thl, then the theory is wrong. The proposed relationships also
imply that monoseriality of the monograptines and the dimorphograptines may be a parallelism
that reflects the highly simplified character of the primordial astogeny of their common ancestor
(an astogeny in which thP was no longer a primordial theca), which established the necessary
preconditions for a shift to uniserial colonies.

This mode of transformation from one primordial astogenetic pattern to another is not unique
to the Monograptidae. The shift from a Pattern A to a Pattern G astogeny, for example, can be
explained as the consequence of the acceleration of the upward-growing flange from the ontogeny
of th2^ (where it had fused with the crossing canal of th2' to form the pair of foramina through
which the prothecae of th2' and th2^ arose) into the ontogeny of thF (where it fused with the
downward-growing crossing canal of thF to form the metatheca of thF and the foramen from
which th2' developed). The result is a simpler, less crowded proximal end in which there are now
only three primordial thecae rather than four. The sequence of changes leading from a Pattern C to
a Pattern D and thence to a Pattern E primordial astogeny (and likewise from Pattern G to Pattern
F) also appears to have required only the relatively straightforward acceleration, mutatis mutandis.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

395

to earlier astogenetic stages of the upward-growing flange of th2' and the consequent reduction or
suppression of the corresponding descending crossing canal. Such a shift in the timing of primordial
astogenetic features, and particularly their transferral to an earlier stage in astogeny, appears to be
the principal means by which the Diplograptacea achieved simplified astogenetic patterns.

Subfamily glyptograptinae subfam. nov.

Diagnosis. Monograptids with glyptograptid to climacograptid or pseudoclimacograptid thecae;
median septum straight to complexly folded and with a Pattern H primordial astogeny.

Generic group taxa. Glyptograptus Lapworth, Clinoclimacograptus Bulman and Rickards, Cystograptus Hundt,
Litinianograptus Paskevicius, Metaclimacograptus Bulman and Rickards, Paraclimacograptus Pfibyl, Pseudo-
glyptograptus Bulman and Rickards [= ‘iComograptus Obut and Sobolevskaya]. May include other taxa such
as Hedrograptus Obut, but the phylogenetic status of these taxa remains to be established.

Genus glyptograptus Lapworth, 1873, emend.

Type species. Diplograpsus tamariscus Nicholson, 1868, Birkhill Shale, Southern Uplands, Scotland; Llan-
dovery Series (M. cyphus to M. turriculatus zones).

Diagnosis. Species with glyptograptid to climacograptid thecae having relatively narrow geniculum
and nearly straight supragenicular wall; proximal end generally narrow and fusiform, with strongly
alternating thecae; generally septate with straight median septum and th2^ or some later theca
dicalycal, but may be aseptate; sicula simple, generally short and broad, lacking antivirgellar spines.

Species included. Taxa assigned to this genus are too numerous to list but include: G. euglyphus, G. sinuatus,
G. tenuissimus, G. persculptus, ‘C.’ brevis, ‘C.’ rotundatus, and "C.' scalaris.

Discussion. The genus is here expanded to encompass the Ordovician and Silurian species formerly
included in Climacograptus that possess a Pattern H astogeny (see discussion of Monograptidae,
above), but restricted to apply only to those species with a Pattern H astogeny. This and other
features distinguish the taxon from similar looking species in Eoglyptograptus gen. nov., Hustedo-
graptus gen. nov., Climacograptus, and Arnheimograptus gen. nov. (see discussion of these taxa).
Glyptograptus differs from Paraclimacograptus Pfibyl in its lack of prominent genicular flanges.

Subfamily retiolitinae Lapworth, 1873, emend.

Diagnosis. Monograptids with sharply acicular proximal end based on a Pattern I astogeny among
forms with fully sclerotized proximal end or with ancora-based retiolitid astogeny; ancora common.
Primitively with orthograptid thecae but elaborated to glyptograptid, climacograptid, or to a stylized
clathrial framework.

Generic group taxa. Subfamily comprises three generic groups: Group 1 (petalograptids), Petalograptus Suess
and Cephalograptus Hopkinson; Group 2 (retiolitids), Retiolites Barrande, Pseudoplegmatograptus Pfibyl,
Sinostomatograptus Huo Shih-Cheng, and Stomatograptus Tullberg; Group 3 (plectograptids), Plectograptus
Moberg and Tornquist, Agastograptus Obut and Zaslavaskaya, Gothograptus Freeh, Holoretiolites Eisenack,
Paraplectograptus Pfibyl, and Spinograptus Boucek and Miinch.

Genus petalograptus Suess, 1851, emend.

Type species. Prionotis folium Flisinger, 1837, Rastrites Shale?, Sweden; Llandovery Series (M. leptotheca and
M. convolutus zones).

Emended diagnosis. Monograptids with orthograptid thecae disposed at a high to moderate angle
to the colony axis and with extensive overlap. Thecal apertures everted. Thecae commonly with
concave ventral walls that may lead to apertural isolation. Distally, thecal inclinations commonly

396

PALAEONTOLOGY, VOLUME 30

increase and the rhabdosome becomes broad and tabular. Ancora commonly present and some
species exhibit additional spines on thecal apertures.

Species included. Representative species include P. folium, P. ovatoelongatus, P. elongatus, ‘O.’ eberleini, ‘O.’
insect if ormis, and ‘O.’ mutabilis.

Discussion. Genus is here expanded to include the Silurian species with a Pattern I astogeny formerly
referred to ''Orthograptus' [= Dittograptus Obut and Sobolevskaya].

Subfamily dimorphograptinae Elies and Wood, 1908

Diagnosis. Monograptids with thF reduced or absent, with proximally uniserial rhabdosomes.
Length of the uniserial portion variable. Rhabdosome commonly septate with straight median
septum. Astogeny of Pattern I. Sicula commonly with ancora.

Generic group taxa. Dimorphograptus Lapworth, Akidograptus Davies, Parakidograptus Le and Gei, Rhaphido-
graptus Bulman.

Discussion. In the light of studies by Li (1985) and Rickards et al. (1977), which suggest that several
of the genera in this family are polyphyletic, and the several additional taxa that have been proposed
for various species with different thecal shapes (e.g. Buhnanograptus Pfibyl, Agetograptus Obut and
Sobolevskaya, and Metadimorphograptus Pfibyl) the phylogenetic status of the entire Dimorpho-
graptinae needs to be re-examined.

Subfamily eoglyptograptidae subfam. nov.

Diagnosis. Archaic monograptids with glyptograptid to climacograptid thecae, straight to undulat-
ing median septum, and Pattern B primordial astogeny.

Generic group taxa. Eoglyptograptus gen. nov. and Undulograptus Boucek, 1973, emend.

Genus eoglyptograptus gen. nov.

Type species. Fucoides dentatus Brongniart, 1828, Upper Levis Shale, Point Levis, Quebec; Whiterockian
Series {Isograptus and P. etheridgei zones).

Diagnosis. Monograptids with glyptograptid thecae having a gentle geniculum located about half-
way along the theca. Thecae overlap about one half their length and commonly bear cuspate
apertures. Narrow, gradually widening rhabdosomes are septate with a straight median septum.
The dicalycal theca may be th2^ or a later theca. The strongly asymmetric proximal end is broadly
rounded and exhibits a Pattern B astogeny. ThU may possess a subapertural spine or the proximal
end may be without spines apart from the virgella.

Species included. ‘G.’ dentatus Brongniart, " Pseudoclimacograptus' jaroslovi Boucek, and ‘G.’ cernuus Jaanus-
son.

Discussion. Skevington’s (1965, fig. 6\a) illustration of the E. dentatus specimen, 01 1228, is inaccur-
ate: the specimen does not possess a th2^ crossing canal where shown on this figure. The illustrated
structure is wholly incompatible with his fig. 62a and with Bulman’s (1936, 1963^r) wax model
reconstructed from serial sections. Text-fig. 3j is a new illustration of 01 1228.

Eoglyptograptus gen. nov. differs from other GlyptograptusAfkQ taxa principally in the form of its
proximal end and primordial astogeny. Species of Glyptograptus sensu stricto lack the apertural
cusps present on the thecae of E. dentatus and E. cernuus and exhibit a narrower and more
fusiform proximal end based on a Pattern H astogeny. Among the Orthograptidae, species of
Arnheimograptus gen. nov. resemble the eoglyptograptids in their rhabdosome form and proximal
end shape, but possess a Pattern F primordial astogeny and an extensively exposed sicula with
antivirgellar spines.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSILICATION

397

Genus undulograptus Boucek, 1973, emend.

Type species. Climacograptus paradoxus Boucek, 1944 [= 1C. pauperatus Bulman, 1953], Sarka Formation,
Krusna hora Mt. region, Bohemia; Llanvirn Series (lower D. bifidus Zone).

Emended diagnosis. Taxon here restricted to forms like the type species, which exhibits climaco-
graptid thecae, a weakly undulating median septum, and a Pattern B astogeny ( = Undulograptus
Jenkins, 1980, par//w).

Species included. If one accepts Boucek’s (1973) determination that ‘C.’ pauperatus is synonymous with the
type species, this taxon is presently monotypic. ‘C.’ pauperatus occurs in the Ogygiocaris Series, Norway and
in the Seby Limestone, Oland; both occurrences are equated with the D. murchisoni Zone.

Discussion. The narrow proximal end and Pattern B astogeny of the type species is quite unlike
those of Oelandograptus austrodentatus and congeners. Jenkins’s (1980) expansion of Boucek’s
taxon to encompass these species was ill advised. However, Boucek (1973) also included several
species in Pseudoclimacograptus {Undulograptus) that are indeed pseudoclimacograptines of the
group P. {Archiclimacograptus), and that exhibit no close similarity to U. paradoxus other than
their possession of an undulating median septum. E. jaroslovi, on the other hand, does seem to
possess a similar proximal end structure but lacks the undulating median septum and has glypto-
graptid rather than climacograptid thecae. Thus, confusion about the scope of this taxon dates
from its inception.

Subfamily monograptinae Lapworth, 1873

Diagnosis. Monograptids with fully uniserial stipes; some with cladia formed by secondary budding
from a mature zooid or sicula; proximal end development highly simplified, having no primordial
thecae; thl grew upwards from a primary porus produced by the metasicula during its ontogeny.

Discussion. Rickards et al. (1977) presented a detailed study of monograptid phylogeny which
indicated that the divisions Monograptidae and Cyrtograptidae of the Monograptina (see Bulman
1955, 1970) are not phyletically meaningful units. They did not present an alternative classification,
however. Thus, the systematic subdivision of the Monograptinae based upon the group’s evolution-
ary history remains to be accomplished.

MACROEVOLUTIONARY PATTERNS

My intention here has been to present a phylogenetic classification of the Diplograptacea. Accord-
ingly, consideration of their evolutionary history forms an integral part of this endeavour. It is not
primarily my intention to review the history of the classification (but see Rigby 1986) or to speculate
about the causes that may have underlain diplograptacean macroevolutionary patterns. Consider-
able work remains to be conducted in deciphering the details of this history. Nevertheless, a number
of large scale features of diplograptacean phylogeny are now apparent. These have implications for
both systematic practice and for future studies of graptolite colonial evolution.

The course of graptolite evolution has generally been traced on the basis of similarity in thecal
characters and in the disposition of the stipes, following the suggestions of Nicholson and Marr
(1895; see also Bulman 1970, p. VI 02). Neither when Nicholson and Marr wrote nor at any time
since has there been any compelling biological justification for this preference among the suite of
characters available for study in flattened graptolites. Rather, the demands of pragmatism, com-
bined with the attractively anti-Darwinian phylogenies that the method generated, led to a general
acceptance of Nicholson and Marr’s proposals among their contemporaries. The major works on
graptolite phylogeny (e.g. Elies 1898, 1922; Bulman 1933a, b, 1936; Boucek and Pfibyl 1951) and
systematics (e.g. Elies and Wood 1901-1918; Ruedemann 1904, 1908; Mu 1950) followed their
recommendations. The conception of graptolite evolutionary history that subsequently emerged
was one characterized by a confusing array of parallel trends, each leading in Lamarkian fashion

398

PALAEONTOLOGY, VOLUME 30

to the progressive improvement of the lineage. These trends manifested themselves not only in
astoundingly similar colony designs within independent lineages but also in their often contempor-
aneous appearance (Elies 1898, 1922; Ruedemann 1904; Bulman 1933(>; Rickards et al. 1977).
Within this framework, primordial astogeny has been seen as simply another of the many features
of graptoloid colonies that underwent extensive parallel change. Thus, Bulman { 1933a, p. 2) while
discussing Elles’s developmental types, cautioned that ‘. . . they represent simply grades of evolution,
probably reached or passed through quite independently in many different lineages. What is here
assembled as a purely morphological series, without strict reference to phylogeny, is believed to
represent an ‘orthogenetic’ trend, comparable with the stipe reduction trend and others described
by Elies. . . .’

Graptolite evolutionary history became one of the prime examples of orthogenesis (Bulman
1933^). Urbanek (1959, p. 326) and others have expressed a similar attitude with regard to the
phylogenetic significance of astogenetic similarities. Although Bulman (1960, 1963a, b) later re-
treated from his statements on the importance of orthogenesis as an explanation for the observed
trends, the present graptoloid classification remains one that is conceptually more compatible with
Osborn’s theory of aristogenesis than with Darwinian theory. This systematic history, combined
with the inherent difficulty of producing a phylogenetic classification of these organisms from their
often inadequately preserved and incomplete remains (see Bulman 19636, pp. 413-416) has pre-
vented the establishment of an integrated graptoloid systematics that is in step with both the group’s
probable evolutionary history and with contemporary evolutionary thought.

The results of the present studies of diplograptacean astogeny and thecal form show that grapto-
lite evolution was strikingly directional and exhibited distinct phases. The major diplograptacean
clades were founded through apparently rapid structural reorganizations. The Diplograptacea differ
from all other virgellinids in a substantial suite of features involving characters of the primordial
astogeny, thecal form, and rhabdosome architecture. The nature of these structural changes, like
those that occurred in the transitions from one primordial astogenetic pattern to another (such as
from Pattern D to Pattern E, or from Pattern A to Pattern C), indicate that they were not gradual
transitions made through a series of intermediate steps. Rather, they were achieved abruptly over a
short interval of time (as in a single allopatric speciation event) compared to the millions of
years over which they remained stable. This is also illustrated by the total lack of any preserved
intermediates between diplograptaceans and non-diplograptaceans (despite nearly ninety years of
searching) or between the groups of species characterized by the nine diplograptacean astogenetic
patterns.

Eollowing several of these rapid structural reorganizations (as in the case of the Orthograptidae,
following the origin of the Pattern G astogeny), the new clade apparently underwent an evolutionary
radiation. During its radiation the clade’s members achieved a substantial diversity of thecal form
and colony design, often exhibiting close analogy with species of other clades. The radiations of the
Orthograptidae and the advanced climacograptines in the late Llandeilo and early Caradoc appear
to have coincided with the waning of their predecessors among the primitive diplograptaceans
(particularly Oelandograptus gen. nov. and Hustedograptus gen. nov.) and among the pseudoclima-
cograptids and diplograptines (see text-fig. 17). The appearance is one of a relay in which a dominant
and diverse clade or set of clades is succeeded by another set which is itself succeeded. Hence, the
faunas of the late Arenig to late Llandeilo were dominated by the archaic diplograptids (the
Diplograptidae and, to a lesser degree, Oelandograptus and Hustedograptus). The late Llandeilo to
latest Ashgill witnessed the proliferation of the advanced Orthograptidae, Climacograptinae
(especially in the form of Orthograptus, Amplexograptus, and Climacograptus), and the Dicrano-
graptidae. Following the near total extinction of diplograptaceans the Monograptidae underwent
an explosive evolutionary diversification in the Llandovery. These intervals of successive clade
dominance are more or less equivalent to the diplograptid subfaunas that Bulman described
(1970, p. V99).

In the course of these three successive major radiations, homeomorphism arose in thecal form
and rhabdosome architecture with a bewildering frequency. Furthermore, this pattern of radiations

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

399

TEXT-FIG. 17. Evolutionary tree show-
ing pattern of descent among the
generic group taxa of the Diplograp-
tacea in the interval of the upper Ar-
enig to the base of the Wenlock Series.
The nemagraptids and retiolitids are
shown as family group taxa because
of the remaining uncertainty about
evolutionary relationships within
them. The relationships among the
generic group taxa of the Mono-
graptinae are beyond the scope of the
present study. Patterning within the
range bars of each taxon indicates
the astogenetic pattern exhibited by
its constituent species and its family
membership according to the inset
key. The absence of patterning in
range bars among the archiretiolitids
(Orthograptidae) and retiolitids
(Monograptidae) corresponds to the
highly stylized clathrial architecture
of these taxa that consequently does
not fit within the astogenetic patterns
defined herein. The plotted strati-
graphic ranges of the taxa are ap-
proximate due to uncertainty about
their species membership.

400

PALAEONTOLOGY, VOLUME 30

and rampant homeomorphism was not confined to the Diplograptacea. A number of authors (but
especially Cooper and Fortey 1982) have recently completed work on the Arenig graptoloids that
has revealed a surprisingly complex history. Early Arenig (Bendigonian Stage) faunas are dominated
by Pendiograptus and an early proliferation of Pseudophyllograptus species. These are succeeded in
the Chewtonian to early Castlemainian (Cal) by pendent didymograptids (mostly D. {Didymograp-
tellus) with an isograptid primordial astogeny) and Phyllograptus sensu stricto. Later in the Castle-
mainian, as the isograptids begin their main radiation, the pendent didymograptids vanish and
Phyllograptus is succeeded by a second radiation of Pseudophyllograptus. Finally, there occurs the
now well known sudden re-emergence, just prior to the beginning of the Darriwilian, of pendent
didymograptids in the form of D. (Didymograptus). This time, however, the rhabdosomes of these
‘tuning fork’ graptolites appear to be based for the most part on an artus-type primordial astogeny
(see Cooper and Fortey 1983). To what degree this seemingly endless playing out of variations on a
few themes reflects the action of either adaptive or constructional constraints (producing conver-
gence), or channelling by historical constraints (leading to parallelisms) is an important area for
further research — an area that may shed as much light on the processes of evolution as on the
palaeobiology of graptoloids.

The nature of the causal connection, if any,, between the waxing of one clade and the waning of
another is unknown. This issue is likely to be intimately related to the source of the overall direc-
tional history of diplograptacean evolution. Ostensibly, their history exhibits a strong birth-bias
in favour of more simplified astogenetic patterns. Following the establishment of the superfamily
and the Pattern A primordial astogeny, ten of the eleven transitions to new primordial astogenetic
patterns among the three lineages of fully scandent diplograptaceans (including the transitions to
the clathrial astogenies of the archiretiolitids and retiolitids) resulted in astogenies less complex
than the patterns that preceded them. (Relative complexity may be gauged by comparing the
number of crossing canals and primordial thecae, as well as the mode of thecal construction, vis.
the ‘direct’ growth pattern of the ontogeny of thF in Pattern F compared to the ‘indirect’ mode of
construction seen in Pattern G.) Only Pattern C is no simpler than its predecessor. Pattern A, but
neither is it more complex. Furthermore, this trend toward greater astogenetic simplicity affected
all three of the dominant Ordovician families: the Orthograptidae, Diplograptidae, and Monograpti-
dae. Accepting the cladogenetic history depicted in text-fig. 13a, the Dicranograptidae underwent
little change in primordial astogenetic structure during their range, except to give rise to the
Nemagraptinae with their right-handed origin of th 1 Transitions to astogenetic patterns of greater
complexity either did not occur among the diplograptaceans or were so unsuccessful that they left
no known record. Thus, the source of the variance that underlay the directional trends in astogeny
and colonial architecture of the Diplograptacea was strongly channelled by directed speciation.
Apart from the observation that loss of complexity is in some way ‘easier’ to achieve than is an
increase in complexity (consider the host of extant albino creatures, from cave crickets and white
rabbits to the Indian pipe, Monotropa uniflora, and the multitude of independent paths by which
they arrived at this lack of pigmentation), we have only speculative answers to the question of why
this bias should exist. Nevertheless, I am convinced by the frequent coincidence of a radiation in
thecal form and an increase in the clade’s diversity with the origin of a new, less complex astogeny
that these astogenetic changes were associated with a selective advantage in favour of graptolites
with a simplified pattern.

Differential rates of origination or extinction, or both, may also have contributed to the observed
replacement of clades with a complex astogeny by clades with a less complex astogeny. During
the course of diplograptacean evolution the changes in character distribution that accompanied the
astogenetic trends involved characters for which variance existed only at the clade level. It is now
clear that, phenomenologically at least, these trends seem to be the result of sorting among clades.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSI EIC ATION

401

TABLE 2. Key to the identification of diplograptid astogenetic pattern on the basis of features visible in mature
rhabdosomes well preserved in semi- or full relief, or in isolated preparations. These features are not a
substitute for the study of isolated growth stages but are guides to the distinctive features of the modal
proximal-end architecture associated with each astogenetic pattern. Sketches in the panels labelled II to VIII
provide illustrations, in otherwise similar rhabdosomes, of the contrasting features under consideration in the
corresponding key statement. For example, statement II asks the user to decide whether or not the prothecae
of th2^ and th2^ arise as a symmetrical pair from the descending portion of the crossing canal of th2^; the
sketches in panel II illustrate proximal ends with a paired and an unpaired th2' and th2^. Abbreviations: a,
apertural thecal spine; av, antivirgellar spines on sicular aperture; m, mesial thecal spine; v, virgella. Diagonal
or horizontal ruling highlights the theca or thecae that are the focus of comparison; vertical ruling highlights

the sicula.

402

PALAEONTOLOGY, VOLUME 30

APPENDIX

Application of classification to non-isolated graptolites

The circumstances described above make a thorough study of the evolutionary history of graptolites essential
and justify a revision of their classihcation, despite the temporary hardship it may impose in the day to day
practice of classification. Furthermore, I am confident that, with patience and attention given to the descrip-
tions and figures provided here, most well-preserved flattened and semi-relief graptolite specimens can be
placed within the new taxonomic system.

There are several features of the present situation that should ease the application of this classification to
non-isolated material. First, although the diplograptacean astogenetic patterns initially could only be recogni-
zed through the study of isolated graptolite growth stages, now that they have been defined it is clear that
each primordial astogenetic pattern exhibits a number of reliable morphological correlates visible in well-
preserved, semi-relief, and flattened, mature specimens. Table 2 presents a key to these features and their
correspondence with the diplograptacean astogenies. Given a working knowledge of the diplograptacean
astogenetic patterns, it is possible to recognize many of their distinctive features in the flattened growth stages
and sub-mature rhabdosomes that generally accompany fully developed colonies on shale surfaces. This is
especially true if the age of the specimens at hand is known. Since many of the new and redefined taxa have
shorter stratigraphic ranges than the previous form genera, this knowledge can be used to considerably narrow
the range of possible astogenetic patterns that a species might possess (see text-fig. 17).

Secondly, scientists with the good fortune to have among their collections material preserved in semi- or
full relief should begin the process of re-examining and re-illustrating the proximal ends and growth stages of
the species so represented. Through the publication of such studies the list of species with known astogenetic
patterns (Table 1) can be augmented. Although Table 1 does not include a majority of diplograptacean
species, it nevertheless does represent a broad spectrum of diplograptacean diversity. By the association of
morphologically similar but less well-preserved species with those in Table 1, the new classification can be
extended to these other species.

Thirdly, the present situation is similar in many respects to the continuing reorganization of conodont form
taxonomy. An interim device, like that used by conodont workers, can be employed in cases where the
astogenetic pattern, and hence the generic classification, of a particular species is uncertain: where the generic
affiliations of a species are ambiguous or unknown, the species name can simply be combined with a name
corresponding to one of the traditional diplograptacean form-taxa, e.g. " Diplograptus' compactus (with the
generic name enclosed in quotation marks to indicate that the name is being used as a form-taxon rather than
in its phylogenetic sense). In cases where uncertainty about the generic assignment remains, but where the
author considers it probable that the assignment of a species to a particular phylogenetic taxon is correct, this
may be indicated as Arnplexograptusl arctus. With these conventions, authors should be able to describe any
diplograptacean fauna that is well enough preserved to support specific identification.

Finally, note that this revision only affects genera and higher taxonomic units. Astogenetic features are
seldom of importance in the definition of graptolite species: not because astogeny is non-adaptive or incon-
sequential to the biology of the species, but because little variance in astogeny exists at this level in the
taxonomic hierarchy. Thus, individual species will be identified as before and their use in biostratigraphy will
remain unchanged.

Acknowledgements. The present study is part of a project begun as my doctoral dissertation at Harvard
University. I thank Mr Anton Kearsley for his willingness to discuss work in progress on Orthograptus and
diplograptacean systematics, and also Drs Anthony Arnold, Roger Cooper, Peter Crowther, Stan Finney,
Richard Fortey, Valdar Jaanusson, Michael Melchin, John Riva, and Henry Williams for their discussions of
some of the ideas and data presented, and their willingness to share their own ideas and data with me. I also
thank Drs Jaanusson, Barrie Rickards, and Isles Strachan for their help and for allowing me to examine
collections of isolated, three-dimensionally preserved graptolites in the collections of the Naturhistoriska
Riksmuseet, Sedgwick Museum, Cambridge, and the University of Birmingham, respectively. I thank my
dissertation advisor. Dr Steven J. Gould, for guidance and encouragement. My dissertation research was
supported by National Science Foundation Dissertation Improvement Grant EAR 8115100 and by grants
from the Department of Geological Sciences, Harvard University and the Royal Society of London.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

403

REFERENCES

ALBERSTADT, L. p. 1973. Articulate brachiopods of the Viola Formation (Ordovician) in the Arbuckle Moun-
tains, Oklahoma. Bull. Oklageol. Surv. 117, 1-90.

BARRASS, R. 1954. Graptolites from Anticosti Island. Q. Jl geol. Soc. Loud. 110, 55-75.

BATES, D. E. B. and KIRK, N. H. 1984. Autecology of Silurian graptoloids. Spec. Pap. Palaeont. 32, 121-139, pis.
1-3.

BERRY, w. B. N. 1964. The middle Ordovician of the Oslo region. No. 16. Graptolites of the Ogygiocaris Series.
Norsk, geol. Tiddskr. 44, 61-170.

BOUCEK, B. 1973. Lower Ordovician graptolites of Bohemia, 182 pp., 24 pis. Czechoslovak Academy of Science,
Academic Press, Prague.

and PRiBYL, A. 1951. Taxonomy and phylogeny of some Ordovician graptolites. Bull. Int. Acad, tcheque

Sci. 52, 265-283.

BULMAN, o. M. B. 1932. On the graptolites prepared by Holm, Pt. 1. Certain diprionidian graptolites and their
development. Ark. Zool. 24A (8), 1-46, quarto pis. 1-9.

1933a [dated 1932]. Notes on the evolution and morphology of certain graptoloids. Ibid. (13), 1-37.

1933Z). Programme-evolution in the graptolites. Biol. Rev. 8, 31 1-334.

1936. On the graptolites prepared by Holm, Pt. 7. The graptolite fauna of the Orthoceras Limestone of

Halludden, Oland and its bearing on the evolution of the lower Ordovician graptolites. Ark. Zool. 28A
(17), 1-107, quarto pis. 1-4.

1945-1947. Monograph of the Caradoc (Balclatchie) graptolites from the limestones in Laggan Burn,

Ayrshire. Palaeontogr. Soc. [Monogr.], 78 pp., 10 pis.

1953. Some graptolites from the Ogygiocaris series (4aot) of the Oslo district. Ark. Miner. Geol. 1, 509-

518.

1960. Some morphologically intermediate genera in graptolite phylogeny. 21st. Int. geol. Congr. Norden,

sect. 22, 65-70.

1962. On the genus Amplexograptus Lapworth, Elies & Wood. Geol. Mag. 99, 459-467.

1963a. On Glyptograptus dentatus (Brongniart) and some allied species. Palaeontology, 6, 665-689.

1963^>. The evolution and classification of the Graptoloidea. Q. Jl geol. Soc. Lond. 119, 401-418.

1968. Graptolithina. In moore, r. c. et a!. Developments, trends and outlooks in paleontology. J. Paleont.

42, 1327-1377.

1970. Graptolithina with sections on Enteropneusta and Pterohranchia. In teichert, c. (ed.). Treatise on

Invertebrate Paleontology, Part V, (2nd edn.), 163 pp. Geological Society of America and University of
Kansas Press, Boulder, Colorado and Lawrence, Kansas.

and RICKARDS, r. b. 1968. Some new diplograptids from the Llandovery of Britain and Scandinavia.

Palaeontology, 11, 1-15.

CHURKIN, M. and carter, c. 1970. Early Silurian graptolites from southeastern Alaska and their correlation
with graptolite sequences in North America and the Arctic. Prof. Pap. U.S. geol. Surv. 653, 1-51.

COOPER, R. A. and fortey, r. a. 1982. The Ordovician graptolites of Spitsbergen. Bull. Br. Mus. nat. Hist.
{Geol.), 36, 157-302, 6 pis.

1983. Development of the graptoloid rhabdosome. Alcheringa, 7, 201-221.

ELLES, G. L. 1898. Graptolite faunas of the Skiddaw Slates. Q. Jl geol. Soc. Lond. 54, 463-539.

1922. The graptolite faunas of the British Isles. A study in evolution. Proc. Geol. Ass. 33, 168-200.

and WOOD, e. m. r. 1901-1918. A Monograph of British graptolites. Palaeontogr. Soc. [Monogr.],

clxxi -I- 539 pp., 52 pis.

FINNEY, s. c. 1980. Thamnograptid, dichograptid and abrograptid graptolites from the middle Ordovician
Athens Shale of Alabama. J. Paleont. 54, 1184-1208.

1985. Nemagraptid graptolites from the middle Ordovician Athens Shale, Alabama. Ibid. 59, 1100-1137.

FORTEY, R. A. and COOPER, R. A. 1986. A phylogenetic classification of the graptoloids. Palaeontology, 29, 631-
654.

and JEFFERIES, R. p. s. 1982. Fossils and phylogeny— a compromise approach. In joysey, k. a. and Friday,

A. E. (eds.). Problems of phylogenetic reconstruction. Spec. Vol. Syst. ,4^5. 21, 197-234.

GOULD, s. J. 1977. Ontogeny and phylogeny, 501 pp. Harvard University Press, Cambridge, Mass.

HOLM, G. 1895. On Didymograptus, Tetragraptus and Phyllograptus. Geol. Mag. 32, 433-441, 481-492, pis. 13
and 14.

HUTT, J., RICKARDS, R. B. and SKEViNGTON, D. 1970. Isolated Silurian graptolites from the Bollerup and
Klubbedden Stages of Dalarna, Sweden. Geologica Palaeont. 4, 1-23.

404

PALAEONTOLOGY, VOLUME 30

JAANUSSON, V. 1960. Graptoloids from the Ontikan and Viruan (Ordovician) limestones of Estonia and
Sweden. Bull. geol. Instn Univ. Upsala, 38, 289-366.

JENKINS, c. 1980. Maendrograptus schmalemeei and its bearing on the origin of the diplograptids. Lethaia, 13,
289-302.

KEARSLEY, A. 1982. Morphology and systematics of species of Orthograptus Lapworth. Newsl. Grap. Wkg Gp
Ini. Paleont. Ass. 3, 10-13.

1985. A new phylogeny of diplograptid graptoloids, and their classihcation based on proximal and thecal

construction. Ibid. 6, 8-22.

LAPWORTH, c. 1873. On an improved classification of the Rhabdophora. Geol. Mag. 10, 500-504, 555-560.

1876. The Silurian System in the south of Scotland. In Armstrong, j., young, j. and robertson, d.

Catalogue of Western Scottish fossils, xxiii + 164 + 24 pp., 4 pis. Blackie & Son, Glasgow.

1877. On the graptolites of County Down. In Systematic lists illustrative of the flora, fauna, palaeontology

and archaeology of the North of Ireland by the members of the Belfast Naturalists Field Club, Vol. 1, Append.
IV, 126-144, pis. v-vii.

1880. On new British graptolites. Ann. Mag. nat. Hist. (ser. 5), 5, 149-177, pis. 4 and 5.

LI ji-jin. 1985. On the origin of dimorphograptids. In bjerreskov, m. and stouge, s. (eds.). Abstr. Third Int.
Grap. Conf. Copenhagen, 1985, 57.

MITCHELL, c. E. 1981. Preliminary morphometric studies of astogeny and colony form in Climacograptus Hall
and the incorporation of astogenetic data into phylogenetic hypotheses. Progm Abstr. Second Int. Grap.
Conf. Cambridge, 1981, 19.

1986. Morphometric studies of Climacograptus Hall and the phylogenetic significance of astogeny. In

HUGHES, c. p. and rickards, r. b. (eds.). Palaeoecology and biostratigraphy of graptolites. Spec. Publ. geol.
Soc. 20, 119-129.

and CARLE, K. J. 1986. The nematularium of Pseudoclimacograptus scharenbergi (Lapworth) and its

secretion. Palaeontology, 29, 373-390, pis. 28 and 29.

MU en-zhi (= MU, A. T.). 1950. On the evolution and classification of graptolites. Geol. Rev. Beijing, 15, 171-
183.

et al. 1974. In A handbook of the stratigraphy and paleontology in South-west China, (ed. Nanjing) Inst.

Geol. and Paleontol., Academia Sinica. Science Publ. Co. Nanjing, 154-221. [In Chinese.]
mOller, a. h. and schauer, m. 1969. Uber schwebeenrichtungen bei Diplograptidae (Graptolithina) aus dem
Silur. Freiberger ForscliHft. C245, 5-26.

NICHOLSON, H. A. and MARR, J. E. 1895. Notes on the phylogeny of the graptolites. Geol. Mag. 42, 529-539.
OBUT, A. M. 1949. Polevoi atlas rukovodyashchikh graptolitov verkhnego silura Kirgizskoi SSR. [Field atlas
of leading graptolites of the Upper Silurian of Kirgiz SSR.] Trudy geol. Inst., Frunze, 1949, 1-56, pis. 1-7.

and SOBOLEVSKAYA, R. F. 1968. In obut, a. m., sobolevskaya, r. f. and merkureva, a. p. Graptolity

Llandoveri v kernakh burovykh skvazhin Norilskogo rayona. [Graptolites of the Llandovery in core samples
from boreholes in the Norilsk District.] Trudy Inst. Geol. Geofiz. Sib. Otd. 1-126, pis. 1-35.

PRIBYL, A. 1947. On the classification of the genus Climacograptus Hall, 1865. Bull. int. Acad, tcheque Sci. 48,
1-12, pis. 1 and 2.

RICKARDS, R. B. and HUTT, J. E. 1970. The earliest monograptid. Proc. geol. Soc. Fond. 1663, 115-119.

and BERRY, w. B. n. 1977. Evolution of the Silurian and Devonian graptoloids. Bull. Br. Mus. nat.

Hist. (Geol.), 28, 5-120, 6 pis.

and koren', t. n. 1974. Virgellar meshworks and sicular spinosity in Llandovery graptoloids. Geol. Mag.

Ill, 193-204.

RIGBY, R. 1986. Critique of graptolite classification, and a revision of the Suborders Diplograptina and
Monograptina. In hugfies, c. p. and rickards, r. b. (eds.). Palaeoecology and biostratigraphy of graptolites.
Spec. Publ. geol. Soc. 20, 1-13.

RiVA, J. 1976. Climacograptus bicornis bicornis Hall, its ancestors and likely descendants. In bassett, m. g.
(ed.). The Ordovician System, 589-619. Proceed. Paleont. Symp. Birmingham, 1974, Univ. Wales Press,
Nat. Mus. Wales, Cardiff, Wales.

ruedemann, r. 1904. Graptolites of New York, Pt. I. Graptolites of the lower beds. Mem. N.Y. St. Mus. nat.
Hist. 7, 457-803, pis. 1-17.

1908. Graptolites of New York, Pt. II. Graptolites of the higher beds. Ibid. 11, 1-583, pis. 1-31.

skevington, d. 1965. Graptolites from the Ontikan limestones (Ordovician) of Oland, Sweden. II. Grapto-
loidea and Graptovermida. Bull. geol. Instn Univ. Upsala, 43, 1 -74.
skoglund, r. 1963. Uppermost Viruan and lower Harjuan (Ordovician) stratigraphy of Vastergotland and
lower Harjuan graptolite faunas of Central Sweden. Ibid. 42, 1-55.

MITCHELL: GRAPTOLOID EVOLUTION AND CLASSIFICATION

405

STRACHAN, I. 1974. Further isolated graptolites from the Ordovician of Girvan. In rickards, r. b., jackson,
D. E. and HUGHES, c. p. (eds.). Graptolite studies in honour of O. M. B. Bulman. Spec. Pap. Palaeont. 13,
99-105.

WALKER, M. o. 1953. The development of Monograptu.s duhius and Monograptus chimaera. Geol. Mag. 90, 362-
373.

wiMAN, c. 1895. liber die graptoliten. Bull. geol. Instn Univ. Upsala, 2, 239-316, pis. 9-15.

URBANEK, A. 1959. Studies on graptolites, II. On the development and structure of the graptolite genus
Gymnograptus Bulman. Acta palaeont. pol. 4, 279-336, pis. 1, 2.

1973. Organization and evolution of graptolite colonies. In boardman, r. s., cheetham, a. h. and Oliver,

JR., w. A. (eds.). Animal colonies: development and function through time, 441-514. Dowden, Hutchinson
and Ross, Stroudsburg, Penn.

CHARLES E. MITCHELL

Department of Geological Sciences
State University of New York at Buffalo

Typescript received 26 March 1986 4249 Ridge Lea Road

Revised typescript received 18 September 1986 Amherst, New York 14226

TRACE FOSSILS FROM RHAETIC SHORE-FACE
DEPOSITS OF STAFFORDSHIRE

by A. D. WRIGHT and m. j. benton

Abstract. The basal Rhaetic deposits of the Needwood Basin, Staffordshire, are of sandstone facies, in
marked contrast to the well known bone beds and dark shales of the transgressive Westbury Formation
which crop out elsewhere. The ochreous yellow sandstones contain abundant but poorly preserved bivalves
(Eotrapezium sp.) together with an extensive ichnofauna which is dominated by Pelecypodichmis but also
includes Arenicolites, Kouphichniimu Palaeophycus, Plauolites, and Rusophycus in addition to various other
trails and burrows. Although this trace fossil assemblage suggests an environment which could range from
non-marine to marginal marine, the latter environment is indicated for these shallow water sands by the
presence of the marine Eotrapezium. The Needwood trace fossils and facies show close similarities to the
German Rhatsandstein, a sequence of offshore and marginal marine sandstones.

The presence of trace fossils in the uppermost Triassic rocks of Needwood Forest, an area of high
ground to the west of Burton-on-Trent, Staffordshire, was first noted by Molyneux (1869) who
recorded the Tracks of annelides’ in the basal sandstones of the Rhaetic succession. Molyneux’s
observations were cited in the Burton-on-Trent memoir by Stevenson and Mitchell (1955, p. 65)
but not enlarged upon further. In 1975 one of us (A.D.W.) collected a variety of trace fossils from
this horizon and presented the preliminary findings at the annual meeting of the Palaeontological
Association, held that year at Newcastle. Of particular interest was the occurrence of the abundant
trace fossil Pelecypodichmis along with moulds of bivalves for although these traces had been
interpreted by Seilacher (1953) as being the work of bivalves (and hence his name Pelecypodichnus),
no body fossils corresponding to the traces had previously been found which, as pointed out by
Hallam (1970, p. 195), is the usual situation. Bromley and Asgaard have since reported that
radiographs have revealed bivalve body fossils in beds containing Pelecypodichmis in the Trias of
East Greenland (1979, p. 45), while Broadhurst et al. (1980, p. 646) and Eagar et al. (1985, p. 130)
have figured Carbonicola in association with escape shafts and Pelecypodichnus from the Upper
Carboniferous of northern England. A second trace of interest was a problematic one and consisted
of some paired impressions (2-5 mm long) which were interpreted as being produced by the append-
ages of one limb of an arthropod. As there was no sign of the other limb, and no indisputable track
sequence, this interpretation was uncertain and suggestions elicited from Association members
ranged from claw traces of dragonfly naiads to the impressions of ostracode valves. The collection
of additional material now enables a satisfactory explanation for these tracks to be given.

GEOLOGICAL SETTING

The bed rock of Needwood Eorest consists predominantly of the red ‘Keuper Marl’ of the Mercia
Mudstone Group, which contributes economically to the district firstly through the contained
Tutbury Gypsum horizon, extensively mined under the north side of the Forest, and secondly in
the contribution of its ground water which, on the east side, blends with that flowing along the
Trent corridor to produce an ideal water for the brewing of pale ale at Burton-on-Trent. In addition
to the Keuper Marl, the Triassic succession of Needwood Forest is capped by two outliers of
Rhaetian age in which the Tea Green Marl (Blue Anchor Formation) is overlain by Rhaetic Beds

I Palaeontology, Vol. 30, Part 2, 1987, pp. 407-428, pis. 49-53.)

© The Palaeontological Association

408 PALAEONTOLOGY, VOLUME 30

(Penarth Group), the lower part of which is correlated with the Westbury Formation by Warrington
et al. (1980, p. 41).

Structurally this succession is in the centre of the Needwood Basin where the beds are essentially
horizontally disposed so that the highest ground is underlain by the highest parts of the succession.
Exposure is very poor, generally being confined to old marl pits, for not only is the area heavily
covered by drift, but the Mercia Mudstones and the Rhaetic shales are soft and readily weathered.
The basal deposits of the Westbury Formation take the form of variably bedded fine grained
ochreous weathering soft grey sandstones which are nevertheless hard enough to form a well defined

TEXT-FIG. 1. Map showing localities (asterisked) from which the fossiliferous material was collected from the
base of the essentially horizontal Rhaetic succession. The broken line indicates the junction between the basal
Rhaetic sandstones (stippled) of the Westbury Formation and the underlying Tea Green Marl (Blue Anchor
Formation) along the wooded north-facing scarp of Needwood Forest. Inset map indicates the location within

central England.

WRIGHT AND BENTON: RHAETIC TRACE FOSSILS

409

escarpment on the northern edge of the Forest. Sections logged by Molyneux (1869, p. 172) record
up to six feet of this sandstone and although many of the outcrops available to him were overgrown
when the area was mapped by Stevenson and Mitchell (1955), the rock is still visible in rather
scrappy exposures in their north-western outlier (1955, p. 63). The trace fossils described herein
are from these basal sandstones, the best material coming from the three localities indicated in
text-fig. 1.

Locality 1 (SK 121 82709) is an old marl pit 400 yards (370 m) SSE of Tomlinson’s Corner Farm
and some 2 miles east of the Bagot’s Park borehole (Stevenson and Mitchell 1955, p. 66). As was
noted by these authors, the pit shows ‘about 2 ft. 6 in. of fossiliferous sandstone resting on Tea-
green Marl. The sandstone yielded "" PuUastrd" arenicola Strickland.’ The new material was collected
in 1975; the present sandstone outcrop is limited to small exposures in tree roots at the edge of
the pit.

Locality 2 (SK 106328 13) is at the head of a small valley cutting the scarp edge 300 yards (275 m)
to the west of Buttermilk Hill. The exposed section here shows the uppermost 10 m or so of the
Blue Anchor Formation at the top of which is a 3 cm grey-green clay immediately underlying the
sandstones of the Westbury Formation. Only the basal half metre of the latter is exposed and
consists initially of thinly bedded, rippled, and undulating fine sandstones, typically of half centi-
metre thickness but with lenses of up to 3 cm; 22 cm above the base is a thicker (6 cm) sandstone
succeeded by thin flags (2 cm or so) to complete the 53 cm of exposure. This locality yielded some
very fossiliferous material, most slabs being collected from or slightly above the thicker sandstone,
the upper surfaces being rippled and with bivalve moulds, the lower surfaces with Pelecypodichmis,
usually on a thin clay seam of up to 5 mm in thickness. Ripple orientation is extremely variable
with one orientation of 135° T being succeeded by one of 42° T on the overlying slab (see below).

Locality 3 (SKI 1072827) is adjacent to a spring some 200 yards (180 m) east of Buttermilk Hill
where the junction of the Blue Anchor Formation and the Westbury Formation is seen in a thin
sequence of strata. Exposed above the mudstones of the Blue Anchor Formation are some 35 cm
of fine grained sandstones. These are for the most part thinly bedded, 1 -2 cm in thickness and again
separated by very thin shale partings, but at the top of the exposure sandstone units of 4, 3, and
5 cm were recorded. Once more the upper surfaces are rippled and from the lower surfaces Pelecypo-
dichnus protrude down into the sticky clay or shale partings.

The sandstones and their sedimentary environment

The sandstones are formed from well-sorted yellow-grey micaceous quartz sand, typically of very
fine grain size but grading to fine sand and occasionally medium sand size. One 3 cm thick sandstone
unit shows a distinctly coarser lower surface with scatters of 0-5 mm quartz grains and cavities of
up to 5 mm in size from which clasts of mudstone, sand, and carbonate have been weathered out.
The yellow-grey sediment is commonly iron stained to a greyish or dark yellowish orange, with
lower surfaces and joint faces not infrequently encrusted with brown iron oxides; sparse scatters of
carbonaceous fragments of up to 1 mm in length occur infrequently in the sand.

On the upper surfaces the ripples are symmetrical and low crested (PI. 49, fig. 2) with ripple crests
peaked or rounded in roughly equal proportions. In most cases the crests are formed of the finer
sediment although on the specimen noted above the upper surface shows very variable grain size,
and has abundant medium sized quartz grains on the crests with patches of finer sand on the slopes
of the troughs. The ripples are generally straight or slightly sinuous and bifurcate in places. Preserved
ripple height varies from 5 to 12 mm, and ripple length from 40 to 110 mm; ripple indices vary
between 6 and 9. In some specimens ladder ripples, with wavelengths ranging from about 25 to
70 mm, are preserved between the main ripple sets (PI. 49, fig. 3). These features all indicate a
wave origin for the ripples, with the ladder ripples providing evidence of emergence in this shallow
water beach environment.

In cross-section, the coarse units are dominated throughout by bioturbated ripple and parallel
cross-lamination (PI. 49, fig. 1). The cross-laminations occur in troughs, some of which are narrow
and well defined, with steep foreset laminae. Other troughs are broader, and the cross-laminae are

410

PALAEONTOLOGY. VOLUME 30

long, and nearly parallel to the base of the trough. Parallel lamination is also seen below the troughs
or below low ripples in some specimens. In terms of the categories of wave generated sediments
described by de Raaf et al. (1977), the sandstone units show some features of the M2 (lenticular
beds) along with the Sj (parallel and cross-laminated sandstones). The units are from 1 to 6 cm
thick with a grain size of very fine to medium, which falls into the M2 category, but the parallel
lamination is an important structure within the units and corresponds to the S, patterns. The mud
flasers of the M2 are moreover essentially missing, the thin clay horizons being on true bedding
surfaces which serve to separate the sandstone units. The M2 and Sj lithotypes indicate, respectively,
moderate but continuous wave action and conditions of considerable wave activity. Parallel and
cross-lamination and symmetrical ripples occur in submerged and emergent sand bars lying a little
offshore (de Raaf et al. 1977, p. 479), in an environment periodically sufficiently sheltered to allow
not only bioturbation but also the accumulation of argillaceous seams.

Bioturbation is clearly seen in cross-section (PI. 49, fig. 1) with some horizons heavily disturbed
by burrowing. Vertical escape structures are observed breaking through thicknesses of up to
25 mm of cross-laminated sand from the more intensively burrowed horizons.

The sedimentary structures on the lower surfaces also include the impressions of symmetrical
ripple marks, although these are much less common than the ripples on the upper surfaces and are
present on only two of the sixteen slabs collected from locality 2 for their well preserved trace
fossils. The ripple wavelength ranges from 60 to 82 mm, the mean of 69 mm in a sample of 6 being
very similar to that of the upper surface with a mean of 65 mm in a sample of 29; the ripples are,
however, much lower. These ripples are simply the impressions of those on the upper surface of the
underlying bed, with the subdued form resulting from the muddy horizon coating the ripples below
and the subsequent burrowing of the bivalves down into this mud. The apparent ripple index of up
to 18 for these shallow ripple impressions should not therefore be interpreted as an indication of
the development of current ripples. It is, however, of interest that there is a relative difference of
70° and 75° in the ripple orientations of the lower and upper surfaces of these two slabs, showing
again the marked variation noted above from the outcrops.

Several bed bases display abundant tool marks in the form of small, shallow groove marks,
0-4 mm wide and 2-10 mm long. On any particular surface, these tool marks generally show two
dominant directions and they are frequently interspersed with, and cut by, small burrows and other
trace fossils.

BODY FOSSILS

The only body fossils present in the sandstone are the bivalves alluded to as "Pullastra' arenicola
Strickland by Stevenson and Mitchell (1955, p. 66). Strickland, however, never figured this species;
there are no type specimens; and as his original description (1843, p. 17) is too generalized to be
diagnostic, the name would appear to be unusable. No figures of Strickland’s species have been
traced prior to those of Phillips (1871, pi. 7, figs 6-12) but these show a wide range of shape and

EXPLANATION OF PLATE 49

Fig. 1. Vertical section through a block of sandstone showing ripple and parallel cross lamination with escape
structures, and two heavily bioturbated bands across the middle with abundant Arenicolites burrows.
SM.X.8948. Locality 1. xO-8.

Fig. 2. Upper surface of sandstone block showing ripples, the internal moulds of Eotrapeziiim (SM.X.8951),
paired pits marking the surface expression of Arenicolites burrows (SM.X.8953) and a broad ridge trail
(SM.X.8955). Locality 1. xO-5.

Fig. 3. Ladder ripples on the upper surface of a sandstone block. SM.X.8960. Locality 2. x 0-5.

Figs. 4 and 5. Small sandstone block from Locality 1. x 1. 4, lower surface of block showing the close
association of Eotrapezium valves (concave) (SM.X.8966) with convex hyporeliefs of Pelecypodichmis
(SM.X.8968). 5, lateral view of this block showing Eotrapezium sp. in life position in a burrow (SM.X.8970).

PLATE 49

;'wyB.

WRIGHT and BENTON, Rhaetic trace fossils

412

PALAEONTOLOGY, VOLUME 30

Structure. It would appear best to refer the forms in the sandstone to Eotrapezium sp., a small
bivalve present in mudstones of this age which could probably leave similar moulds in sandstone
(Dr H. Ivimey-Cook, personal correspondence). Variations in outline and proportions do exist and
it may well be that more than one species is present; but the specimens, although numerous, occur
in the form of very poorly preserved moulds which simply indicate overall shape and, on the external
moulds, the presence of concentric growth lines. The convex internal moulds give no indication at
all of the form of the hinge structures.

The moulds, ranging in length from 3 to 1 1 mm, occur both on the upper and lower bedding
surfaces and within the body of the sandstones. Those on the upper surfaces are scattered across
the troughs of the rippled sand surface as disarticulated valves. The overwhelming majority are in
the form of convex internal moulds with only the occasional concave external mould; the ratio of
22: 1 counted in the ripple trough of one slab appears fairly typical. One upper surface is unusual
in displaying the concave impressions of the external dorsal hinge region of half a dozen articulated
shells in addition to the flat lying disarticulated valves.

Disarticulated valves, equally poorly preserved and with the same convex up orientation are
found on bedding surfaces within the sandstone units; the valves of one individual both lie flat on
the surface, indicating entombment in the sediment before decay of the ligament. Occasional
specimens are seen within the body of the sediment orientated perpendicular to the sediment surface
in life position (PI. 49, fig. 5).

On the lower surfaces the disarticulated valves are preserved essentially as concave external
moulds, i.e. with the valves again orientated convex side up. While such surfaces may be covered
with these moulds and lacking Pelecypodichnus traces, and other surfaces covered with these trace
fossils but lacking bivalves. Eotrapezium and Pelecypodichnus were found together on ten out of
the sixteen slabs from locality 2. These were sometimes very closely associated (PI. 49, fig. 4),
although the relative proportions of the shells and the traces varied considerably.

Thus apart from the abundance of bivalves and Pelecypodichnus on the opposite surfaces of the
sandstone units, this preservation of the two on the same surface and the clear evidence of burrowing
through the sand provide abundant circumstantial evidence that the Pelecypodichnus traces were
the product of these bivalves burrowing down through the sand until the underlying clay horizon
was reached.

TRACE FOSSILS

A range of trace fossils was observed m the sandstone units. These indicate a diverse fauna moving
over the surface of the sand (Koiiphichnium, groove and ridge trails of various kinds), living within
the sand (ArenicoUtes and bivalve escape burrows), and living temporarily, if not continuously, at
the lower surface along the sand-clay interface (Palaeophycus, Planolites, Rusophycus, and small
trails).

EXPLANATION OF PLATE 50

Fig. I. Kouphiclmiwn trackway (SM.X.8971) trending obliquely across a ripple trough (the two partially
preserved ripple crests cross the top and bottom of the figure), with an indeterminate track of sand mounds
pushed away by some appendage extending to the right just above the middle and an oblique angular
furrow (SM.X.8972) terminating in an oval structure to the left. Locality 2. x 1.

Fig. 2. ArenicoUtes burrows (SM.X.8977, 8) on the broken edge of a sandstone block. Locality 1. x 1-5.

Fig. 3. Koiiphichnium tracks (SM.X.8973) on the upper surface of a sandstone block clearly showing the
impressions of the hind limb. Locality 2. x 2.

Fig. 4. Koiiphichnium track (SM.X.8981 ) progressing across a ripple (top right corner to about mid-bottom)
in the forms of pits and a median drag mark (right-hand side), then down the bivalve mould-covered longer
slope to the left. Here the track is in the form of groups of positive epireliefs best preserved in the top (right-
hand) track with a well preserved group of the bottom (left-hand) track recording the basic chevron pattern.
Locality 2. x 2.

PLATE 50

WRIGHT and BENTON, Rhaetic trace fossils

414

PALAEONTOLOGY, VOLUME 30
Ichnogenus Arenicolites Salter 1857

Pairs of small pits from 0-8 to 2 0 mm in diameter and separated by about 3 to 10 mm are seen on
the upper surfaces of the sandstone slabs. These are the surface expression of U-shaped burrows
and are quite common on some surfaces (PI. 49, fig. 2). One fortuitously broken slab (PI. 50, fig. 2)
shows a vertical section through two of these burrows. They extend from 7 to 10 mm in depth, and
transect the laminae of the sand but do not disturb it otherwise. The absence of spreiten indicates
Arenicolites.

Cut blocks of sandstone show horizons heavily permeated by the Arenicolites burrows (PI. 49,
fig. 1), the activities of which were evidently terminated by an influx of sand that contains only the
occasional burrow and escape structure.

Environmental interpretation. Arenicolites is generally regarded as typical of the shallow marine
Skolitlios and Glossofungites ichnofacies, although it is clear that such structures can also be
produced in freshwater deposits (see Bromley and Asgaard 1979, p. 43).

Ichnogenus Koiiphiclmium Nopsca 1923

Arthropod-like trackways have been observed on ten of the collected blocks, mostly as poorly
preserved trackways (PI. 50, fig. 1) with only sporadic sharply defined impressions (PI. 50, fig. 3).
The trackways are between 15 and 20 mm wide, and in places show a central drag mark 0-2 to
0-5 mm wide (PI. 50, fig. 4). The impressions of the tips of the appendages generally take the
form of short furrows or pit-like scratch marks. The trackway which crosses the crest of the ripple
in PI. 50, fig. 4 develops into a series of somewhat irregularly disposed groups of 5 to 6 rounded
positive epireliefs each up to about 0-7 mm across, arranged en echelon, with only a single series on
the left side of the trail to produce a chevron shape. The well defined prints are about 3 to 4 mm
long and 2 to 4 mm wide (PI. 50, fig. 3) and show three or four ‘toes’, with a small mound of sand
medianly which has been pushed back by the appendage.

Discussion. The evidence from the rather variably preserved tracks points clearly to Kouphichnium,
the trackway of a limulid. The chevron series noted indicates the direction of locomotion of the
animal as being down this slope of the ripple (Caster 1944, p. 77). The positive reliefs here, as
opposed to the negative reliefs elsewhere, are regarded simply as preservation by the compaction of
the sand beneath the appendages; the presence of six impressions, rather than five, and their irregular
disposition is interpreted as the result of stumbling down the slope; other isolated impressions where
some object has pushed into and displaced the sand on a down slope are not uncommon. The toe-
like impressions are typical of the limulid hind foot, with its four or five moveable flat spines which
serve to push the animal forward; while the central furrow, seen from time to time, is interpreted as
a drag mark of the telson.

Environmental interpretation. Limulid tracks are typical of beach and lagoonal deposits, modern
Limulus migrating from subtidal to supratidal environments with maximum densities of adults at
depths of 5 to 6 m on sandy substrates (Rudloe 1979). The classical fossil occurrence is in the
lagoonal Solenhofen Limestone (Walther 1904), and other marginal marine records include the
Upper Devonian of Pennsylvania (Goldring and Seilacher 1971), the Upper Triassic of Arizona
(Caster 1944), and the Rhaetic of Germany (Goldring and Seilacher 1971). Other records of
xiphosuran tracks are in sediments interpreted as non-marine in the Upper Carboniferous of
England and North America (King 1965; Hardy 1970; Goldring and Seilacher 1971; Chisholm
1983; Eagar et al. 1985), the Lower and Middle Triassic of Germany (Goldring and Seilacher 1971;
Pollard 1981), and the Upper Triassic of Germany and North America (Goldring and Seilacher
1971). This is further discussed under Rusophycus (below).

WRIGHT AND BENTON: RHAETIC TRACE FOSSILS

415

Ichnogenus Palaeophycus Hall 1847
Palaeophycus, Type A

Small horizontal and subhorizontal burrows, the sediment of which is the same as the matrix, occur
as convex hyporeliefs on the bases of the sandstone units, and occasionally as endichnial burrows
(PI. 51, fig. 3). The burrows are roughly cylindrical, 1-0-2-5 mm in diameter and run in straight,
slightly curved, or sinuous lines which may branch on occasions, and which disappear upwards
into the sandstone. Some surfaces are particularly heavily burrowed by Palaeophycus, with speci-
mens crossing and overlapping in a complex mass (PI. 51, fig. 4).

Discussion. These burrows are smaller than typical Palaeophycus, which are generally between 3
and 15 mm in diameter (Hantzschel 1975, p. W89).

Palaeophycus, Type B

Many of the Palaeophycus burrows are transversely annulated. These annulated burrows range
from 1 to 3 mm in diameter and occur as short, slightly curved or sinuous convex hyporeliefs,
extending for up to 2 cm before passing upwards into the sandstone bed. The annotations are
spaced at between 0-5 and 0-9 mm intervals and appear to circumscribe the outer surface of the
burrows.

Discussion. The annulations are in general poorly preserved, appearing clear and regular in only a
few cases (PI. 52, fig. 1). Typically only a few clear annulations are seen, with the rest of the burrow
having a irregular undulose surface. It could be that in most cases the grain size of the sandstone
on the lower surfaces is too coarse to preserve the fine detail of the annulations.

Palaeophycus Type A and Type B are differentiated on the basis of having either a smooth or an
annulated surface. It may be that these two types simply represent different preservational aspects
of the same trace fossil; certainly specimens which show good annulations in one part apparently
lack them in another.

Pemberton and Frey (1982, p. 853) recognized five species of Palaeophycus, and anticipated a sixth
species {P. 'annulatus') characterized by continuous annulations along the burrow. The Needwood
specimens would appear to fit into this ichnospecies. While the late Precambrian trace fossil de-
scribed as Torrowangea by Webby (1970, p. 100) shows a comparable narrow diameter (1-2 mm)
to the present specimens, that form has more crudely developed annulations with constrictions at
1 -4 mm intervals, and the larger and strongly meandering trails suggest that it is the product of a
quite distinct trail maker.

Environmental interpretation. Palaeophycus, along with Planolites (below) is recorded from ‘virtually
all sedimentary facies’ (Pemberton and Frey 1982, p. 849).

Ichnogenus Pelecypodichnus Seilacher 1953
(= Lockeia James 1879 nomen ohiitum)

The case for using Seilacher’s generic name and regarding James’s genus Lockeia as a nomen ohiitum
rather than a senior synonym has been put by Hakes (1977, p. 222) and is accepted herein.

Pelecypodichnus is the most abundant trace fossil recovered, and occurs on the lower surfaces of
the sandstone units as almond-shaped convex hyporeliefs, widely interpreted as the resting traces
of bivalves and, as noted above (p. 412), closely associated with the Eotrapezium sp. The hyporeliefs
typically measure between 5 and 12 mm long and 4 to 5 mm wide. There is, however, a considerable
size range, from 2-5 mm up to 16 mm in length. Some specimens are quite broad, being almost as
wide as long (PI. 51, fig. 1).

Poorly preserved specimens appear to be rather rounded, but most examples show some detail
produced by their former occupants. While the resting traces are somewhat irregular, they do show
pointed ends and a clear longitudinal ridge-like structure over the deepest part, which may stand

416

PALAEONTOLOGY, VOLUME 30

away from the surface by up to 4 mm. Variably disposed furrows may define the ridge laterally,
and may be interpreted as impressions made by the valve margins; otherwise the surfaces are
smooth.

Discussion. The density of the Pelecypodichnus varies enormously, with some slabs showing only
the occasional specimen while on others they are very densely packed (PI. 51, fig. 1); counts of up
to 40 within a 5 cm x 5 cm square have been made on this block. Although up to 35 are present
within the same area on the block illustrated in Plate 51, fig. 2, the impression here of reduced
density is simply a reflection of the smaller size of the majority of the individual traces. As was
noted by Osgood (1970, p. 308), although there is no preferred orientation, individual traces may
be aligned one behind the other (see also Seilacher 1955, fig. 5; Hiintzschel 1975, p. W6). These
short chains of up to five or six traces need not necessarily represent the same number of animals
but suggest that, particularly where one trace partly truncates another, they may simply reflect a
repositioning of a single animal. Thus as far as density is concerned, any data on the traces should
not necessarily be interpreted as an indication of the actual numbers of bivalves burrowing down
to that surface.

From the two population samples referred to, it would appear that there are different size classes,
perhaps representing cohorts of different ages. The lengths of 132 specimens from five blocks were
measured to the nearest 0-5 mm and a frequency polygon plotted (text-fig. 2). There may be up
to five size classes represented here, with modes equal to 3-5 mm, 5-5 mm, 8 0 mm, 1 TO mm, and
12-5 mm.

Environmental interpretation. Pelecypodichnus is a poor indicator of environment since it is known
from the distal shelf Cruziana Ichnofacies (e.g. Osgood 1970) to the Scoyenia Ichnofacies (e.g.
Bromley and Asgaard 1979).

Ichnogenus Planolites Nicholson 1873

One block has been collected which shows a horizontal Planolites burrow (PI. 52, fig. 3). The burrow
is preserved as two essentially perpendicularly disposed short segments, each about 20 mm long
and with a diameter of 3-5 mm. One is a roundedly convex hyporelief fading at each end as the
burrow passes upwards into the sandstone, and apparently crossing over the other flatly convex
hyporelief. The latter has evidently been abraded, so that the contact with the adjacent sand shows
the difference between infill and matrix particularly clearly. The infill is finer, better cemented, and
paler, lacking the iron-stained particles of the surrounding rock.

Discussion. The burrow is unlined, with an infill which is structureless, without backfills, and
markedly different from the surrounding rock. Thus despite being the sole burrow of this form

EXPLANATION OF PLATE 51

Fig. 1. Densely packed Pelecypodichnus (SM.X.8956) on the lower surface of a sandstone block. Locality 1.
X 0-5.

Fig. 2. Lower surface of a sandstone block largely covered by Pelecypodichnus (SM.X.8987) and concave
external moulds of Eotrapezium sp. (SM.X.8990). Note size and shape contrast of these Pelecypodichnus
with those of fig. 1 . Locality to the west of Locality 2 as indicated on text-figure 1. x 0-5.

Fig. 3. Endichnial Palaeophycus Type A (SM.X.8993). Lower half of the figure is the upper surface of a ripple
trough; the Palaeophycus is seen medianly emerging on an internal sandstone surface from beneath this and
extending towards the top of the figure. Fine ridge trails (SM.X.8994) and bivalve internal moulds
(SM.X.8996) also present. Locality 2. x 1.

Fig. 4. Lower surface of sandstone block dominated by Palaeophycus Type A (SM.X.8961). Downward facing
internal sandstone surface (top left) has a scatter of external moulds of Eotrapezium sp. (SM.X.8964).
Locality 2. x 0-5.

PLATE 51

WRIGHT and BENTON, Rhaetic trace fossils

418

PALAEONTOLOGY, VOLUME 30

TEXT-FIG. 2. Size-frequency distribution of the lengths of 132 specimens of Pelecypodichmis measured on five

sandstone blocks.

collected, it clearly falls within the scope of PlanoUtes as redefined by Pemberton and Frey (1982,
p. 865).

Ichnogenus Rusophycus Hall 1853

Several specimens of roughly bilobate convex hyporeliefs are preserved. On one block (PI. 52,
fig. 4), the traces are 18-21 mm wide and 36 mm long. There are two lateral lobes, on either side of
a median furrow, each marked by irregular and well spaced obliquely transverse furrows. At the
anterior, there is a broad ‘head shield’ mark, a crescent shaped ridge that cuts across the bilobed
shape behind. The buckler is 18-21 mm wide, and 8-9 mm long.

EXPLANATION OF PLATE 52

Fig. 1. Palaeophycus Type B (SM.X.8998) showing characteristic annulations, associated with Pelecypodichnus
(SM.X.9001 ) on the lower surface of a sandstone block. Locality 2. x 2.

Fig. 2. Broad ridge trail (SM.X.8955) on the upper surface of a sandstone block (detail of PI. 49, fig. 2). x 1.
Fig. 3. PlanoUtes burrows (SM.X.9009) seen as a convex hyporelief (mid-field), with a flatly convex segment
extending to the right and parallel to the top of the figure from about mid-length of the convex relief
Pelecypodichnus resting trace (SM.X.901 1) at bottom left. Locality 2. x 1-5.

Fig. 4. Rusophycus (SM.X.9003) with Pelecypodichnus (SM.X.9006) and occasional external moulds of
Eotmpezium sp. (SM.X.9008) on the lower surface of a sandstone block. Locality 2. x 0-5.

Fig. 5. Rusophycus (SM.X.8986) on the lower surface of a sandstone block. Locality 2. x 1.

PLATE 52

WRIGHT and BENTON, Rhaetic trace fossils

420

PALAEONTOLOGY, VOLUME 30

One other specimen (PI. 52, fig. 5) shows only the buckler (25 mm wide, c. 12 mm long), associated
with a few shallow ‘scratch’ marks behind.

Discussion. These traces are interpreted as shallow burrows made on a muddy surface by the
limulids responsible for the Kouphichnium traces described above. The size of the Rusophycus (18-
21 mm wide) compares well with the Kouphichnium track width (15-20 mm). Further, the head
shield trace matches the form and curvature of a limulid head shield and is again of the correct
size, being as wide as, or wider than, the trackway width.

Rusophycus, characteristic of the distal shelf Cruziana Ichnofacies, is normally assumed to be a
trilobite trace fossil (Osgood 1970; Hantzschel 1975, pp. WlOl-102) and typically shows the same
characteristics as the present specimens; bilobed, irregularly wrinkled convex hyporeliefs on bed
bottoms; occasional presence of head shield markings; association with trackways of the supposed
producer. Typical Rusophycus is recorded from the Upper Precambrian to the Devonian (Hantzschel
1975, p. W102), and it would not be expected after the extinction of the trilobites in the late Permian.
Traces which, like the present ones, are morphologically and ethologically equivalent to Rusophycus
have nevertheless been recorded from Triassic rocks.

The general problem of trilobite traces of Triassic age has been commented on by several authors.
Seilacher (1953, p. 89) referred them to the work of phyllopod crustaceans, while Bromley and
Asgaard (1972, 1979) ascribed some examples from Greenland, and Pollard (1985) some examples
from England and Scotland, to notostracan branchiopods. Several authors (e.g. Osgood 1970,
p. 303; Hantzschel 1975, p. W74; Trewin 1976) attempted to restrict the names Cruziana and Ruso-
phycus to marine trilobite-produced forms, and to name the non-marine examples as Isopodichnus.
There are great problems with this approach, since many non-marine and Triassic examples of
arthropod-produced traces are indistinguishable from species of Cruziana and Rusophycus
(Bromley and Asgaard 1972, 1979; Frey and Seilacher 1980; Seilacher 1985). The problems in
terminology are shown by Hantzschel who himself ascribed a bilobate resting trace from the
Upper Triassic of Germany to Rusophycus (1975, p. W6). Pollard (1985, p. 283) considers the nomen-
clatorial problems of Cruziana- Rusophycus- Isopodichnus and prefers to retain the last name for
Triassic bilobates, but admits that there are no clear morphological criteria for its identification.

We follow Bromley and Asgaard (1979) in naming the Needwood bilobates as Rusophycus on
strictly morphological criteria. The specimens from Greenland (Bromley and Asgaard 1972, 1979)
and England (Pollard 1985) that were ascribed to notostracans differ from the Needwood specimens
in being much smaller, in having well preserved oblique striations on the lobes, and in lacking the
crescent-shaped head shield impressions. Closer to the present specimens in this respect are the
xiphosurid traces described from the Upper Carboniferous of Northern England by Hardy (1970)
and by Eagar et al. (1985). These consist of several crescent-shaped casts of the head shield,
associated with occasional footprints and telson drag marks, which were found in shallow water,
probably non-marine, sandstones associated with Pelecypodichnus. Seilacher also described similar
bilobates from the marine Rhaetic Sandstone of Pfrondorf, SW Germany, although his figured
specimen (1985, p. 233) is a cruzianiform trail rather than a rusophycoid resting trace. He noted
the arcuate impressions of the head shield which result from the xiphosurian tendency to burrow in
a head-down (prosocline) manner. Further Cruziana-Wkc trace fossils, probably made by xiphosu-
rids, have been described from fluvial sequences in the Lower Triassic of South Africa (Shone 1978).

The limulid resting trace recently described by Miller (1982) as Linmiicubichnus is one which
occurs as a shallow concave epirelief on the upper surface of the sand. The tear-shaped outline of
this trace is characterized by a poorly defined anterior margin which is ascribed by Miller to the
slumping in of sand after burrow formation. Eagar et al. (1985) ascribe a selection of xiphosurid
resting traces from the Carboniferous of England to Limulicubichnus. These specimens are convex
hyporeliefs, like the Needwood examples, but they are lunate in shape, mirroring the shape of the
xiphosurid prosoma fairly precisely. It would be inappropriate to use the name Limulicubichnus for
the present examples both in terms of their shape and in terms of their ethological interpretation.
The Needwood specimens of Rusophycus have a deep frontal definition which reflects the burrowing

WRIGHT AND BENTON: RHAETIC TRACE FOSSILS

421

of the animal down into the clay horizon, and the infilling of this mould with sand when the burrow
was abandoned.

Environmental interpretation. Although Rusophycus is normally regarded as typical of the Cruziana
Ichnofacies (marine: distal shelf), with those of non-trilobite origin noted above of the fresh-water
terrestrial environment (Scoyenia Ichnofacies), the Rusophycus traces produced by limulids indicate
a shallow marine to terrestrial environment on the basis of modern limulid distribution, and by
comparison with other fossil occurrences (e.g. Miller 1982; Eagar ei al. 1985).

Furrows

Several slabs show angular furrows with V-shaped profiles, up to 2 mm deep and a width generally
between 1-5 and 2-5 mm, scored across the upper surface of the sandstone. The furrows have
preserved lengths of up to 80 mm and follow broadly curved shapes, with sand grains heaped up to
form an irregular ridge on either side of the furrow. In four examples the furrow ends with an
expansion into an oval area 5 mm or so in width. One slab shows a group of furrows radiating
from a ripple crest to give the fortuitous impression of an irregular star shaped trace (PI. 53, fig. 4).

Discussion. These marks are too deep and too irregular in their direction to be tool marks, and
were evidently formed by some suspended object, 2 mm or more in width, which ploughed through
the sand with some force to leave the V-shaped furrow and to push the displaced sand out sideways.
The oval structure shown in PI. 53, fig. 2 gives a clear indication of the furrow maker changing
direction and pushing the sand to the right as it moved to the left and upwards away from the sand
surface.

The furrows occur on surfaces with Kouphichniwn traces, and while the furrows are more deeply
impressed than the telson drag marks running between the Kouphichnium trackways already noted,
they may well have been formed by some such intermittent trace maker. Very closely comparable
traces have been recorded from fluviatile deposits by Turner (1978), essentially differing only from
the Needwood specimens in their larger size (up to 3 mm deep, and with constant widths of between
7 and 12 mm). These traces were tentatively identified as Scolicia? and the suggestion made that
they could have been produced by snails or bivalves. Similar trails made by modern bivalves have
been figured by Schafer (1972, pi. 17b), Chamberlain (1975, figs. 19. 4K, 19. 8B), and Hakes (1976,
pi. 4, fig. lb).

Beaded furrow trails

A furrow on the upper surface of one slab is unusual in having a ‘beaded’ appearance (PI. 53,
fig. 1); three, and possibly six, other furrows may well be of the same form but are indifferently pre-
served compared with the figured specimen.

The furrow is 10-1 -2 mm wide and 1-0 mm deep, rounded at the bottom; along each edge is a
series of discrete, subspherical lumps with a count of 1 1 in 10 mm and which apparently alternate
on either side. Of 0-6-0-8 mm diameter, they increase the overall width of the trail to a total of 2-4-
2-9 mm. The whole trail is 65 mm long, following a slightly sinuous course and shallowing at one
end where it seems to pass into a beaded ridge, which may indicate the fading of the furrow and
one beaded margin.

Discussion. There are numerous trails on this surface, particularly of the fine ridge trails noted
below, several of which have been cut through by this beaded trail. Although the form of the beads
is commonly indistinct, this is to be expected as the beads themselves are formed from the same
relatively friable sandy sediment as the adjacent surface. This similarity of sediment further mitigates
against their having been produced by a sediment ingestor. Where well preserved, the beads appear
to be regularly developed, and alternate in position on either side of the furrow. The trail seems to
have been formed by an organism ploughing through the surface sediment and displacing fine sand
to either side. The beaded appearance was probably formed by motions of lateral limbs or setae, or

422

PALAEONTOLOGY, VOLUME 30

by jerking locomotory movements, rather than by any passage of sediment through the body of the
organism. Although showing some resemblance to Chevronichnus from the Carboniferous of Kansas
(Hakes 1976, p. 22, pi. 3), the Needwood specimens are distinct in lacking the continuous V-shaped
markings across the furrow that characterize the larger Chevronichnus.

Fine ridge trails

On the upper surfaces of the sandstone are convex epireliefs in the form of narrow rounded ridges
(PI. 53, figs. 1 and 5) which range from 0-3 to 0-8 mm wide and have a preserved length typically
between 5 and 30 mm. Although some show irregularities, most are essentially straight or gently
curving, and tend to fade into the surface at their ends. The ridges do not show branching, but
cross in places, either normally or obliquely. In these cases it becomes apparent that the ridge is
flanked on either side by a flattened or depressed area, each of about the same width as the ridge
itself. The degree of definition of the flanking areas varies from depressions so shallow and poorly
defined as to be virtually imperceptible, to depressions sufficiently deep as to be defined by slight
ridges along the lateral edges of the depression. At intersections where the depressed areas are well
developed, the relative age of the ridges becomes particularly clear as the earlier ridge is cut through
by the flanking depressions of the later ridge. The best preserved ridges are in general located in the
troughs of the ripples. There is no clear preferred orientation on the floor of the troughs, and
although some ridges on the edge of the troughs are aligned normally to the crest, the preserved
remains of others trend more or less parallel to or even along the crests.

Discussion. The fine ridges with their flanking depressions and not infrequent cross-overs are
probably crawling trails produced on or very close to the surface. The trails are suggestive of those
produced by a small snail, possible one similar to the Recent Hydrohia (Schafer 1972, pi. 18a). No
shells of such a snail have yet been found in this sediment, although if their size were comparable
to the Recent forms from Strangford Lough whose spires are about 2 mm in height, recognizable
preservation would appear to be unlikely in view of the poor preservation of the much more
substantial associated bivalves.

Hydrohia is in fact recorded from the Permian fresh water deposits of the Karroo System in
Rhodesia, and is ecologically interesting in that the species H. jenkinsi Smith is known to have
changed from brackish to fresh water habitat within historic time (Cox 1960, p. 187).

Similar types of trails, but with a width from five to ten times as great, were described by Osgood
(1970) from the Upper Ordovician of Ohio. These trails show considerable variation in transverse
profile, with one form passing into another (1970, p. 382). In addition to forms with a median ridge,
forms with a median furrow also developed, which is what would be expected if the snail were
ploughing across the surface of the sand. The use of the foot to draw sand grains inwards to leave

EXPLANATION OF PLATE 53

Fig. 1. Beaded furrow trail (SM.X.9012) cutting across several fine ridge trails (SM.X.9013) on the upper
surface of a sandstone block. Internal moulds of Eotrapeziiim (SM.X.901 5) also present. Locality 2. x 1.

Fig. 2. Angular furrow (SM.X.9016) passing down a ripple slope and terminating towards top of figure as the
furrow maker changed direction, pushing sand to the right as it moved to the left and upwards away from
the sand substrate. Locality 2. x 1.

Fig. 3. Sinuous fine groove trail (convex hyporelief, SM.X.901 9), crossing two external moulds of Eotrapeziiim
sp. (SM.X.9020) on the lower surface of a sandstone block. Locality 2. x 1 .

Fig. 4. Angular furrows (SM.X.9026) on the upper surface of a sandstone block, fortuitously arranged in an
irregular star shape on a ripple crest and associated with numerous poorly preserved internal moulds of
Eotrapeziiim sp. (SM.X.9029). Locality 2. x 1.

Fig. 5. Fine ridge trails (SM.X.8995), crossing in places, on a ladder rippled upper surface of a sandstone
block. Locality 2. x 1.

PLATE 53

WRIGHT and BENTON, Rhaetic trace fossils

424

PALAEONTOLOGY, VOLUME 30

the marginal depressions, and then to organize this sand into a median (presumably mucus coated)
ridge astern, seems a less obvious type of development.

Some small burrows produced by arthropods are similar in appearance. On modern tidal flats
and lake shores, small arthropods such as beetles and mole crickets may produce narrow sinuous
burrows beneath the surface of the sand which look like epichnial trails (Chamberlain 1975, figs.
19. 3E, 19. 4C, 19.5C-E). The trace fossil ''ITrichichnus', described by Hakes (1976, p. 36, pi. 10,
fig. 2) from the Carboniferous of Kansas appears to be very similar to the Needwood specimens.
It is 0-5 mm wide, and occurs as short, fairly straight ‘halbform’ burrows that cross each other.

Environmental interpretation. On the assumption that these traces are produced by gastropods, and
as gastropods live in such diverse environments, they have little to contribute environmentally.
Although the more complex Scolicia type trails have been extensively reported, records of the very
thin trails of simple form described here are not known to the authors.

Fine groove trails

Convex hyporeliefs, similar in form and dimensions to the fine ridge trails noted above, are preserved
in some abundance on the lower surfaces of several sandstone blocks (PI. 53, fig. 3). The convex
ridges are 0'4-0-6 mm wide, with a rounded profile and observed specimens range from 4 to 15 mm
in length. The ridges are straight or curved, less common and generally less well preserved than
the morphologically comparable epireliefs; flattened areas lateral to the ridge are suggested by
occasional rather poorly defined specimens.

Discussion. Although morphologically closely comparable to the convex epireliefs, these convex
hyporeliefs represent the infill of grooves developed on the top of the underlying clay seam or
alternatively may be the infill of burrows of a small organism moving along the base of the sand at
the clay interface. The latter seems unlikely, as the sediment of the epireliefs does not visibly differ
from the adjacent sand on the base of the blocks, and in one sample the ridge contains relatively
coarse quartz grains of up to 0-3 mm, again comparable to scattered coarse grains in adjacent
sediment. These features would appear to rule out the ingestion of sediment by an organism feeding
at the interface. The problem remaining is that the convex epireliefs represent furrows rather than
ridges as on the comparable structure of the upper surface. As already noted, the form of similar
trails is known to vary in this regard (Osgood 1970, p. 382); in the present case the difference may
simply be a reflection of differing substrate lithologies, with the presumed snail leaving a groove on
the mud surface and a ridge on the sand surface.

The trails are rare on surfaces with abundant Pelecypodiclmus and Palaeophychus burrows and
are typically found on relatively smooth lower surfaces with bivalve moulds and with only the
occasional Pelecypodiclmus.

Broad ridge trail

A single convex epirelief in the form of a relatively broad unbranching ridge is present on the upper
surface of a slab with numerous Arenicolites burrows (PI. 52, fig. 2). The epirelief has a low rounded
convex profile, width varying from 5-5 to 7-5 mm, and is variably well defined for about 6 cm where
it stands out as an arcuate ridge on a ripple slope. Where it crosses the ripple crest definition is
poor, but may be traced into the adjacent ripple trough for about 4 cm from one end, mainly by
the continuation of the weathered groove at one side, and apparently all the way across the trough
at the other end, although the trail is here both ill defined and interrupted (PI. 49, fig. 2).

Discussion. Where well defined, the ridge is of smoother appearance than the adjacent sandstone as
a result of the sand grains being packed more closely and better cemented. Further, where the trail
is completely formed there is less mica on the surface; where abraded, mica flakes appear to be as
common as elsewhere in the sand surface. This arrangement is suggestive of a smooth lined burrow
such as Palaeophycus tuhularis Hall (see Pemberton and Frey 1982, p. 859). However, there is no

WRIGHT AND BENTON: RHAETIC TRACE FOSSILS

425

sign of the trace having the cylindrical form of a burrow and it seems to be simply a ridge produced
on the sand surface rather than at a clay-sand interface comparable to the Palaeophycus hyporeliefs
at the sand-clay interface noted above.

Accordingly the ridge appears most obviously interpreted as the trail of a gastropod moving
across the rippled sand surface; it may be that the sedimentary characteristics of the well defined
parts of the trail resulted from compression and/or mucus secretion from the foot of the animal.

ENVIRONMENTAL INTERPRETATION

The Rhaetic succession of the British Isles has long been regarded as representing deposition in a
shallow sea transgressing rapidly across an area of low relief from the south and south-west. In the
lower beds of most areas the shale-mudstone facies is important (Audley-Charles 1970, p. 74), so
that the most familiar lithologies at the base of the Westbury Formation are those of the Rhaetic
Bone Bed and the Rhaetavicula contorta Shales. Thus the clean-washed ochreous sandstones at the
base of the sequence of Needwood are unexpected, and hard to match elsewhere in the British Isles.
Comparisons could be made with the bioturbated sandstones and siltstones at the base of the
Williton Member of North Somerset, a unit within the Blue Anchor Formation that immediately
preceded the Rhaetic. The trace fossils include ArenicoHtes, Diphcraterion, Miiensteria, Planolites,
Rhizocorallium, and Siphonites in addition to several genera of bivalves (including Eotrapezium)
which together indicate shallow marine conditions (Mayall 1981). Another sandy shoreline facies,
recorded only from boreholes along the western margin of the London Platform, is termed the
‘Twyford Beds’ (Donovan et al. 1979, p. 163; Poole 1979, p. 304; Warrington et al. 1980). In this
last paper, the brief description is of a ‘pale arenaceous and rudaceous marginal facies’ (p. 41);
detailed descriptions of the type ‘Twyford Beds’ are to be published in the Chipping Norton Memoir
referred to in these papers.

The Needwood outcrop is clearly well removed from the London platform, and also the other
land areas of low relief in northern England, Wales, the Mendips, and south-west England widely
depicted on palaeogeographic maps (e.g. Audley-Charles 1970, Poole 1979), so that these thin sands
would appear to have had a more local source.

The trace fossils (text-fig. 3) described herein are not too helpful with regard to the depositional
environment of the sands as many are well known facies crossers. These include Palaeophycus,
Planolites, and the various ridges and trails. Other trace fossils restrict the ichnofacies to Cruziana-
Scoyenia (shelf to non-marine: Pelecypodichnus) or to Skolithos-Scoyeuia (marginal marine to non-
marine: Arenicolites, Kouphicimium, and (post-Permian) Rusophychus). The bivalve Eotrapezium,
which is clearly demonstrable as living in these sandy deposits, is a marine form, so that the total
fossil evidence points to a marginal marine environment.

A possible analogue of the present sediments and trace fossils is found in the Lower Rhaetic
sandstones of south-west Germany. The Rhatsandstein occurs in several deposits around Basel,
Metz, Tubingen, Niirnberg, and Coburg. The sandstone is light coloured and fine-grained and it
occurs in long trough-like deposits which pass into clay-flaser sandstones and dark sandy clays in
places. Both the sandstones and clays contain Rhaetavicula contorta, indicating marine conditions
(Geyer and Gwinner 1968, p. 51; Aepler 1974).

The Rhatsandstein of south-west Germany has yielded an ichnofauna consisting of resting traces
such as Asteriacites and Pelecypodichnus, small Cruziana {' Isopodichnus'), 'Cruziana-MV-c' limulid
furrows, limulid tracks {Kouphicimium), and U-shaped burrows {Arenicolites, and forms with
spreiten structures) (Linck 1942; Seilacher 1953, 1955; Goldring and Seilacher 1971; Aepler 1974;
Pollard 1981; Seilacher 1985). The Rhatsandstein contains three main facies types, which have been
interpreted (Aepler 1974) as elements of a prograding delta sequence which contains evidence of
longshore drift, and episodic radial sedimentation during floods. The Needwood sandstones could
be compared with either the Schonbuch or the Pfrondorf facies. The Schonbuch facies consists of
relatively thin (50-200 mm) sandstone beds with U-shaped burrows, resting traces {Asteriacites)
and casts of bivalve shells, together with bone beds. The sandstones are regarded as high energy

426

PALAEONTOLOGY, VOLUME 30

littoral deposits with sediment input from nearby rivers, particularly in the bone beds. The Pfrondorf
facies, which seems most like the Needwood sandstones, consists of thick (several metres) and thin
(50-200 mm) sandstone units inter-bedded with mudstones. The thin units are heavily bioturbated,
with Pelecypodichiius, U-shaped burrows, horizontal burrows and trails, and limulid traces {Kou-
pliiclmium, "Cniziana). This facies has been interpreted (Aepler 1974) as an emergent offshore bar
near to the mouth of a major river.

A number of non-marine ichnofaunas {Scoyenia ichnofacies: Frey and Seilacher 1980) also show
similarities with the Needwood ichnofauna. For example, the Upper Triassic Schilfsandstein of
south-west Germany has yielded Rusophycus (?notostracan), SagUtichnus (?bivalve resting trace),
Pelecypodicimus, Kouphichnium, Merostomicivntes, Planolites, Biformites (small burrows), Cylindri-
cum (vertical burrows), and various vertebrate footprints and unnamed simple trails (Seilacher
1955; Pollard 1981, 1985). The Schilfsandstein is interpreted (Pollard 1981, 1985) as a deltaic deposit
with channel and interdistributary lagoonal facies.

TEXT-FIG. 3. Reconstruction to show the ichnofaunal elements present in the basal Rhaetic sands of Need-
wood Forest. \ —Arenicolites; 2 — Kouphichnium; 2~Palaeophycus; 4 — Pelecypodicimus (with Eotrapezium);
5— Planolites; 6 — Rusophycus; 1 — furrows; 8 — beaded furrow trail; 9 — fine ridge trails; 10— fine groove

trails; 1 1— broad ridge trail.

Another possible non-marine analogue to the Needwood trace fossils is the Pelecypodichnus-
Kouphichniiim- Arenicolites Assemblage described by Eagar et a!. (1985, p. 127) from the Upper
Carboniferous of the Central Pennine Basin. Pelecypodicimus occurs in association with the non-
marine bivalve Carbonicola, and other trace fossils such as Planolites, Cochlichnus, Limulicubichnus
{"Kouphichnium'), and Arenicolites. The environments in which these trace fossils occur are inter-
preted as shallow non-marine interdistributary bay zones which were periodically flooded during
monsoon-like rains and exposed during dry spells.

Thus while the trace fossils in themselves could indicate either marginal marine or non-marine
conditions, the presence of marine bivalves suggests that the sandy facies at the base of the
Needwood Rhaetic is possibly most clearly comparable with the offshore sand bar deposits of the
German Rhatsandstein. In this case the present material gives evidence of the post-Palaeozoic
Rusophycus (and potentially Cruziana) occurring in marine as well as non-marine sediments (cf.
Frey and Seilacher 1980, p. 203).

WRIGHT AND BENTON: RHAETIC TRACE FOSSILS

427

Repository. All specimens figured herein have been deposited in the collections of the Sedgwick Museum,
Cambridge.

Acknowledgements. It is a pleasure to record our thanks to Dr Hugh Ivimey-Cook for his helpful comments
on the bivalves, to Miss Libby Lawson for drawing the text-figures, and to Mr Sean Watters for his photo-
graphic work. We are also grateful to Dr John Pollard for the benefit of his most constructive observations.

REFERENCES

AEPLER, R. 1974. Der Rhatsandstein von Tubingen— ein kondensiertes Delta. Neiies Jb. Geol. Paldont. Abh.
147, 113-162.

AUDLEY-CHARLES, M. G. 1970. Triassic Palaeogeography of the British Isles. Q. Jl geol. Soc. bond. 126, 49-89.
BROADHURST, F. M., SIMPSON, 1. M. and HARDY, p. G. 1980. Seasonal sedimentation in the Upper Carboniferous
of England. J. Geol. 88, 639-651.

BROMLEY, R. G. and ASGAARD, u. 1972. Notes on Greenland trace fossils, I. Freshwater Cruziana from the
Upper Triassic of Jameson Land, East Greenland. Rapp. Gronl. geol. Unders. 49, 7-13.

and 1979. Triassic freshwater ichnocoenoses from Carlsberg Fjord, East Greenland. Palaeogeog.

Palaeoclimat. Palaeoecol. 28, 39-80.

CASTER, K. E. 1944. Limuloid trails from the Upper Triassic (Chinle) of the Petrified Forest National Monu-
ment, Arizona. Am. J. Sci. 242, 74-84.

CHAMBERLAIN, c. K. 1975. Recent Lebensspuren in nonmarine aquatic environments. In prey, r. w. (ed.). The
Study of Trace Fossils, 431-458. Springer Verlag, New York.

CHISHOLM, J. I. 1983. Xiphosurid traces, Kouphichnium aff. variabilis (Linck), from the Namurian Upper
Haslingden Flags of Whitworth, Lancashire. Rep. Inst. Geol. Sci. 83/10, 37-44.
cox, L. R. 1960. General characteristics of Gastropoda. In moore, r. c. (ed.). Treatise on Invertebrate Paleonto-
logy, Part I, 184-11 69. Geological Society of America and University of Kansas Press, Boulder, Colorado
and Lawrence, Kansas.

DONOVAN, D. T., HORTON, A. and iviMEY-cooK, H. c. 1979. The transgression of the Lower Lias over the
northern flank of the London Platform. J. geol. Soc. bond. 136, 165-173.

EAGAR, R. M. C., BAINES, J. G., COLLINSON, J. D., HARDY, P. G., OKOLO, S. A., and POLLARD, J. E. 1985. Trace foSsil
assemblages and their occurrence in Silesian (Mid-Carboniferous) deltaic sediments of the Central Pennine
Basin, England. In curran, a. h. (ed.). Biogenic sedimentary structures; Spec. Pubis Soc. econ. Paleont.
Miner. 35, 99-149.

FREY, R. w. and SEiLACHER, A. 1980. Uniformity in marine invertebrate ichnology. bethaia 13, 183-207.

GEYER, o. F. and GWINNER, M. p. 1968. Einfulirung in die Geologie von Baden-Wiirttemberg. E. Schweitzer-
bart’sche Verlagsbuchhandlung, Stuttgart. 228 pp.

GOLDRING, R. and SEILACHER, A. 1971. Limulid undertracks and their sedimentological implications. Neues Jb.
Geol. Paldont. Abh. 137, 422-442.

HAKES, w. G. 1976. Trace fossils and depositional environment of four clastic units. Upper Pennsylvanian
megacyclothems, northeast Kansas. Paleont. Contr. Univ. Kansas Pap. 63, 1 -46.

1977. Trace fossils in Late Pennsylvanian cyclothems, Kansas. In crimes, t. p. and harper, j. c. (eds.).

Trace Fossils 2, Geol. J. Spec. Iss. 9, 209-226.

HALLAM, A. 1970. Gyrochorte and other trace fossils in the Forest Marble (Bathonian) of Dorset, England. In
crimes, t. p. and harper, j. c. (eds.). Trace Fossils, Geol. J. Spec. Iss. 3, 189-200.

HANTZSCHEL, w. 1975. Trace fossils and Problematica. In teichert, c. (ed.). Treatise on Invertebrate Paleonto-
logy, Part W, .Miscellanea Suppl. 1. W1-W269. Geological Society of America and University of Kansas
Press, Boulder, Colorado and Lawrence, Kansas.

HARDY, p. G. 1970. New xiphosurid trails from the Upper Carboniferous of Northern England. Palaeontologv
13, 188-190.

KING, A. F. 1965. Xiphosurid trails from the Upper Carboniferous of Bude. Proc. geol. Soc. bond. 1626, 162-
165.

LINCK, o. 1942. Die Spur Isopodichnus. Senckenbergiana 25, 232-255.

mayall, m. j. 1981. The Late Triassic Blue Anchor Formation and the initial Rhaetian marine transgression
in south-west Britain. Geol. Mag. 118, 337-383.

miller, m. f. 1982. bimulicubichnus: a new ichnogenus of limulid resting traces. J. Paleont. 56, 429-433.
molyneux, w. 1869. Burton-on-Trent, its History, its Waters and its Breweries. Triibner & Co., London.
264 pp.

428

PALAEONTOLOGY, VOLUME 30

OSGOOD, R. G. 1970. Trace fossils of the Cincinnati area. Paleoiitogr. am. 6, 281-444.

PEMBERTON, s. G. and FREY, R. w. 1982. Trace fossil nomenclature and the Planolites-Palaeoplivcus dilemma,
y. Pu/cont. 56, 843-881.

PHILLIPS, J. 1871. Geology of Oxford and the Valley of the Thames. Clarendon Press, Oxford. 523 pp.

POLLARD, j. E. 1981. A comparison between the Triassic trace fossils of Cheshire and South Germany.
Palaeontology 24, 555-588.

1985. Isopodichnus, related arthropod trace fossils and notostracans from Triassic fluvial sediments.

Trans. R. Soc. Edin., Earth Sci. 76, 273-285.

POOLE, E. G. 1979. The Triassic-Jurassic boundary in Great Britain. Geol. Mag. 1 16, 303-31 1 .

RAAF, J. F. M. de, BOERSMA, J. R. and GELDER, A. VAN. 1977. Wave-generated structures and sequences from a
shallow marine succession. Lower Carboniferous, County Cork, Ireland. Sedimentology 24, 451-483.
RUDLOE, A. 1979. Limidus polyphemus: a review of the ecologically significant literature. In cohen, e. (ed.).

Biomedical Applications of the Horseshoe Crab (Limiilidae). A. R. Liss, New York, pp. 27-35.

SCHAFER, w. 1972. Ecology and Palaeoecology of Marine Environments. Oliver & Boyd, Edinburgh.

SEiLACHER, A. 1953. Studien zur Palichnologie II. Die fossilen Ruhespuren (Cubichnia). Neues Jh. Geol.
Paldont. Ahh. 98, 87-124.

1955. Spuren und Lebensweise der Trilobiten. In schindewolf, o. h. and seilacher, a. Beitrage zur

Kenntnis des Kambriums in der Salt Range (Pakistan). Ahh. math.-naturw. Kl. Akad. Wiss. Mainz 10,
86-116.

1985. Trilobite palaeobiology and substrate relationships. Trans. R. Soc. Edin., Earth Sci. 76, 231-237.

shone, r. w. 1978. Giant Cruziana from the Beaufort Group. Trans, geol. Soc. S. Afr. 81, 327-329.

STEVENSON, I. p. and MITCHELL, G. H. 1955. Geology of the Country between Burton-upon-Trent, Rugeley and
Uttoxeter. Mem. geol. Surv. G.B. 178 pp.

STRICKLAND, H. E. 1843. On Certain impressions on the surface of the Lias Bone Bed in Gloucestershire. Proc.

geol. Soc. Land. 4, 16-18 (also printed in Ann. Mag. nat. Hist. 11, Suppl., 511-513).

TREWiN, N. K. 1976. Isopodichnus in a trace fossil assemblage from the Old Red Sandstone. Lethaia 9, 29-37.
TURNER, B. R. 1978. Trace fossils from the Upper Triassic fluviatile Molteno Formation of the Karoo (Gond-
wana) Supergroup, Lesotho. J. Paleont. 52, 959-963.

WALTHER, J. 1904. Die Fauna der Solnhofener Plattenkalke, bionomisch betrachtet. Denkschr. med.-naturw.
Ges.Jena 11, 133-214.

WARRINGTON, G., AUDLEY-CHARLES, M. G., ELLIOTT, R. E., EVANS, W. B., IVIMEY-COOK, H. C., KENT, P. E., ROBINSON,
p. L., SHOTTON, F. w. and TAYLOR, F. M. 1980. A Correlation of Triassic rocks in the British Isles. Spec. Rep.
geol. Soc. Loud. 13, 78 pp.

WEBBY, B. D. 1970. Late Precambrian trace fossils from New South Wales. Lethaia 3, 79-109.

A. D. WRIGHT and M. J. BENTON

Department of Geology
The Queen’s University of Belfast,
Belfast BT7 INN,
Northern Ireland.

Typescript received 4 April 1986
Revised typescript 7 July 1986

b

■s

I

.V IIIIL. llKh'''L.

t

■ » ■

■S':

If

7

1

>

II

it-

f:.

II

5

: ** A

J

*y. .i:.f-

\ .

H’.

it" '■

.'>

.1

;

'<5

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform in
style to those already published in this journal, and should be sent to Dr. Dianne Edwards, Department of Plant Science,
University College, P.O. Box 78, Cardiff CFl IXL, who will supply detailed instructions for authors on request (these are
published in Palaeontology 1985, 28, pp. 793-800),

Special Papers in Palaeontology is a series of substantial separate works conforming to the style of Palaeontology.

SPECIAL PAPERS IN PALAEONTOLOGY

In addition to publishing Palaeontology the Association also publishes Special Papers in Palaeontology. Members may
subscribe to this by writing to the Membership Treasurer: the subscription rate for 1987 is £33 (U.S. S58) for Institutional
Members, and £16-50 (U.S. S30) for Ordinary and Student Members. A single copy of each Special Paper is available to
Ordinary and Student Members only, for their personal use, at a discount of 25% below the listed prices. Non-members
may obtain copies, but at the listed prices, from Marston Book Services, P.O. Box 87, Oxford 0X4 ILB, England.

RECENT PALAEONTOLOGICAL ASSOCIATION

PUBLICATIONS

Special Papers in Palaeontology

Numbers 1 19 are still in print and are available (post free) together with those listed below:

20. (for 1977): Fossil Priapulid Worms, by s. conway morris. 155 pp., 99 text-figs., 30 plates. Price £16 (U.S. S24).

21. (for 1978): Devonian Ammonoids from the Appalachians and their bearing on International Zonation and Correlation,
by M. R. HOUSE. 70 pp., 12 text-figs., 10 plates. Price £12 (U.S. $18).

22. (for 1978, published 1979): Curation of Palaeontological Collections. A joint Colloquium of the Palaeontological
Association and Geological Curators Group. Edited by m. g. bassett. 279 pp., 53 text-figs. Price £25 (U.S. $38).

23. (for 1979): The Devonian System. A Palaeontological Association International Symposium. Edited by m. r. house, c. t.
SCRUTTON and m. g. bassett. 353 pp., 102 text-figs., 1 plate. Price £30 (U.S. $45).

24. (for 1980): Dinoflagellate Cysts and Acritarchs from the Eocene of Southern England, by j. p. bujak, c. downie, g. l.
EATON andc. L. williams. 100 pp., 24 text-figs., 22 plates. Price £15 (U.S, $23).

25. (for 1980): Stereom Microstructure of the Echmoid Test, by a. b. smith. 81 pp., 20 text-figs., 23 plates. Price £15
(U.S. $23).

26. (for 1981): The Fine Structure of Graptolite Periderm, bv p. r. crowther. 1 19 pp., 37 text-figs., 20 plates. Price £25
(U.S. $38).

27. (for 1981): Late Devonian Acritarchs from the Carnarvon Basin, Western Australia, by G. playforu and r. s. dring.
78 pp., 10 text-figs., 19 plates. Price £15 (U.S. $23).

28. (for 1982): The Mammal Fauna of the Early Middle Pleistocene cavern infill site of Westbury-sub-Mendip, Somerset, by
M. j. BISHOP. 108 pp., 47 text-figs., 6 plates. Price £25 (U.S, $38).

29. (for 1982): Fossil Cichlid Fish of Africa, by j. a. h. van couvering. 103 pp., 35 text-figs., 10 plates. Price £30 (U.S. $45).

30. (for 1983): Trilobites and other early Arthropods. Edited by d. e. g. briggs and p. d. lane. 276 pp., 64 text-figs., 38 plates.
Price £40 (U.S. $60),

31 . (for 1984): Systematic palaeontology and stratigraphic distribution of ammonite faunas of the French Coniacian, by w. j.
KENNEDY. 160 pp., 42 text-figs.. 33 plates. Price £25 (U.S. $38).

32. (for 1984): Autecology of Silurian organisms. Edited by m. g. bassett and j. d. lawson. 295 pp., 75 text-figs., 13 plates.
Price £40 (U.S. $60).

33. (for 1985): Evolutionary Case Histories from the Fossil Record. Edited by j. c. w. cope and p. w. skelton. 202 pp., 80 text-
figs., 4 plates. Price £30 (U.S. $45).

34. (for 1985): Review of the upper Silurian and lower Devonian articulate brachiopods of Podolia, by o. i. Nikiforova,
T. L. MODZALEVSKAYA and M. G. BASSETT. 66 pp.. 6 text-figs., ]6 plates. Price £10 (U.S. $15).

35. (for 1986): Studies in palaeobotany and palynology in honour of N. F. Hughes. Edited by d. j. batten and
D. E. G. BRIGGS. 178 pp., 29 plates. Price £30 (U.S. $50).

36. (for 1986): Campanian and Maastrichtian ammonites from northern Aquitaine, France, by w. j. Kennedy. 145 pp., 43
text-figs., 23 plates. Price £20 (U.S. $35).

Field Guides to Fossils

1. (1983): Fossil Plants of the London Clay, by m. e. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $12).

Other Publications

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $30).

1985. Atlas of Invertebrate Macrofossils. Edited by j. w. Murray. Published by Longman in collaboration with the
Palaeontological Association. xiii-r24l pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24.95.

© The Palaeontological Association, 1987

Palaeontology

VOLUME 30 ■ PART 2

CONTENTS

Idamean (Late Cambrian) trilobites from the Denison Range,
south-west Tasmania

J. B. JAGO 207

The preservation of conifer wood: examples from the Lower
Cretaceous of Antarctica

T. H. JEFFERSON 233

Early Devonian conodont faunas from Buchan and Hindi,
Victoria, Australia

R. MAWSON 251

The oldest ammonoid ‘colour’ patterns: description, compari-
son with Nautilus, and implications

R. H. MAPESand D. A. SNECK 299

Early Cretaceous belemnites from southern Mozambique
P. DOYLE 311

Taxonomy, evolution, and functional morphology of southern
Australian Tertiary hemiasterid echinoids

K. J. MCNAMARA 319

Evolution and phylogenetic classification of the Diplo-
graptacea

C. E. MITCHELL 353

Trace fossils from Rhaetic shore-face deposits of Staffordshire
A. D. WRIGHT and M. J. BENTON 407

Printed in Great Britain at the University Printing House, Oxford
by David Stanford, Printer to the University

ISSN 0031-0239

iSNi NviNOSHilws S3iyvyan libraries Smithsonian institution NoiiniiiSNi nvinoshui/

rr, O

>

rn ^ m p<^J^ ^ m 'Vg\DC^ xiruaro^ m

C/) E C/> — CO ± (/)

^lES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS SaiaVBan LIBRARIES SMITHSONl’

CO 2 . CO Z CO Z CO

< . 2 .< s , < .oix V s:

t:

2 x-i^ > 5 >

* ^ u) ~ :z. in z CO * z

liSNI NVINOSHIIWS S3iavaaiT LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHUI

r; <o — in CO —

RIES SMITHSONIAN INSTITUTION NOliniliSNI NVlNOSHiilAIS S3iaVBan LIBRARIES SMITHSONl

CD
70
>

30 \r>

m ^ x^Ast;^ rn - rn

CO _ in — in

NviNOSHiii\s S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshii

CO -z. ^ (n ^ z . CO z

liSNi NviNosHiiws S3iavaan libraries Smithsonian institution NoiiniiiSNi nvinoshui

z

o

CD

> N

TO

[Ti ^ m ^ m in pri

CO X E — CO

RIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS S3iavyail LIBRARIES SMITHSONl
in 2 ,., CO z CO z CO

- ^xS ? V- - y A

I
in

o

Cl \CAS»»»CiP/ Cl " ^

> 2 2 ''XLSS-^ > 5 ^x >

z if) ‘‘ z in z CO * z

liSNI NVlNOSHiilAIS S3IBVyan libraries SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHII

— , , CO 2" Z ^

‘Xx Lu CO ^ . CO z

_ '§J VC '^^4^ -1

^ ^ O 'X^y^^Z ^ Q ovj^ _ xc4ua»^ Q

_ _ Z _| Z -I z"

RIES SMITHSONIAN INSTITUTION NOliniliSNI NVINOSHilWS SSIBVaaiT LIBRARIES SMITHSONl

z

o

c

o z o

—J * 2 — I 21

Hvaan libraries Smithsonian institution NoiiniiiSNi nvinoshhins ssidvaan libr;

Jj'jTiivii?*' n pc ^ ni 'so^oc — / w PI

(S) £ 05 £ CO

riTUTiON NoiiniiiSNi NvmosHiiiMS S3iavaan libraries Smithsonian institution Noiin

,v, 00 z ^ W Z ,<V- ^ Z 05

-■ ,/ y-y^ ' Z -H /Wi —

05

X

CO

o

z

«s

4A f %% i

Z </5 z C/5 ‘ 2: (O - z

avaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiws saiavaan libr

5 t/’ 2 ^ ^ ^ ^ X 05

00 ^ y ,y ♦ vV<^N> -

—4 /Ky^wwVJw^^X .✓ >..' ^ y^»./i. — ^ 1 r z\ /jv^tfa w T _J .'X V 'Vx^ / w Vlfek*^

I) <

cn

SOcHT^ O '■^^ONoy^ 5 XOjvDcy' “ 5 O

riTUTioN NoiiniiiSNi“'NviNosHiiws ssiavaaiT u b rar i es^smithsonian'^institution Noun

z t-, z r- z ^ z

m y<rra?ox 2 m 2 z^

oo

m ■ X \,>vA^>A>y m xojosyz ^ m

X 05 — CO £ 05

avaan libraries Smithsonian institution NoiiniiiSNi nvinoshiiiais saiavaan li

2 , W Z » CO Z OO Z

I ^

^ .V —4 /^^(4Iwk1Bu. lA _v _ .1 I

\,

X
CO

> \ii:A2n>y ^ .-^^A- -, ^ 5’^ 5

05 ' 2 CO •*■" Z CO Z 05

ITUTION NOIiniliSNI_NVINOSHillMS S 3 I a V a a ll_ L I B R A R I E S SMITHSONIAN_INSTITUTION NOliO

<

QC
CD

c

Q X^QSV^

2 -J

avaan libraries smithsonian institution NoiiniiiSNi nvinoshiiims S3i avaan libr

Z r- z r- z i~

Ui

riTUTION NOlinillSNI NVINOSHillNS S3iaVaaiT LIBRARIES SMITHSONIAN INSTITUTION NOliH

5 -V- ^

— ^ ZuSVAtX ..■

* Z CO 2 CO* Z (O'Z

avaan libraries smithsonian institution NoiiniiisNi nvinoshiiws saiavaan libR)

CO X CO ~ , CO —

Za^^sov^X ixJ £ ojj y^^SlvTTx ^ ^

DC Zj

c/

ITUTION NOliniliSNI “‘nvinoshiiiais ^S 3 I a V a 8 IT ~' L I B R A R I E S^ SMITHSONIAN"*! NSTITUTION NOliO

LIBRARIES

3 9088 01375 6994