CD
33
>

^

m x»/vpc>^ — - -33^ m \«i/ "'•S^nS^x m

CD _ CD _ CD \ ± CD

TUTION NOlinillSNI NVINOSH1IINS S3IHVaai1 LIBRARIES SMITHSONIAN INSTITUTION NOlinill

\

X
CD

0

1 2 3,

ivaan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIWS^S 3 I B V a 8

UJ

1_ LI B RAR

1 3

o pc y — o _ '<^Un^y o x> " Xj gUusg X Q X2 n_dc,

TUTION^ NOlinillSNI ^NVINOSHIIWS SBIHVaan^LIBRARI E S*^ SMITHSONIAN - * INSTITUTION 2 NOlinill

Z r~ \- 2: r- z r* 2

vaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3iavaail L I B R A R

W%

z

<

A y

o

CD

X

2

CD

TUTION NOlinillSNI NVIN0SH1IWS S3IHVaai1 LIBRARIES SMITHSONIAN INSTITUTION NOlinill

O

_i z _j x -J

ivaan libraries Smithsonian institution NOlinillSNI nvinoshiiims saiavaan librar

ijussX m vt i* pc ^z .u m — xduo^x m

CD — CD £ CD

ITUTION NOlinillSNI NVIN0SH1IWS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOlinilJ

... CD Z _ ^ CD Z - ^ Z AV-, 00

Jjvl. 2 .< s E ^ ^ '<

JVd a ll _ LI B RAR I ES^SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIIMS S3iavaa

CD ~ CD — CD

1_ LI B RAR
z

UiiS^X o Xg\Tc " '‘SJg®’" o pAX Z NgpciTsg^ o N ' — o X^!_dS^

ITUTION ^ NOlinillSNI ^NVINOSHIIIMS ^S 3 IHVaail^LIBRARI E S^ SMITHSONIAN INSTITUTION NOlinill

i w \i n 1

m v \\y ^ asv^ x m

co — co

IDDADIFQ QMITUCOMIAM IMCTITIITIHM MHI I D1 I ICKII

’ MWIMOQU I 11AJQ QlDIVWfln LIBRAR

z

o

CXI
70
>

70 „

~ S5? - n»- w „

rn xjov/ w n^ojTx^o/ m w i/> x ^ousg z m E2 Yy /%j0?' m x ^£*dc

co _ to \ fE to — co

noshiiws saiavaan libraries Smithsonian institution NouniusNi nvinoshiiws s 3 i y v ;

CO Z CO 2 co Z

,/ 2 — 1 z ,wv. h ./ *.y?L

co W^m. x CO co

1 r y/& t

2 " > vguiI<<X 5 ^ 2 >

z CO * Z y> *• Z CO Z

THSONIAN INSTITUTION NOliniUSNI NVINOSH1IWS S3IUVHai1 LIBRARIES SMITHSONIAN _ INSTITl
— co — co 2 to 2

, CO

q v ymiy O XJ^dC/ '<5$^'" O

_ z " _i Z -J Z

M0SH1IINS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOliniUSNI NVIN 0 SH 1 II/MS S3IBV
z r- z f~ Z ^ ^ [2 v ^

O ^5> . ~ /ToaT^x o mV —

CXI

70

> l=S

to \iy

m EJ m XQvoiv'V' m

co £ co _ co

THSONIAN INSTITUTION NOliniUSNI NVIN0SH1MS S 3 I a V a 0 IT _L I B R A R I E S ^SMITHSONIAN _l NSTITl
7T . to z co

•5 v \X V o» ' 2 X^vosBiZ > Z >

^ > Z CO Z CO ' Z co

NOSHiiws S3iavaan libraries Smithsonian institution NouniusNi_NvmosHims saiav

co — co — co

- W m ^

% ce /~2 -» /^ywyix ~ Wmjto h

§|k <C (gfc 1

^ \^r 'VZL •?/ ( — \H -rafll 1 "• J&f'/f'J

oc VT^x.iJrP/ 5 ^ y-y r

co

O ® r o 5 X Jvas* $Z O

_ Z _l Z — I z _J

ITHSONIAN INSTITUTION NOliniUSNI NVINOSHIII/MS S3iaVHaiT LIBRARIES SMITHSONIAN I NST1T

z
o

CD
70
>

33

— VC.

m
co

noshiiws S3iavaan l

co Z

BRARIES SMITHSONIAN INSTITUTION NOliniUSNI NVIN0SH1II/MS S3 I a V

ITHSONIAN INSTITUTION NOliniUSNI NVIN0SH1IIMS S3IBVyail LIBRARIES SMITHSONIAN _ INSTIT

o

NOSHIIWS S3iavyail LIBRARIES SMITHSONIAN^INSTITUTION ~N0lini!lSNI~NVIN0SHlllMS_S3 I y V

§ ” / 4 pl^\ § # 5 ^ ” ? /^P^\ 5 ^ ■' ^ ^

/,# *a g (if ' -

m % ’ ■" m w m >W" z

CO _ co

i r* o OMiti ire

Palaeontology

VOLUME 28 ■ PART 3 AUGUST 1985

Published by

The Palaeontological Association • London

Price £21 -50

THE PALAEONTOLOGICAL ASSOCIATION

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

COUNCIL 1 984- 1985

President : Professor C. Downie, Department of Geology, University of Sheffield, Sheffield SI 3JD

Vice-Presidents'. Dr. J. C. W. Cope, Department of Geology, University College, Swansea SA2 8PP
Dr. R. Riding, Department of Geology, University College, Cardiff CF1 1XL
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. R. Lord, Department of Geology, University College, London WC1E 6BT
Secretary: Dr. P. W. Skelton, Department of Earth Sciences, Open University, Milton Keynes MK7 6AA
Circular Reporter: Dr. D. J. Siveter, Department of Geology, University of Hull, Hull HU6 7RX
Marketing Manager: Dr. R. J. Aldridge, Department of Geology, University of Nottingham, Nottingham NG7 2RD

Editors

Dr. D. E. G. Briggs, Department of Geology, Goldsmiths’ College, London SE8 3BU
Dr. P. R. Crowther, Leicestershire Museums Service. Leicester LEI 6TD
Dr. L. B. Halstead, Department of Geology, University of Reading, Reading RG6 2AB
Dr. R. Harland, British Geological Survey, Keyworth, Nottingham NG12 5GG
Dr. T. J. Palmer, Department of Geology, University College of Wales, Aberystwyth SY23 2AX

Other Members

Dr. E. N. K. Clarkson, Edinburgh Dr. C. R. C. Paul, Liverpool

Dr. D. Edwards, Cardiff Dr. A. B. Smith, London

Dr. P. D. Lane, Keele Professor T. N. Taylor, Columbus

Dr. A. W. Owen, Dundee

Overseas Representatives

Australia: Professor B. D. Webby, Department of Geology, The University, Sydney, N.S.W., 2006
Canada: Dr. B. S. Norford, Institute of Sedimentary 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. Cuffey, Department of Geology, Pennsylvania State University, Pennsylvania
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 1985 are:

Institutional membership
Ordinary membership
Student membership
Retired membership

£45-00 (U.S. $68)
£21-00 (U.S. $32)
£11-50 (U.S. $18)
£10-50 (U.S. $16)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr. A. R. Lord,
Department of Geology, LIniversity College, Gower Street, London WC1E 6BT, England. 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, Gosta Green, 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 1985 will receive Palaeontology , Volume 28, Parts 1-4.
All back numbers are still in print and may be ordered from Marston Book Services, P.O. Box 87, Oxford OX4 1LB, England,
at £21-50 (U.S. $33) per part (post free).

Cover: The dinoflagellate cyst Impagidinium patulum (Wall) Stover and Evitt, from bottom sediments in the North Atlantic
Ocean. British Geological Survey specimen M.P.K. 4234. x 2500.

PALAEONTOLOG Y REVIEW

THE ORIGINS AND AERODYNAMICS OF FLIGHT
IN EXTINCT VERTEBRATES

by KEVIN PADIAN

Abstract. Active (flapping, powered) flight has evolved in only three groups of vertebrates: pterosaurs (late
Triassic), birds (late Jurassic), and bats (early Tertiary). Gliding has arisen many times in vertebrates, is a
separate adaptation from flying, and does not appear to be a prerequisite for active flight. Skeletal features
that distinguish flyers from gliders include modifications of the pectoral and forelimb apparati, elongation of
the distal part of the wing skeleton for thrust, and certain physiological modifications that often leave clues in
the skeleton. Soaring evolved in birds and pterosaurs secondarily, after powered flight was well established in
both groups: it is a necessary result of phyletic size increase outstripping the ability to meet power requirements
for sustained flapping.

The origin of flight can be approached through a combination of phylogenetic, functional, and aerodynamic
evidence. A basic question is whether flight evolved in the trees or on the ground. Of the three groups of
active flyers, two (pterosaurs and birds) show no trace of gliding antecedents and appear to have evolved
flight directly from the ground. Bats show many morphological and phylogenetic indications of an arboreal,
gliding ancestry and are very different in all such respects from pterosaurs and birds. The theory of an arboreal
origin of flight in birds so far lacks support from phylogenetic and functional-morphologic evidence; arguments
in favour of this theory have invoked hypothetical selective advantages of features that either cannot be tested
or apply equally to a terrestrial origin. Most of these features were already present in the coelurosaurian
dinosaur ancestors of birds. Pterosaurs were structurally and functionally convergent on birds in many
locomotory respects, and show prima-facie evidence of a cursorial, non-gliding origin of flight.

Aerodynamic considerations of extinct vertebrates have mainly focused on two animals: Archaeopteryx (the
first known bird) and Pteranodon (a specialized Cretaceous pterosaur). Functional inferences from skeletal
evidence imply that Archaeopteryx was capable of flapping flight, though most 'modern' avian flight features
were not developed; it does not seem well built for gliding. Pteranodon (a soarer, not a glider), like many large
birds, was capable of active flight but probably only used it to take off, gain altitude, and avert danger. All
pterosaurs were strong, active fliers and only large size constrained this ability. Aerodynamics of Pteranodon
have commanded much productive interest, but nearly all models have been based either on ( 1 ) a morphologic
analogy to bats, which is structurally incorrect, or (2) an aerodynamic analogy to certain low-speed aircraft
or hang-gliders, which is both structurally and aerodynamically incorrect. Reappraisal of the anatomy and
aerodynamic parameters indicate that Pteranodon s flight range was higher and that it was more active and
manoeuverable than previous studies have suggested, and so more comparable to modern soaring birds.
Studies of flight in extinct organisms cannot rely solely on engineering models or presumed selective advantages
or pressures; they must take into consideration all aspects of phylogeny, function, and aerodynamics.

Powered flight is a difficult and complex adaptation which commands attention as a truly 'major
feature’ of adaptive evolution. Aerodynamic requirements are severe, and they constrain the
kinematics of the flight stroke (the defining feature of powered flight) and the morphology of the
wing into adaptive channels that have converged in several vertebrate lineages. For this reason
the problem of the origin of flight is especially approachable in macroevolutionary terms (Padian
1982). By comparing the morphology, phylogeny, and ecology of various kinds of flying verte-
brates the evolutionary origins of flight may be studied. This approach shows that, although many
features of vertebrate flight are common to all flyers, the differences in morphology and

| Palaeontology, Vol. 28, Part 3, 1985, pp. 413-433.)

414

PALAEONTOLOGY, VOLUME 28

ecology suggest different evolutionary pathways to the same ends (Pennycuick 1972; Rayner 1981;
Padian 1983/ff.

The purpose of this work is to assess the flying abilities of extinct animals. The first step is to
review briefly the several modes of air travel, and to show how skeletal indicators of these modes
may appear in the fossil record. A basic introduction to the known fossil flyers is followed by a
consideration of the origins of flight in birds, pterosaurs, and bats. Finally, treatments of the
aerodynamics of extinct vertebrates are reviewed, and new interpretations suggested.

It should be noted that in the present context ‘flight’ denotes active, powered, flapping flight only.
Other modes of air travel that are not self-powered (gliding, parachuting, and soaring) are thereby
differentiated; unfortunately space does not permit extensive treatment of the latter. The scope of
this review is by no means exhaustive, but is an introduction to current issues. Important problems
such as energetics, neurology, and migration must be largely omitted, as there is little direct evidence
of them from the fossil record.

TYPES OF VERTEBRATE FLIGHT

Air travel in living vertebrates has often been divided into four modes: parachuting, gliding, flapping,
and soaring (Lull 1906; Savile 1962; Hildebrand 1982). These have recently been defined and
contrasted by Rayner (1981) in an excellent review of flight adaptations in living vertebrates.
Parachuting is usually distinguished from gliding in two ways: the force responsible for the majority
of aeronautic support (lift for gliding, drag for parachuting), and the angle of descent (shallower
than 45° for gliding, steeper than 45° for parachuting). This can be reduced to a consideration of
lift/drag ratio. In evolutionary terms, parachuting may require less morphologic modification, for
it is merely a way of slowing a fall; gliding implies a greater horizontal component and a longer
time in the air, often with a relatively precise pre-selected landing point. Some authors believe
that only radially symmetrical organisms can really qualify as parachuters, though others define
parachuting as drag force exceeding lift force. Flapping flight, as its name implies, is defined by the
flight stroke, which imparts power in the form of forward thrust. When the resulting increase in air
speed from the flight stroke is applied over the surfaces of an aerodynamically efficient airfoil, a
pressure differential creates thrust, enabling the animal to gain altitude regardless of assistance from
winds or a high starting point. Soaring is a secondary adaptation in large birds and (apparently)
large pterosaurs that evolved from flappers and still retain some capacity to flap. Soaring, which
has often been likened to falling down an up escalator, allows the animal to make use of thermals
(convection rings of rising air) to gain energy that offsets the animal’s weight. Such energy may
also be gained from winds, as well as from wind speeds that vary with altitude (Pennycuick 1972;
Brower 1983). Soaring is both energetically inexpensive and advantageous to predators with acute
long-distance vision; the low energetic cost appears to have enabled many soaring birds and
pterosaurs to grow phyletically to a size at which flapping for extended periods is energetically
impossible. One would think that there is no reason why gliding animals, incapable of powered
flight, could not directly evolve a soaring habit. However, gliders have a wing of poor aspect ratio
for optimal soaring performance, they cannot easily avoid various kinds of aeronautical hazards,
and there are few ecological advantages of soaring to them because they are not visually oriented
predators. Rayner (1981) has pointed out that bats do not soar because there are no convective air
currents at night. In the absence of selective pressure to become diurnal, seek large prey, and
develop acute vision, soaring probably would not be useful for bats. Gliding animals, by virtue of
their ecology and diets, also have no reason to soar and it is not surprising that this mode of air
travel is largely restricted to a secondary adaptation in groups of predatory and scavenging flappers.

Skeletal correlates of aerial adaptations can provide insight into the aerodynamic abilities of
extinct vertebrates. Gliding adaptations are difficult to recognize in the fossil record because they
often leave no skeletal clues, and airfoils are seldom preserved. In the absence of an airfoil,
adaptations to some form of air travel may be recognized by analogy to modern forms (text-fig. 1).
For example, the hyperelongated ribs of the lizard-like Kuehneosaurus (upper Triassic, Bristol

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

415

Channel: Robinson 1962) and the closely related Icarosaurus (upper Triassic, New Jersey: Colbert
1966, 1970) are quite similar to those of the modern agamid lizard Draco , which uses its ribs to
support a gliding membrane (Colbert 1967). Morphologic features can also indicate functional,
aerodynamic, and even physiological abilities and limitations in fossil forms. Several such criteria
are available to distinguish active flyers from passive gliders (Padian 19836).

text-fig. 1 . Three reptiles that modified their ribs as gliding organs, a, Weigeltosaurus , upper
Permian (after Evans); b, Icarosaurus , upper Triassic; c, the living agamid lizard Draco. Scale bars

represent 2 cm.

1. The defining feature of a flying vertebrate is its flight stroke. Adaptations related to the
generation of the power stroke include an expanded bony sternum or breastbone (including a
pronounced median keel) for anchoring the flight muscles, a shoulder girdle that is braced to the
sternum, an enlarged deltopectoral crest on the humerus for the insertion of flight muscles, and a
shoulder articulation that limits forearm movement to activities compatible with the flight stroke
(text-fig. 2). Flying vertebrates have developed these features to a relatively greater or lesser degree
(bats rather less than birds and pterosaurs in some respects), but such features are never found in
the skeletons of gliders.

2. The forelimb proportions in aerial vertebrates are often greater than in non-aerial relatives.
Flying squirrels, for instance, have humerus and forearm segments significantly elongated over
those of other squirrels (Bryant 1945; Thorington and Fleaney 1981). True flyers take this a step
further in that the outermost segment of the wing, comprising the wrist (birds), hand (bats), or one
finger (pterosaurs), is hypertrophied, which never happens in gliders (text-fig. 2). This is the area
that provides thrust (forward motion) in active flight, whereas the inner two wing segments provide
lift (on which a glide depends for aerial support: see Rayner 1981).

3. The airfoil in gliders is normally a simple extension of skin and superficial muscle, stretched
by bony elements (limbs and bone spars in most gliders, but ribs in others), and without any (or
with only rudimentary) internal support structures (Jepsen 1970; Thorington and Fleaney 1981;
Novacek 1982). In flyers the airfoil is always stiffened by anteroposteriorly oriented structural
elements. These are the feather shafts of birds, the fingers of bats, and the intercalated wing fibres
of pterosaurs (Zittel 1882; Wellnhofer 1975; Padian 19836). These may reduce spanwise tension in
the wing membrane, a problem noted by Bramwell and Whitfield (1974) for pterosaurs, and equally
applicable to bats. Such structural elements certainly act in all flyers to give camber to the wing
and to provide competence of the airfoil during the flight stroke.

4. Gliders, which are all arboreal, retain most locomolory abilities of their non-gliding relatives.
This is not true for flyers, whose limb structures have been modified to accommodate the kinematics
of the flight stroke by reducing unrelated mobility at certain joints, lightening and strengthening

416

PALAEONTOLOGY, VOLUME 28

bones, and eliminating unnecessary muscle weight from the wings. Limitations of normal mam-
malian locomotion are obvious in the bats (except Desmodus , the vampire bat, which is secondarily
modified for walking and jumping: Altenbach 1979). Ostrom (19766) demonstrated the restrictions
of certain wrist and hand movements in post -Archaeopteryx birds, compared to their coelurosaurian
dinosaur ancestors. And the limitations on the motion of pterosaur forelimbs, documented by
Hankin and Watson (1914), Bramwell and Whitfield (1974), Wellnhofer (1978), and others, have
been shown to be modifications related to the down-and-forward flight stroke common to birds
and bats (Padian 1983a, 6). This stroke and its aerodynamic effects are well understood in living
forms (Pennycuick 1972; Rayner 1979).

5. Pneumatic foramina (holes in the skeleton for expansion of respiratory surface into the bone
cavities, used to cool the blood), and thermoinsulatory coverings such as feathers and fur, indicate
a level of metabolism necessary to sustain active flight. Bats lack pneumatic foramina, though birds
and pterosaurs have them. Therefore, whereas such foramina are not necessary for flight, their
presence leads to only one inference (Seeley 1870). Pterosaurs and birds, unlike other known
diapsids (crocodiles, squamates, and Sphenodori), evolved a thermoinsulatory covering. Bats, of
course, are furred, but so are other mammals, so this feature by itself does not relate to flight.
Indeed, it is not even clear that these features evolved strictly in the context of flight in the other
two groups, but knowledge of such fossilized structures is poor.

TRUE VERTEBRATE FLYERS

Pterosaurs have their earliest records in the Norian (upper Triassic) of Italy (Wild 1978), birds in
the upper Jurassic of Germany (with apparent occurrences in France and Utah [Jensen 1981]), and
bats in the Eocene of North America (Jepsen 1970) and Germany. Pterosaurs died out at the close
of the Cretaceous Period, along with all the dinosaurs except birds, which by their coelurosaurian
ancestry are properly considered theropod dinosaurs (see below). Pterosaurs coexisted with birds,
and have a more extensive and diverse fossil record than birds, throughout the Cretaceous. During
this time the record of birds is virtually restricted to open-water forms such as the ternlike Ichthyor-
nis , the diving, flightless Hesperornis, and its relative Baptornis (Marsh 1880). At the close of the
Cretaceous some other forms appear, but these are only poorly known (Brodkorb 1963) and only
tenuously linked to living orders. Cretaceous pterosaurs are also almost entirely found in marine
facies, and have commonly been presumed pelican- or gull-like in their habits. Because the known
Cretaceous representatives of these groups evidently had similar ecologies, on face value it may be
inferred that pterosaurs did not die out from competition with the birds, but rather from a failure
to keep diversifying and replacing taxa, a necessary component of evolutionary persistence. Bats
seem to have occupied a nocturnal, originally insectivorous adaptive zone since sometime in the
early Tertiary, after the pterosaurs were gone and the birds already well established.

Pterosaurs. Pterosaurs have recently been reviewed and revised by Wellnhofer (1970, 1974-1975,
1978: anatomy and diversity) and Padian (1979, 1980, 1983a, b\ locomotion and flight). In light of
recent work a summary of salient features of pterosaurs may be given. They were active flyers with
a shoulder girdle strongly buttressed to the sternum, which was widely expanded over the ventral
thorax and keeled in the midline (Wellnhofer 1978). The sternum anchored the flight muscles, and
as in birds the recovery stroke of flight, powered by the M. supracoracoideus equivalent, was effected
by a pulley mechanism involving the acrocoracoid process of the coracoid, which changed the
primitive action of the muscle from an adductor to an elevator (Padian 19836). The wing could be
folded, but the joint separating the second and third of the three major functional units of the wing
was between the fourth metacarpal and its phalanx, not at the wrist as in birds (text-fig. 2). The
wing was a membrane of skin with a network of closely intercalated ‘fibres’ that provided strength
and camber; these ‘fibres’ are never found folded but are always gathered, so their structural in-
tegrity is evident (Zittel 1882; Wellnhofer 1975; Padian 19836). They may have been modified scales,
and were presumably keratinous. The wing was brought forward through a down-and-forward path

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

417

bird

pterosaur

bat

c a

text-fig. 2. Diagrammatic comparisons of the thoracic regions and forelimbs of the three groups of
vertebrate flyers. Thoracic regions (above) are seen from the front; right forelimbs (below) in dorsal
view. Structurally, the coracoids of pterosaurs and birds and the clavicles of bats appear to be analogous,
as do the bird’s furcula, the cristospine of pterosaurs, and the manubrium of bats; the last two structures
are situated at the anterior extreme of the sternum. Abbreviations: c, carpus; ca, calcar; clav , clavicle;
cor, coracoid; furc, furcula; hum, humerus; me, metacarpus; pt, pteroid; r, radius; sc, scapula; ster,
sternum; u, ulna; I-V, numbered digits. Not to scale.

418

PALAEONTOLOGY, VOLUME 28

during the flight stroke, and retracted by an up-and-backward motion, as in birds and bats (Padian
1983/;). The hindlimbs were sufficient for bipedality to be the only means of terrestrial locomotion;
pterosaurs could not walk on all fours because the forelimbs could not rotate past the limit of the
forward flight stroke (Padian and Olsen 1984; Padian 1983/;). The femur was held in a diagonal to
horizontal position nearly parallel to the body midline, as in birds and most dinosaurs, and the gait
was parasagittal and digitigrade (Padian 1983r/, b ). The Pterosauria comprise some forty genera,
traditionally divided between the paraphyletic Rhamphorhynchoidea and their monophyletic
descendants, the Pterodactyloidea; diagnostic differences are reviewed in Wellnhofer (1978) and
Padian (1980). Pterosaurs, though not dinosaurs, were very closely related to them and share with
them many synapomorphies (Gauthier 1984); their closest known sister taxon appears to be the small
ornithosuchian archosaur Scleromochlus (von Huene 1914;Padian 1980, 1984; Gauthier 1984).

Birds. The fossil record of birds begins with Archaeopteryx, from the upper Jurassic Solnhofen
limestones of Bavaria. The history of the five known specimens (plus the original feather) has often
been reviewed (see e.g. Ostroin 1979 and references therein). Even more frequently repeated is the
concept of the mosaic ‘half-reptile, half-bird’ morphology of Archaeopteryx , although until recent
years there was no convincing picture of which ‘reptiles’ included the direct ancestors of birds.
Ostrom, in a series of papers (1973-1979), established that birds were descended from small co-
elurosaurian theropod dinosaurs, on the basis of a series of unusual and generally overlooked
characters that were unique to these dinosaurs and birds (represented by Archaeopteryx). Padian
(1982) formalized Ostrom’s evidence and arguments, along with additional evidence, into a testable
cladistic framework in which some fifty synapomorphies of Archaeopteryx and coelurosaurs were
recognized; Gauthier (1984) has expanded this list to over 120 (Gauthier and Padian, in press).
Critics of Ostrom’s theory (e.g. Walker 1977; Tarsitano and Hecht 1980; Martin et at. 1980; Martin
1983) have quarrelled with interpretations of individual characters or have argued that some
proposed synapomorphic features are ‘not similar’, and have pointed out resemblances of either
Archaeopteryx or modern birds to other selected fossil archosaurs. However, these critics have
not recognized or addressed the structure of cladistic methodology, which is a valuable tool for
reconstructing phylogeny precisely because it transforms mere lists of characters (e.g. Martin 1983)
into hierarchical distributions of nested sets of characters, thereby forming a more robust logical
structure than a list. In this context, non-hierarchical claims of ‘similarity’ or ‘dissimilarity’ have no
objective meaning. Statements about sister-group relationships between two taxa must be gauged
against a third in order to establish phylogenetic homology; the latter is deductive, not declarative,
and so, logically, are statements about ‘dissimilarity’ (non-homology at a given phylogenetic level).

As Martin (1983) and others have pointed out the coelurosaurian hypothesis is based almost
exclusively on an extensive series of post-cranial characters (though the skull of Archaeopteryx is in
all recognizable respects coelurosaurian), which ostensibly evolved independent of the others.
Ostrom compared these across a wide range of coelurosaurs and other archosaurs. Walker (1973)
and Martin el al. (1980), who advocate a common origin of crocodiles and birds (among as yet
unspecified ‘thecodonts’), have mainly relied upon certain features of the ear region which are
unfortunately not preserved in the crucial coelurosaurs. Because two-thirds of a comparison cannot
establish anything with respect to the other third, at present the evidence of the ear region is only
tantalizing, though potentially quite valuable. Critics of Ostrom’s post-cranial theory though have
yet to demonstrate or even propose that any other specified taxon is closer in these respects to
Archaeopteryx', consequently, it must be provisionally accepted that birds are descended from
theropod dinosaurs— in fact, they are the closest sister-group of deinonychosaurian coelurosaurs
( senso Colbert and Russell 1969; see Padian 1982, and Gauthier and Padian, in press).

This phylogenetic premise is necessary to understand which skeletal features conventionally
regarded as ‘avian’ are really avian and not simply dinosaurian, theropodan, or coelurosaurian.
The typical textbook litanies of ‘avian’ characters include hollow bones, a lightly built skeleton, long
forelimbs, fused clavicles, and a keeled sternum; yet all these features are already synapomorphies of
coelurosaurian dinosaurs, and the last is not preserved in Archaeopteryx (Gauthier and Padian, in

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

419

press). At present the only known character distinguishing Archaeopteryx as a bird is the flight
feathers, which demonstrates that Archaeopteryx was an active flier (Feduccia and Tordoff 1979).
Later birds are distinguished from Archaeopteryx by the fused and reduced wrist and fingers, and
the ossified contact of the coracoid with the expanded sternum, both modifications for flight; and
by the fusion of pelvic elements, reduction of teeth, and various cranial features. These, however,
must be regarded as ancillary to the origin of birds; in an adaptive sense they can be viewed as
fine-tuning the flight mechanism and the avian life-style.

Bats. Evidence for the origin of bats is indirect. The earliest fossil bat is Icaronycteris, from
the Green River (Bridgerian: middle Eocene) of Wyoming (Jepsen 1970). Its complete skeleton,
magnificently preserved, has been subjected to some phylogenetic debate, but most workers prefer
to assign it, with some reservations, to the Microchiroptera. It has a long tail and many other
primitive features, but it is in all respects a flying bat, with fully developed wings. Possible dental
records of bats from the early Eocene are provocative but shed no light on the question of bat
origins; spectacular skeletal fossils from the Middle Eocene (Lutetian) Messel pits of the Darmstadt
region of Germany represent several species of primitive bats. Recent re-evaluation of several
ancient, generalized, closely related placental groups (including the Scandentia or tree shrews,
bats, primates, and lipotyphlan insectivores) suggest that dermopterans and bats are sister groups
(Novacek 1982). If this view prevails the understanding of plesiomorphic characters and ecological
factors in the origin of bat flight may fit a cohesive evolutionary pattern (see below).

ORIGINS OF FLIGHT

‘Origins’ has been left plural because mounting evidence suggests that pathways to vertebrate flight
have differed, depending as much on phylogenetic constraints as on aerodynamic ones. Powered
flight is defined here by the common use in all flyers of the down-and-forward flight stroke of the
wings. The wing produces both lift (inner segment) and thrust (outer segment). How such wings
evolve is not so clear-cut. In order to be convincing, explanations of the evolution of flight must be
consistent with empirical knowledge. Evolutionary theory can support many kinds of adaptive
explanations, but only a fusion of many independent lines of evidence can suggest which historical
explanations are more appropriate in a given case (Padian 1982).

Bird flight. Far more attention has been paid to birds than to bats and pterosaurs with regard to
the origin of flight. Ostrom (1974, 1979) reviewed the old dichotomy between the terrestrial, cursorial
origin (‘from the ground up’: Williston 1879; Nopcsa 1907, 1923)— obscure and unsupported by
evidence for nearly a century— and the arboreal, gliding origin (Marsh 1880; Bock 1965), which has
predominated in evolutionary thought. The latter is intuitively more convincing, perhaps because
the images of climbing, leaping, parachuting, gliding, and finally flapping sail past our eyes like
cartoons in a flipbook. The question is not whether this theory is possible; I will argue later that it
is, at least for bats. The question is really whether it is supported by evidence; and so far the answer
is surprisingly negative, at least for birds. This is odd considering the reliance most modern birds
place on arboreal life; however, to study the origin of flight is not to deal with why modern birds
live in trees, but how ancient birds got into the air. Post hoc arguments are not applicable.

Bock (1965) championed the arboreal theory by showing that it was consistent with the neo-
Darwinian Modern Synthesis of evolution — invoking no teleology and no inadaplive stages, with
each adaptive level self-sufficient and derivable from the previous one by small steps. Selective
advantages of each intergrading stage were self-evident, on the basis of their survival and success in
modern forms. But as valid as these assumptions may be, they hold true regardless of any empirical
evidence that may be brought to bear on the origin of birds and their flight. Bock's theory, in
the absence of evidence, reduces to a statement of belief that flight evolved in accordance with the
Modern Synthesis; but many possible explanations can be accommodated by the Synthesis. The
point at issue is to discover which factors could discriminate between alternate theories.

There are at least three kinds of factors at work; phylogenetic, functional, and aerodynamic

420

PALAEONTOLOGY, VOLUME 28

B

traditional model

'streamlined' model

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

421

(Padian 1982). Phylogenetically the most robust hypothesis is that birds began from small, active,
carnivorous, bipedal theropod dinosaurs. As Ostrom (1974) realized, any satisfactory theory of bird
origins has to begin with Archaeopteryx ; and any satisfactory theory of Archaeopteryx must consider
the habits of its closest sister-group. This, the predatory Deinonychosauria, included agile, terres-
trial, cursorial bipeds with long arms, large brains, and stereoscopic vision (Hopson 1980; Russell
1980; Ostrom 1980). Trends toward these features in the evolution of coelurosaurs is clear (Padian
1982; Gauthier 1984; Gauthier and Padian, in press). Though perhaps nothing about the skeletons
of these animals, including Archaeopteryx , precludes at least the smaller forms from climbing trees,
no evidence favours it (Ostrom 1979)— despite a wealth of skeletal correlates in modern birds for
scansorial and arboreal adaptations (Feduccia 1973; Bock and Miller 1959). Therefore, unless
compelling arguments can be adduced for an arboreal stage in the evolution of flight, it seems to
me that the phylogenetic and palaeoecological evidence of terrestrial ancestry must be taken at face
value.

Coelurosaurs had many structural features that later figured in the evolution of flight. These
attributes may be adduced from the extensive skeletal comparisons of coelurosaurs and other
archosaurs by Ostrom ( 1 974- 1 976). Some of these have a direct relationship to the later development
of the flight stroke in birds. In addition to the anatomical features mentioned above, Ostrom (1969)
noted the semilunate wrist joint in deinonychosaurs (text-fig. 3a), which allowed them to flex their
long hands laterally against the forearm— in fact, to fold them exactly as a bird folds its wing
(Ostrom 19766). Archaeopteryx shows no specialization of the forelimb bones beyond those of other
deinonychosaurs, and the bones themselves are only slightly longer proportionally than in its
larger, non-volant relatives. Direct palaeontologic evidence indicates that deinonychosaurs used the
forelimbs to grasp prey while attacking it with the teeth and clawed feet (Kielan-Jaworowska
1975)— just as Ostrom (1969) had predicted. To seize prey from a retracted position the humerus
must be protracted and adducted, the forearm extended, and the hand extended by swinging forward
and mediad, pushed in part by differential movement of the radius relative to the ulna (text-fig. 3a).
(When the elbow is flexed the radius slides forward over the ulna and flexes the long hand against
the forearm, as in birds.) The shoulder joint is a ball and socket with only partial restriction of
movement, the elbow is a hinge, and the wrist another hinge; the mobility of these joints is equivalent
to or greater than those of birds, and restricted in similar ways. Therefore only the smallest
conceivable modification is necessary to change the functional repertoire of the deinonychosaurian
forelimb to accommodate the down-and-forward motion of flight. Because these structures and
functions were useful in a very different context for terrestrial predators, it cannot be argued that
they evolved specifically for flight. They were co-opted from a predatory function, and this would
only have been possible in a terrestrial setting, as Ostrom (1974) explained:

Climbing and flying involve different sets of muscles and require very different movements of the various
forelimb components. In all probability, selective forces that tended to perfect one activity would not have
been optimal for the other. And while we can rationalize the advantages of climbing into trees as a necessary
precursor to the earliest stages of the evolution of flight, from a functional anatomical aspect the two activities
are unrelated.

text-fig. 3. a, left pectoral girdle and forelimb of the coelurosaurian dinosaur Deinonychus, in lateral view:
left , with arm folded, right, with arm extended. The semilunate carpal synapomorphic of deinonychosaurs
and birds is at the base of the three digits, b, right hindlimbs in lateral view: the fruit bat Pteropus , the
pterosaur Dimorphodon , the small theropod dinosaur Compsognathus , the first known bird Archaeopteryx,
and the pigeon Columba. Note the differences in orientation, femoral head, fibular location, and metatarsal-
phalangeal structure between the bat and the four archosaurs; the former hangs upside-down in trees, whereas
the latter are and were presumably active terrestrial bipeds, c, dorsal view of the Cretaceous pterosaur
Pteranodon in flight: left side, the traditional ‘wide-winged’ model, after Bramwell and Whitfield 1974; right
side, the revised ‘streamlined’ model based on new analysis of the forelimb and hindlimb articulations (Padian
19836), and on wing impressions preserved in the genus Rhamphorhynchus.

422

PALAEONTOLOGY, VOLUME 28

What of the role of feathers? Most workers have accepted Regal's (1975) contention that feathers
evolved initially as modified scales to enhance thermoregulation, though whether to shed heat,
retain it, or both, depends on answers to palaeoenvironmental and palaeophysiological questions
that may never be resolved. In any case. Regal’s thoughtful analysis applies only to body feathers
(down and contour); flight feathers are clearly specially modified contour feathers, and their in-
sulatory function when folded is almost certainly secondary in an evolutionary sense to their
development as a flight organ. Ostrom (1975a, b) proposed, following the predatory ecology of the
coelurosaurs, that hypertrophied feathers on the forelimbs were selected as an aid to batting down
flying insects. Although this use is consistent with generally distributed predatory traits in theropods,
it is difficult to see how the improvement of a predatory function such as this would have paved the
way for the development of flight, a locomotory function (Padian 1982). (Ostrom produced a
parallel argument, quoted above, to dispel the arboreal theory’s putative connection between
climbing and flying.) Martin (1983) commented that a solid mesh of feathers was a poorly designed
'net' that probably would have only blown the insect prey farther away. Caple et al. (1983) showed
that the 'insect nets’ would have generated severe instability and loss of balance. They proposed
instead that if the earliest birds and their immediate forerunners caught prey with their teeth instead
of their hands, the arms would have been very effective bilateral stabilizers during a jump into the
air. Even a forelimb surface expansion capable of lifting 1% of the animal’s body weight would
have had a significant effect on stability. A greater surface area would result in even greater stability,
which, combined with faster takeoff speed, would result in increased lift, a longer time in the air,
and presumably a more successful insect forage. Ostrom and most workers have since conceded
the advantages of this model (Lewin 1983).

Caple et al. (1983) made the terrestrial flight model a strong contender by overcoming the
objection that when the winged proto-bird leapt into the air it would immediately lose speed from
its only source of power (the legs). The authors set up the basic requirements for the evolution of
the flight stroke itself, but did not pursue it to the specific case of birds; I have shown that this
stroke is almost fully evolved already in deinonychosaurs, though nearly inconceivable in any other
contemporary animals. Any protraction and extension of an airworthy, feathered forelimb would
have increased lift and time in the air — whether useful for pursuit of prey (Ostrom 1975a, b ), escape
from predators (Harrison 1976), or simply more agile running over broken ground (Padian 1982),
is not important. Any repetition of such a stroke sustains the animal in flight even longer. From
these modest beginnings the flight of birds evolved by steps no less adaptive, incremental, or
self-sufficient than those of the arboreal scenario. The difference is that ecological stages for which
no evidence exists are not invoked. (See Harrison 1976 for several perceptive comments on this
issue.)

The fossil record indicates that the immediate ancestors of birds were terrestrial, agile, bipedal,
cursorial, and predatory. It does not indicate that they were arboreal, climbers, parachuters, or
gliders. Bock (1983) suggested that feathers would have evolved to advantage in treetops, where
heat loss is allegedly greater than on the ground; that stereoscopic vision would have been useful
for proto-birds clambering through branches; and that long feathers would have helped break an
accidental fall from the trees. These hypothetical advantages have yet to be supported by evidence
for arboreality. Martin (1983) asserted that Archaeopteryx 'was more adapted for moving about in
the trees than for a life in the open plain’, based on an analogy to primates. His contention that
Archaeopteryx could not run or even stand fully erect hinges on an interpretation that the proximal
femur is obliquely oriented in the acetabulum; however, this orientation applies to dinosaurs,
modern birds, pterosaurs, and most mammals, all of which walk parasagittally (text-fig. 3b). This
advanced condition is sharply contrasted with the ‘semi-erect’ condition in 'thecodonts’ and croco-
diles, all of which have a primitive sigmoid femur with a head that is continuous with the shaft, not
set off by a distinct neck. The mobility and anatomy of every joint in the hindlimb of Archaeopteryx
must be considered; Martin (1983) dealt only with the long proportions of the hindlimb, which he
suggested was an adaptation to jumping. He did not detail his scenario in which 'a small arboreal
reptile with a tendency toward bipedality . . . [which] was improved by vertical climbing

PAD1AN: FLIGHT IN EXTINCT VERTEBRATES

423

and leaping’, developed flight, nor how fully terrestrial abilities might have re-evolved. It seems,
on balance, that many unnecessary steps must be invoked only to get the trees in there somehow.
Though both hypotheses demand further work, most workers in the recent literature seem to
have accepted Ostrom’s ideas on the anatomy, phylogenetic relationships, and functional morpho-
logy of Archaeopteryx (e.g. Bakker and Galton 1974; Wellnhofer 1974; Desmond 1975; Padian
1983rt, b; Thulborn and Hanley 1982; Gauthier 1984; Cracraft 1977; Colbert 1980; Russell 1980;
Hotton 1980; Bakker 1980; Halstead and Halstead 1981; McGowan 1980; Caple et al. 1983; etc.).
The aerodynamic model of Caple et al. (1983) promises to be highly productive in further in-
vestigations of the evolution of bird flight (Lewin 1983).

Bat flight. The question of the origin of bat flight is in some ways at about the same stage as the
question of bird flight was a decade ago, perhaps because so little is known about the ancestry of
bats. Once Ostrom proposed a specific origin of birds the question of the origin of their flight
assumed a whole new dimension, because models could be constructed on actual taxa. This was
particularly important with birds because they have no living relatives that are the least bit like
them ecologically.

Though the fossil record does not reveal much about the origin of bats, they share a close
common ancestry with other orders of small mammals of nocturnal, arboreal, insectivorous, or
omnivorous habits (lipotyphlan insectivores, Dermoptera, and Scandentia). Dermopterans are at
least as ancient as bats, if the fossil record gives any indication. However, no one would propose
that the modern dermopteran is a plausible Urtyp for the earliest bat: for reasons laid out by Jepsen
(1970) the colugo is highly specialized for its inverted, fruit-eating lifestyle. But the forests in which
the colugo now lives were certainly not always of their present compositions, and therefore it is
reasonable to assume that dermopterans have changed with their environment, as bats have. It is
highly probable that in the early Eocene or Palaeocene the members of the two groups looked more
like each other than their modern representatives do.

Perhaps from these considerations a general idea of proto-bat ecology may be extrapolated. Let
ns assume, as nearly all workers on the problem have, that the ancestors of bats had the ecological
characteristics noted above. As in many primitive mammals, there may also have been a rudimentary
sense of echolocation, though perhaps the mechanism was not homologous to the organs used in
chiropteran echolocation. Because these features are generally distributed among the sister-groups
of bats, no special explanation of the adaptive value of these characters to bats is necessary. At this
point, to go further in the investigation of the origin of bat flight requires a more specific statement
about the closest sister-taxon of bats. If, for example, the Dermoptera were so established (Novacek
1982) the investigation is reduced to three alternatives: (1) bats did not go through a gliding stage,
and evolved powered flight completely independently of the dermopterans’ evolution of gliding;
(2) the common ancestors of bats and dermopterans went through a gliding stage, and the two
lineages subsequently diverged; (3) bats and dermopterans independently evolved a gliding stage,
and the bats went on from there to evolve powered flight. The phylogenetic milieu is a powerful
source of information about the context of functional evolution. Without a better fossil record of
the earliest bats and proto-bats the most promising line of evidence for the origin of bat flight
would seem to be analysis of the interrelationships of known orders and the trends that mark their
histories. However, some interesting ideas and approaches of previous workers suggest that even in
the absence of phylogenetic information, the functional problem can be explored.

Jepsen (1970) proposed three stages in the evolution of bat flight. Stage 1, the pre-bat, was much
like the animals described above, except that Jepsen postulated ’large (and, possibly, webbed) front
feet’ useful in leaping after prey, with hind legs and feet that ‘could be extended outward (laterally)
from the body when it moved around in crevices’. It also could hang by its hind feet, as flying
squirrels can, and could leap from this posture to a nearby target. Stage 2, the sub-bat, had ‘webbed
large hands (or small wings) which were used principally in catching flying prey’, and the proto-wings
of skin ‘enabled the sub-bat to be very briefly sustained in the air by rapid flapping’ after prey. The
legs were now fixed laterally. Stage 3 is the essentially modern bat, with fully grown wings and

424

PALAEONTOLOGY, VOLUME 28

refined skeletomuscular adaptations. It is important to note that Jepsen did not believe that bats
ever passed through a gliding stage: he regarded gliding as a separate evolutionary plateau (or dead
end). Instead, like Ostrom, he placed great reliance on the hypertrophy of the hands as prey-catchers.

Two problems with this are: (1) how did a (normally) quadrupedal mammal get around in trees
with these large webbed hands, and (2) once again, why should the improvement of a predatory
function pave the way for a locomotory function? The evolution of inverted posture is indeed very
important, and (as Sam McLeod once suggested to me) may have evolved well before the other
features: otherwise, how would the forelimbs be freed for flight? Once suspended upside down, it is
easier to drop to a lower target than to climb up, so presumably the advantage of a gliding ability
is not eliminated (see below).

Smith (1977), working from the model of a typical gliding proto-bat, regarded the expansion of
the wrist and hand membranes almost as a developmental by-product of elongating the digits along
with the other forelimb bones. (Fair enough, but why then do not other mammalian gliders have
hypertrophied hands?) He suggested that ‘the continued development of the wing, in this manner,
eventually would have produced an ungainly and clumsy structure that necessitated movement as a
wing rather than a fixed gliding device’. The theory stresses the random nature of raw materials
upon which selection may act, but the development of the flight stroke and the form-function
complexes of bat limbs and girdles is left unexplained. Smith, however, concluded that bats passed
through a gliding stage. His view contrasts with that of Pirlot (1977), who suggested that bat flight
began as brief periods of hovering while jumping at insects from the ground, again without a gliding
stage. Clark (1977) argued against this because the curve of power requirements for increasing flight
speed is U-shaped; therefore it would have been far less costly for bats to begin with medium-speed
flight, because hovering is as expensive as high-speed flight. Clark concluded that ‘it is more
reasonable to suggest that bat ancestors were gliders which gradually evolved the capacity for
sustained (and controlled) flight at speeds where power requirements were minimal'.

One argument in favour of a gliding origin for bats is that, if their ancestors were indeed arboreal,
they would almost have had to have been gliders first: an animal that experiments with powered
flight in the treetops risks mortal danger at each outing without some kind of airfoil to break the
fall. Evidence for such a glider-type design is found in the configuration of the wing in bats, the
only flyers to incorporate the hindlimbs into the airfoil, as all mammalian gliders do (Padian 1982).
Without a gliding stage, it must be postulated that the legs became incorporated into the wing only
after flapping flight evolved, which did not happen in birds or pterosaurs. An alternative is that
enlargement of the hand, and evolution of the flight stroke, occurred in bats after the gliding habit
was established. The gliding membrane could have been the ‘safety net' for the evolution of flight
in an arboreal setting.

Pterosaur flight. The section is quite brief because there is almost no discussion of the origin of
pterosaur flight in the literature. This is hardly surprising, as most writers have considered pterosaurs
mere gliders, and their exact phyletic origins have not been well understood. Von Huene (1914)
suggested that pterosaurs evolved from small arboreal ‘thecodonts’ like Scleromochlus , which
jumped from branch to branch, then developed parachuting, gliding, and flapping flight. Romer
repeated von Huene’s origin of flight theory nearly verbatim in all editions of his Vertebrate
Paleontology , but leaving out mention of Scleromochlus , which he considered a dinosaur. Elsewhere
I argue the opposite (Padian 1984): that von Huene got the phylogeny right, but the scenario wrong.
Scleromochlus is the closest known sister-group to pterosaurs, as von Huene thought (Padian
1980; Gauthier 1984). But in locomotory adaptations it was a small, light, bipedal runner, and so
were pterosaurs for nearly the first hundred million years of their existence (upper Triassic-upper
Cretaceous).

Pterosaurs parallel birds in so many adaptive respects that every argument applicable to the
terrestrial theory given above for birds also applies to pterosaurs (Padian 1983/?). They stood, held
their limbs, and moved their joints in almost exactly the same ways (text-fig. 3b), and such
adaptations as the acrocoracoid process of the shoulder girdle, the restricted glenoid fossa, the

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

425

coracoids buttressed to the sternum, the narrow wings unconnected to the feet, the pelvic con-
figuration, the reduced fibula, and the mesotarsal ankle suggest, even with the regrettable paucity
of supporting fossils, that pterosaurs evolved flight in a cursorial, terrestrial context, without a
gliding stage. They never developed an avian-style perching foot, their hind claws were never
sharply curved (unlike their fore claws), and they always kept a low femur/tibia ratio and a high
metatarsal/tibia ratio characteristic of lightly built, active animals (Coombs 1978), and un-
characteristic of non-avian arboreal forms. Perhaps they could climb trees; but as in the earliest
birds, no evidence currently supports this point, whereas ample evidence indicates high proficiency
as terrestrial bipeds. The origin of the first group of vertebrate flyers, unfortunately, is far more
poorly known than the origins of flight in the other two groups, and their comparative biology far
more difficult to approach.

AERODYNAMIC PERFORMANCE OF FOSSIL VERTEBRATES

Apart from calculations made by Colbert (1966, 1967, 1970), Evans (1982), Thorington and Heaney
(1981), and others of the weight, wing area, and wing loadings of various gliders, studies of
aerodynamic performance in fossil vertebrates have centred on two animals: Archaeopteryx and
Pteranodon. In both, estimates have been made of gliding performance, with some consideration of
minimal power requirements for flapping flight.

Archaeopteryx. It is important to remember that most work on the aerodynamics of Archaeopteryx
preceded Ostrom's hypotheses of theropod ancestry and terrestrial origin of flight. In this aero-
dynamic work it was assumed that Archaeopteryx was arboreal and mainly a glider, which flew
weakly if at all. If Ostrom’s ideas (later modified and developed mathematically by Caple et al.
1983) are correct, a gliding stage would have been aerodynamically obviated, because gliding from
the ground up is so ineffective. There is no way to tell how much, if any, gliding Archaeopteryx did,
but there is certainly value in estimating its gliding performance, as well as its power requirements.

Flying animals can glide at a range of speeds, merely by varying the incidence of the glide. They
also flex the wings at high speeds in order to obtain a range of glide performance (J. M. V. Rayner,
pers. comm.). Gliding performance is maximized when the gliding angle (proportional to the sinking
speed) is low, because the lift is high relative to drag. But if the lift is too high, the animal slows
until it stalls. The minimum flying speed (F min ) is achieved when lift is maximized (C L max: just before
stalling) and is inversely correlated with it. This is expressed by the formula

V 2 = 2 WlpSC L

in which W, the weight (mass x gravity), approximates the lift ( L ) in a steady glide, p is the density
of air, and 5 the area of the airfoil. Cz, max is best calculated empirically, and ranges from 1-3 to 1-6 in
modern birds and bats (Pennycuick 1972). Because p is usually assumed, the critical biologic
variables are the weight and wing area, the quotient of which is called the wing loading, and is
roughly proportional to gliding speed.

The aerodynamic analysis, then, begins with calculations of weight and wing area. Heptonstall
(1970) confirmed Jerison’s (1968) estimate of the former at 500 g, and calculated the wing area at
373 cm 2 , exclusive of body and tail surfaces. Bramwell (1971) and Yalden (1971a) argued for lower
weights (200-250 g) and larger lift areas (479 cm 2 , including 91 cm 2 on the body between the wings).
Yalden (19716) compared Archaeopteryx to birds of similar wing span (58 cm) and found a weight
range of 170-300 g; the same range was found for mammals of similar head-body length (21-8 cm).
Yalden used estimates of 150, 200, and 250 g in his calculations, favouring the intermediate value.

Heptonstall (1970) used the formula given above to calculate what he believed to be the maximum
flying speed of Archaeopteryx at 20-9 m/s. Heptonstall calculated L by estimating maximum bending
moments possible on a humerus with a tensile strength commensurate with experimental results,
thus deriving the maximum lift possible. In his formula the expression L was equated with W. a
common practice when calculating performance in a steady glide, in which maximum lift is generated

426

PALAEONTOLOGY, VOLUME 28

(Pennycuick 1972). However, Bramwell (1971) argued that the formula only works for minimum
speed (Fmi„): maximum lift is not generated with maximum speed, and vice versa. Yalden (1971/?),
using a similar formula from Pennycuick (1969), calculated the minimum power speed at 6-9, 7-6,
and 8-2 m/s, depending on the weights listed above. Because in modern birds some 15% of body
weight is pectoral muscle, Yalden took 30 g as the available weight for the power stroke, and
derived a power requirement of 105-140 watts/kg. All these figures are well within avian range,
but Yalden noted that this is not surprising, as he based all estimated values on those observed in
modern birds. He concluded that if Archaeopteryx approached modern birds in muscle physiology,
it probably could have flown.

The important point to be made about all the work discussed above is that the calculated values
are within the ranges of modern birds. Despite differences in estimates and derivations the results
agree to within a factor of two, and usually much closer, and are therefore reliable if not precise. In
this context it is interesting to consider Yalden’s estimates of minimum power speed, the speed at
which the least work has to be done. Once again, the U-shaped curve relating power requirements
to flight speed shows that very low and very high speeds are most expensive. Starting flight from a
standing position, then, requires a lot of energy to be expended initially, before getting up to an
economical speed. A running takeoff can minimize the effort required by the wings to build up this
speed, and if the wings are merely spread significant lift can be generated. Caple et al. (1983) based
calculations on a cursorial model of 100 g, shaped as a cylinder 15 cm long and 3 cm diameter, and
applied a ground speed of 3-4 m/s derived from empirical observations by Taylor (1973). Because
maximum running speed is observed to vary proportionally with mass, it is clear that an animal the
size of Archaeopteryx , endowed with the cursorial skeletal adaptations of coelurosaurs, would have
had no trouble bringing its ground speed up to minimum flying speed.

Once in the air, the flapping performance can be estimated only if the physiology is known or
assumed. The gliding performance depends on wing loading and aspect ratio (the shape of the wing:
wingspan squared divided by wing area). Of great importance is the ability of a flying animal to
land. To land slowly and easily, most animals reorient the body and beat the wings vigorously to
achieve minimum speed flight. (This discussion applies only to landing on the ground, as Caple et
al. ( 1983) have shown the difficulty of landing on a branch to an animal that is not already extremely
sophisticated in its flight. ) Heptonstall ( 1 970) reckoned that the high wing loading and sinking speed
of Archaeopteryx would have made landing very rough, though Bramwell (1971) figured that the
tail would have reduced the stalling speed; Yalden (19716) and Bramwell (1971) both calculated
lower wing loadings. In view of the cursorial adaptations of Archaeopteryx , it may be surmised that
a running landing was possible, so that the airspeed need only have been reduced to the minimum
flying speed in order for the legs to take over. Heptonstall (1970) calculated optimal gliding speeds
at 10-15 m/s, but the lower wing loading suggested by other authors would have reduced this figure
considerably. By my calculations, stalling speed for Archaeopteryx would have been on the order
of 5-6 m/s, and a short burst of flapping (generating enough lift to slow airspeed) would have
enabled a running landing at a speed of 3-4 m/s. (Use of the alula in landing was probably not
available to Archaeopteryx.)

Pterosaurs. Fascination with the aerodynamics of these extinct archosaurs began well before man
invented powered flight. Early work particularly reflected the hope that pterosaurs would reveal
possibilities for human flight, though as soon as workable aircraft were invented interest in ptero-
saurs quickly cooled. In the past decade it has been rekindled by the opposite hope, that modern
advances in aviation might reveal how pterosaurs flew.

Most attention by far has centred on the crested pterodactyloid, Pteranodon. Rhamphorhynchus,
an earlier, smaller, long-tailed form, was studied by von Kripp (1943) and a flapping model built
and flown by von Holst (1957); however, the model would work properly only when the leaf-shaped
vane at the end of the tail, preserved in several fossil specimens, was oriented horizontally, whereas
its true orientation is vertical. Apart from these studies, modern work has been almost entirely
devoted to Pteranodon, and a general overview of this work will now be given.

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

427

The problem initially faced was to determine the upper size limit of Pteranodon , because until
1975 it was believed to be the largest flying creature of all time. Most studies drew material from
the worn and broken bone ends of the Pteranodon- like Ornithocheirus in the Cambridge University
collection, or from the more complete but thoroughly crushed material of Pteranodon itself in the
Yale University collections (described by Eaton 1910). A typical wing span used in the modern
aerodynamic work is about 7 m (Heptonstall 1971; Bramwell 1971; Bramwell and Whitfield 1974;
Brower 1983; etc.), though Eaton described a partial radius and ulna that, if projected isometrically,
would have yielded a wing span of 816 m (Heptonstall 1971 ). My study of the Yale collections and
Eaton’s work reveal that the largest size for which complete wings exist is about 5 m. Even at this
size, measurements from many incomplete specimens must be pooled in order to arrive at a mean
figure. The exact size of the largest Pteranodon , however, matters little because in 1975 Lawson
described remains of a pterodactyloid he later named Quetzalcoatlus northropi , which had a wing-
span initially projected at 15-5 m. This estimate proved to be too high, and a figure of about 12 m
(35-40 ft), based on additional material and further proportional comparisons, is now generally
accepted (Langston 1981). Numerous remains of apparent juveniles of this species, almost
exactly half the length of the larger form in all dimensions, have a wing span of approximately
5.8 m (19 ft).

An up-to-date review of aerodynamic assessments of Pteranodon and its relative Nyctosaurus was
given by Brower (1983), which obviates long discussion here; I will summarize only the major
outlines and conclusions of other authors (Table 1), and point out possible directions for future
research. The most influential work, of course, is the classic monograph by Bramwell and Whitfield
(1974) on the biomechanics of Pteranodon , which wedded the early functional-morphologic work
of Hankin and Watson (1914) to modern concepts of aerodynamics in a beautifully written and
lucid paper. According to their findings, echoed by Heptonstall (1971), Stein (1975), Sneyd et al.
(1982), Brower (1983), and others, Pteranodon was a superb low-speed soaring animal that had
difficulty flying in high winds and landing, but had a low sinking speed, an excellent lift/drag pro-
file, a light wing loading, low turning radius, high manoeuvrability, and optimal performance at
7 10 m/s. It was presumed to spend most of its time gliding at sea, trapping fish at the surface
in its great beak. However, its existence must have been marginal, because it was so large that

table 1. Calculated aerodynamic performance of the cretaceous pterosaur pteranodon

Author

W

Weight

(kg)

S

Wing area
(m 2 )

Wing

Loading

(kg/m 2 )

B

Wingspan

(m)

X

Mean

Chord

(m)

AR

Aspect

ratio

V

Speed

(m/s)

von Kripp 1943*

30

3-5

8-5

7

1

14

15

15

2-25

6-6

6

0-75

16

13-27

Heptonstall 1971

22-7

3-44

6-6

6-8

[0-51]

13

7

Bramwell 1971
Bramwell and Whitfield

18

5-8

3-1

8-2

[0 70]

11-7

6-7

1974t

16-6

4-62

3-6

6-95

[0 66]

10 5

7-7-8-0

Brower 1 983$

14-94

2-53

5-9

6-95

[0-36]

19-1

9-5

Values I calculated from data in other works are in brackets.

* Two configurations were given, based on different aerodynamic performance models.

t Bramwell and Whitfield (1970) considered three weight estimates (11-36, 18, and 25 kg) and used 18; in 1974 they
considered 12-8, 16-6, and 23-8, accepting the intermediate value. This is very close to Brower’s (1983) estimate, and
reflects a range comparable to those of large modern birds.

| Brower calculated that Pteranodon could fly as fast as 17 m/s under certain conditions.

428

PALAEONTOLOGY, VOLUME 28

it was only barely capable of level flight; how it managed to catch fish, recover from the weight
of the prey, and overcome the sudden strain on the neck to rise above the water’s surface was
not clear.

These studies have been very constructive in their efforts to determine parameters of flight mode,
and they continue to be productive. It is worth noting, however, that most recent work on the
palaeobiology of pterosaurs post-dated the aerodynamic work, and the former has important
implications for the latter. Far from suggesting that aerodynamic approaches should be abandoned,
I would like to provoke further discussion and investigation into the basis of understanding the
aerodynamics of extinct vertebrates by raising the following questions and alternate interpretations
to previous work. For some of these interpretations there is good evidence; others require only a
change of attitude or modification of assumptions which are, I believe, as plausible as those of
other authors.

Many palaeobiological problems, usually overlooked, relate to the structure, function, and ecol-
ogy of Pteranodon , which was phylogenetically only a bizarre sideline of 140 million years of
pterosaur evolution. Its smaller forerunners were active fliers; and Pteranodon, though neither the
last nor the largest of pterosaurs, and not a sustained flapper, retained both limited flapping ability
and full bipedal terrestrial locomotion. Any trade off of flapping ability for increased size in this
lineage must have been conditional on great advantages to that way of life, and it is unlikely that
millions of years of biological and aerodynamic fine-tuning would have been sacrificed in the
process. Analyses of Pteranodon s flight have so far ignored this legacy and have overlooked or
misinterpreted many important morphological factors. For example, Bramwell and Whitfield (1974)
described a ‘locking mechanism’ in the shoulder joint of Pteranodon that, they argued, would have
enabled the wing to be fixed in gliding position without expending much energy. This idea has been
picked up by many later authors. The form of the glenoid facet, however, is not especially modified
as they claim; it merely reflects the suture of the scapula and coracoid, which are fused in most
adult diapsids. The corresponding ‘ridge’ they identify on the head of the humerus is not present
on the articular surface but at its margin; this ridge is formed where the surficial laminar bone of
the shaft gives way to the porous epiphyseal surface which was covered by cartilage, as was the
glenoid. Therefore the ‘locking joint’ is very questionable. Also, the glenoid fossa faces postero-
laterally, not anterolaterally as their Figure 22 shows. These factors greatly influence inter-
pretation of the articulation and movement of the wing, which in reconstructions is almost always
swept too far forward.

The structure of the wing membrane is also very important. Most aerodynamic treatments have
overlooked palaeobiological considerations of the fine, stiff, intercalated ‘fibres’ that permeate the
wing membrane (Zittel 1882; Wellnhofer 1975; Padian 1979, 1980, 19836). Brower (1983) notes that
they are ‘approximately parallel to the wingspan [and to] the major direction of tensioning of the
membrane’, which is incorrect. These strong, rodlike ‘fibres’ of keratin or perhaps collagen are
never found bent or folded, and their orientation parallels that of the feather shafts of birds and
the fingers of bats (text-figs. 2, 3c). Hence they are precisely perpendicular to the direction of
spanwise tension at any point along the wing, not parallel to it. Brower, like most other analysts,
concluded that pterosaur wings, lacking internal structure, could not have been as manoeuvrable as
those of bats and could not have flapped at low speeds. But it is precisely the overlooked internal
structure of the wing that suggests the opposite: because the spanwise tension would have been
resisted by the network of ‘fibres’, the wing could have been collapsed much more than in bats
without loss of aerodynamic competence. Ability to draw in the wing (that is, to reduce its surface
area, thereby increasing wing loading) without losing the aerodynamic competence has been denied
by most authors, and so most aspects of flight performance observed in modern soaring birds have
not been applied to pterosaurs. The flight of Pteranodon at all speeds and under all conditions has
therefore been modelled as if the animal needed to keep its wings fully outstretched, which is
unrealistic in view of the animal’s biology.

The configuration of the pterosaur wing, as discussed earlier, was narrow and the aspect ratio
high, as in a gull or albatross (Zittel 1882). Most aerodynamic analyses of the flight of Pteranodon

P ADI AN: FLIGHT IN EXTINCT VERTEBRATES

429

(Bramwell and Whitfield 1974; Stein 1975; McMasters 1975; Sneyd et al. 1982) have used a model
in which the comparatively broad wings were attached to the feet, as in bats, and most of these
authors have agreed that a strong tendon along the trailing edge of the wing, anchored to the feet,
would have been necessary to control the membrane (text-fig. 3c, left). In fact, preserved wings of
pterosaurs show inconlrovertibly both that the hindlimbs were free of the wing and that no trailing
wing tendon existed: further evidence for the structural integrity of the ‘fibres’. Without these
fibres, a strong trailing tendon would have been necessary, and in that case it would indeed, as
aerodynamicists have argued, been very difficult for Pteranodon to draw in the wing at all without
collapsing it. Furthermore, the extreme forward sweep of the wings commonly pictured (e.g.
Bramwell and Whitfield 1974; Brower 1983; Sneyd et al. 1982) was impossible, being based on
incorrect anatomical interpretations and the assumption of a wide wing. With a narrow wing and
correct articulations, the centre of lift is further back, and the wing profile slimmer and more
laterally directed (text-fig. 3c, right).

What are the aerodynamic consequences of these considerations? Only Brower (1983) has used a
narrow wing configuration in aerodynamic calculations, and his calculated wing area is 55% of that
used by Bramwell and Whitfield and others. (My own estimate is closer to 45%. ) The wing loading
is then effectively doubled, which has a significant effect on calculated flying and sinking speeds,
polar curves, turning radius, mass distribution, and flapping performance. These will be considered
in detail elsewhere, but it may be noted for instance that the wing chord and induced drag are
halved, and wing profile drag is probably no longer comparable to that of the Gottingen 417a
airplane, which lacks the large leading spar of the pterosaur’s wing and is very dissimilar to it in
aspect and cross-section.

Accepting that the wing did not lose its shape (i.e. its aerodynamic competence) when partially
drawn in, Pteranodon need no longer be considered only in fully extended position. The larger
wing loadings, higher flying speeds, and lower sinking speeds that result are characteristic of the
performances of modern soaring birds. For instance, Bramwell (1971) calculated that in a typical
thermal Pteranodon would gain half a mile in altitude in five minutes. But what if Pteranodon
wished to use the thermal to search for food without gaining altitude? One can now see that it had
only to flex the wings to achieve lower wing area, higher wing loading, greater airspeed, and even
lower sinking speed. Using Bramwell’s configurations and calculations, a thermal rising at 4T m/s
would have carried Pteranodon up at a rate of about 3-45 m/s. This value is also her calculated
sinking speed at a flying speed of 16-5 m/s. Therefore Pteranodon could have flown level in a
thermal at 16-5 m/s. This is more than twice the calculated ‘optimal’ flying speed in still air, and
the wings are fully outstretched. How would Pteranodon s performance improve if the ability to
flex the wings to control flight were considered?

In general, a palaeobiological view of Pteranodon s flight appears to give greater ranges for
most calculated flight variables: higher speeds, more manoeuvrability, and better take off and
landing performance. The typical calculated polar curve of Pteranodon (e.g. Bramwell and
Whitfield 1974; Brower 1983) is much more attenuated compared to those of birds and aircraft;
but of course, the latter curves were discerned by empirical observation, not by calculation.
What would happen if values for albatrosses and falcons were calculated based on the kinds of
data estimates used for Pteranodon , and then compared with empirical results? Until this is done
there is no way to judge the accuracy of approaches that have so far been taken to pterosaur
flight. Pterosaurs were not aircraft, and their wings were in no way comparable to those of
hang-gliders ( contra Brower 1983 and McMasters 1975) or sailplanes ( contra Bramwell and
Whitfield 1974); nor were their wing skeletons inflexible spars with the membranes under con-
siderable spanwise tension ( contra these authors and Sneyd et al. 1 982). Their wings were comparable
in biological and aerodynamic respects to the wings of birds and bats, with a design inherited
from their ancestors, shaped by natural selection, and fine-tuned by evolutionary constraints and
opportunities. Until these factors are considered in engineering approaches, we shall probably lack
a realistic view of the flight performance of pterosaurs, and continue to view them as inferior
precursors to birds and bats.

430

PALAEONTOLOGY, VOLUME 28

CONCLUSION

Over the past two decades knowledge of animal flight has deepened considerably. Much more is
known of the mechanics of flight, and the ability to calculate flight energetics accurately (see
Pennycuick 1972) augurs well for the understanding of the evolution of flight and for the significance
of the differences among modern flyers in the physiology of flight. Preliminary results of the attempts
to fuse engineering with palaeontology to arrive at realistic appraisals of the flight performances of
extinct taxa have been pioneering in their approaches. Yet most of this work has yet to take
advantage of recent palaeobiological advances in the phylogenetic and functional understanding of
these animals. If, as current research indicates, birds evolved from small theropod dinosaurs and
developed flight from the ground up, the utility of studying gliding performance in Archaeopteryx
is uncertain; the earliest birds may seldom, if ever, have glided. The main problem is still getting up
in the air and staying up, which requires a realistic analysis of the evolution of flapping. In
pterosaurs the efficiency of the wings as flying organs seems to have been underestimated, because
considerations of the aerodynamics and functional morphology of the large soaring form Pter-
anodon have neglected the functional and phylogenetic evolution of pterosaurs. Bats, unfortunately,
remain largely shrouded in mystery with respect to the means by which they evolved flight; it can
be hoped that as their phylogenetic relationship with other mammals is clarified, the characteristics
clearly distinguishing bats from these groups may shed light on the evolution of flight in these most
unusual of flying vertebrates.

Acknowledgements. I am most grateful to G. R. Caple, W. A. Clemens, D. Yalden, J. A. Gauthier, E. J.
Laitone, S. McLeod, M. J. Novacek, J. H. Ostrom, C. Pennycuick, J. M. V. Rayner, M. K. Smith,
P. Wellnhofer, and R. Wild for helpful discussions of their work and comments on my own. Their concordance
with my views is, of course, not implied, and any errors or misinterpretations are purely mine.

REFERENCES

altenbach, j. s. 1979. Locomotor morphology of the vampire bat, Desmodus rotundas. Amer. Soc. Mamm.
Special Publ. 6, I 137.

barker, r. t. 1980. Dinosaur heresy- dinosaur renaissance: why we need endothermic archosaurs for a
comprehensive theory of bioenergetic evolution. In thomas, r. d. k. and olson, e. c. (eds.). A cold look at
the warm-blooded dinosaurs , 351-462. AAAS Selected Symposium Series 28, Westview Press, Boulder,
Colorado, 514 pp.

— and galton, p. m. 1974. Dinosaur monophyly and a new class of vertebrates. Nature , Lond. 248, 168-
172.

bock, w. j. 1965. The role of adaptive mechanisms in the origin of higher levels of organization. Syst. Zool.
14, 272-287.

1983. On extended wings. The Sciences , 23 (2), 16-20.

-and miller, a. h. 1959. The scansorial foot of the woodpeckers, with comments on the evolution of
perching and climbing feet in birds. Amer. Mas. Novit. 1931, 1-45.
bramwell, c. d. 1970. Flying speed of the largest aerial vertebrate. Nature , Lond. 225, 660.

— 1971. Aerodynamics of Pteranodon. Biol. J. Linn. Soc. 3 (4), 313-328.

and whitfield, G. R. 1974. Biomechanics of Pteranodon. Phil. Trans. R. Soc. Lond. B 267, 503-581.
brodkorb, p. 1963. Birds from the Upper Cretaceous of Wyoming. Proc. 13th Int. Orn. Congr. 1962, 1, 55-
70.

brower, j. c. 1983. The aerodynamics of Pteranodon and Nyctosaurus, two large pterosaurs from the Upper
Cretaceous of Kansas. Jour. Vert. Paleo. 3 (2), 84 124.
bryant, m. d. 1945. Phylogeny of Nearctic Sciuridae. Amer. Midi. Nat. 33 (2), 257-390.
caple, G. R., balda, r. t. and willis, w. r. 1983. The physics of leaping animals and the evolution of pre-flight.
Am. Nat. 121, 455-467.

Clark, b. d. 1977. Energetics of hovering flight and the origin of bats. In hecht, m. k., el at. (eds.). Major
Patterns of Vertebrate Evolution , 423-425. Plenum Press, New York.

PADIAN: FLIGHT IN EXTINCT VERTEBRATES

431

Colbert, E. H. 1966. A gliding reptile from the Triassic of New Jersey. Amer. Mus. Novit. 2246, 1 -23.

— 1967. Adaptations for gliding in the lizard Draco. Ibid. 2283, 1-20.

— 1970. The Triassic gliding reptile Icarosaurus. Bull. Amer. Mus. Nat. Hist. 143 (2), 87 142.

— 1980. The evolution of vertebrates (3rd edn.). John Wiley and Sons, New York.

and russell, d. a. 1969. The small Cretaceous dinosaur Dromaeosaurus. Amer. Mus. Novit. 2380,
1-49.

COOMBS, w. p. 1978. Theoretical aspects of cursorial adaptations in dinosaurs. Quart. Rev. Biol. 53,
393-418.

CRACRAFT, J. 1977. John Ostrom’s studies on Archaeopteryx , the origin of birds, and the evolution of flight.
Wilson Bull. 39, 488-492.

DESMOND, A. J. 1975. The hot-blooded dinosaurs , 238 pp. Blond and Briggs, London.

Eaton, G. F. 1910. Osteology of Pteranodon. Mem. Conn. Acad. Arts Sci. 2, 1-38.
evans, s. e. 1982. The gliding reptiles of the Upper Permian. Zool. Jour. Linn. Soc. 76, 97-123.
feduccia, a. 1973. Evolutionary trends in the Neotropical ovenbirds and woodhewers. Orn. Monogr. 13,
1-69.

and tordoff, H. B. 1979. Feathers of Archaeopteryx', asymmetric vanes indicate aerodynamic function.
Science, 203, 1021-1022.

Gauthier, j. a. 1984. A cladistic analysis of the higher systematic categories of the Diapsida. Ph D. thesis,
Department of Paleontology, University of California.

and padian, k. (in press). Phylogenetic, functional, and aerodynamic analyses of the origin of birds and
their flight. In hecht, m. k., ostrom, j. h., viohl, g. and wellnhofer, p. (eds.). The Beginnings of Birds.
Jura Museum, Eichstatt.

hankin, E. H. and watson, d. m. s. 1914. On the flight of pterodactyls. Aeronaut. Jour. 18, 324 335.

Harrison, c. J. o. 1976. Feathering and flight evolution in Archaeopteryx. Nature Lond. 263, 762-763.
heptonstall, w. b. 1970. Quantitative assessment of the flight of Archaeopteryx. Ibid. 228, 185-186.

1971. An analysis of the flight of the Cretaceous pterodactyl Pteranodon. Scott. J. Geol. 7(1), 61-78.
Hildebrand, M. 1982. Analysis of Vertebrate Structure (2nd edn ). John Wiley and Sons, New York.
holst, E. von. 1957. Der Saurierflug. Palaeont. Z. 31, 15-22.

hopson, j. a. 1980. Relative brain size in dinosaurs: implications for dinosaurian endothermy. In thomas, r.
d. k. and olson, e. c. (eds.). A cold took at the warm-blooded dinosaurs, 287-310. AAAS Selected Symposium
Series 28, Westview Press, Boulder, Colorado.

huene, f. von. 1914. Beitrage zur Geschichte der Archosaurier. Geol. Pal. Abh. N.F. 13, 1 -53.
jensen, J. A. 1981. Another look at Archaeopteryx as the world’s oldest bird. Encyclia, 58, 109 128.
jepsen, g. l. 1970. Bat origins and evolution. In wimsatt, w. a. (ed. ). Biology of Bats, Vol. I, 1-64. Academic
Press, New York.

jerison, h. j. 1968. Brain evolution in Archaeopteryx. Nature, Lond. 219, 1381.

kielan-jaworowska, z. 1975. Late Cretaceous mammals and dinosaurs from the Gobi Desert. Amer. Sci. 63
(2), 150-159.

kripp, F. von. 1943. Ein Lebensbild von Pteranodon ingens auf flugtechnischer Grundlage. Nova Acta Leap.
N.F. 12 (83), 217-246.

langston, w., jr. 1981. Pterosaurs. Sci. Am. 244, 122-136.

lawson, d. a. 1975. Pterosaur from the Latest Cretaceous of West Texas: discovery of the largest flying
creature. Science , 187, 947-948.

lewin, r. 1983. How did vertebrates take to the air? Science, 281, 38-39.
lull, r. s. 1906. Volant adaptation in vertebrates. Am. Nat. 40 (476), 537-566.

mcmasters, J. h. 1975. An analytic survey of low-speed flying devices— natural and manmade. Technical
Soaring, 3 (4), 17-39.

marsh, o. c. 1880. Odontornithes: a monograph of the extinct toothed birds of North America. Rept. U.S.
Geol. Explor. 40th Parallel , 7, I 201.

martin, l. d. 1983. The origin of birds and of avian flight. Current Ornith. 1, 105-129.

— Stewart, j. d. and whetstone, k. n. 1980. The origin of birds: structure of the tarsus and teeth. The
Auk, 97, 86-93.

nopcsa, f. von. 1907. Ideas on the origin of flight. Proc. Zool. Soc. Lond. 1907, 223-236.

1923. On the origin of flight in birds. Ibid. 1923, 463-477.
nov acek, m. j. 1982. Information for molecular studies from anatomical and fossil evidence on higher
Eutherian phylogeny. In Goodman, m. (ed.). Macromolecular Sequences in Systematic and Evolutionary
Biology, 3-42. Plenum Press, New York.

432

PALAEONTOLOGY, VOLUME 28

ostrom, j. h. 1969. Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous
of Montana. Bull. Peabody Mas. Nat. Hist. Yale , 30, 1 1 65.

— 1973. The ancestry of birds. Nature , Load. 242 , 136.

— 1974. Archaeopteryx and the origin of flight. Quart. Rev. Biol. 49, 27-47.

1975a. The origin of birds. Ann. Rev. Earth Planet. Sci. 3, 55-77.

— 19756. On the origin of Archaeopteryx and the ancestry of birds. Colt. Int. CNRS , 218 , 519-532.

1976u. Some hypothetical anatomical stages in the evolution of avian flight. In olson, s. l. (ed.). Collected
Papers in Avian Paleontology. Smiths. Misc. Coll. 27 , 1-21.

19766. Archaeopteryx and the origin of birds. Biol. Jour. Linn. Soc. 8, 91-182.

1979. Bird flight: how did it begin? Amer. Sci. 67 (1), 46-56.

1980. The evidence for endothermy in dinosaurs. In thomas, r. d. k. and olson, e. c. (eds. ). A cold look
at the warm-blooded dinosaurs, 15-60. A A AS Symposium Series 28, Westview Press, Boulder, Colorado.

radian, K. 1979. The wings of pterosaurs: a new look. Discovery ( New Haven), 14 , 20-29.

1980. Studies on the structure , evolution , and flight of pterosaurs. Ph.D. thesis, Yale University. 309 pp.

— 1982. Macroevolution and the origin of major adaptations: vertebrate flight as a paradigm for the
analysis of patterns. Proc. 3rd N. Amer. Paleont. Conv. 2 , 387-392.

— 1 983c/. Description and reconstruction of new material of Dimorphodon macronyx (Buckland) (Ptero-
sauria: Rhamphorhynchoidea) in the Yale Peabody Museum. Postilla, 189, I 44.

— 19836. A functional analysis of flying and walking in pterosaurs. Paleobiology , 9 (3), 218-239.

— 1984. The origin of pterosaurs. In reif, w.-e. and westphal, f. (eds.). Proceedings , Third Symposium on
Terrestrial Mesozoic Ecosystems , 163-168. ATTEMPTO, Tubingen.

— and olsen, R. e. 1984. The fossil trackway Pteraichnus : not pterosaurian, but crocodilian. J. Paleont. 58 ,
178-184.

pennycuick, c. J. 1969. The mechanics of bird migration. Ibis, 111 , 525-556.

— 1972. Animat Flight , 68 pp. Edward Arnold, London.

pirlot, p. 1977. Wing design and the origin of bats. In hecht, m. k., et al. (eds.). Major Patterns of Vertebrate
Evolution , 375-410. Plenum Press, New York.

rayner, I. m. v. 1979. A new approach to animal flight mechanics. Jour. Exp. Biol. 80, 17-54.

1981. Flight adaptations in vertebrates. In day, m. h. (ed.). Vertebrate Locomotion , 137-172. Academic
Press, New York.

regal, p. J. 1975. The evolutionary origin of feathers. Quart. Rev. Biol. 50 , 35-66.

robinson, p. L. 1962. Gliding lizards from the Upper Permian of Great Britain. Proc. Geol. Soc. Lond. 1601 ,
137-146.

RUSSELL, D. A. 1980. Reflections of the dinosaurian world. In Jacobs, l. l. (ed.). Aspects of Vertebrate History ,
257-268. Museum of Northern Arizona Press, Flagstaff.
savile, d. b. o. 1962. Gliding and flight in the vertebrates. Amer. Zoo!. 2 , 161-166.

seeley, h. G. 1870. Remarks on Professor Owen's monograph on Dimorphodon. Ann. Mag. Nat. Hist. (4) 6,
129-152.

smith, j. d. 1977. Comments on flight and the evolution of bats. In hecht, m. k., et at. (eds.). Major Patterns
in Vertebrate Evolution, 427-437. Plenum Press, New York.
sneyd, a. d., bundock, m. s. and reid, d. 1982. Possible effects of wing flexibility on the aerodynamics of
Pteranodon. Am. Nat. 120 , 455-477.

stein, r. s. 1975. Dynamic analysis of Pteranodon ingens: a reptilian adaptation to flight. J. Paleont. 49 , 534-
548.

tarsitano, s. and hecht, m. k. 1980. A reconsideration of the reptilian relationships of Archaeopteryx. Zool.
Jour. Linn. Soc. 69 , 149-182.

taylor, c. R. 1973. Energy costs of animal locomotion. In botis, l., et ah (eds.). Comparative Physiology, 24
42. North Holland, Amsterdam.

thomas, R. d. K. and olson, e. c. (eds.). 1980. A cold look at the warm-blooded dinosaurs , 514 pp. AAAS
Selected Symposium Series 28. Westview Press, Boulder, Colorado.
thorington, r. w., jr. and heaney, l. r. 1981. Body proportions and gliding adaptations of flying squirrels
(Petauristinae). J. Mamm. 62 , 101-1 14.

thulborn, r. a. and hanley, T. l. 1982. The reptilian relationships of Archaeopteryx. Aust. J. Zool. 30, 611-
634.

walker, a. d. 1973. New light on the origin of birds and crocodiles. Nature Lond. 237 , 257-263.

1977. Evolution of the pelvis in birds and dinosaurs. In Andrews, s. m., et al. (eds.). Problems in Vertebrate
Evolution , 319-358. Academic Press, London.

PAD IAN: FLIGHT IN EXTINCT VERTEBRATES

433

wellnhofer, p. 1970. Die Pterodactyloidea (Pterosauria) der Oberjura Plattenkalke Suddeutschlands. Abb.
Bayer. Akad. Wiss. N.F. 141 , 1 133.

1974. Campylognathoides liasicus (Quenstedt), an Upper Liassic pterosaur from Hozmaden. Arm. Car-
negie Mus. 45 (2), 5-34.

1975. Die Rhamphorhynchoidea der Oberjura Plattenkalke Suddeutschlands. Palaeontographica A 148 ,
1-33, 132-186; A 149 , 1-30.

- 1978. Pterosauria. Handbuch der Palaeolierpetologie , Ted 19. Stuttgart: Gustav Fischer Verlag.
wild, R. 1978. Die Flugsaurier (Reptilia, Pterosauria) aus der Oberen Trias von Cene bei Bergamo, Italien.
Boll. Soc. Pal. Ital. 17 (2), 176-256.

williston, s. w. 1879. ‘Are birds derived from dinosaurs?' Kansas City Rev. Sci. 3 , 457 460.
yalden, D. w. 1971 £7. Flying ability of Archaeopteryx. Nature , Lond. 231 , 127.

19716. The flying ability of Archaeopteryx. Ibis , 1 13 , 349-356.
zittel, k. a. von. 1882. Ueber Flugsaurier aus dem lithographischen Schiefer Bayerns. Palaeontogr. 29 , 47

80.

Typescript received 28 June 1984

Revised typescript received 13 November 1984

KEVIN PADIAN

Department of Paleontology
University of California
Berkeley
California 94720
U.S.A.

THE MICROSTRUCTURE OF TOOTH ENAMEL
IN MULTITUBERCULATE MAMMALS

by G. FOSSE, Z. KIELAN-JAWOROWSKA and S. G. SKAALE

Abstract. The enamel microstructure of single teeth and teeth in situ in whole jaws of late Jurassic, late
Cretaceous, and Palaeocene multituberculates belonging to the Plagiaulacoidea, Taeniolabidoidea, Ptilo-
dontoidea, and Meniscoessus (Cimolomyidae, suborder indet.) is examined by incident light microscopy,
scanning electron microscopy (SEM), and polarized light microscopy. For comparison one docodont tooth
and some single late Cretaceous and Recent eutherian teeth are included. The enamel of the Plagiaulacoidea
like that of the docodont tooth is not prismatic, but consists of radially arranged, closely packed 5 pm thick
columns of crystals, which diverge from the central axis of each column towards the outer enamel surface.
The Asian as well as the North American Taeniolabidoidea have gigantoprismatic enamel, the numerical
density of prisms per unit area being four to five times lower than in the Ptilodontoidea and Eutheria. In most
taeniolabidoid jaws the prism density is somewhat higher in the molars than in the incisors. The oldest
gigantoprismatic enamel was found in some undescribed multituberculate teeth from the early Cretaceous of
Asia. As Meniscoessus (suborder indet.) has gigantoprismatic enamel, it is suggested that this feature may be
useful in establishing the taxonomic position of some multituberculate groups.

This paper examines the tooth enamel of multituberculate mammals from the upper Jurassic,
Kimmeridgian to the late Palaeocene, in order to discover differences in enamel microstructure
among the suborders Plagiaulacoidea, Ptilodontoidea, and Taeniolabidoidea, to elucidate their
relationships.

Moss (1969) found that multituberculate enamel was a non-prismatic, continuous structure, and
that it contained tubules which followed a zigzag course. Fosse et al. (1973) confirmed the presence
of these zigzag tubules but challenged Moss’ view on the non-prismatic enamel structure on the
basis of six unidentified multituberculate teeth from the late Cretaceous in which the enamel was
discontinuous and prismatic, the prisms being extremely large in comparison with those of other
mammals. Subsequently Fosse et al. (1978) investigated the teeth of four identified multituberculate
species, Catopsalis joyneri and Stygimys kuszmauli (Taeniolabidoidea) and Mesodma thompsoni and
M. formosa (Ptilodontoidea). The number of prisms per mm 2 ranged from 3650 to 5860 in the
members of the Taeniolabidoidea and from 26 600 to 27 200 in the Ptilodontoidea. The
taeniolabidoid prism density was the lowest observed in any group of mammals.

Sahni (1979) studied the enamel microstructure in several late Cretaceous North American
Ptilodontoidea and Taeniolabidoidea, as well as Eutheria (from the Hell Creek Formation of
Montana, Sloan and Van Valen (1965)) and found large prisms in all the multituberculate genera.
According to Sahni the prism density per mm 2 in Mesodma and Meniscoessus (suborder indet. -
see Hahn and Hahn 1983) was 9000 and 7400 respectively, while in Catopsalis and Stygimys it was
8700 and 4500. The data for Mesodma differ considerably from those obtained by Fosse et al.
(1978). Using the scales given on the Mesodma micrographs in Sahni’s paper we calculated a mean
prism density of 21 400 per mm 2 .

In view of the differences between the results obtained by Fosse et al. (1978), by Sahni (1979,
Table 1), and by us on Sahni's micrographs, concerning Mesodma, we decided to examine once
more isolated teeth of Mesodma sp. and to compare them with a molar of Meniscoessus sp.
(Cimolomyidae, suborder indet.), all from the late Cretaceous of North America, and with various
taeniolabidoid teeth from the late Cretaceous of Asia (Kielan-Jaworowska 1970, 1974a), the Late
Palaeocene of Asia (Matthew and Granger 1925), and the late Cretaceous of North America (Sloan

IPalaeontology, Vol. 28, Part 3, 1985, pp. 435-449, pis. 48-50.|

436

PALAEONTOLOGY, VOLUME 28

and Van Valen 1965). All the North American late Cretaceous material comes from the Hell Creek
Formation of Montana (Clemens et al. 1979). The late Cretaceous Asian material which forms the
bulk of the material comes from the Djadokhta and Barun Goyot formations, or the stratigraphic
equivalent of the latter: the red beds of Khermeen Tsav. We tentatively accept, after Gradzinski et
al. ( 1977) that the Djadokhta Formation belongs to the upper Santonian and/or lower Campanian
Stage, while the Barun Goyot Formation (and the red beds of Khermeen Tsav) belong to the middle
Campanian Stage.

We also examined the enamel microstructure of isolated teeth from the early Cretaceous of Asia,
which are at present being studied by Kielan-Jaworowska, Dashzeveg, and Trofimov. Some of these
teeth (genus Arginbaatar) were assigned to the Taeniolabididae by Trofimov (1980), while Hahn
and Hahn (1983) erected the family Arginbaataridae within the Plagiaulacoidea. Consequently at
present we assign all the early Cretaceous Mongolian multituberculates to a suborder indet.

The earliest multituberculate teeth examined by us belong to the suborder Plagiaulacoidea and
come from the late Jurassic, Kimmeridgian, of Portugal (Hahn 1969, 1971, 1978). From the same
location we also included an unidentified docodont tooth for comparison, as well as teeth of late
Cretaceous and Recent eutherian mammals.

ABBREVIATIONS

GI Institute of Geology, Academy of Sciences of the Mongolian People’s Republic, Ulan Bator.
IAUB Institute of Anatomy, University of Bergen, Bergen.

PIFU Palaeontologishes Institut, Freie Universitat, Berlin.

UM University of Minnesota, Minneapolis, U.S.A.

ZPAL Institute of Palaeobiology, Polish Academy of Sciences, Warsaw.

I incisor dp deciduous premolar P permanent premolar M molar.

MATERIAL

Docodonta

Docodontidae

gen. et sp. indet., Kimmeridgian, Portugal, Leiria, Guimarota: PIFU no number (a molar)
Multituberculata
Plagiaulacoidea

Family, gen. et sp. indet., Kimmeridgian, Portugal Leiria, Guimarota: PIFU no number (a molar)
Paulchoffatidae (all from the Kimmeridgian of Portugal, Leiria, Guimarota):

Paulchoffatia sp.: PIFU VJ 270-155 (dp 1 ); PIFU VJ 272-155 (P 1 ); PIFU VJ 273-155 (P 1 )

Kuehneodon sp.: PIFU VJ 303-155 (P s ); PIFU VJ 308-155 (P 5 )

Suborder indet. (all from the ?Aptian or Albian Guchin beds of Mongolia, Guchin Us):

Arginbaataridae

Arginbaatar dimitrievae Trofimov: GI PST 10/11 (P 4 ); GI PST 1 0/ 1 3(P 4 )

Family, gen. et sp. indet.: GI PST 10/29 (I 1 ); GI PST 10/23 (P 4 )

Taeniolabidoidea

Eucosmodontidae

( Chulsanbaatar , Nemegtbaatar , and Kryptobaatar are from the upper Cretaceous of Asia, Gobi Desert,
Stygimys from the upper Cretaceous, Hell Creek Formation, North America, Montana, Bug Creek):
Chulsanbaatar vulgaris Kielan-Jaworowska, Barun Goyot Formation, Khulsan: ZPAL MgM-I/62 (U,
M,); ZPAL MgM-I/157 (I,, P 4 , Md; Barun Goyot Formation, Nemegt: ZPAL MgM-I/111 (U);
red beds of Khermeen Tsav, Khermeen Tsav II: ZPAL MgM-I/108 (U); ZPAL MgM-I/109 (U, P 4 )
Nemegtbaatar gobiensis Kielan-Jaworowska, red beds of Khermen Tsav, Khermen Tsav II: ZPAL
MgM-I/81 (U, M,); ZPAL MgM-I/82 (U, P 4 , MU

Kryptobaatar dashzevegi Kielan-Jaworowska, Djadokhta Formation, Bayn Dzak: ZPAL MgM-1/7
(P 4 ); ZPAL MgM-I/9) (P 4 , MU; ZPAL MgM-I/37 (I,, MU; ZPAL MgM-I/53 (U, P 4 , M,)
Stygimys kuszmauli Sloan and Van Valen, Hell Creek Formation, Bug Creek: UM no. 5 (U, M 2 )

FOSSE ET AL.: M U LTIT U BE RCLI L ATE ENAMEL

437

Taeniolabididae (all from Asia, Gobi Desert):

Kamptobaatar and Catopsalis are from the upper Cretaceous, Prionessus from the Upper Palaeocene
Kamptobaatar kuczynskii Kielan-Jaworowska, Djadokhta Formation, Bayn Dzak: ZPAL MgM-I/38

(P 4 )

Catopsalis catopsaloides (Kielan-Jaworowska), red beds of Khermeen Tsav, Khermeen Tsav II: ZPAL
MgM-I/78 (L, M t ); ZPAL MgM-1/80 (MU
Prionessus lucifer Matthew and Granger, Naran Bulak: ZPAL MgM-II/67 (L, M,, M 2 )

Ptilodontoidea

Neoplagiaulacidae (all from the upper Cretaceous Hell Creek Formation, North America, Montana, Bug
Creek):

Mesodma thompsoni Clemens: UM no. 3 (P 4 )

Mesodma sp.: ZPAL MK-I/7 (P 4 ); ZPAL MK-I/8 (P 4 ); ZPAL MK-I/3 (M 1 ); ZPAL MK-1/6 (M 1 )
Suborder indet.

Cimolomyidae, Bug Creek, Montana, Hell Creek formation:

Meniscoessus sp.: ZPAL MK-I/9 ( M 0
Eutheria
Proteutheria

Kennalestidae, Djadokhta Formation, Gobi Desert, Bayn Dzak:

Kennalestes gobiensis Kielan-Jaworowska: ZPAL MgM-I/3 (P 3 )

Rodentia

Muridae, Recent, Europe:

Rattus norvegicus (Berkenhout): IAUB no number (L, MU
Primates

Hominidae, Recent, Europe:

Homo sapiens L.: IAUB no number (P 2 ).

METHODS

The Asian multituberculate material investigated in Warsaw consisted of whole mandibles with teeth in situ.
The right or left mandible was positioned in plasticine on a microscope slide under a dissection microscope in
such a way that a selected region of enamel on the tooth to be studied was the highest point of the whole
dentition. This region was then carefully planed horizontally by hand using 0/2 and then 0/4 grit emery paper
(Buehler Ltd., Evanston, 111., U.S.A.), each grade having been glued and trimmed to either of the two long,
narrow sides of a 10x2x03 cm rectangular, planed wood stick. The horizontal, tiny but relatively ilat enamel
surface, less than 1 mm in diameter formed in this way was then etched with a very small amount of 0-37 N
HN0 3 applied by a line-pointed brush. The etching was interupted after 5 sec with plain water, using a similar
brush. Next, to micrograph the etched surface, the microscope slide with the specimen still in the original
position on it was transferred to a Leitz Laborlux microscope equipped with camera, an Ultropak incident
light condensor and a U-O-l 1 objective. A Leitz microscale with 10 pm divisions was micrographed with the
same magnification.

At IAUB the films of the etched surfaces with cross-cut enamel prisms were copied on 23 x 30 cm film
sheets with a standard magnification.

The smallest unit that describes the number of cross-sectioned prisms per mm 2 (numerical prism density) is
a triangle consisting of central distances between three adjacent prisms. Determining the prism density in the
enamels consisted of measuring the distances between centres of adjacent prisms in several such triangular
units within each micrographed enamel area (Fosse, 1968a).

Table 1 presents prism densities (a), mean central distances between adjacent prisms (D), and the new
parameter A which signifies the theoretical mean cross-sectional area in pm 2 of the enamel producing end of
the ameloblasts (Fosse 1968 d\ Fosse et al. 1973, et al. 1978). The prism density values presented were calculated
from incident light micrographs of superficially planed and etched natural outer enamel surfaces which are
nearly planoparallel with an original layer of ameloblasts (Fosse et al. 1973). As it is still generally believed
that each prism rod is produced by one ameloblast (Fosse et al. 1978), the number of prisms per mm 2 in such
planes should reflect the number of original enamel producing ameloblasts per mm 2 in that plane, irrespective
of its angle with the prism rods underneath (Fosse 1968c).

Two multituberculate lower jaws, Chulsanbaatar vulgaris Kielan-Jaworowska, ZPAL MgM-I/62, and Kryp-
tobaatar dashzevegi Kielan-Jaworowska, ZPAL MgM-I/53, as well as docodont and plagiaulacoid single teeth

438

PALAEONTOLOGY, VOLUME 28

(gen. et sp. indet.), and five plagiaulacoid single teeth determined at generic level, all from the Kimmeridgian,
Leiria, Guimarota, Portugal, were sectioned by a wire string saw (Fosse et al. 1974) to obtain cut surfaces or
thin sections for transmitted light microscopy. For the SEM micrographs a Jeol T-200 instrument was used.
The specimens were covered by gold-palladium before SEM micrography. A Leitz Ortholux Pol microscope
was used for transmitted light micrographs of thin tooth sections.

RESULTS

In the material in Warsaw, clusters of prisms were usually quickly recognized in the planed and
etched enamel surfaces. Text-fig. 1 a represents P 4 of K. dashzevegi Kielan-Jaworowska, ZPAL
MgM-I/7. With the same magnification text-fig. 1 b, c, and d respectively show the prism densities
in Mesodma sp. (P 4 ), Stygimys kuszmauli (Ii), and Meniscoessus sp. (M 2 ) enamels, all three from
the late Cretaceous, Lancian (Maastrichtian), North America, Montana, Bug Creek. In incident
light planed and etched enamel surfaces of Kimmeridgian plagiaulacoid and docodont teeth showed
regularly packed structures in a pattern resembling cross-cut prism rods, (text-fig. \e, f). Their
numerical density and mean interproximate central distance were of a magnitude between those of
the eutherians: late Cretaceous Kennalestes gobiensis and Recent Rattus norvegicus , (text-fig. 1 g;
Table 1 ). Human enamel had considerably larger prisms than R. norvegicus. The prism density near
the cusp on the outer surface of a human premolar was about the same as that of Mesodma sp.,

text-fig. 1. Incident light micrographs of planed and etched surfaces of various enamels reproduced
with the same magnification, x 250. a represents the enamel surface of a tooth micrographed in situ ,
c and / represent the cut and etched enamels of sectioned teeth, all the others represent superficially
planed and etched outer enamel surfaces of teeth embedded in plastic blocks, a, Kryptobaatar
dashzevegi Kielan-Jaworowska, ZPAL MgM-I/7, P 4 , occlusal edge above, b, Mesodma sp., ZPAL
MK-I/8, P 4 , occlusal edge above, c. Stygimys kuszmauli Sloan and Van Valen, UM no. 5, T, showing
longitudinal section through medial enamel facet seen in the lingual direction, dentine below.
d , Meniscoessus sp.. ZPAL MK-I/9, M 2 , occlusal surface above, e, Paulchoffatia sp., PILU VJ 273-1 55,
P 1 , showing prism-like pattern of structures without distinct borders, occlusal surface to right.
/, docodont molar (gen. et sp. indet.), PILU no number, obliquely cut through one of the cusps,
showing prism-like pattern, montage, g. Rattus norvegicus (Berkenhout), IAUB no number, Mi,
occlusal surface above. It. Homo sapiens L., IAUB no number, P 2 , occlusal surface above.

FOSSE ET A L.\ MULTITUBERCULATE ENAMEL

439

table 1. The number of prisms per mm 2 (a), the mean central distance in microns between adjacent prisms
(D), and the theoretical mean cross-sectional area in /nn 2 of the enamel producing cells (A) (see Methods) in
the enamel of some specimens of European, Asian, and North American multituberculates from different
ages, of one undetermined docodont molar, one Cretaceous, and two recent eutherian species. The ZPAL
specimens were represented by whole jaws from which the enamel parameters of more than one tooth were

usually available, see Material.

Species/specimen

a

D

A

Docodont, gen. & sp. indet.

PIFU no number

34 514

5-78

28-9

M

Paulchoffatia sp.

PIFU VJ 273-155

43 080

5-17

23-2

P i

PIFU VJ 272-155

59 987

4-38

16 6

pi

Kuehneodon sp.

PIFU VJ 303-155

50 983

4-75

19-6

P 5

Arginbaatar dimitrievae Trofimov

GI PST 10/11

7123

12-73

140-3

P 4

GI PST 10/13

5426

14-58

184-2

P 4

Multituberculata subord. fam. gen. and sp. indet.

GI PST 10/29

4891

15-36

204-4

I 1

GI PST 10/23

11 365

10-07

87-9

P 4

Chulsanbaatar vulgaris Kielan-Jaworowska

ZPAF MgM-I/62

5812

14-09

172-0

I 1

10 001

10-74

99-9

M

ZPAF MgM-I/157

4960

15-25

201-5

I,

5734

14-18

174-3

P 4

11 623

9-96

86-0

M

ZPAL MgM-I/1 1 1

6519

13-30

153-4

I,

ZPAL MgM-I/108

7219

12-64

138-5

Ii

ZPAL MgM-I/109

6258

13-58

159-7

I.

9520

11-01

105-0

P 4

Nemegtbaatar gobiensis Kielan-Jaworowska

ZPAL MgM-I/81

5399

14-62

185-1

I,

12 133

9-75

82-4

M

ZPAL MgM-I/82

5435

14-57

183-9

I,

4241

16-49

235-7

P 4

8271

11-81

120-9

M

Kryptobaatar daszevegi Kielan-Jaworowska

ZPAL MgM-I/7

6314

13-52

158-3

P 4

ZPAL MgM-I/9

3379

18-48

295-9

I.

6349

13-48

157-4

P 4

ZPAL MgM-I/10

4464

16-08

223-9

Ii

5271

14-79

189-6

P 4

ZPAL MgM-I/21

3705

17-65

269-8

Ii

4908

15-33

203-7

P 4

3980

17-03

251-2

M,

ZPAL MgM-I/37

4292

16-40

232-9

I,

3753

17-53

2664

M,

ZPAL MgM-I/53

3415

18-38

292-7

I,

5812

14-09

172-0

P 4

6039

13-82

165-5

M,

Stygimys kuszmauli Sloan and Van Valen

UM no. 5

3860

17-29

259-0

I,

Kamptobaatar kuczynskii Kielan-Jaworowska

ZPAL MgM-I/38

6776

13-05

147-5

P 4

440

PALAEONTOLOGY, VOLUME 28

Table 1 ( cont .)

Species/specimen

a

D

A

Catopsalis catopsaloides (Kielan-Jaworowska)

ZPAL MgM-I/78

4740

1 5 60

210-9

Ii

6650

1317

150-3

Mj

ZPAL MgM-I/80

6063

13-79

164-9

M t

Prionessus lucifer Matthew and Granger

ZPAL MgM-II/67

6440

13-38

155-2

u

5776

14-13

173-1

M,

6138

13-71

162-9

M,

Mesodma sp.

ZPAL MK-I/7

28 627

6-35

34-9

P 4

ZPAL MK-I/8

26 321

6-62

37-9

P 4

ZPAL MK-I/3

26 694

6-58

37-5

M 1

ZPAL MK-I/6

21 557

7-32

46-4

M 1

Meniscoessus sp.

ZPAL MK-I/9

4088

16-80

244-5

m 2

Kennalestes gobiensis Kielan-Jaworowska

ZPAL MgM-I/3

31 599

6-04

31-6

P 3

Rat tits norvegicus (Berkenhout)

IAUB no number

65 703

4-19

15-2

Ii

67 095

4-14

14 9

M|

Homo sapiens L.

IAUB no number

25 335

6-75

39 5

P 2

(text-fig. 1 b, h: Table 1 ). In most ZPAL taeniolabidoid jaws with more than one tooth micrographed,
the prism density of the incisor was lower than that of P 4 or Mi, see Table 1. The higher density in
the molars apparently is caused not so much by smaller prisms as by less interprismatic enamel
(PI. 48, figs. 1 and 2). The great difference between the microstructure of taeniolabidoid and Mesodma
enamels is demonstrated in Plate 48, figs. 1-4, where it is also seen that the crystal structure of the
Kryptobaatar enamels was coarser than that of the Mesodma enamels. Near the outer surface of
human enamel there is very little interprismatic substance (PI. 48, fig. 5). The human prism diameters

EXPLANATION OF PLATE 48

Figs. 1-5. SEM micrographs of superficially planed and etched outer enamel surfaces reproduced with the
same magnification, x 2900, documented by the automatically recorded scale divisions of 10 /mi having
been retained in the micrographs. 1, Kryptobaatar dashzevegi Kielan-Jaworowska, ZPAL MgM-I/53, I t ,
showing that the prisms are widely separated by interprismatic enamel consisting of crystals being normal
to the surface and that the crystals of the prism cores are obliquely orientated relative to the surface and
inclined in an incisal direction above. 2, M, from the same dentition as I, in fig. 1, showing that the spatial
arrangement of the crystals in the prisms and interprismatic enamel is less distinct and that the prisms are
nearly of the same size but more closely packed than in L. 3, Mesodma sp., ZPAL MK-I/8, P 4 , showing
that the prisms are smaller and their numerical density per unit area considerably lower than in Kryptobaatar
enamel, also that the crystals are more delicate and densely packed. In the interprismatic enamel the crystals
are normal to the surface, while those of the prism cores are inclined in a cuspal direction to the left.
4, Mesodma sp., ZPAL MK-I/3, M 1 , not belonging to the same individual as P 4 in fig. 3, but showing similar
prism size, numerical density of prisms, and crystal orientation. Cuspal direction is to the left. In the upper
half are some openings of enamel tubules. 5, Homo sapiens L., IAUB no number, P 2 , showing that prisms
are nearly as large as in Kryptobaatar enamel (fig. 1), but that their numerical density approximates that of
the Mesodma enamels (figs. 3 and 4). Cuspal direction is to the right.

PLATE 48

FOSSE, KIELAN-JAWOROWSKA and SK.AALE, multituberculate enamel

442

PALAEONTOLOGY, VOLUME 28

are about as large as those of the Kryptobaatar (PI. 48, figs. 1 and 2) and Chulsanbaatar enamels
(PI. 49, fig. 3), whereas the distances between centres of adjacent prisms equal those of the Mesodma
enamels (PI. 48, figs. 3 and 4; Table 1). Thus there is no interdependence between prism diameters
and number of prisms per unit area.

Longitudinal and transverse sections of taeniolabidoid incisors showed cross-cut prisms; in longi-
tudinal sections when they passed through the medial enamel facets where the prism rods were
inclined in a dorsomedial (mesiolingual) direction in a transversal plane relative to the incisors, in
transversal sections in the ventrolateral facets where the prism rods were inclined in an anterior
(incisal) direction in a sagittal plane relative to the tooth. Regardless of the orientation of the enamel
surfaces represented by Plate 48, figs. 1-4 and Plate 49, figs. 1 and 3, they all demonstrate that the
crystals in the interprismatic enamel are orientated with their long axis nearly normal to the natural
outer enamel surface. The crystals of the prism cores are parallel with the prism rods, and the apices
of the arcades are pointing in the direction of the acute angle between prism rods and the dentine
enamel junctional surface.

In the SEM discrete enamel prisms in the Kimmeridgian enamels could not be discerned. Plate
49, figs. 2 and 5 show plagiaulacoid and docodont enamels at the same magnification. A certain
regular pattern in the crystal orientation may be observed. This pattern seemed to consist of 5 pm
thick, closely packed columns of crystals, the latter diverging from the central axis of each column
towards the external enamel surface. In Table 1 are given the values for three plagiaulacoid teeth
and one docodont molar from the Kimmeridgian, Portugal, based on measurements in incident
light micrographs.

Longitudinal sections, about 80 pm thick, were prepared from three plagiaulacoid teeth of which
two were determined on the generic level, and one docodont molar (gen. et sp. indet.) from the
Kimmeridgian. In the microscope one of the plagiaulacoid enamels (gen. et sp. indet.) showed large
black spots along lines that might correspond to the course of growth lines (striae of Retzius,
PI. 50, fig. 1). In polarized transmitted light with crossed polars and the dentine enamel junction at

EXPLANATION OF PLATE 49

Figs. 1-5. SEM micrographs of various multituberculate enamels reproduced with the same magnification,
x 2900, documented by the automatically recorded scale divisions of 10 pm having been retained along the
right margins of the micrographs. Figs. 1, 2, and 3 represent sectioned and etched enamel surfaces, figs. 4
and 5 superficially planed and etched outer enamel surfaces. I, Stygimys kuszmauli Sloan and Van Valen,
UM no. 5, L, enlargement of the same enamel surface as figured in text-fig. lc, but rotated 90°, dentine
enamel border at right. The prisms are large and widely separated by interprismatic enamel where the
crystals are orientated with their long axes in the figured plane, from left to right, e.g. perpendicularly to
the outer enamel surface, while the crystals of the prisms are normal to the figured surface. Towards the
dentine at right the borders of two prisms consist of an amorphous material. 2, Paulchojfatia sp., PIFU VJ
270-155, dp 1 , showing oblique section through cusp where the enamel consists of crystals without pre-
ferential orientations in prisms and interprismatic material. In some places it may be seen that the crystals
are arranged in fan-shaped clusters. Dentine in lower left corner. 3, Chulsanbaatar vulgaris Kielan-
Jaworowska, ZPAL MgM-I/62, I,, showing transversal section through ventrolateral enamel facet. The
prisms are somewhat smaller and their numerical density higher than in Kryptobaatar (PI. 48, fig. 1) and
Stygimys (fig. 1 ) enamels. The interprismatic enamel consists of crystals orientated with their long axes in
the figured plane from top to bottom, e.g. normal to the outer enamel surface while the crystals of the prism
cores are nearly normal to the figured plane. Arcade shaped grooves surround the prisms, the apices of
which point towards the dentine at top. 4, Meniscoessus sp., ZPAL MK-I/9, M 2 , showing arcade shaped
grooves surrounding the large prisms, the apices of which point in the cuspal direction at left. Crystals of
interprismatic enamel are generally normal to the figured surface, while the crystals of the prisms are inclined
to the left. 5, docodont molar (gen. et sp. indet.), PIFU no number. There is no organization of crystals in
prisms and interprismatic enamel, but the uneven appearance of the etched surface indicates the presence
of crystal clusters about 5 pm wide.

PLATE 49

FOSSE, KIELAN-JAWOROWSKA and SKAALE, multituberculate enamel

444

PALAEONTOLOGY, VOLUME 28

a small angle with the polarizer axis, poorly defined band-like 5-6 /an thick structures could be seen
in plagiaulacoid as well as docodont enamels, running radially from the inner to the outer surface
(PI. 50, figs. 2-5). Enamel tubules were very scarce. In polarized light the longitudinally cut Cluil-
sanbaatcir , Kryptobaatar , and Stygimys enamels showed discrete broad and straight bands of a
regular width, running at an angle of approximately 45° to the outer enamel surface. Enamel tubules
were abundant, coursing from the dentine enamel junction along the bands for short distances, but
mostly crossing them, running mainly in a radial direction (PI. 50, figs. 6-8). Longitudinally sec-
tioned Mesodma premolar enamel like that of the taeniolabidoid enamels showed discrete bands of
a regular but much narrower width. A few enamel tubules were seen (PI. 50 fig. 9). Black spots like
those seen in the plagiaulacoid enamel, but irregularly arranged, were observed in some sections of
taeniolabidoid enamels. The bands of the docodont and plagiaulacoid enamels were most distinctly
seen when their long axes were parallel with one of the polarizer planes. They were negatively
birefringent when positioned with their long axes diagonally in the field of vision; in this position
these enamels seemed structureless. The bands of the taeniolabidoid and Mesodma enamels shown
in Plate 50, figs. 6-9, were also negatively birefringent with respect to their length, and most
distinctly seen by maximum prism extinction which occurred when they were inclined a little,
relative to one of the polarizer planes; from 0° to 20° for the different sections and different enamel
areas within each section. This maximum extinction was obtained by rotating the stage with the
section in the direction of the prism inclination towards the cusp from the position where the prisms
were parallel with one of the polarizer planes.

DISCUSSION

Poole (1956) found that tooth enamel in synapsid reptiles was non-prismatic as the crystals were
arranged in closely packed cylindrical groups that were normal to the enamel surface. They were
called pseudo-prisms. Poole (1957) stated that prismatic enamel generally originated in primitive
mammals. Moss (1969) studied fossil therapsid, non-therian and therian enamels, including the
enamel of fossil marsupials and placentals, and concluded that therapsid and all non-therian enamels
are continuous, but with a banded appearance in longitudinal thin sections when viewed in the
polarizing microscope. According to the same author true prismatic enamel which is characteristic

EXPLANATION OF PLATE 50

Figs. 1-9. Thin sections of enamel of longitudinally sectioned teeth micrographed in transmitted light with
the same magnification, x 875. Excepting fig. 1 where normal light was used, the sections were micrographed
in polarized light with crossed filters. The incisal/cuspal direction is to the left, dentine below. 1, late
Jurassic, Kimmeridgian plagiaulacoid molar (gen. et sp. indet.). Black spots within the enamel lie in rows
probably along growth lines (striae of Retzius). A few enamel tubules are seen, section thickness 60 pm.

2, the same section showing indistinct band-like structures of irregular width normal to the dentine where

hair-pin bends of dentinal tubules are seen. 3, late Jurassic, Kimmeridgian docodont molar (gen. et sp.
indet.), showing band-like structures normal to the dentine enamel junction, section thickness 55 pm.

4, Paulchoffatia sp., PIFU VJ 273-155, P 1 2 * 4 , showing band-like structures nearly normal to the dentine, section
thickness 90 pm. 5, Kuehneodon sp., PIFU VJ 308-155, P 5 * * 8 , showing band-like structures normal to the

dentine, section thickness 90 pm. 6, Chulsanbaatar vulgaris Kielan-Jaworowska, ZPAL MgM-I/62, U,

showing distinct, broad bands of regular width inclined about 45° to the dentine enamel junction, enamel

tubules cross the bands, section thickness 80 pm. 7, Kryptobaatar dashzevegi Kielan-Jaworowska, ZPAL
MgM-I/53, P 4 , showing distinct bands of similar width and orientation as in fig. 6, section thickness 90 pm.

8, Stygimys kuszmauU Sloan and Van Valen, UM no. 5, M 2 , showing slightly broader bands but of similar
orientation as in figs. 6 and 7, enamel tubules crossing bands, section thickness 60 pm. 9, Mesodma thompsoni
Clemens, UM no. 3, P 4 , showing distinct curved bands, but of a narrower width than in figs. 6, 7, and 8, a
few enamel tubules are seen, section thickness 50 pm.

PLATE 50

FOSSE, KIELAN-JAWOROWSKA and SKAALE, multituberculate enamel

446

PALAEONTOLOGY, VOLUME 28

only for Theria first appeared in the early Cretaceous (Albian) forms. Poole (1971) studying Jurassic
dryolestids, suggested that prismatic enamel originated in Theria. Poole and Cooper (1971) found
prismatic enamel in the extant agarnid Uromastix however, and concluded that enamel prisms are
not confined to mammals. Osborn and Hillman (1979) studied by polarizing microscopy the enamel
of the pelycosaur Dimetrodon , the therapsids Thrinaxodon , Probainognathus , Probelesodon , Dia-
demodon, and Massetognathus , the early Jurassic primitive triconodont Eozostrodon (? = Mor-
ganucodon ), a late Cretaceous dryolestid, and therian mammals. They found that prismatic enamel
only appeared in the Cretaceous non-therian and therian mammals. In the Permian Dimetrodon
the enamel was micromorphologically homogeneous with regard to crystal orientation, whereas all
Triassic-early Jurassic enamels are characterized by ‘an arrangement of close-packed hexagonal
columns of crystals (Osborn and Hillman 1979, p. 58). In a longitudinal section of Diademodon
enamel a column of crystals was about 5 /<m wide. There was no interprismatic enamel. These
observations concerning crystal orientation in Triassic-early Jurassic enamels are consistent with
the results of our light and SEM microscopic study of the late Jurassic plagiaulacoid and docodont
enamels as radially orientated bands about 5 /a n wide were seen in longitudinal sections (PI. 50,
figs. 2-5) and prism-like structures with interproximate central distances of about 5 /mi in tangential
planes (text-fig. 1 e,f). In SEM the late Jurassic enamels studied by us showed crystals spraying out
towards the outer enamel surface from the centre of 5 /<m thick closely packed ‘columns’ (PI. 49,
figs. 2 and 5).

Grine et al. (1979) by SEM studies found discrete prisms with interprismatic enamel between
them in Eozostrodon teeth. Grine and Vrba (1980) also demonstrated prismatic enamel in the
cynodont Pachygenelus. Interprismatic central distances according to magnifications were about
5 /mi. Frank et al. (1984) described ‘preprismatic’ enamel in late Triassic haramiyids and so did
Sigogneau-Russell et al. (1984) in the early Jurassic therian Kuehneotherium. Preprismatic enamel
according to these latter authors consists of radially arranged columns of crystals similar to those
described by Osborn and Hillman (1979) and by us in this paper. Excepting the results of Grine et
al. (1979) and Grine and Vrba (1980), it may be concluded at present from the reports cited above
that prismatic enamel originated in non-therian mammals and not before the Cretaceous as stated
by Osborn and Hillman (1979), whereas preprismatic enamel where crystals are arranged in 5 /im
thick closely packed columns originated in cynodonts, therapsids, and persisted in some non-
therians and therians from the Triassic through the late Jurassic. In light of this the findings of
Grine et al. (1979) and Grine and Vrba (1980) are perplexing and should encourage further studies
on enamel microstructure of non-therian forms of pre-Cretaceous age.

Our findings support the hypothesis of Fosse et al. (1978) that the multituberculate suborder
Taeniolabidoidea was characterized by remarkably large and widely separated enamel prisms.
Osborn and Hillman (1979) have confirmed the existence of such large prisms in Catopsalis sp. and
Stygimys sp. Carlson and Krause (1982) with a few exceptions found large prisms in Taenio-
labidoidea. The oldest ZPAL multituberculates derive from the late Cretaceous, Gobi Desert,
Djadokhta Formation, which is of ?late Santonian and/or ?early Campanian age (see Material).
However, due to the courtesy of Dr. Demberlyin Dashzeveg we were able to examine by incident
light microscopy the enamel microstructure of two identified and two unidentified multituberculate
teeth from the early Cretaceous (?Aptian or ?Albian) of Guchin Us, Gobi Desert, Mongolian
People’s Republic (see Clemens et al. 1979). They are the specimens GI PST 10/1 1, GI PST 10/13,
GI PST 10/29, and GI PST 10/23 housed in the Institute of Geological Sciences of the Mongolian
Academy of Sciences in Ulan Bator. These teeth show the same size and numerical density of prisms
as do the Taeniolabidoidea from late Cretaceous of Asia (Table 1 ). The Guchin Us multituberculates
are currently being investigated by Kielan-Jaworowska, Dashzeveg, and Trofimov and if it can be
demonstrated that the Early Cretaceous Asian multituberculates are close to the ancestors of late
Cretaceous Taeniolabidoidea, which the enamel structure indicates, it may be concluded that this
peculiar prismatic enamel structure was established and persisted in Taeniolabidoidea through a
time span ranging from Aptian or Albian to late Palaeocene.

Our finding that Mesodma (suborder Ptilodontoidea) enamel had a mean prism density of 25800

FOSSE ET A L. : MU LTITUBERCU LATE ENAMEL

447

per mm 2 agrees with the results of Fosse et al. (1978), but disagrees with the conclusion of Sahni
(1979) with regard to Mesodma enamel. By our method we calculated the densities 24120, 22300,
and 19840 respectively in his figured Mesodma enamels (Sahni 1979, pi. 1, figs. 2, 3, and 6), which
represent cross-cut prisms in Mesodma enamel in his paper. Therefore, we submit that in Mesodma
enamel, there is a significantly higher prism density and smaller prisms than in the representatives
of Taeniolabidoidea. Carlson and Krause (1982) in other ptilodontoid taxa also found prism
diameters similar to those we observed in Mesodma enamels. It seems that Ptilodontoidea do not
differ significantly with respect to numerical prism density from many recent representatives of
Eutheria (Fosse 1968/?, d) and Metatheria (Fosse et al. 1973), nor from the late Cretaceous eutherians
Protungulatum donnae (Fosse et al. 1978) and Kennalestes gobiensis (Table 1 in this paper). Thus the
enamel in Taeniolabidoidea is gigantoprismatic, meaning that the prism density per unit area is four
to five times lower than in the ptilodontoid Mesodma sp. and in all other known mammals. Poole
(1956) suggested that during amelogenesis each column in pseudoprismatic enamel may be the
product of one ameloblast. Assuming that columns in preprismatic enamel and prisms in ‘normal’
as well as gigantoprismatic enamel are the products of single ameloblasts, the present documentation
concerning the range of column and prism dimensions, see text-fig. 1, demonstrates the enormous
diversity in ameloblast diameters in mammals. It may be speculated that each prism and half the
thickness of interprismatic enamel surrounding it were formed by more than one ameloblast in
taeniolabidoid enamel. Considering the similarity of crystal orientation in prisms and interprismatic
material between gigantoprismatic and ‘normal’ enamel (PI. 48, figs. 1, 3, and 4), this seems
improbable and would represent a very unique and special organization of the active ameloblasts
in Taeniolabidoidea compared with Ptilodontoidea and other extinct and all Recent mammals.

Fosse et at. (1973, Figs. 1, 12, and 13) in the medial facet (not explicitly named so) of a
multituberculate incisor from the Hell Creek Formation, Montana, later identified as belonging to
the Taenilabidoidea (Fosse et al. 1978), observed that the prism rods deviated in a transversal plane
relative to the tooth. In the ventrolateral facet (named buccodistal. Fosse et at. 1973) of the same
tooth the prism rods in the inner two thirds of the enamel deviated towards the tip of the crown. In
the present study the same general arrangement of prism rods was observed in sections of the
incisors of Chulsanbaatar vulgaris , ZPAL MgM-I/62 (PI. 49, fig. 3), and S. kuszmauli, UM no. 5
(PI. 49, fig. 1), and we tentatively suggest that this prism rod orientation in incisor enamel may be
common to all late Cretaceous Taeniolabidoidea.

Fosse et al. (1973, Fig. 6) described the crystal orientation in multituberculate enamel, stating
also that the prisms are arcade shaped in cross-section, and that the apices of the arcades point in
the direction of the inclination of the prism rods relative to the dentine enamel junction. The same
orientation of the crystals and the same morphology of the cross-cut prism rods were found in
the present study of taeniolabidoid and ptilodontoid late Cretaceous enamels by SEM (Pis. 48 and
49). Polarized light microscopy indicated that the c-axes of the crystals in the prisms are inclined
slightly in a cervical direction relative to the long axes of the prism rods, as the acute angle between
prism rods and polarizer plane in the extinction position was cervically positioned relative to the
prism rods.

In twelve mandibles from the ZPAL collection we were able to compare the prism density of
incisors, premolars, and molars and in ten of them the numerical density was lower in incisors
than in premolars or molars; in two of these it increased as much as twofold from incisor to molar
(Table 1). It should, however, be stressed that the lowest prism densities of Mesodma or other
mammalian teeth studied were about twice as high as the highest in taeniolabidoid molars.

Like Sahni (1979, Table 1 and pi. 3, fig. 6) we found gigantoprismatic enamel in the late
Cretaceous North American Meniscoessus. If other genera of Cimolomyidae are distinguished by
gigantoprismatic enamel this may indicate a relation of that family to Taeniolabidoidea, although
the lower incisors in Cimolomyidae (Archibald 1982) are completely covered by enamel, a feature
characteristic of Ptilodontoidea.

In view of the documented large difference between the microstructure of taeniolabidoid
and ptilodontoid enamel, we believe that future studies of enamel microstructure may assist in

448

PALAEONTOLOGY, VOLUME 28

establishing the systematic position and phylogenetic relationships of some poorly known multi-
tuberculate groups regarded at present as incertae sedis.

Acknowledgements. We wish to express our gratitude to Dr. B. Krebs and the late Dr. S. Henkel
(Palaeontologisches Institut, Freie Universitat, Berlin) for the loan of the Kimmeridgian multituberculate and
docodont specimens from Portugal and the permission to study them, and to Dr. D. Dashzeveg (Institute of
Geological Sciences of the Mongolian People’s Republic, Ulan Bator) for the premission to study the enamel of
early Cretaceous multituberculates from Guchin Us in Mongolia. We are grateful to Dr. D. Sigogneau-Russell
(Institut de Paleontologie, Museum National d’Histoire Naturelle, Paris) and Professor G. Hahn (Department
of Geology, University of Marburg) for reading the manuscript and offering useful comments. The second
author thanks Professor M. C. McKenna (The American Museum of Natural History, New York) for
donating to the Institute of Paleobiology in Warsaw a sample of Hell Creek Formation teeth, some of which
are studied in this paper.

We also wish to thank the photographer at ZPAL Mr. M. Dziewinski and the photographer at IAUB Mr.
R. Jensen for their assistance in the preparation of the micrographs.

REFERENCES

Archibald, j. d. 1982. A study of Mammalia and geology across the Cretaceous-Tertiary boundary in Garfield
County, Montana. Univ. Calif. Publ. Geol. Sci. 122, 1-286.
carlson, s. j. and krause, d. w. 1982. Multituberculate phylogeny: evidence from tooth enamel ultrastructure.
Ann. Meeting 1982 Geol. Soc. Amer ., Abstr ., 460.

clemens, w. a., lillegraven, J. a., lindsay, e. h. and simpson, G. G. 1979. Where, when, and what— a survey
of known mammal distribution. In lillegraven, j. a., kielan-jaworowska, z. and clemens, w. a. (eds.).
Mesozoic mammals: the first two-thirds of mammalian history , 7-58. University California Press. Berkeley,
Los Angeles, London.

fosse, g. 1968a. The calculation of prism diameters and number of prisms per unit area in dental enamel.
Acta odont. scand. 26, 315-336.

— 1968fi. Prism density and pattern on the outer and inner surface of the enamel mantle of canines. Ibid.
501-543.

— 1968c. The vertical compression of the prism pattern on the outer enamel surface of human permanent
teeth. Ibid. 545-572.

— 1 968<Y. The numbers of cross-sectioned ameloblasts and prisms per unit area in tooth germs. Ibid. 573-
603.

risnes, s. and holmbakken, n. 1973. Prisms and tubules in multituberculate enamel. Calc. Tiss. Res. 11,

133-150.

— roli, J- and knudsen, h. 1974. A sectioning machine for teeth and other brittle materials. Acta odont.
scand. 32, 299-304.

— eskildsen, o., risnes, s. and sloan, r. e. 1978. Prism size in tooth enamel of some Late Cretaceous
mammals and its value in multituberculate taxonomy. Zool. Scripta, 7 (1), 57-61.

frank, r. m., sigogneau-russell, d. and voegel, j. c. 1984. Tooth ultrastructure of Late Triassic Haramiyidae.
Jl dent. Res. 63 (5), 661-664.

gradzinski, r., kielan-jaworowska, z. and maryanska, t. 1977. Stratigraphy of the Upper Cretaceous
Djadokhta, Barun Goyot and Nemegt formations of Mongolia. Acta Geol. Polonica , 27 (3), 281-318.
grine, f. e. and vrba, e. s. 1980. Prismatic enamel: a pre-adaptation for mammalian diphyodonty? South Afr.
Jl Sci. 76, 139-141.

— and cruickshank, a. r. i. 1979. Enamel prisms and diphyodonty: linked apomorphies of mammalia.
Ibid. 75, 114 120.

hahn, G. 1969. Beitrage zur Fauna der Grube Guimarota Nr. 3. Die Multituberculata. Palaeontographica , A,
133(1-3), 1-100.

1971. The dentition of the Paulchoffatidae (Multituberculata, Upper Jurassic). Memoria Serv. Geol.
Portugal, (n.s.), 17, 7-39.

1978. Die Multituberculata, eine fossile Saugetiere-Ordnung. Sonderbd. Naturwiss. Ver. Hamburg , 3, 61-
95.

— and hahn, r. 1983. Multituberculata. In westphal, f. (ed.). Fossilium Catalagus , I: Animalia , pars 127,
1 -409. Kugler Publications, Amsterdam.

FOSSE ET AL.: MULTITUBERCULATE ENAMEL

449

kielan-jaworowska, z. 1970. New Upper Cretaceous multituberculate genera from Bayn Dzak, Gobi Desert.
In kielan-jaworowska, z. (ed.). Results Polish Mongolian Paleont. Expeditions, IE Palaeontologia Polon-
ies 21 , 5-41.

1974a. Multituberculate succession in the late Cretaceous in the Gobi Desert (Mongolia). In ibid. V.
Palaeontologia Polonica, 30 , 23-44.

matthew, w. d. and granger, w. 1925. Fauna and correlation of the Gashato Formation of Mongolia.
Amer. Mus. Novi rates, 189 , 1-12.

moss, m. L. 1969. Evolution of mammalian dental enamel. Ibid. 2360, 1-39.

osborn, j. w. and hillman, j. 1979. Enamel structure in some therapsids and Mesozoic mammals. Calcif.
tissue Int. 29, 47-61 .

poole, d. f. G. 1956. The structure of the teeth in some mammal-like reptiles. Quart. Jl. Micr. Sci. 97 , 303-
312.

— 1957. The formation and properties of the organic matrix of reptilian tooth enamel. Ibid. 98, 349-367.

— 1971. An introduction to the phylogeny of calcified tissues. In dahlberg, a. a. (ed.). Dental morphology
and evolution, 65-79. Chicago University Press, Chicago and London.

— and cooper, j. s. 1971 . Prism structure in the enamel of a reptile. Jl dent. Res. 50 , 68 1 .

sahni, a. 1979. Enamel ultrastructure of certain North American Cretaceous mammals. Palaeontographica,
A, 166(1-3), 37- 49.

sigogneau-russell, d., frank, r. m. and hemmerle, j. 1984. Enamel and dentine ultrastructure in Early
Jurassic therian, Kuehneotherium. In Patterson, c. (ed.). Commemorative volume of Professor K. A.
Kermack. Zool. Jl Linn. Soc. Lond. 82, 207-215.

sloan, r. and van valen, l. 1965. Late Cretaceous mammals from Montana. Science, 148 (3667), 220-227.
trofimov, b. a. 1980. M ultituberculata i Symmetrodonta iz nizhniego miela Mongolii. Dokl. Akad. Nauk
SSSR , 251 ( 1 ), 209-212. [In Russian.]

G. FOSSE
S. G. SKAALE

Institute of Anatomy
University of Bergen
5000 Bergen, Norway

Z. KIELAN-JAWOROWSKA
Zaklad Paleobiologii
Polska Academia Nauk
al. Zwirki i Wigury 93

Typescript received 5 March 1984 02-089 Warszawa

Revised typescript received 31 July 1984 Poland

MICROPALAEONTOLOGY OF THE LATE
PROTEROZOIC VETERANEN GROUP,
SPITSBERGEN

by ANDREW H. KNOLL and KEENE SWETT

Abstract. Shales and siltstones of the Upper Proterozoic Veteranen Group, Spitsbergen, contain abundant and
well-preserved microfossil populations. Plankton assemblages from open coastal deposits include a number of
taxa previously known from Upper Riphean sequences in Scandinavia, the Soviet Union, and North America.
Microfossils from more restricted coastal environments are dominated by small coccoidal unicells and
filamentous sheaths of probable cyanobacteria, with locally abundant rod-shaped fossils of blue-greens or other
bacteria. Biostratigraphic considerations indicate that the Veteranen Group, which comprises the earliest
unmetamorphosed sedimentary sequence in Spitsbergen, was deposited between 800 and 900 Ma ago; this nearly
4000 m sequence was deposited within a relatively brief interval during the early stages of subsidence in a basin
that eventually opened to become Iapetus. The preservation of delicate prokaryotic microfossils in lagoonal
black shales contributes to the resolution of systematic problems arising from the development of two distinct
research traditions in Precambrian palaeontology.

During the past decade, it has become clear that the Late Proterozoic earth supported a
taxonomically diverse and ecologically heterogeneous biota of micro-organisms, as well as, in its later
stages at least, early representatives of the multicellular (tissue grade) kingdoms. This diversity is well
documented in the sedimentary successions of the Svalbard archipelago, where beautifully preserved
microfossil assemblages are found at numerous stratigraphic horizons throughout a 7000 m sequence
that spans some 400 million years of Late Proterozoic to Early Cambrian time. Both planktic and
microbenthic fossils are found in rocks representing a variety of sedimentary environments (Knoll
1982a, b , 1984; Knoll and Calder 1983), and this record is augmented by stromatolites and
microphytolites which are widely distributed in carbonate portions of the succession (Golovanov
1967; Golovanov and Raaben 1967; Raaben and Zabrodin 1972; Swett and Knoll 1985). During the
1981 and 1982 field seasons we conducted detailed stratigraphic and sedimentological studies of
several localities in north-eastern Spitsbergen (text-figs. 1, 2). In this paper we describe the
distribution of microfossil assemblages preserved in the lowermost unmetamorphosed units of the
Spitsbergen Proterozoic succession, the Veteranen Group.

GEOLOGICAL SETTING

The late Precambrian and Palaeozoic geological history of the North Atlantic region centres on the
inception, development, and ultimate destruction of the Iapetus Ocean. The closing of this ocean
basin is documented by the igneous intrusions, metamorphism, sedimentary patterns, and structural
features associated with the Caledonian orogeny. Its early development is reflected in the late
Proterozoic sedimentary sequences of the region. In north-eastern Spitsbergen, some 6600 m of
unmetamorphosed sedimentary rocks lie beneath the lowermost Cambrian strata of the island; this
succession is in turn underlain by a comparably thick sequence of metasedimentary and metavolcanic
rocks (Harland 1959; Harland and Wright 1979). The thickness and almost uniformly shallow
marine nature of the unmetamorphosed sequence indicates that it was deposited near the margin of a
slowly subsiding trough, genetically linked to, but perhaps antedating, the opening of a substantial
ocean basin. Fairchild and Hambrey ( 1 985) interpret palaeocurrent data from glaciogenic rocks near

IPalaeontology, Vol. 28, Part 3, 1985, pp. 451-473, pis. 51-53.)

452

PALAEONTOLOGY, VOLUME 28

text-fig. 1. Map of Svalbard showing the location of the study area in northern Spitsbergen. The
shaded study area is enlarged in text-fig. 2.

the top of the Proterozoic section as indicating glacial flow from the south-west, a finding that
supports the concept of a late Vendian opening for the ocean basin (Harland, pers. comm. 1984).
Except for a few hundred metres of well-laminated graded siltstone and shale couplets that may
reflect fault control of local deposition, and two thin lava flows at the same general stratigraphic level
(Wilson 1958), there is little to suggest an early rifting stage of basin evolution. Evidence for such an
event may lie in the metamorphic rocks at the base of the sequence.

In north-eastern Spitsbergen the unmetamorphosed Proterozoic section comprises three easily
contrasted groups: the immediately sub-Cambrian Polarisbreen Group, approximately 800-900 m of

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

453

text-fig. 2. Enlargement of shaded area in text-fig. I .
Stippled areas represent outcrop; white areas indicate
glacial ice, except for Lomfjorden and the Hinslopen-
stretet (upper right edge of figure) which are water
covered. Detailed examinations of the Veteranen
Group were undertaken at Faksevagen (a) and
Cavendishryggen (b). The upper part of the group was
also studied in the Polarisbreen area (c).

predominantly fine-grained siliciclastic rocks with extensive tillites (Wilson and Harland 1964;
Hambrey 1982); the underlying Akademikerbreen Group, up to 2000 m thick and consisting largely
of carbonates (Wilson 1961); and, at the base of the succession, the sequence of greatest interest
here— the Veteranen Group (Wilson 1958).

The Veteranen Group has been divided into four formations (in ascending stratigraphic order):
Kortbreen, Kingbreen, Glasgowbreen, and Oxfordbreen (Harland et al. 1966). Basing our field
studies on the thorough stratigraphic description of this group by Wilson (1958), we measured
sections along the eastern wall of Veteranen Glacier at Cavendishryggen and in the Faksevagen area
west of Lomfjorden (text-figs. 2, 3). The oldest Veteranen strata located are thinly laminated and
occasionally cross-laminated calcisiltites of the lower Kortbreen Formation, exposed in the core of
an anticline at Faksevagen. These limestones are tectonically deformed, but Wilson’s (1958)
thickness estimate of 300 m agrees well with our observations. Tan-weathering dolomitic units are
common in this section, as are thin intercalations of quartzarenite and dark shales. The limestones
become increasingly shaly in their uppermost 20 m and are overlain abruptly, but with apparent
conformity, by the thick quartzarenites of the upper Kortbreen Formation. These sandstones are
well-rounded, fine to medium grained quartzarenites, variously white or tinted pink, buff, maroon, or
light green. Ripple marks with wavelengths of a few to 6 cm abound, as do cross-beds with set
thicknesses of up to about 15 cm. Mudcracks occur intermittently throughout the unit, and

454

PALAEONTOLOGY, VOLUME 28

fossiliferous black shales also occur as thin interbeds. Wilson (1958) characterized these sandstones
as ‘typically shallow water deposits’. We concur and further suggest that they accumulated in
shallow, tidally influenced environments marginal to the exposed continent.

Conformably overlying the Kortbreen quartzarenites is the Kingbreen Formation. This formation
is 1 200 m thick at Faksevagen and slightly thicker in the wall of Cavendishryggen; it has been divided
into three members (Harland et al. 1966). At Cavendishryggen the basal Galoistoppen Member
contains a thick sequence of millimetre-laminated dark siltstones and shales. As noted above the
laminae in this unit comprise distinct graded couplets that are quite similar to beds from the some-
what younger Ocoee Supergroup of Tennessee. Keller (1979) has interpreted the Ocoee couplets as
indications of the waning stages of fault influenced sedimentation, prior to the transition from a rift

CL

3

O

CE

O

<

cr

>

METERS

SAMPLE HORIZONS

• F-2200

O ® F-1605

9 F-1450

9

9 _ F-9 4 4

• * F-8 56

• ? F-575

• w F-346

9 F-l 5

F-1777

F-l 3 45
F-8 95
F-7 I 2
F-500
F- 160

• ? F2-32

• A F2-27
9 W F 2-25

F2-33

F2-29

F2-26

F2-22

F2-I 8
F 2-1 6

F2-23

F2-I7

O F 2- I 3

O F2-I

O F2-4

O F2-5

• F2-9

9 F2-II

9

F2-I2

O

F2-40
F2-4 2
F2-35

F 2 39
to 38

text-fig. 3. Stratigraphic column of the Veteranen Group at Faksevagen, based on
Wilson (1958) and our own measurements. Palynological sample horizons (with
sample numbers) are indicated by circles. Filled circles indicate well-preserved
microfossil assemblages; half-filled circles indicate poorly preserved assemblages;
samples marked by open circles are barren. In the column, dashed lines indicate
shales and siltstones; dotted lines signify sandstones; and rectangles and rhombic
patterns indicate limestones and dolostones, respectively. Circles indicate ooids and
concave downward arcs signify stromatolitic carbonates.

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

455

basin setting to a passive margin continental shelf. Bedding surfaces of the lower Galoistoppen
sequence contain textural lineations that indicate unidirectional flow.

In the upper portion of the Galoistoppen Member, quartzose sandstone interbeds become
increasingly common, and the dominant character of the sequence shifts to quartzarenites and
interbedded black shales, with conspicuous calcareous units that contain low angle cross-beds,
shallow channels, and intraformational conglomerates. The section at Cavendishryggen is more
conspicuously calcareous than the correlative beds at Faksevagen, where variegated and often
mudcracked shales and dolomitic shales are the dominant lithologies.

The overlying Bogen Limestone Member contains frequent oolite and microphytolite-bearing
limestones, as well as cross-bedded (including herringbone cross-beds) calcarenites and intercalated
siliciclastic units. The uppermost Cavendishryggen Quartzite Member contains massive, cross-
bedded quartzarenites with thinner interbedded units of sandy and silty flagstones and shale.

The Kingbreen Formation is conformably overlain by the Glasgowbreen Formation, a 900 m unit
that bears many similarities to the underlying Cavendishryggen sequence. This formation contains
massive thicknesses of pink, green, and buff quartzarenites. Like those near the base of the Veteranen
Group, the Glasgowbreen sandstones are generally compositionally mature, well rounded quartz-
arenites, marked by conspicuous ripple marks, interference ripples, megaripples, and cross-bedding
(including herringbone cross-beds). Interbedded carbonaceous shales are common in some horizons
within the formation. Wilson ( 1 958) referred to these carbonaceous units as greywackes, but it is clear
that they are discrete shales and sandy shales environmentally related to the tidal sandstones with
which they are intercalated. The succeeding Oxfordbreen Formation contains proportionally fewer
sandstones and more shales, and is distinguishable by the presence of carbonate interbeds (including
oolitic and stromatolitic units). The stromatolites are low profile, laterally linked hemispheres up to
I m high and 1 -7 m across. Maximum synoptic relief developed during deposition was about half this
height. Red shales are common in the upper part of the formation; these are generally flat-laminated,
but can contain ripples, cross-laminations (in intercalated sandstones), mudcracks, and ripped-up
mud flake clasts. These sedimentary features are particularly conspicuous at a locality along the north
face of Polarisbreen Glacier (text-fig. 2). The top of the formation is fixed at the base of the massive
dolomites and limestones that characterize the Akademikerbreen Group.

It can be seen from the foregoing summary that the rocks of the Veteranen Group document a
variety of sedimentary environments, but that variation occurs within fairly strict limits. Almost the
entire sequence records shallow marine depositional environments, ranging from below wave base to
tidally influenced conditions and quiet lagoons marked by occasional subaerial exposure and storms.

Available radiometric dates from Precambrian rocks of north-eastern Spitsbergen reflect
Caledonian overprints of isotopic systems. On the basis of stromatolite and microphytolite
distributions, Milstein and Golovanov (1979) have suggested that the Veteranen Group is late
Riphean (950-670 Ma) in age. Acritarchs in this group, coupled with those in the correlative
Franklinsundet and Celsiusberget groups of Nordaustlandet (Knoll 1 982c/) and those found in
overlying units, indicate that the group as a whole probably exceeds 800 Ma in age, but is unlikely to
be much older than 900 Ma (see discussion below).

MICROFOSSIL ASSEMBLAGES

Extensive collections of Veteranen Group samples were made from both the Cavendishryggen and
Faksevagen sections. The Cavendishryggen area is characterized by more severe diagenetic
alteration, approaching chlorite grade metamorphism. Strata are more highly indurated; they
commonly develop a platy aspect, and a weak slaty cleavage is developed in the lower units. As a
consequence the organic matter in Cavendishryggen rocks is black and fossil preservation is poor. In
contrast, twenty-eight of the thirty-six samples taken from an approximately 3500 m section at
Faksevagen contained identifiable fossils. Preservation at this locality is poorest in the youngest
formation examined, the Oxfordbreen Formation, despite the abundance of reddish-brown organic
matter in dark green to black shales. Preservation is much better in the Glasgowbreen and Kingbreen

456

PALAEONTOLOGY, VOLUME 28

formations, where microfossils are often abundant and relatively diverse. The colour of microfossil
walls changes systematically as one descends the local stratigraphic column, but identifiable
microfossils can still be seen in the dark grey to black organic residues isolated from grey shales within
the carbonates at the base of the Kortbreen Formation.

Small (< 25 pm)coccoidal unicells are the most abundant fossils in every sample examined (PI. 51,
figs. 8, 11-18). Variation in wall thickness, wall ornamentation, and clustering patterns suggests that
these unicell populations are taxonomically heterogeneous, an inference that can also be drawn from
the size frequency distributions of sample populations (text-fig. 4a, b). Many taxonomic names have
been applied to previously described fossils of similar aspect (e.g. Timofeev 1969) but, because
diagnoses have often been based on characters likely to be diagenetic in origin and because published
illustrations are often difficult to interpret, it would be hazardous to apply names without first
making a careful study of the type materials. In so far as many of the observable differences between
these small unicells may be related to intraspecific and/or diagenetically introduced variation, the
formal recognition of each discernible morphotype would lead to an overestimate of species
diversity. On the other hand the simple morphology of these fossils and the inevitable loss of
taxonomically important characters (such as pigment complements) during post-mortem degrada-
tion must lead to an underestimate of original species diversity (Knoll and Golubic 1979). We have
for these reasons not attempted to quantify the preserved diversity within this size class, preferring
instead to treat these fossils as a unit, while acknowledging their probable biological heterogeneity.

(F-712)

5 10 15 20 25 30 >30 [f/m)

DIAMETER

(F2-27)

5 10 15 20 25 30 >30 (pm)

DIAMETER

text-fig. 4. Size frequency distribution of small coccoidal unicells from samples F-712, Glasgowbreen
Formation (a), and ¥2-21 , Kingbreen Formation (b). For both a and b, N — 500.

Among larger spheroidal microfossils, Leiosphaeridia asperata (Naumova) Lindgren, 1982
( = Kildinella hyperboreica Lindgren, 1982) is easily the most abundant acritarch (PI. 52, figs. 4-8, 10,
1 1 ). It is found in all samples analysed and is common in most. L. asperata is morphologically simple,
consisting of flexible, smooth-walled vesicles approximately 15-80 pm in diameter (usually 25-
45 pm) that apparently dehisced by means of a median split mechanism. It is likely that these fossils
are the remains of algal cysts (e.g. Lindgren 1981; Vidal and Knoll 1983) but, as Lindgren (1981,
1982) has pointed out, a variety of taxonomically and functionally different algal remains could be
lumped within this form species.

Other common microfossils include Kildinosphaera chagrinata Vidal in Vidal and Siedlecka, 1983
( = in part Kildinella sinica , according to Vidal and Siedlecka 1983), Synsphaeridium sp., and cf.
Stictospheridium spp. ( sensu Vidal 1976). More complex acritarchs are rare; these include
Kildinosphaera granulataWded in Vidal and Siedlecka, 1983, Tasmanites riphejicus Jankauskas, 1978,
Satka colonialica Jankauskas, 1979n, Bavlinella faveolata (Schepeleva) Vidal, 1976, Favoso-

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

457

sphaeridium sp., and fragmentary remains of very large acritarchs, probably Chuaria circularis
Walcott, 1899 (PI. 51, figs. 1, 2). Tubular microfossils, the probable extracellular sheaths of
filamentous cyanobacteria, occur in many samples (PI. 51 , fig. 19; PI. 52, figs. 13, 12), and small rod-
shaped remains are common in sample F-712 (PI. 51, figs. 5, 6, 9, 10). Within any one sample,
morphological diversity is low (text-fig. 5), a characteristic of late Riphean acritarch assemblages
previously noted by Vidal and Siedlecka (1983).

' SAMPLE HORIZON

TAXON ~ " —

KO

KI

GL

ox

~n

r 3

kb

~n

no

s

~n

ro

~nI

~n

r\>

ro

GJ

~n

ro

ro

~Nl

~n

ro

GJ

ro

~n

GJ

CD

~n

cn

O

O

~n

CJl

-nI

cn

~n

~n)

ro

~n

CD

cn

CD

T1

OJ

4-

<T>

~n

ro

ro

O

O

Bavlmella foveolata (Schepeleva) Vidal

R

R

Eosynechococcus sp.

R

R

R

c

Tasmamtes rifejicus Jankauskas

R

R

R

R

? Chuaria circularis Walcott

R

R

R

R

R

R

R

Favosophaeridium sp.

R

KUdmosphaera chagrinata Vidal

R

R

R

C

R

R

C

R

K Udinosph aera granule ta Vidal

R

R

C

R

Leiosphaeridia asperata (Naumova) Lmdgren

C

C

R

C

R

R

c

C

C

R

c

C

C

Sotko colonialica Jankauskas

R

R

R

R

R

R

cf. Stlctosphoeridium sp. ( sensu Vidal)

C

c

C

R

c

Synsphoeridium sp.

R

R

c

R

R

R

R

R

R

Small Unicells

c

C

A

A

A

C

A

A

A

A

A

C

C

Filaments

c

R

A

A

A

c

R

R

R

C

C

R

R

Number of taxa

4

4

4

6

7

5

7

11

11

7

11

4

8

text-fig. 5. Chart showing the distribution of microfossil taxa within selected samples of the
Veteranen Group. Samples are arranged stratigraphically, with the oldest sample on the left.
KO, Kortbreen Formation; KI, Kingbreen Formation; GL, Glasgowbreen Formation; OX,
Oxfordbreen Formation. R, rare; C, common; A, abundant.

Despite the thickness of the Veteranen Group, there is no evidence for stratigraphically significant
changes in assemblage composition within the sequence. This is not to say that all recovered
assemblages are identical; certainly, they are not. Gray and Boucot (1975) have noted, however, that
variation in assemblage composition can reflect taphonomic and/or palaeoecological, as well as
stratigraphic differences. Relatively poor preservation through much of the Oxfordbreen Formation
may mask evolutionary changes in the plankton biota: fossils in the overlying Akademikerbreen
Group do differ in ways that are most likely a consequence of evolutionary turnover. On the other
hand, one well-preserved assemblage from the Oxfordbreen Formation (sample F-2200) is indis-
tinguishable from assemblages found in the lower formations of the group.

Much of the variation evident in the relative abundance of microfossii groups represented is
thought to reflect palaeoenvironmental variation within the coastal marine sedimentary deposits.
Filamentous sheaths provide a case in point. In several samples (e.g. F2-27), filaments are
conspicuously abundant, comprising some 40% or more of all preserved specimens. The filaments
are long (often more than 100 /urn in length) and not infrequently intertwined with one another. Large
acritarchs ( > 30 /urn diameter) comprise only 3-5% of the assemblage, and biostratigraphically
significant taxa are generally rare or absent. Filament rich assemblages are found in thin, black to
greenish shale interbeds within predominantly cross-bedded and rippled quartz sandstones or oolitic
and cross-bedded calcarenites. We interpret these assemblages as the remnants of very near-shore

458

PALAEONTOLOGY, VOLUME 28

coastal to lagoonal communities. The filaments are thought to represent autochthonous or nearly
autochthonous, probably benthic, cyanobacterial populations.

At the other end of the spectrum are assemblages in which filaments occur only as rare, fragmented
and probably allochthonous individuals. In these assemblages, large acritarchs are relatively
common (10-15% of all individuals in sample populations from sample F-575) and diverse. Such
assemblages tend to occur within thicker, green to black siltstone and shale units, with only minor
sandstone intercalations. We interpret these biological and sedimentary assemblages as representing
the most off-shore or ‘normal marine’ setting within the Veteranen Group. Most if not all of the
spheroidal microfossils found in these assemblages are thought to have been planktic.

DISCUSSION

Biostratigraphy

The Veteranen biota contains several taxa that have long stratigraphic ranges and hence limited bio-
stratigraphic utility. Among these fossils are: Leiosphaeridia asperata, cf. Stictosphaeridium sp.,
Synsphaeridium sp., and the small unicells, filaments, and rods. More useful are Kildinosphaera
granulata, Tasmanites riphejicus , and Satka colonialica which to date are known only from Upper
Riphean rocks. Along with co-occurring L. asperata, K. chagrinata , and probable Clmaria, these
fossils form an assemblage similar to previously described assemblages from Upper but not
uppermost (Kudashian, possibly equivalent to Vidal’s (1976) Lower Vendian) Riphean rocks
elsewhere in the Northern Hemisphere.

Two forms, Favososphaeridium sp. and Bavlinella faveolata, are morphologically similar to taxa
best known from uppermost Riphean to Cambrian rocks. The occurrence of Bavlinella is particularly
interesting because this distinctive fossil has sometimes been considered an index fossil for Vendian
rocks. Its actual range is now known to extend from the latest Riphean to the Cambrian, and this new
population confirms earlier reports of its questionable occurrence in Upper Riphean rocks (see
discussion in Vidal 1976). Bavlinella is often an abundant constituent of Vendian assemblages
associated with glaciogenic rocks and has been interpreted as an opportunistic taxon by Knoll et al.
(1981). Such an interpretation carries with it the prediction that B. faveolata should occur as a minor
component of earlier and later microfloras, and this seems to be the case.

The apparent stratigraphic homogeneity of the Veteranen assemblages may reflect a slower rate of
morphological evolution among early algae, a rapid rate of deposition for the Veteranen Group, or
both. Comparable thicknesses of miogeosynclinal sediments are known to have accumulated during
the approximately 70 million year long Cambrian Period along the eastern and western margins of
North America (Bond et al. 1983; Cook and Bally 1975). The relatively rapid subsidence required for
such accumulation is thought to be related to the cooling of the lithosphere following rifting
(McKenzie 1978; Bond et al. 1983; Armin and Mayer 1983). It is also true, however, that our present
understanding of late Riphean palaeontology is such that it is only possible to divide the period into a
handful of assemblage zones, each the same order of length as the Cambrian (Vidal and Knoll 1 983).
This may reflect slow rates of morphological (but not necessarily physiological) evolution. Thus, the
stratigraphic indivisibility of the Veteranen Group may well be a consequence of both rapid deposi-
tion and slow morphological change.

Absolute age estimates for the Veteranen assemblages must be inferred from radiometric dates
assigned to comparable microbiotas from other areas. Biotas containing the distinctive elements that
characterize the Veteranen Group are found in the Klubbnes and Andersby formations of the Vads m
Group, East Finnmark (Vidal 1981); the lower Batsfjord Formation of the upper Barents Sea Group,
also in East Finnmark (Vidal and Siedlecka 1983); the Chuar Group, Arizona (Vidal and Ford 1985);
the Red Pine Shale of the Uinta Mountain Group, Utah (Vidal and Ford 1985); the ‘type’ Upper
Riphean beds of the southern Urals (Jankauskas 1982); and the lower part of the Upper Visingso
Beds, Sweden (Vidal 1976; Vidal and Ford 1985).

With the exception of a K-Ar determination recalculated as approximately 640 Ma for dolerite
dykes that cut the Barents Sea Group in northern Norway (Beckinsale et al. 1975), available radio-

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

459

metric dates on Upper Proterozoic sequences in Scandinavia are largely Rb-Sr whole rock analyses
of shales (reviewed in Vidal and Knoll 1983). Klubbnes Formation shales have yielded dates of
807 + 19 Ma (recalculated by Vidal from Sturt et al. 1975), while Visingso occurrences are overlain by
shales dated at 707 + 37 Ma (Vidal 1981). Western North American assemblages antedate a structural
disturbance bracketed at 820-770 Ma (reset K-Ar dates on basalts) and postdate an episode of
basaltic extrusion dated at 1070 + 70 Ma by Rb-Sr whole rock analyses (Elston and McKee 1982).
Rb-Sr whole rock determinations of Red Pine Shale specimens yield an age of 950-925 Ma for that
formation (Crittenden and Peterman 1975; Chaudhuri and Hanson 1980). K-Ar dates for Upper
Riphean sedimentary rocks in the Urals suggest a depositional age of 850-940 Ma (Keller 1982).

Whole rock chronometric analyses of detrital sedimentary rocks have been the subject of much
debate, as have many K-Ar determinations of presumably early diagenetic glauconites from
Precambrian sediments. None the less, available radiometric data, coupled with corroborative
tectonic and palaeomagnetic considerations (Elston, pers. comm. 1984) suggest that assemblages of
the type found in the Veteranen Group very likely fall in the 800-900 Ma range. While it is clear that
planktic microfossils are of demonstrated value in Proterozoic biostratigraphy (e.g. Vidal and Knoll
1983), it is also apparent that the full stratigraphic potential of early plankton will be realized only
when better radiometric, palaeoecological, and biogeographic control is available.

Palaeoecology

The palaeoenvironmental variability of late Proterozoic microfossil assemblages has frequently been
discussed in the literature (Vidal 1976, 1981; Knoll 1981, 19826, 1984; Knoll and Calder 1983; Vidal
and Knoll 1983). Recognition of palaeoecological patterns of distribution is important for both
biological and geological reasons: biological because ‘lateral’ variation must be taken into account in
any evolutionary interpretation of the fossil record (e.g. Strother et al. 1983); and geological in that
Proterozoic fossils, like their Phanerozoic counterparts, are potentially valuable as indicators of
sedimentary environment. In the Veteranen Group, both the diversity and the relative abundance of
large ( > 30 jum) acritarchs increase along a gradient from inshore, often lagoonal deposits to more
open coastal siltstones and shales. As noted above, the abundance and preservational quality of
filament populations decreases along the same gradient. (Environments are established on the basis
of sedimentary structures, textures, and bedding sequences.) Similar distributions have been
recognized in other Upper Proterozoic sequences, both carbonate and siliciclastic (reviewed by Vidal
and Knoll 1983); the Veteranen observation serves to increase one’s faith in the generality of the
pattern.

Palaeoecological and biogeographic distributions are also relevant to biostratigraphic determina-
tions. For example, Trachysphaeridium laminaritum Timofeev, 1966 is an important constituent of
many late Riphean microfloras but has not been found in the Veteranen Group. Vidal and Ford
(1985) have observed that in rocks of the Grand Canyon and elsewhere this species and
Kildinosphaera chagrinata have mutually exclusive distributions. K. chagrinata is relatively common
in the Veteranen Group, so perhaps some poorly defined environmental parameter excluded T.
laminaritum from the Veteranen sea. A more problematic example concerns the distinctive micro-
fossil K. lophostriata (Jankauskas) Vidal in Vidal and Siedlecka, 1983 which occurs in late Riphean
assemblages from the Soviet Union, Scandinavia, and North America, but which has not been
recognized in Veteranen assemblages. Does this absence indicate that K. lophostriata had not evolved
at the time of Veteranen deposition, that it was extinct by the time the Svalbard shelf originated, or
that K. lophostriata is missing for reasons of ecology, biogeography, or chance? We simply do not
know, it being impossible to eliminate any one of these possible explanations on the basis of present
evidence. Once again this underscores the need for further investigations of late Proterozoic micro-
biotas conducted within a framework of strict stratigraphic and sedimentological control.

Systematics and palaeobiology

Proterozoic microfossils are generally studied by the petrographic examination of silicified
carbonates or by the palynological maceration of siliciclastic rocks. In a critique of the systematic

460

PALAEONTOLOGY, VOLUME 28

problems created by this dichotomy of approach. Diver and Peat (1979) have suggested that many of
the apparent differences between ‘chert’ and ‘shale’ biotas are illusory, the result of separate research
schools with independently evolved taxonomic practices. Systematic traditions certainly pose a
serious problem for comparative studies of Precambrian microbiotas, as do differences in
preservation potential that may separate shales and silicified carbonates. On the other hand, some of
the observed differences among assemblages are real and reflect the ecological partitioning of
Proterozoic environments. Recognition of this aspect of the problem suggests a path toward the
solution of the ‘chert-shale’ dilemma.

Most of the known biostratigraphically useful, open shelf acritarch assemblages come from
carbonaceous siltstones and shales (e.g. Timofeev 1969; Vidal 1976); however, not all open coastal
rocks are siliciclastic. Compacted and compressed acritarchs have been isolated from carbonaceous
limestones and dolomites of the Upper Visingso Beds of Sweden (Vidal 1976) and the Batsfjord
Formation of East Finnmark, Norway (Vidal and Siedlecka 1983); and open coastal carbonates from
the uppermost Riphean Hunnberg and Rysso formations of Nordaustlandet, Svalbard, contain
microfossils preserved by early diagenetic silicification (Knoll 1984; Knoll and Calder 1983). The
assemblages in these cherts are comparable to contemporaneous ‘shale’ biotas and demonstrate that,
if present in silicified carbonates, large acritarchs such as Trachysphaeridium , Kildinosphaera ,
Chuaria , or Trachyhystrichosphaera species can be recognized and identified. Silicified carbonates
from lagoonal and intertidal facies of the same formations do not contain abundant large acritarchs.

Several assemblages from lagoonal shales of the Kingbreen (samples F2-27, F2-29) and
Glasgowbreen (F-712) formations of the Veteranen Group exemplify the converse distribution.
Here, colonies of small spheroidal unicells, filaments, and rods — the stuff of most ‘cherty’
microbiotas— constitute the fossil populations. This demonstrates that early diagenetic silicification
is not a prerequisite for the preservation of delicate prokaryotic remains, and it provides a second
opportunity to compare assemblages from similar physical environments preserved in different ways.
In the relative abundance and size frequency distribution of unicell and filament populations (text-
figs. 4, 6), the Veteranen lagoonal assemblages closely resemble those from lagoonal silicified
carbonates of the slightly younger Draken and Rysso formations of Svalbard, as well as the subtidal
associations of the approximately contemporaneous Bitter Springs Formation, Australia (Schopf
1968; Knoll 1981). They do not closely resemble microbiotas from open coastal shales or intertidal to
supratidal silicified carbonates.

Similar preservation of delicate ‘cherty’ microfossils in shales has been reported from the
approximately 1300 Ma old Roper Group of northern Australia (Peat et al. 1978) and the Upper
Riphean of the southern Urals (Jankauskas 1982). Thin-section examination of microbiotas

2 4 6 8 10 (pm)

CROSS- SECTIONAL DIAMETER

text-fig. 6. Size frequency distribution of filamentous sheaths from samples F-7 1 2,
Glasgowbreen Formation (a) and F2-27, Kingbreen Formation (b). For both a and

b, N = 200.

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

461

preserved in shales has been successful in a few instances (Moorman 1 974; Peat et al. 1 978; Horodyski
1980, el a!. 1980; Knoll et al. 1981;Chauvel and Mansuy 1981) but most siliciclastic microfloras are
best observed in maceration residues. A principal disadvantage of this procedure is that it loses the
important information of spatial distributions of populations within rock. Such data are valuable in
both systematic and palaeoecological studies (Knoll and Golubic 1979; Knoll 1981, 19826). Of
particular significance to the present discussion is the fact that when populations can be defined
spatially their size frequency distribution, degradational variability, and (sometimes) cell division
sequences can be determined. Comparisons of small unicell populations from the Draken,
Hunnberg, Rysso, and Bitter Springs formations with those from the Veteranen Group and other
siliciclastic units suggests that, although different names have been applied to permineralized and
compressed populations, they are very similar in their morphological modes and ranges of variation.
The same is true of filament and rod populations. In these cases, parallel taxonomic schemes hinder
biological comparisons, and some systematic revision is necessary. However, drawing a lesson from
analogous comparative studies of Carboniferous coal ball and compression floras, any revisions
must be undertaken on a case by case basis using all available fossil materials. Simplistic taxonomic
solutions will obscure biological differences as surely as the present systems hide similarities.

In summary, most Proterozoic open coastal and shelf microbiotas are preserved as compressions
or compactions in siltstones and shales, while most late Precambrian petrifactions come from
restricted coastal environments where early diagenetic silicification was most likely. Many algae that
were thus precluded from restricted lagoonal and intertidal habitats are found as acritarchs in shales,
while some mat building micro-organisms (e.g. Eoentophvsalis Hofmann, 1976) characteristic of
intertidal zones are preserved only or predominantly in silicified carbonates. Because the correspon-
dence of environments and preservational modes is not absolute, silicified shelf biotas and com-
pressed or compacted lagoonal assemblages allow hypotheses of differential preservational effects to
be tested. The limited observations available to date suggest that the most serious problems of
taxonomic obfuscation concern populations of small unicells and, to a lesser extent, filaments.

A geophysical aside

Although taxa differ in the rates at which their remains change colour with increasing temperature, colour
characterization of fossil pollen, spores, and algal cysts is useful in studies of organic thermal maturation
(Gutjahr 1966; Staplin 1969; Hunt 1979). Acritarchs in the Hecla Hoek succession of north-eastern Spitsbergen
vary systematically from amber in Cambrian shales to an opaque black in the oldest parts of the Veteranen
Group. Within the Veteranen succession exposed at Faksevagen, colours range from a slightly reddish Moderate
Brown (Kelly and Judd 1976) in sample F-2200 from the upper Oxfordbreen Formation to a dark Brownish
Black to Black in the lowermost Kingbreen and Kortbreen formations. In so far as the Faksevagen section
comprises the limb of an open folded anticline without apparent intrusions in the immediate vicinity, one can
hypothesize that observed colour changes are primarily due to maximum burial depth achieved in the
Ordovician Period, just prior to Caledonian deformation. Oxfordbreen sample F-2200 sits beneath some 4300 m
of preserved late Precambrian and Cambro-Ordovician strata. Assuming that any additional (now eroded)
thickness of Upper Ordovician and Silurian strata was minimal, and assigning a maximum temperature to this
rock of 125-150°C based on the organic maturation index (Hunt 1979, p. 324), one can estimate an early
Palaeozoic geothermal gradient of about 1 °C per 29 35 m. Estimates based on the occurrence of Dark Greyish
Brown acritarchs and filaments in the upper Kingbreen Formation (samples F2-17 to 33) and nearly black
materials in the Kortbreen Formation (F2-12, and slightly lighter in F2-40) fall in the same range. This falls
within the normal range for continental geothermal gradients and suggests that at Faksevagen the temperature
history of the Veteranen Group was controlled primarily by burial. At Cavendishryggen and in correlative rocks
on Nordaustlandet other factors, probably related to Caledonian tectonism, intensified the thermal regime.

SYSTEMATIC PALAEONTOLOGY

No new or emended taxa are here proposed and we have limited our systematic discussions to brief
remarks designed to complement existing data and interpretations of the taxa found in the Veteranen Group.
All samples and prepared slides are housed in the Paleobotanical Collections of the Botanical Museum,
Harvard University.

462

PALAEONTOLOGY, VOLUME 28

Kingdom monera Haeckel, 1866

Division cyanophyta (Sachs) Pascher, 1931
Class coccogonae Thuret, 1875
Order pleurocapsales Geitler, 1925
Family unknown

Genus bavlinella (Schepeleva) Vidal, 1976
Type species. Bavlinella faveolata (Schepeleva) Vidal, 1976.

Bavlinella faveolata (Schepeleva) Vidal, 1976
Plate 51, figs. 3, 4

Discussion. This distinctive fossil occurs in small numbers in samples F2-27 and F-856. Multisphere
size is 5- 1 2 pm, and individual microsphere units are less than 1 pm in diameter. Chauvel and Mansuy
( 1981 ) noted that in the Brioverian of Normandy and Brittany, B. faveolata specimens with small unit
cells characterize older ( > 670 Ma) portions of the sequence, while Vendian (640-580 Ma) deposits
additionally contain many specimens having a significantly larger unit cell size. (Chauvel and
Mansuy used the name Sphaerocongregus variabilus Moorman, regarded by Vidal (1976) as a junior
synonym of B. faveolata.) Veteranen specimens are consistent with these unit cell size observations.

B. faveolata reached its acme in the Vendian, when it apparently expanded opportunistically with
the climatic restriction and extinction of other previously dominant taxa. Its complete stratigraphic
range runs at least from the late Riphean (this paper) to the early Cambrian (Vidal 1981 ). Moorman
( 1 974) and Knoll et al. (1981) compared B. faveolata with pleurocapsalean cyanobacteria, noting the
close correspondence in size frequency distribution, unit cell or baeocyte size frequency, and multiple
fission pattern of reproduction (see Waterbury and Stanier 1978). Mansuy and Vidal (1983)
suggested a chroococcalean origin for B. faveolata, comparing the fossil populations with species of
the colonial chroococcalean genera Gomphosphaeria, Coelosphaerium , and Microcystis. Although
most species of these genera differ from Bavlinella in colony size and shape, colony architecture, unit
cell size frequency distribution, or unit cell shape (see Geitler 1932; Desikachary 1959), some species
of Microcystis do form tightly packed spheroidal colonies. Ecologically, Bavlinella does resemble
some modern colonial chroococcaleans in its inferred planktic mode of life and tendency to bloom
under eutrophic conditions (Knoll et al. 1981; Mansuy and Vidal 1983).

The critical data bearing on the affinities of B. faveolata concern patterns of cell division.
Pleurocapsalean multispheres and tightly packed chroococcalean colonies may have similar
morphologies, but the development of these morphologies occurs quite differently in the two orders.

explanation of plate 51

For each figure, slide number (which includes a sample number from text-fig. 3), stage co-ordinates (where ‘x’ on
slide F500-14 = 23-7 x 102-2) and Harvard University Paleobotanical Collection number are given. Bar in
fig. 2 = 40 pm for figs. 1, 2, and = 10 pm for all other figures.

Figs. 1, 2. 'IChuaria circularis Walcott. Fragmentary remains; note small unicell (comparable in size to fig. 18) in
fig. 1. 1.F575-5, 52-2 x 107-8, 60754. 2, F575-5, 45-5 x 95-6, 60755.

Figs. 3, 4. Bavlinella faveolata (Schepeleva) Vidal. 3, F2-29-2, 55 x 96-5, 60756. 4. F2-29-2, 33-3 x 102-3, 60757.

Figs. 5, 6. 9, 10. Eosynechococcus sp. 5, F500-13, 54-9 x 106-2, 60758. 6, F712-4, 42-4 x 1 1 0-4, 60759. 9, F712-4,
13 x 109-6, 60760; note second, small specimen at the upper right corner of the figure. 10, F712-6, 33-4 x 107,
60761.

Fig. 7. A coccoidal unicell distorted by diagenetic crystal growth. Similar morphologies have been described as
Octoedryxium spp., but genuine O. truncation specimens (e.g. Vidal 1976) do not resemble tins specimen.

Figs. 8, 1 1 -18. Small coccoidal unicells. Note small blebs of internal organic matter in 11, 17, and 18. 8, ¥2-21-2,
29-4x113-7,60762. 1 1, F2-23-1, 30-6 x 101-7, 60763. 12, F856-1, 38-9 x 108-5, 60764. 13, F500-13,

49-8 x 108-8, 60765. 14, F575-5, 38-2 x I 14-9, 60796. 15, F712-6, 52-8 x 105-5, 60766. 16, F575-6, 39 x 107-6,
60767. 17, F575-5, 50-5 x 114-3, 60768. 18, F575-7, 44 x 106-1, 60769.

Fig. 19. Filamentous sheath containing degraded remnants of cells (arrows). F500-14, 45-2 x 96, 60770.

PLATE 51

KNOLL and SWETT, Late Proterozoic microfossils

464

PALAEONTOLOGY, VOLUME 28

In modern pleurocapsalean cyanobacteria, baeocyte clusters are formed by repeated binary fissions
of a large initial cell, without intervening cell growth (Waterbury and Stanier 1978). In Microcystis
and related genera, colonies arise by cell divisions interspersed with growth, so that at no point in the
life cycle are any individual cells significantly larger than the cells that are found in the multicellular
colony. Thus, if B. faveolata has pleurocapsalean affinities, its multispheres should co-occur with
unicells, dyads, and tetrads having the same approximate total volume. If Bavlinella is a
chroococcalean blue-green, no large cells should occur in the population and colony size might be
expected to be somewhat variable. In samples processed by maceration, it is risky to infer that unicells
are part of the life cycle of multispheres in the same assemblage, especially when the multispheres
themselves are as rare as they are in the Veteranen Group. Bavlinella- rich assemblages from the
Hector Formation, Alberta (Moorman 1974), the Mineral Fork Formation, Utah (Knoll et al. 1981),
and the Brioverian of France (Chauvel and Schopf 1978; Chauvel and Mansuy 1981; Mansuy 1983)
have been studied in thin-section; in at least two of these assemblages (Hector and Mineral Fork),
Bavlinella multispheres occur in intimate spatial association with unicells and rare spheroidal dyads
and tetrads whose maximum diameters are about the same size as the B. faveolata multispheres. It can
be argued correctly that close spatial association of planktonic populations need not imply
taxonomic identity. On the other hand the cell morphologies predicted by the ‘pleurocapsalean
hypothesis’ do occur, while micron-sized unicells and variable colony sizes that might represent a
chroococcalean growth series have not been observed. Therefore, pending further morphological
data, we prefer to keep this fossil among the Pleurocapsales.

Order chroococcales Wettstein, 1924
Family chroococcaceae Nageli, 1849
Genus eosynechococcus Hofmann, 1976

Type species. Eosynechococcus moorei Hofmann, 1976.

Eosynechococcus sp.

Plate 51, figs. 5, 6, 9, 10

Description. Isolated rod-shaped vesicles; 2-9 /xin long and 1-3 /xm wide (mean dimensions = 3-6 x 1-8 /xm,
s x — 1 ^m, s y = 0-5 /xm, N = 50).

Discussion. Eosynechococcus is a form genus for small rod-shaped microfossils, most of which are
probably, but not demonstrably, cyanobacteria (see Knoll 1982a). Although many previously
described Eosynechococcus populations occur as dense aggregations in stromatolitic laminae,
occurrences of non-clustered solitary rods and dyads have been described (Hofmann 1976; Knoll
1982a; Jankauskas 1982; Strother et al. 1983), and some of these are known to occur in non-
stromatoliticmuds. The generic diagnosis of Eosynechococcus Hofmann, 1976 specifically states that
individuals may be solitary or clustered, so there is no question that this generic name is appropriate
for the Veteranen fossils. In its size frequency distribution, the Veteranen population resembles
E. moorei Hofmann, 1976 and E. brevis Knoll, 19826, but maceration has destroyed the evidence of
spatial distribution and division patterns that would allow these two species to be distinguished.
Thus, we have elected to use the designation Eosynechococcus sp. That these are fossils and not
modern contaminants can be demonstrated by the colour of their walls, which indicates thermal
alteration under conditions of deep burial.

Class hormogonae Thuret, 1875

Order oscillatoriales Copeland, 1936 or nostocales Geitler, 1925
Filamentous microfossils
Plate 51, fig. 19; Plate 52, figs. 1-3, 12

Discussion. Filamentous microfossils are abundant constituents of several Veteranen samples,

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

465

particularly those from the Bogen Member of Kingbreen Formation. Specimens are uniformly
non-septate and are interpreted as extracellular sheaths; however, internal patterns of thickening
and attenuation often faithfully record the dimensions of trichome cells that originally occupied
sheath interiors (PI. 52, fig. 2). These ‘ghosts’ suggest that the micro-organisms had a single
trichome composed of undifferentiated cells, much like modern Lyngbya. In rare instances, partially
degraded trichome fragments are preserved inside the sheaths (PI. 51, fig. 19). Size frequency
distributions (text-fig. 6) indicate that several taxa are represented. Like silicified microbiotas
from the Bitter Springs (Schopf 1968; Knoll 1981), Draken (Knoll 19826), and Sukhaya Tunguska
(Mendelson and Schopf 1982) formations, the Veteranen populations fall into several size classes
having modes in the 1, 2-4, 4-8, and 10-12 fim diameter ranges. The widest sheath observed is
16 ftm in diameter.

In terms of permineralization taxonomic practice, these populations can be described as species of
Tenuofilum , Eomycetopsis , and Siphonophycus , with the generic distinctions based largely on size.
Macerated filaments have been assigned to these genera (e.g. Jankauskas 1982) or to the form genus
Taeniatum. Biologically, most filamentous sheath populations were probably produced by oscil-
latorialean or nostocalean cyanobacteria. The absence of well-preserved trichomes precludes further
comparison, although ‘ghosts' of cells impressed on sheath interiors do indicate an oscillatorian
affinity for many specimens.

Kingdom protoctista Copeland, 1956 emend. Margulis, 1971
Division prasinophyta Round, 1971
Order pterospermatales Schiller, 1925
Family tasmanitaceae Sommer, 1956 ex. Tappan, 1980
Genus tasmanites Newton, 1875

Type species. Tasmanites pimctatus Newton, 1875.

Tasmanites rifejicus Jankauskas, 1978
Plate 53, fig. 1 1

Discussion. Specimens of T. rifejicus are rare in the Veteranen Group, but they are easily
distinguished by the numerous pores that perforate the vesicle. One Veteranen specimen (112 /am)
falls in the size range for the species observed by Jankauskas (1978, 1982) and Vidal and Ford (1985),
but others range from 43 to 52 /xm. Despite their smaller size, these specimens have been placed in
T. rifejicus by virtue of their wall structure.

Group acritarcha Evitt, 1963
Genus ?chuaria Walcott, 1899

Type species. Chuaria circularis Walcott, 1899.

1 Chuaria circularis Walcott, 1899
Plate 51, figs. 1 , 2

Discussion. No complete specimens of C. circularis have been observed in Veteranen material, but
large fragments of robust, and often differentially coalified, spheroidal vesicles are common in some
samples. Individual fragments range up to 140 in maximum dimension, and the curvature of these
broken specimens indicates that the original vesicles were often in excess of 300 ^m diameter. Several
species of Trachysphaeridium and Kildinosphaera exceed 200 /un in diameter but, given the robust
nature of the walls and the commonness of the fragments, we suggest that the large shards belong to
C. circularis. Chuaria is known to occur in correlative beds in Nordaustlandet, Svalbard (Knoll
1 982n), and East Greenland (Vidal 1 979), and indeed is common in Upper Riphean rocks from many
localities throughout the world (Ford and Breed 1973; Hofmann 1977).

466 PALAEONTOLOGY, VOLUME 28

Genus favososphaeridium Timofeev, 1959, ex Timofeev, 1966
Type species. Favososphaeridium scandicum Timofeev, 1966.

Favososphaeridium sp.

Plate 53, figs. 7, 10

Discussion. A single specimen 95 /am in diameter was observed in sample F-575. The size and
irregularly reticulate surface ornamentation of this specimen (PI. 53, fig. 10) are comparable to
those characterizing the specimens of Timofeev described by Vidal (1976) from the Visingso Beds,
Sweden. Poor preservation makes specific comparison impossible.

Genus kildinosphaera Vidal, 1983
Type species. Kildinosphaera chagrinata Vidal, 1983.

Kildinosphaera chagrinata Vidal, 1983
Plate 53, figs. 1 -3

Discussion. Vidal (in Vidal and Siedlecka 1983) cut a Gordian knot in Proterozoic acritarch
taxonomy by creating the new genus Kildinosphaera and describing as its type species K. chagrinata.
The new genus was necessitated by Lindgren’s (1982) transfer of Kildinella hyperboreica (the type
species of Kildinella) to Leiosphaeridia Eisenack, and the species was described to impart rigor to a
troubled system of morphologically overlapping form species, most prominently K. sinica Timofeev,
1966. The Veteranen specimens have flexible, easily foldable walls with a subdued chagrinate surface.
Size range is 26-78 pm (x = 40 /am, s x = 10-6 /am, N = 50).

Kildinosphaera gramdata Vidal, 1983
Plate 53, figs. 9, 12-14

Discussion. Veteranen specimens of K. gramdata have flexible walls with a conspicuous, finely
granulate surface texture. Size range is 30-85 pm (x = 52-2 /am, s x = 15-2 /am, N = 20). This species is
common only in the most diverse, open coastal assemblage, sample F-856 and, less so, F-575.

Genus leiosphaeridia Eisenack, 1958
Type species. Leiosphaeridia baltica Eisenack, 1958.

Leiosphaeridia asperata (Naumova) Lindgren, 1982
Plate 52, figs. 4-8, 10, 1 1

EXPLANATION OF PLATE 52

For each figure, slide number (which includes a sample number from text-fig. 3), stage co-ordinates (where V on
slide F500-14 = 23-7 x 102-2) and Harvard University Paleobotanical Collection number are given. Bar in
fig. 12 = 10 /am for figs. 1-3, 12, and = 20 /am for all other figures.

Figs. 1-3, 12. Filamentous microfossils. 1, F500-12, 43-1 x 103-4, 60771. 2, F2-17-2, 48-5 x116-5, 60772.
3, F2-29-2, 40-2 x 104-3, 60773. 12, F712-2, 42 x 1 12-4, 60774.

Figs. 4 8, 10, 11. Leiosphaeridia asperata (Naumova) Lindgren. 4, F712-6, 21 x 107-2, 60775. 5, F500-12,
49 x 95-5, 60776. 6, F500-6, 50-5 x 95-7, 60777. 7, F500-12, 48 x 90-1, 60779. 8, a cluster of vesicles, F2200-I,
47-7 x 105-3, 60780. 10, F500-10, 31-4 x 109-1, 60781. 1 1, F712-6, 34 x 100-3, 60795.

Fig. 9. Synsphaeridium sp. F575-5, 58-4 x 1 10-6, 60782.

PLATE 52

KNOLL and SWETT, Late Proterozoic microfossils

468

PALAEONTOLOGY, VOLUME 28

Discussion. Lindgren (1982) synonymized the well-known Proterozoic acritarch species Kildinella
hyperboreica Timofeev, 1 966 with L. asperata. Veteranen L. asperata specimens have smooth, flexible
walls 20-52 pm in diameter (x = 32-9 ^m, s x — 7-7 pm. N = 52).

Genus satka Jankauskas, 1979a
Type species. Satka favosa Jankauskas, 1979a.

Satka colonialica Jankauskas, 19796
Plate 53, figs. 4-6, 8

Discussion. Vidal and Ford (1985) have interpreted S. colonialica as a thin, chagrinate to finely
granular envelope whose outline reflects the dimensions of the numerous, smaller ellipsoidal cells
about which it was compressed. Veteranen specimens support this interpretation; both in the
Spitsbergen population and in previously described colonies, the ellipsoidal internal bodies are
generally absent. Individual S. colonialica specimens reach 100 pm in length, but most fall in the
40-50 jum range. Ellipsoidal internal bodies are 7-14 pm long.

Genus stictosphaeridium Timofeev, 1962
Type species. Stictosphaeridium podolense Timofeev, 1962.

cf. Stictosphaeridium sp. sensu Vidal, 1976

Discussion. Very thin-walled, spheroidal vesicles with diameters of 25-63 pm are common in several
Veteranen samples.

Genus synsphaeridium Eisenack, 1965
Type species. Synsphaeridium gotlandicum Eisenack, 1965.

Synsphaeridium sp.

Plate 52, fig. 9

Discussion. Microfossils assigned to Synsphaeridium differ from other small unicells of the Veteranen
Group in their possession of a thick (but not brittle) psilate wall that compresses to yield rounded
folds rather than the sharp, pleated folds seen in many other taxa. Specimens have diameters of
13-17 pm and commonly occur in clusters.

explanation of plate 53

For each figure, slide number (which includes a sample number from text-fig. 3), stage co-ordinates (where ‘x’ on
slide F500-14 = 23-7 x 102-2) and Harvard University Paleobotanical Collection number are given. Bar in
fig. 14 = 20 /im for figs. 1 -6, 8-14, and = 60 pm in fig. 7.

Figs. 1-3. Kildinosphaera chagrinata Vidal. 1 , F575-7, 34 x 95-8, 60783. 2, F500-10, 57 x 95-2, 60784. 3,F575-5,
44-2 x 114-9, 60785.

Figs. 4-6, 8. Satka colonialica Jankauskas. 4, F500-12, 42-3 x 104, 60786. 5, F500-10, 57x95-2, 60787.

6, F575-5, 44-2 x 1 14-9, 60788. 8, F500-12, 45-5 x 102, 60789.

Figs. 7, 10. Favososphaeridium sp. 7, F575-1, 42-4 x 109-7, 60790. Fig. 10 is an enlargement of a portion of fig. 7
showing the structure of the vesicle wall.

Figs. 9, 12 14. Kildinosphaera granulata Yida\. 9,F500-10,51 x 1 10-4, 6079 1 . 12, F2200-1, 39-5 x 1 13 1, 60792.

13, same specimen as fig. 12, photographed using interference contrast. 14, F500-10, 58-2 x 97-4, 60793.

Fig. 1 1. Tasmanites rifejicus Jankauskas. F575-7, 34-3 x 110-1, 60794.

PLATE 53

KNOLL and SWETT, Late Proterozoic microfossils

470

PALAEONTOLOGY, VOLUME 28

MICRO-ORGANISMS INCERTAE SEDIS

Small coccoidal unicells

Plate 51, figs. 8, 1118

Discussion. Numerous taxonomic names have been proposed for the classification of the small
(3-25 ,u m), psilate to granular, spheroidal vesicles that occur as common constituents of late
Proterozoic microbiotas (Timofeev 1966, 1969). If found as permineralizations, most of these popu-
lations would be assigned to the genus Myxococcoides Schopf, especially M. minor Schopf, 1968 or
M. cantabrigiensis Knoll, 1982a. Because maceration destroys original spatial relationships and often
scatters individuals from loose clusters, some of the characters that are important in the recognition
of Myxococcoides species are lost in the acid-resistant residues herein under consideration. Jankauskas
(1982) treated comparable materials from the Upper Riphean of the southern Urals as species of the
genera Synsphaeridium , Leiosphaeridia , Margominiscula , Arctacellularia, Leiominuscula , and
unnamed spheroids. Pending the opportunity to examine type specimens in the Soviet Union, we
prefer to treat these populations informally. Size frequency distributions of two sample populations
are shown in text-fig. 4.

Acknowledgements. We thank the Norsk Polarinstitutt and the Cambridge Spitsbergen Expedition for logistical
and intellectual co-operation, E. Burkhardt and S. Goldberg for assistance in preparing the plates and text-
figures, and G. Vidal and W. B. Harland for helpful criticisms of our manuscript. This research was supported
in part by NSF Grants DPP 80-19998 and DPP 83-01226.

REFERENCES

armin, r. a. and mayer, l. 1983. Subsidence analysis of the Cordilleran miogeocline: implications for the timing
of late Proterozoic rifting and amount of extension. Geology , Boulder, Colo. 11, 702-705.
beckinsale, r. d., reading, h. g. and rex, d. c. 1975. Potassium-argon ages for basic dykes from East Finnmark:
stratigraphical and structural implications. Scott. J. Geol. 12, 51-65.
bond, G. c., rominz, m. a. and devlin, w. j. 1983. Thermal subsidence and eustasy in the Lower Paleozoic
miogeocline of western North America. Nature, Lond. 306, 775-779.
chaudhliri, s. G. and hanson, w. r. 1980. Rb-Sr ages of the Uinta Mountain Group of Utah and Colorado.
Abstr. Progm geol. Soc. Am. 12 (6), 269.

chauvel, j. j. and mansuy, c. 1981. Micropaleontologie du Proterozoique du Massif Armoricain (France).
Precamb. Res. 15, 25-42.

— and schopf, j. w. 1978. Late Precambrian microfossils from Brioverian cherts and limestones of Brittany
and Normandy, France. Nature, Lond. 275, 640-642.

COOK, T. D. and bally, a. w. (eds.). 1975. Stratigraphic Atlas of North and Central America, 271 pp. Princeton
University Press, Princeton.

copeland, j. l. 1936. Yellowstone thermal Myxophyceae. Ann. N.Y. Acad. Sci. 36, 1-232.

Crittenden, M. d. and peterman, z. e. 1975. Provisional Rb/Sr age of the Precambrian Uinta Mountain Group,
northeastern Utah. Utah Geol. 2, 75-77 .
desikachary, t. v. 1959. Cyanophyta, 686 pp. New Delhi.

diver, w. l. and peat, c. j. 1979. On the interpretation and classification of Precambrian organic-walled
microfossils. Geology, Boulder, Colo. 7, 401-404.

eisenack, a. 1958. Tasmanites Newton 1875 und Leiosphaeridia n.g. als der Hystrichosphaeridea. Palaeonto-
graphica. A, 110, 1-19.

1965. Microfossilen aus dem Silur Gotlands. Hystrichosphiiien, Problematika. Neues Jb. Geol. Palaont.
Abh. 122,257-274.

elston, d. p. and mckee, e. h. 1982. Age and correlation of the Late Proterozoic Grand Canyon Orogeny,
Northern Arizona. Bull. geol. Soc. Am. 93, 681 699.

evitt, w. r. 1963. A discussion and proposals concerning fossil dinoflagellates, hystrichospheres, and acrit-
archs, II. Proc. natn. Acad. Sci. U.S.A. 49, 298-302.

fairchild, i. j. and hambrey, m. j. 1985. The Vendian succession of northeastern Spitsbergen: petrogenesis of a
dolomite-tillite association. Precamb. Res. 26, 1 1 1-167.
ford, T. D. and breed, w. j. 1973. The problematical Precambrian fossil Chuaria. Palaeontology, 16, 535-550.

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

471

GEITLER, L. 1925. Synoptische Darstellung der Cyanophyceen in morphologischer und systematischer Hinsicht.
Beih. bot. Zbl. 41 , 163.

— 1932. Cyanophyceae. In kolkwitz, r. (ed.). Kryptogamenflora von Deutschland, Osterreich und der
Schweiz, 14 , 1 196 pp. Akademische Verlagsgesellschaft, Jena.

Golovanov, n. p. 1967. Stromatolites from the Riphean deposits of Murchisonfjorden, Nordaustlandet. In
krasil'shchikov, a. a. (ed.). Documents on the stratigraphy of Spitsbergen, 6-20. Trudy NIIGA, Leningrad,
-and raaben, M. E. 1967. Upper Riphean analogues in the Spitsbergen Archipelago. Dokl. Akad. Nauk
SSSR, 173 , 1141-1144. [In Russian !

gray, G. and boucot, a. J. 1975. Color changes in pollen and spores: a review. Bud. geol. Soc. Am. 86, 1019 1033.
GUTJAHR, c. c. M. 1966. Carbonization measurements of pollen-grains and spores and their application. Leids.
geol. Meded. 38 , 1 -29.

haeckel, E. H. 1866. GenereUe Morphologie , 2 vols., 1036 pp. Georg Reimer Verlag, Berlin.
hambrey, m. j. 1982. Late Precambrian diamictites of northeastern Svalbard. Geol. Mag 1 19 , 527-551.
harland, w. b. 1959. The Caledonian sequence in Ny Friesland, Spitsbergen. Q. Jl geol. Soc. Loud. 114 ,
307-342.

— wallis, r. h. and gayer, r. a. 1966. A revision of the Lower Hecla Hoek succession in Ny Friesland.
Spitsbergen. Geol. Mag. 93 , 265-286.

— and wright, n. j. r. 1979. Alternative hypothesis for the pre-Carboniferous evolution of Svalbard. Skr.
norsk Polarinst. 167 , 89 117.

hofmann, h. J. 1976. Precambrian microflora, Belcher Islands, Canada: significance and systematics. ./. Paleont.
50 , 1043-1073.

— 1977. The problematic fossil Chuaria from the late Precambrian Uinta Mountain Group, Utah. Precamb.
Res. 4, 1 11.

horodyski, R. J. 1980. Middle Proterozoic shale-facies microbiota from the lower Belt Supergroup, Little Belt
Mountains, Montana. J. Paleont. 54 , 649-663.

— donaldson, j. a. and kerans, c. 1980. A new shale-facies microbiota from the Middle Proterozoic Dismal
Lakes Group, District of Mackenzie, Northwest Territories, Canada. Can. J. Earth Sci. 17 , 1 166-1 173.
hunt, j. m. 1979. Petroleum geochemistry and geology, 617 pp. W. H. Freeman, San Francisco.
jankauskas, t. v. 1978. Plant microfossils from the Riphean of the southern Urals. Dokl. Acad. Sci. USSR,
Earth Sci. Section, 242 , 913-915. [In Russian.]

— 1 979c/. Lower Riphean microfossils of the southern Urals. Ibid. 247, 1465-1467. [In Russian ]

— 19796. Middle Riphean microbiota from the southern Urals and the Bashkirian Urals. Ibid. 248, 190 193.
[In Russian.]

— 1982. Microfossils from the Riphean of the Southern Urals. In keller, b. m. (ed.). Stratotype of Riphean:
palaeontology and palaeomagnetics, 84-120. Acad. Sci. USSR, Geol. Inst., Nauka, Moscow. [In Russian ]

keller, b. m. (ed.). 1982. Stratotype of Riphean: palaeontology and palaeomagnetism, 175 pp. Acad. Sci. USSR,
Geol. Inst., Nauka, Moscow. [In Russian.]

keller, f. 1979. Stratigraphy and structure of the Ocoee Supergroup in Eastern Tennessee, 320 pp. Ph.D. thesis
(unpublished), Yale University.

kelly, k. l. and Judd, d. b. 1976. Color: universal language and dictionary of names. U.S. Nat. Bur. Standards,
Spec. Pub!. 440 , 184 pp.

knoll, a. h. 1981. Paleoecology of late Precambrian microbial assemblages. In niklas, k. (ed.). Paleobotany,
paleoecology, and evolution, 1 , 17-54. Praeger, New York.

— 1982u. Microorganisms from the late Precambrian Draken Conglomerate, Ny Friesland, Spitsbergen. J.
Paleont. 56 , 755-790.

— 19826. Microfossil based biostratigraphy of the Precambrian Hecla Hoek sequence of Nordaustlandet,
Svalbard. Geol. Mag. 119 , 269-279.

1984. Microbiotas of the Late Precambrian Hunnberg Formation, Nordaustlandet, Svalbard. ./. Paleont.
58 , 131 162.

— buck, n. and awramik, s. m. 1981 . Stratigraphic and ecologic implications of late Precambrian microfossils
from Utah. Am. J. Sci. 281 , 247-263.

— and calder, s. 1983. Microbiotas of the late Precambrian Rysso Formation, Nordaustlandet, Svalbard.
Palaeontology, 26 , 467-493.

— and golubic, s. 1979. Anatomy and taphonomy of a Precambrian algal stromatolite. Precamb. Res. 10 ,
115- 151.

lindgren, s. 1981. Remarks on the taxonomy, botanical affinities, and distribution of leiospheres. Stockh.
Contr. Geol. 38, 1-20.

472 PALAEONTOLOGY, VOLUME 28

lindgren, s. 1982. Algal coenobia and leiospheres from the Upper Riphean of the Turukhansk region, eastern
Siberia. Ibid. 37-45.

mckenzie, d. 1978. Some remarks on the development of sedimentary basins. Earth planet. Sci. Lett. 40 , 25-32.
mansuy, c. 1983. Les Microspheres de Proterozoique Superieur Armoricain (Brioverien): nature, repartition
stratigraphique, affinites biologiques, 76 pp. These (unpublished), Rennes.

— and vidal, G. 1983. Late Proterozoic Brioverian microfossils from France: taxonomic affinity and
implications of plankton productivity. Nature , Load. 302 , 606-607.
margulis, L. 1971. Whittaker’s five kingdoms: minor modifications based on considerations of the origins of
mitosis. Evolution , 25 , 242-245.

mendelson, c. v. and schopf, j. w. 1982. Proterozoic microfossils from the Sukhaya Tunguska, Shorikha, and
Yudoma formations of the Siberian Platform, USSR. J. Paleont, 56 , 42-83.
milstein, v. e. and golovanov, n. p. 1979. Upper Precambrian microphytolites and stromatolites from
Svalbard. Skr. norsk Polarinst. 167 , 219-224.

moorman, m. 1974. Microbiota of the Late Proterozoic Hector Formation, southwestern Alberta, Canada. J.
Paleont. 48 , 524-539.

nageli, c. 1849. Gattungen einzelliger Algen, 139 pp. Friedrich Schutthess, Zurich.
newton, e. T. 1875. On ‘Tasmanite’ and Australian ‘White Coal’. Geol. Mag. (n.s.), 12 , 337-342.
pascher, a. 1931. Systematische Ubersicht fiber die mit Flagellaten in Zusammenhang stehenden Algenreihen
und Versuch einer Einreiheng dieser Algenstamme in die Stamme des Pflanzenreiches. Beih. hot. Zhl. 175 ,
417-428.

peat, c. J., muir, m. d., plumb, K. a., mckirdy, d. m. and norvick, m. s. 1978. Proterozoic microfossils from the
Roper Group, Northern Territory, Australia. BMR J. Aust. Geol. Geophys. 3 , 1-17.
raaben, m. e. and zabrodin, v. e. 1972. Algal problematics of the Upper Riphean (stromatolites, oncolites).

Trudy Inst. Geol. Acad. Sci. USSR , 217 , 130 pp.
round, f. e. 1971. The taxonomy of the Chlorophyta II. Br. phycol. J. 6, 235-264.

schiller, j. 1925. Die planktontischen vegetation des adriatischen meers: B. Chrysomonadina, Heterokontae,
Cryptomonadina. Arch. Protistenk. 53 , 59-123.

schopf, j. w. 1968. Microflora of the Bitter Springs Formation, late Precambrian, central Australia. J. Paleont.
42,651-688.

staplin, f. l. 1969. Sedimentary organic matter, organic metamorphism, and oil and gas occurrence. Bull. Can.
Petrol. Geol. 17 , 47-66.

strother, p. k., knoll, a. h. and barghoorn, e. s. 1983. Micro-organisms from the Late Precambrian
Narssarssuk Formation, North-West Greenland. Palaeontology, 26 , 1-32.
sturt, b. a., pringle, I. r. and Roberts, d. 1975. Caledonian nappe sequence of Finnmark, northern Norway,
and the timing of the orogenic deformation and metamorphism. Bull. geol. Soc. Am. 86, 710-718.
swett, k. and knoll, a. h. 1985. Stromatolitic bioherms and microphytolites from the late Proterozoic Draken
Conglomerate Formation, Spitsbergen. Precamb. Res. (in press).
tappan, h. 1980. The paleobiology of plant protists, 1028 pp. W. H. Freeman, San Francisco.
thuret, g. 1875. Essai de classification des Nostochinees. Ann. Sci. Nat., 1 (6), 372.

timofeev, b. v. 1959. The ancient flora of the Baltic and its stratigraphic significance. Trudy vses. neft. nauchno-
issled. geol.-razv. Inst. 129 , 320 pp. [In Russian.]

-1962. The Theodolite Stage in palaeontology. Ibid. 196 , 601-647. [In Russian.]

1966. Microphytological investigations of ancient formations, 145 pp. Acad. Sci. USSR, Lab. Precamb.
Geol., Nauka, Leningrad. [In Russian.]

1969. Proterozoic sphaeromorphs, 146 pp. Acad. Sci. USSR, Inst. Precamb. Geol. Geochronol., Nauka,
Leningrad. [In Russian ]

vidal, G. 1976. Late Precambrian microfossils from the Visingso Beds in southern Sweden. Fossils Strata, 9 , 1-56.
1979. Acritarchs from the Upper Proterozoic and Lower Cambrian of East Greenland. Bull Gronlands geol.
Unders. 134 , 1-55.

1981. Micropaleontology and biostratigraphy of the Upper Proterozoic and Lower Cambrian sequence in
East Finmark, northern Norway. Norg. geol. Unders. 362, 1-53.

and ford, t. d. 1985. Planktonic microfossils (acritarchs) from the Upper Proterozoic of North America
and their chronostratigraphic implications. Precamb. Res. (in press).

and knoll, a. h. 1983. Proterozoic plankton. Mem. geol. Soc. Am. 161 , 265-277 .

and siedlecka, A. 1983. Planktonic, acid-resistant microfossils from the Upper Proterozoic strata of the
Barents Sea region of the Varanger Peninsula, East Finmark, northern Norway. Norg. geol. Unders. 382,
45-79.

KNOLL AND SWETT: LATE PROTEROZOIC MICROFOSSILS

473

walcott, c. D. 1899. Pre-Cambrian fossiliferous formations. Bull. geol. Soc. Am. 10 , 199-244.
waterbury, j. b. and stanier, r. y. 1978. Patterns of growth and development in Pleurocapsalean cyano-
bacteria. Microbiol. Rev. 42 , 2-44.

wettstein, r. von. 1924. Hancibuch der systematischen Botanik (Third edition), 1017 pp. Wien, Leipzig.
wilson, c. b. 1958. The Lower Middle Hecla Hoek rocks of Ny Friesland, Spitsbergen. Geol. Mag. 94, 305-327.

— 1961. The Upper Middle Hecla Hoek rocks of Ny Friesland, Spitsbergen. Ibid. 98 , 89-1 16.

— and harland, w. b. 1964. The Polarisbreen Series and other evidence of Late Pre-Cambrian Ice Ages in
Spitsbergen. Ibid. 101 , 198-219.

ANDREW H. KNOLL

Department of Organismic and Evolutionary Biology
Harvard University
Cambridge, Massachusetts 02138

U.S.A.

KEENE SWETT

Typescript received 15 May 1984
Revised typescript received 14 January 1985

Department of Geology
University of Iowa
Iowa City, Iowa 52242
U.S.A.

LOWER CRETACEOUS INOCERAMID BIVALVES
FROM THE ANTARCTIC PENINSULA REGION

by J. A. CRAME

Abstract. The occurrence of rich faunas of Lower Cretaceous inoceramid bivalves in the Antarctic Peninsula
region further emphasizes their widespread distribution, and enhances their potential for regional biostrati-
graphic correlations. The Antarctic material is contained in approximately seven of twelve species groups that
are recognized on a worldwide scale. Six of these are assigned to the genus Inoceramus and one to Birostrina. The
comparatively rare genus, Anopaea, is left undivided.

In the Fossil Bluff Formation of Alexander Island, Berriasian representatives of the I. ovatus group (/. cf.
ovatus Stanton and /. sp. aff. ellioti Gabb) are succeeded by A. trapezoidalis (Thomson and Willey) which has
undifferentiated Berriasian-Aptian affinities. This is in turn followed by an Aptian member of the I. neocomiensis
group (I. deltoides sp. nov.) and in the Albian there are occurrences of Anopaea sp. nov. aff. mandibula
(Mordvilko), BP. cf. concentrica (Parkinson) (B. concentrica gp.), I. cf. anglicus elongatus Pergament, /. sp. aff.
bellvuensis Reeside, I. sp. aff. comancheanus Cragin (all I. anglicus gp.), and I. flemingi sp. nov. (I. liwerowskyae
gp.). Aptian-Albian strata on James Ross Island have yielded both /. stoneleyi sp. nov. (/. liwerowskyae gp.) and
Anopaea sp. nov. /3. These are followed by the Albian species I. cf. sutherlandi M'Coy and /. carsoni M'Coy (both
/. carsoni gp.) and the highest Lower Cretaceous specimens within this sequence have been referred to B.
concentrica (Parkinson).

Although specimens of I. cf. heteropterus Pokhialainen (I. heteropterus gp.) and I. annenkovensis sp. nov.
(unclassified) from Annenkov Island are of probable Hauterivian-Barremian age, it is noticeable that there is a
marked lack of Valanginian-Barremian inoceramids in the Antarctic Peninsula region. This gap probably
reflects a period of regional uplift and non-deposition.

Representatives of the I. ovatus and /. heteropterus groups provide a means of correlation between the
Berriasian-Barremian of the Antarctic Peninsula and the North Pacific region. I. deltoides sp. nov. can be closely
matched with Northern Hemisphere Aptian members of the I. neocomiensis group and I. stoneleyi sp. nov. and
I. flemingi sp. nov. have possible counterparts within the Aptian-Albian of Spitzbergen, south-east USSR and
far eastern USSR. Of the various Albian species groups, that based on I. carsoni provides a direct link between
Antarctica and Australia and those based on I. anglicus and B. concentrica facilitate a range of long-distance
correlations. The latter category, in particular, may be one of the first truly cosmopolitan inoceramid groups.

In comparison with their Upper Cretaceous counterparts, Lower Cretaceous inoceramid bivalves
have received little attention from palaeontologists. It is usually assumed that a few meagre lineages
persisted from the Jurassic-Cretaceous boundary through to the beginning of the Upper Cretaceous,
when a remarkable diversification occurred. In an early paper on the evolution of Inoceramus
through the period. Woods (1912) depicted the entire range of English Upper Cretaceous species as
originating from just two Aptian species. Possible Neocomian (i.e. Berriasian-Valanginian) and
Barremian ancestors were not considered.

We now know that there was in fact a considerable diversity of Lower Cretaceous inoceramids. In
the Neocomian they can be traced around the North Pacific margins from California to Kamchatka
(Pokhialainen 1 974) as well as in Northern Siberia (Zakharov 1 966; Zakharov and T urbina 1 979) and
parts of Europe (Gillet 1924). In the Barremian-Aptian there is a notable development of the I.
neocomiensis group, and in the Albian a variety of species is known from regions such as the North
Pacific (e.g. McLearn 1943; Imlay 1961; Pergament 1965), the Russian Platform (Saveliev 1962),
Western Europe (e.g. Woods 1911), and Australasia (e.g. Day 1969; Stevens and Speden 1978). The
distribution of Lower Cretaceous Inoceramus species would seem to be such as to offer considerable
potential for regional biostratigraphic correlations.

IPalaeontology, Vol. 28, Part 3, 1985, pp. 475-525, pis. 54-61.|

476

PALAEONTOLOGY, VOLUME 28

One further region in which Lower Cretaceous inoceramids are well represented is the Antarctic
Peninsula (text-fig. 1 ). On the south-western flanks of the peninsula, along the inner or eastern margin
of Alexander Island (text-fig. 2), they occur through the greater part of the Fossil Bluff Formation
(?Kimmeridgian-Albian). This unit, which is at least 4 km thick and composed of a variety of fine- to
coarse-grained volcaniclastic sediments, accumulated in a fore-arc environment (Taylor et al. 1979).
On the north-eastern side of the peninsula, within the James Ross Island group (text-fig. 3), further
Cretaceous sediments are encountered that were deposited in a rear-arc setting. Recent fieldwork has
shown the lower stratigraphic units mapped by Bibby ( 1 966) in this area to be lower Cretaceous. They
are approximately 1400 m thick, and are composed of lithologies ranging from bioturbated siltstones
to breccio-conglomerates, and contain both ammonites and inoceramids. They pass up into a pre-
dominantly finer grained Upper Cretaceous sequence (the Hidden Lake Beds and Snow Hill Island
Series of Bibby, 1966), that may be in excess of 3 km thick. Further fossiliferous Cretaceous sediments

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

477

text-fig. 2. Locality and geological map of the central east coast of Alexander Island. The Fossil Bluff
Formation comprises the Upper Jurassic and Lower Cretaceous strata. Inset a shows the position of the
enlarged area on the east coast of Alexander Island, inset b is an enlargement of localities in the Succession
Cliffs Fossil Bluff area, and inset c of localities in the Adams Nunatak Mt. Lassell Mt. Phoebe area.

478

PALAEONTOLOGY, VOLUME 28

text-fig. 3. Locality and geological map of NW James Ross Island. The inset shows the position of the enlarged

area on the NW coast of James Ross Island.

can be traced through the Scotia arc in both the South Shetland Islands and the South Georgia Island
group. In the former of these areas, early Cretaceous fossils have been collected from a sequence of
fine- to coarse-grained clastic sediments on Byers Peninsula, Livingston Island (text-fig. 1 ) (Smellie et
al. 1980). These sediments are intimately associated with a series of contemporaneous island-arc
volcanic rocks in a setting similar to that described on Annenkov Island, close to South Georgia (text-

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

479

fig. 1) (Pettigrew 1981). Indeed, similar fossiliferous Lower Cretaceous sediments associated with
volcanic rocks can be traced into the regions of southern Patagonia that were once contiguous with
the island-arc marginal basin systems of the Scotia arc (e.g. Tanner et al. 1981; Tanner 1982).

It is the aim of this paper to describe the Lower Cretaceous inoceramids from the Antarctic
Peninsula region and assess their potential for both local and regional biostratigraphic correlations.
Until a formal re-classification has been completed (Crame, in prep.), it will be necessary to divide the
two principal genera ( Inoceramus J. Sowerby, 1814 and Birostrina J. Sowerby, 1821 ) into a series of
informal species groups (Table 1). These groups are partly based on those of previous authors
(Woods 1911, 1912; Gillet 1924; Maury 1936; McLearn 1943; Hayami 1960; Saveliev 1962;
Pergament 1965; Sornay 1966; Pokhialainen 1969, 1974) and are partly new herein. Inoceramus
is subdivided into nine species groups and Birostrina (which is here taken to be a valid genus,
sensu Kauffman 1977, 1978«-c) into three. In Table 1 these twelve categories are arranged in
approximate stratigraphic order. The comparatively rare genus, Anopaea Eichwald, 1861, is left
undivided.

Wherever possible, the following measurements were taken:

shell length (L) — the length of the valve as measured along the direction of maximum growth (or growth axis).

shell width (W) — the maximum dimension perpendicular to the length.

apical angle (a) — the angle between the hingeline and the anterior margin.

All the specimens are currently held in the collections of the British Antarctic Survey, Cambridge, UK. In the
systematic descriptions the following abbreviations are used: WS — whole specimen (i.e. bi-valved specimen);
LV — left valve; RV— right valve.

SYSTEMATIC PALAEONTOLOGY

Family inoceramidae Giebel, 1852
Genus inoceramus ,J. Sowerby, 1814

Type species. Inoceramus cuvierii J. Sowerby, 1814 from the ‘Upper Chalk’ of Sussex, England; by subsequent
designation (Cox 1969, p. N315).

Inoceramus ovatus group
Inoceramus cf. ovatus Stanton, 1895

Text-fig. 4a

cf. 1895 Inoceramus ovatus Stanton, p. 47, pi. 4, fig. 15.

cf. 1938 Inoceramus ovatus Stanton; Anderson, p. 99, pi. 4, fig. 9.

cf. 1966 Inoceramus ovatus Stanton; Zakharov, p. 98, pi. 35, fig. 3.

1972 Inoceramus sp. a, Thomson and Willey, p. 9, fig. 7c.

Holotype. Inoceramus ovatus Stanton (1895, p. 47, pi. 4, fig. 15); Paskenta group (Berriasian-Valanginian),
Shasta series, California (see Anderson 1938); by monotypy.

Material. One slightly distorted internal mould (WS) KG.719.15) from approximately the 1675 m level in the
Ablation Valley section, Alexander Island (70° 49' S., 68° 28' W .; text-figs. 2 and 10).

Occurrence. As for material; specimen KG.719.15 occurs in association with a Berriasian Haplophylloceras -
Bochianites ammonite assemblage (Thomson 1979, p. 31); Tithonian-Valanginian (and Hauterivian?) range
established in the Northern Hemisphere.

Description. The left valve (text-fig. 4 a), which is the better preserved of the two, clearly has a sub-
symmetrical pyriform outline; the narrow, pointed umbonal region is curved gently forwards and
there are well-rounded anteroventral, ventral, and posteroventral regions. The length of the valve is
62 mm and width is 45 mm (W/L = 0-73). It is moderately convex, with the maximum inflation
occurring in the umbonal region and centre of the valve. There are steep descents to the antero- and
posterodorsal regions, but much shallower gradients in a ventral direction. Ornament consists of

480

PALAEONTOLOGY, VOLUME 28

5 03
.2 g

B «
.2 8

-G x

a ^

013 -G
O ~
4> 73
013 43

-g •-
o-.tQ
2 73

013 C

'12 G

c °

G OQ

G <s)

•r # 0>

8 O
•— 4>

g °-

G _

- £ I

03 t O
0 0-2
'5b c ,o
o

O £ S

•5 S i>

P- »H
2 'G G
2 «

^ G

o 8

c 8
■2 «

3 u

<43

-o X 43 X 73

3 ^ V) ^ c

^ h 3 oo 5

•-ON o 00 **
"G ,— ‘ -G ’”■' >
S. ^ O C/3 O

^ ON X ^

2 VO ' r- > G

^ - r ;« 4 :

^ £ M

■ • ‘G G

d> N

-2 M3 ^
£ 7t ON

<u oo •'-*
B G

£ G O

^ tJ) G

G

N

G

<5 B

_< Q

On ^ '

013

- £
- Z i

C/5

”2 43
O ro

O ON

PC

Dh 3

013 75

<U X
-s: u-

< o

cj

G to

2 P

a

on a,x
*- OX) On
>n .52

I

X 2

c ON 3
*-« G
G ^

N ©

ON

a g

O 7G
o G
O x

2 3

. o

^ cu

On

NO cn

"» ON VO .

; — on

^? ’g B F-

G O
G 00
,G • *■

O

o > bo

■s 2 §)
«•£ "
8 2 S

•g <u
g 3 3

3 '£ 3

4B 8 M
33"

3 °P SO
"U C
2 C 3

f « »

f ^ 333
■ 3 X>

4? £ «

N

2

H .-

o S

^ 3 2

! oo \>

; no ^ >n

: 2 6 C S

! NO G

§24

sl §1

x; .2 ^

: g g ^

N G
1 • - G 1:3
^ IZN rG — :

1 3 <h o

0\ C NO NO
( — 43 NO ON

OV g ON -

G r-
2

G —

^ on

'll

73 ON

G »— •

hJ c/3

- • - 73
1 §

ON 2 ^
. m 13
73 ^

2 o -

> oo
G ON

O ON

G —

3 £>

G ON

r G no
fc o

■ ^ ON

; g ^

^ e

u

G

E Oh

O ON

G

B

Ui

h ;

73

G

G

2 0

.2 S

13

> ^
G H-*
C/3 G
. , W
on G

^ G
ON 5/j
^ J-
r- ^

2 ®-

G

S CN

X NO
<D ON

?<Z S p!

On (o 3-h

^ g:

(U -HH

W) P-I
T3 . .

G

O

Oh ro

vJ G 22

2 2

j'cf .a

: £ 1
^ G
>,C/3
-2 • -
O NO

55 2

W ^

<^4 JO

So U
' — 1 >

<30 O
OJj C/3
73 .

T3 ^

4) . .

B d

yj 013

-go.

X -G

22 d cn

„ s: no
vo g ^
NO C ^

oo o >

— • Nj <U

o ^

G • . >
3 ^ ^

N 7 T1
G OO g
G

U

= 23
gcS

g M C

y ^ 8
B £ l 2
o ^ q3

U 1-1 IZ3 Oh
.2

Lh ’o G
4) (U >-<

p “ S

| 04j

"a

■ . O «3 0<

O

.>•33
•O & O

VO .SP 2
Ov X3 6
— J. o

C P £

O 73 ^
2 • - ON
G^nO
013 <N on
^ OO PH

^ *G

o X
^ 2

Goo >*(UO G^cn «

o£

N

I 73 ■^f

; c <n

G ON

- &

43 G

o

— <3

u 2:
s c

'

00

•p- G

' 2
G ^
O --G
G ^
15 G

« PC B

C/3 G

n-t

c G 73

o x o
"o • »• o

M CO >
. rn >-

S - < 7:

C3 ON

9^ v>

J o,^

U -J o

■C ^

r\ r7

; ON

G O
O ^

1 OQ

1 cci Co G
>■0 0
i O c «

j e < S

I a . „ u

N >cf G

: c« r 4 O

So£^

C G
•- 4)

-G J

^ 2 C

6-1

H^uSuS

G

O

Oh

2

o

O

a

o o

O G ^
o IS ’C
cG < aj

‘SSfi

G hJ c/3

0 -

G B

2 S
5 K

o 4)

2
6 8

’G B

^ °
o. c

O 'B

Br* g

G
X O

G C
O 4)
. p 73
oo O
6 £

Cd O O O

g S

° G

s ^

G o

B 2

43 4)

S b?

£ G

73 >» G
2 44 3
G G 013
013 4) 4)

§ 5 fc

<u

o

JP 1 03 T3

y > u

3-3 3

O

8 Si

<U ;> ^ ■

3 S >
^ 2 8

8 N la
2 § N
1 II.

Qh cd

c n ■£
in 5

D B

c+h C/3

C/5 QJ

G x

w -S

d

c/3 o r t:
D2 o a

4-. B Z C/3 G

U Qh Ch

H-* ^ O

pH

2 £
oo.S

G >

B 'G

G B
> G
I G

g i

75 ^

•g s

L o

4) ,_J

OQ

43

o

73 S

4) *- H

B d
o

: B £

43

cn —

c n x

4) p^S

‘ G
4)

* s

O 0

o p>

- .G cG Q y,

G G Li M ™

OO Ui

Lh

£ B

Lh G

O co

>73 d

O °3
8 2

a 8
=

|S

G 4)

>.§

-I B
w o
•a B

<L) C3.
a o C
cm — u
c T3 P

>. 33
o 'B

G

43

.S G <

is -

B ■= C

■G G 4) '
C™ •— < r -1

G

£ §
G '

N

N

pG

M --H

O -G

Oh ^ =

a p S ^
= a- 2

■2 .2 o n)

L> ^ Qh fvj

. ^ <3 O
*-■*( O- NJ ~C)

X

^ >-
o 2
so H

^3

S «n

5 S, 33

s 3 ^ §

43 33

<s y
3 •§

X 173
O 03

^ a

'G ^

° n°
CU

s Co
-Si a

U

G

75 Oh

'G ^

5 C/3

r2 D

oo ^

<3 y
o E
4 <

W ss

u 2 * o

4) <2 pG G X
w\Lh X r t i .G

013

U < Jo cu

B G

G B

4)

i- 5

43 43

ft O

n 6
>- J o

oo

— ^ 2
.2 2 -3

C '8 ^
“ O O

C/3 Q. X

UJ

> CCS

o®

■J 8
£
o

hJ

PJ3 O

73 s-
4) d

'G GJ

G 43 73
U 4>

^ O £3

t £ 2 -

§ > g 2
9d>-£

o 7,

4) 43
_ G ^

-dB 8 £

4) B 73 2

43

S J5

t; co
O 43

^ S

,s o
s

"u >>

-Jo
g S.y

•a -3J 2

2 g g

33 > CJ

E o

3 1/1

X 00

73 >

4) ^

*7* -J

2 5 G

u"3 2

e O xs
| £ £

3 u ^
— « J

CS > . .

3 '£ >
8 8 *

-o o

43

G

~ E
a g .g
a ,g

Oo G X

■S X p

§• <b L

2 ^ <o
8

^ a a

Oo

a
oo

a g -U
. »•

W -g t:

8 o

CL, Z

I£

G

G

££
G Lh

•o^ S
^ « o
c > L
3 hJ 2
o . , c

U 73 ■ -

2 o >

43 OC

u £ c

a &|

so g <E
O 33

G Lh
1 — 1 X d-/

2 •- o

r< -2 G
G G

3 013 X

X G 4>

<j£ c
^ u £

O G "G

a g B
^ Gh S

O 03

2 B

43 x

•gS

B °

< cu

CRAME: CRETACEOUS INOCE RAMIE) BIVALVES FROM ANTARCTICA

481

s! c i a'ln ■.

grs s go g

n £ 2 -c

1-1 ^ X> — > cd o

•s .a -35 O o £

^ E o c ^ ^

^ b ci n u

Oj 03 O c£3

w cjfs u g

* £ S I £ z

Uh

G

oj

X

Cl ■
CL

D

G o
G X

*- c :

§ 5

•5 £

, . O'

i- T3 .

<D

X) ^

E D ,

T3 £
£ £
a E

52 3

. Qi

-C c/5

S 3

■_ «
o o

‘ CU

« c c

W ct ~ -s
h So g g
r?« o E

bo<

e

2

X)

* < j*

^ ^ 2
x a a
>, 0-2
3 D g

£ <L> g

) <L> ^ (D

■ bU tP ^ c

; c o Jb 4 5

G h-j i "3? 03

; >- O---* c

; <U ^ c/5 3

^■o g S

1) O- o
G 'O — -J £

^ eg g
•S 2 aU

G

.2
*c/5
03 C
X 2
X >< .

> <L> X

^T3g
§ g c

a. g ’>

< £ -c a

I O G 3

C 3 c

3 _U ^

E £ X

<D —

J-i C/5

° C

ax

b/J

G

>

G PJ

■a|

c/5 3

o'

£ 03
1 ) cj
bfi-~

t — -. C4_ QJ ^

X o x < N

b- °I = s

rrt "c^ I— 3 P

S c u
p .. - U | £2

SfeDaZm 2

i QC
•5 C/5

3 C/5 ‘-j

o D 2

G

O x
: — .

c5 <!

£ 3

t; c

2 g S E

G o3

.3 IS

s: 0 °

^ a — 1 c

G

03

= -^ a u

2 C g a

X

<

u

a a 2 <d

(U G-
O. D-

S£ 0 |

4)

O-

O-

D J

^ < .S J

O-

D

|G oj >)>
'0 > <D G
G o x <y
o x 3 <X
a 53 « °

•— ; O "U

= 2 B *

G bo „ ■ -
$*> X C/5 ^

o w "2 g

2 <o b .>

c/5 G X

60’C 5 >

■S - S3

<U o3 P"

a S « g

03 X £ .5

. r "X 3 >,

u u u ^
G q 73 X
X X G W)
3 g . 513

G

£

o

a •

. , ^
c W

■§>■§

2 §
£ «-
j_ y

O £P

'5 43

C 60

§ E
a S3

~ u

O

X

O

O

£

o x
£ ' c
15 .2

<u

^ .£ T3
O o ^
»- OX) 03
G .£ ^
G X .£

d. £

o3

^ blj

X G

'</5 O
(U o
I-

J- T3
03 o3

"3 2

bij x

S5 J >
G XT K

~

03

J* C
<U X
73 >,

G XP

3 <D

P ^ P

o X
+■» G

-2 c -d*

£P 3 15 ^

3 >, G 3
P

« X 03

_T « <L)

X T3 >-
<3 O P

2 E E
£>

E

G v i-
X X D
73 G OD

U O »_
£ ^ 2
o <lT >
^ c x
xxx

03 t3 _blj

£ o 'iz

C/5

o _P G
'O, ° o
03 ^ O
• •• ^ (— <
.S 2 o

bD u X
i3 ^ >
£ G >

£ o' £

u in £
O — »-.
X

o 2

1 >

O

os a>
O- £

z §

>N CJ

X ><

-SP S « -

g a g

]3 Oh > I-
to o3 (U P

£ ’g

E ?■
S o

73 T3

<D <L>

T3 ^
G g
3 £•
p £

X ed _G 5> _o
03 £

03

X X
X x
X P
O Oh

a. c.

2 o

^ u

Sf-S ^

a c- -as jj

io

a

(L) C^-

! §
3

X ^

73 ^

p 3 ^
O- < ^

C 3

.Is

X

60
13

X P

a

S 3

N

3 >

>5 .*> w X 3
§ § -g p X

?s ^

03

1 ®
3 b

G G
N X
03 X

O N

o ^

<D £
> o3
03 bO

^ s

2 Oh

> -C)

C3 G

^ Q X
cu G »-<
t5 *5 G

2 oh
bo Cl,

s. 3 -4

, G

>N >

aj O

“MS

t -%o

00 C . ^*5

S ^ s

5 O

G ^ ^ ^

' ' boc^- O'

G G

S3 P 53 £ 3 3 c-

£ £

o« o- o*

G
<u

j cdj X

2 S ?2
a. -=c 2

03 c
u X
X bJj
O G

S U

^3

O

u

table 1. Classification of Lower Cretaceous inoceramids. The table summarizes the morphological features, and straligraphical and geographical distributions of the
principal Lower Cretaceous species. The genera Inoceramus and Birostrina are subdivided into informal species groups and it is envisaged that these will eventually form
the basis of a scries of subgenera. References used in establishing the various categories arc as follows:

Inoceramus ovatus gp.: Whitcavcs 1883; Stanton 1895; Anderson 1938, 1945; Zakharov 1966, 1968; Pokhialaincn 1974; Zakharov and Turbina 1979. /. proconccnlricus
gp.: Zakharov 1966; Pokhialaincn 1969; Zakharov and Turbina 1979. /. aucella gp,: Trautschold 1865; Hayami I960; Pergamenl 1965; Pokhialaincn 1969, 1974;
Glazunova 1973. /. colonicus gp.: Anderson 1938; Imlay I960; Pcrgament 1965; Zakharov 1968; Pokhialainen 1969, 1974; Jelclzky 1970 . 1 heieropterus gp.: Whitehouse
1924; Anderson 1938; Imlay I960; Jones and Gryc 1960; Pcrgament 1966; Pokhialaincn 1969; Kauffman el al. 1978. /. neocomiensis gp.: d’Orbigny 1846; Whitcavcs 1883;
Wollcmann 1906; Woods 1911; Gillct 1924; Weaver 1931; Hayami I960; Imlay I960; Casey 1961; Sornay 1965; Pokhialaincn 1969; Glazunova 1973; Zakharov and
Turbina 1979. /. Hwerowskyae gp.. Newton 1909; Stollcy 1912; Wellman 1959; Saveliev 1962; Pergamenl 1965; Glazunova 1973. /. anglicus gp.; Woods 1911, 1912;
McLearn 1919, 1933, 1943; Recside 1923; Imlay 1961; Saveliev 1962; Pergamenl 1965; Glazunova 1973; Jelclzky 1977; KaufTman I978a-c, Kauffman el al. 1978;
Matsumotoef a/. 1978. l.carsoni gp.: M’Coy 1865, 1866, 1867; Etheridge 1872; Etheridge Jr 1892, 1901, 1905; Ludbrook 1966. Birostrina Salomon! gp . : Woods 1911, 1912;
Casey 1961; Saveliev 1962; Kauffman 1978a. B. concenlrica gp.: J. Sowerby 1821; Etheridge 1907: Woods 1911, 1912, 1917; Heinz 1930; Maury 1936; Nagao and
Matsumoto 1939; Warren and Stclck 1940; Saveliev 1962; Pcrgament 1966; Glazunova 1973; Kauffman 1977. 1978a -c; Speden 1977; Wicdmann and Kauffman 1978;
Crame 1980. B. sulcata gp.: Parkinson 1819; J. Sowerby 1821; d'Orbigny 1846; Wollcmann 1906; Woods 1911, 1912; Jones 1960; Kauffman 1978a i Anopaea
(undivided): Etheridge Jr. 1901; Saveliev 1962; Pokhialaincn 1969; Thomson and Willey 1972; Crame 1981; Kelly 1984.

Genus and group name

Inoce

; ovatus gp

/. proconccnlricus gp.

Component species Diagnostic features

Stratigraphic range Distribution

/. ovatus Stanton, /, ellioli Gabb,
/. quatsinoensis Whiteaves,

?/. impurus Zakharov, ?/.
carinatus Zakharov, ?/.
bojarkaensis Zakharov, ?/,
gather ti Zakharov and
Turbina (pars)

/. proconcentricus Pokhialaincn.
/. murgalensis Pokhialaincn,

/. pronatus Pokhialainen, /
tuimyricus Zakharov. /.
vassilenkovi Pokhialainen. ?/.
apiatus Pokhialaincn

/. aucella T rautschold, /.
pseudopropinguus Pcrgament,

?/. tnaedae Hayami, ?/.
gagaensis Pokhialaincn

Small to medium; oval to pyriform outline:
generally erect but may include some
obliquely elongated forms; only slightly
inequivalvc; weakly to moderately inflated,
subdued or irregular concentric ornament

Small to medium; oval to well rounded;
moderately to strongly inequivalvc; most
gryphaeoid forms bear superficial resemb-
lance to Birostrina concenlrica ; weakly to
moderately inflated (LV always more con-
vex); subdued to prominent, irregular con-
centric ornament

Medium; pyriform to rounded-elongate;

LVs sometimes have angular (almost trape-
zoidal) outline; moderately to strongly in-
cqui valve; moderately inflated (LV more than
RV); LV umbo only moderately enrolled;
broad, low, regular concentric folds

/. colonicus Anderson, /. Medium to large; pyriform to rounded-

ovatoides Anderson, /. sub- elongate; strongly gryphaeoid; LV always

colonicus Pokhialainen much more inflated than RV; narrower than

l aucella or B. concenlrica; long, narrow RVs
have ‘pinnid’ appearance; long, curving
ligamentat; almost smooth

Tithonian Valanginian; Pacific Coast of

?Hauterivian USA; northern

Siberia

Berriasian Valanginian; Far East of USSR;

Lower Hauterivian northern Siberia

Essentially Upper Hauterivian;
possibly some extension into
both the Lower Hauterivian
and Lower Barrcmian

Upper Hauterivian-Lower
Barrcmian

Far East of USSR.
Japan; Pacific
Coast of North
America; Spitz-
bergen; Russian
Platform; Crimea;
northern Siberia
Far East of USSR,
Pacific Coast of
North America

heieropterus gp.

/. Hwerowskyae

gP-

/. heieropterus Pokhialainen,

/. setnicostaius Pokhialaincn,

/. pelliformis Pokhialainen, I.
solus Pokhialainen, /. lereclwvai
Pokhialainen, ?/. vallejocnsis
Anderson, ?/. dunveganensis
McLearn, ?/. scutulaius White-
house, ?/. procerus Whitehouse

Medium to large; rounded, rounded-triangular,
or oval outline; tapenng dorsally; prominent
narrow, pointed beaks rise slightly above
hingelinc and curve forwards; moderately
inflated; slightly inequivalve ?; RV often with
sleep, shelf-like anterior margin; prominent
ligamentat bearing large rounded-
rectangular resilifers; almost smooth or
faint, broad, concentric ribs

/ neocomiensis d’Orbigny, /.
subneocomiensis Glazunova,

/. volgensis Glazunova, ?/.
quatsinoensis Whiteaves, ?/.
ewaldi Schlutcr, ?/. curacoensis
Weaver, ?/. borealis Glazunova,
?/. ohiusus Glazunova,

?/. gusselkaensis Glazunova,

II. pocltialaynen Zakharov
and Turbina

/. Hwerowskyae Saveliev, /.
kedroviensis Pcrgament. /.
chojfali Newton, I. spilz-
bergensis Stollcy, /. saralo-
viensis Glazunova. ?/. urius
Wellman

/. anglicus Woods, /, subanglicus
Pergamenl, /. anglicus elongalus
Pcrgament, /. anglicus con-
jugulis Pcrgament, ?/. dowlingi
McLearn, ?/. cadollensis
McLearn, ?/. comancheanus
Cragin, 7/. bellvuensis Recside

/. carsoni M’Coy, I. Sutherland i

Small to medium; broad, rounded-triangular
outline; weakly to moderately inflated; LV
slightly larger and more convex than RV;
straight anterior margin; apical angle
approx. 95 -115°; narrow, closely and
evenly spaced ribs which occasionally split
or become irregular; ribs have acute to
rounded summits and arc sometimes super-
imposed on primary low folds

Small, erect, rounded-triangular to rounded-
quadrate outline; equivalve; moderately to
strongly inflated in centre of valve and
umbonal region, often flattening towards
valve margins; prominent, prosogyrous
beaks; faint, narrow concentric ornament
which may become irregular vcntrally;
occasional coarse ribs with acute summits
may also develop vcntrally

Small to very large; broad rounded to rounded-
triangular outline; length approximately
equal to width; anterior region more promi-
nent and apical angle larger than in /. neo-
comiensis group; weakly inflated; slightly
inequivalve?; narrow, regular concentric ribs
usually have well-rounded summits and are
separated by broad, flat interspaces; ribs
typically symmetrical about growth axis and
occasionally weaken or bifurcate

Medium to very large; sub-erect to strongly
obliquely elongated; both narrow and broad
forms exist; equivalve to slightly inequi-
valvc ?; moderately to strongly inflated over
the umbonal region but flattened vcntrally;
pointed umbonal region strongly projecting;
anterior margin often in the form of a shallow
sigmoidal curve; narrow, regular concentric
ornament which may be in two distinct

Essentially Upper Hauterivian;
if /. vallejocnsis is a true
member, then range may
extend down into the
Valanginian; if /. dun-
veganensis, I. scutulaius, and
/. procerus arc true members,
then the range may be ex-
tended up to the Upper Albian,
and possibly even Middle
Cenomanian

Barrcmian Aptian, with a
possible downward extension
into Hauterivian and
Valanginian

Far East of USSR;
?Anlarclic Penin-
sula; possible ex-
tension to Japan;
Pacific Coast of
North America;

Australia

Europe; south-
western USSR,
?Far East of
USSR; Pacific
Coast of North
America,
southern
Argentina

Upper Aptian to Middle or South-western
Upper Albian USSR; Far East

of USSR. Spits-
bergen; South
Africa; ?New
Zealand

Essentially Middle Upper
Albian, but may also range
into both Lower Albian and
Lower Cenomanian

Europe; south-
western USSR;
Far East of
USSR; Japan;
Pacific Coast of
North America;
East Greenland;
eastern Mexico

Upper Albian

North-cast

Australia

CRAME: CRETACEOUS INOCERAMID BIVALVES FRO

TABLE I ( coni .)

Genus and group nam<

2 Component species

Diagnostic features

Stratigraphic range

Distribution

B. sulonwni gp.

/. saloinoni d'Orbigny, /
captensis Casey

Small to medium; rounded-rectangular to
obliquely elongated; moderately to strongly
incquivalvc; LV moderately inflated, RV
only weakly so; LV umbo strongly incurved;
broad, shallow sulcus on LV variably
developed; simple, regular concentric

Essentially Lower Albian, but
1. sulonwni may range into the
Middle Albian

Although origin-
ally restricted to
southern England,
/. captensis may

South Africa and
Angola; 1. sala-

Europc. south-
western USSR.
Far East orUSSR,
Pacific Coast of
USA. ?Pcru

B. conccnlrica gp.

/. conccntricus Parkinson. /.
conceit trials nipponicus Nagao
and Malsumoto (pars ?), /.
conccntricus costatus Nagao
and Malsumoto (pars ?), 11.
zuvoljiensis Glazunova, 11.
volviunibonatus Etheridge, 11.
warakius Wellman, 11. tenuis
Mantel), 11. corpulentus
McLearn (pars), 11. incelebratus
Pergament, 11. reduncus
Pergamcnl

Small to medium, occasionally large;
obliquely oval to pyriform; noticeably
longer than wide; short hingeline; posterior
wing missing; moderately to very strongly
incquivalvc; LV always more strongly in-
flated than RV; narrow, prominent, and
strongly enrolled umbo LV gives gryphacoid
appearance; thin prismatic shell layer;
simple regular concentric ornament which
may be traversed by weakly developed radial
folds and sulci

Essentially Middle and Upper
Albian, but with the possi-
bility of a few Lower Albian
records; may also extend well
into the Cenomanian

Europe, south-
western USSR;
Far East of
USSR; Pacific
Coast of North
America; eastern
Mexico; Brazil;
?Pcru, ?southcrn
Argentina; South
Africa: New
Zealand;
Antarctic
Peninsula

B. sulcata gp.

1 sulcalus Parkinson, /. sub-
sulcatus Wiltshire. I. sulcatoides
Saveliev, 11. subsulcatiforniis
Bose

Similar to small B. concentrica but with strong
radial plicae covering all or part of shell

May well be restricted to
Upper Albian, but there arc a
few possible Middle Albian
records

Europe; south-
western USSR;
Far East of
USSR: Japan;
Pacific Coast of
North America;
eastern Mexico;
South Africa

Anopaca

A. brachowi (Rouillicr), A.
Irapczoidalis (Thomson and
Willey), A. constrictus
(Etheridge), 11. niandibula-
formis Pokhialaincn, 11.
nwndibula Mordvilko

Small to medium; elongate-pyriform outline,
with high rounded posterior and narrow,
pointed anterior; deep anterodorsal lunulc;
variably developed anterior sulcus; long,
gently sloping hingelinc; cquivalve to
slightly incquivalvc; slightly to moderately
inflated; thin prismatic shell layer; ornament
of low commarginal folds with super-
imposed secondary growth lamellae

Bcrriasian Albian

Eastern England;
?Far East of
USSR; ?south-
western USSR;
Queensland:
Antarctic
Peninsula

PALAEONTOLOGY, VOLUME 28

T3

, O G
c ^
'5b'# w)

O t> W •

x is c

Ifl

C G

0^

- CO 1

Oh £

co D

co •

g ^
# <

^ **2
s $ ?

1.1
bij sr
C
<

O C

U Oh

<d p_) cG

. 0 ^

X CO

aS 1

o D

pj ^

. CO

5 &G?rt^Ooo 3 ^b£ t Gr 4 s--'*-<D 9 a 3 >--^G^G^

PJ ^UhQ^D # ^PhhJ^Z

CD G

" s
5'z

= ^ -a —

G G £ ^ c

<u o ,g g .E

W) ‘C a 3 G C

. ffJ
-C CO
3 CO

o D

3 L,'
UJ 06

.. oo

u u

“ ? “
o £ 5
U < ^

o j- E
E r u

o ^

W

*0
G
G o 3
.2 wi
? ^~~ l G

< w

JO 3

3 a
5 1

JO c2

3 c/?

W D
S3 t/3

-D

3 M 3

c H |

cd g .E

D

D S

xS o

OC Oh

c

o

a

6

o

U

3 S

.2 <u

JO 60

< s

I— >-

<u

£ G G

o P 2

J -JH £

ET 1 o <

^ XJ

q -a

03 •*-

T3

w> £
o G > ^

Z 8* 3 .

g "55 « c

Ft O u '^,-g

c G bXJ '0

g a 2

« G O >
o 3 "Z +3 >

to •a E o -

•d c 2 g °

c n o S «

G • *• ^ G

0 D O s O

°S E j s

1 e J g ^

3 |

5—3 s I ’m
o a > g h
3 g > to

.sr §■ >-

X) # C I-
o .5 o-c

<D c

Oh <D

© e

0> G
> G
<L> t-
T 3 O

a

Oh

D

G CD

G ,G

£ Z

£ 'O

<D

G

_G 0 O

S |

15 £

Ic3

<D

G

C ' S

a> £

s <

w

TO ^

S S 2

■K 2 3

"O

‘o

<D

O O'

B £
3 15

>> E

3 £ >

£ Z'h

E

c/a

o &

c -a 3

c ^
-br* ^ E -

M*; u
g 3 “>
g u w
5 G W
03 'G >
D C ‘OD

o 2 >
» 5 o
g fc E

■I 3 3
S = 3

> ^ u

%, g

> § S

g J3 >%

•g ri w

, o ’O c
cr « o

g G S

.5 qG

-G ^

o 2

Vh £ 0J
^ Oh

O' G o

G hj — i

H CJ

— G >

<D G W
JO 3 TO

0

H-> CD 'G
G ' ^ G
2 ti w

1 g 5
■C g _g>

Oh G ^

G ° *0
• G CD (D

5 s S|

§ ag|

G D *j ^
G J-h C

G ^ X) g

a&S'-o
g.| so

G C /3 G Ch-h

'G G

§ a

.Is

s o

2 E

^ ^ rt"

g S2 >

o G o

3 3 §

C/3 Q N

GO ^
T3 .

5 G
W 2

CD O

a

Hh h^

3 CH.

a. h

8 s i?-

3 "E
o c

<o G CD G

C3 O 2 ^

G hJ

G CD

>^-h CD CD

4 0 0

G G

o 8 £

CD £

■ 2 3 S

C jd 2

? h-5

0)

1 -H G

<D

^ £ S

G2 ^ ,/5

2 in o g

O* 0- > u

G ^ ^ CD

■S x

K

•® i

a o-

•h o

s: ^

^ G

G U
G G

C /3 (J

O £

•G -O

• G 2

CO

| S ■§

’tz i 2 ^

G G
G G
W) W)

G

co CQ

<

G

£ G bb

O § G

b -3 o o

G

>a CD
CD £

3 7 ^

•g O 3

Sg p

T3

CD O

■I E

/H •“ 1 *— I

9-od

? s 8- 3

CL OJ 3
3 TO 3

Sh >>G^
•— : a)
" -G -g ■

<D W) ^ .

.s =S -B'

<D • r G ■

S 1 > £

1) 1)
O- G
g g:

C /3 1 )

jo E

.GJ G

s G

a >

JO 3 TO

,E J -a Zp

£ "O

JO 3 .2

1-503
: ? cl >

’ G G -r-

> fe

W)'G ^

.£ cr .£

Oh « -G

09 ^

^ G c 3
G tJ) qG
D cG G

5 6

TG O

o H

^ C^I

- ; o
3 ^

3

^ JO §

X MO 5

a- ^

u u *S
"% “ 1, !

G

Oh

G

o

PQ

<

H

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

483

text-fig. 4. Inoceramus cf. ovatus Stanton and /. sp. aff. ellioti Gabb from the Fossil Bluff Formation of
Alexander Island, a, internal mould of a whole specimen of/, cf. ovatus (KG.719. 1 5) viewed from the left
side; specimen from Ablation Valley, b, internal mould of an incomplete left valve of /. sp. aff. ellioti
(KG. 401.51) from Tombaugh Cliffs. Both specimens x I.

poorly developed primary concentric folds with a spacing of 5-7 mm. Superimposed on this initial
pattern, especially in the dorsal half of the valve, are finer, closely spaced secondary ribs (text-fig. 4a).

The right valve has been partially crushed and its true form is uncertain. Flowever, it probably
agrees closely in size and shape with the left and it can be concluded that the specimen was, at most,
only slightly inequivalve. The hinge appears to have been short and oblique and the apical angle is
approximately 73°.

Remarks. The pyriform outline, narrow pointed umbones, subdued ornament, and near equality of
the valves all suggest affinity with the I. ovatus group, and more especially with /. ovatus itself.
However, poor preservation prevents positive identification with this species.

Inoceramus sp. aft', ellioti Gabb, 1869
Text-fig. 4b

cf. 1938 Inoceramus ellioti Gabb; Anderson, p. 99, pi. 7, fig. 1 .

1972 Inoceramus sp. a; Thomson and Willey, p. 9, fig. lb.

Holotype. Inoceramus ellioti Gabb; ?Paskenta group (Berriasian-Valanginian), Shasta series, California; re-
figured in Anderson (1938, pi. 7, fig. 1).

Material. One incomplete internal mould (with traces of a thin prismatic shell layer) (WS) (KG.401.51); 30 m
level at Tombaugh Cliffs, Alexander Island (71° 04' S. 68° 18' W.; text-figs. 2 and 10).

Occurrence. As for material. Associated fossils include Berriasian ammonites, belemnites, and bivalves (Taylor
etal. 1979, p. 36; Crame 1982, text-fig. 9). /. ellioti has a probable Tithonian-Valanginian age-range in California
(Anderson 1938; Zakharov 1968).

Description. Although the anterior and ventral regions of this specimen are incomplete, it can be
judged to have had a slightly oblique, oval outline. The better preserved left valve (text-fig. 4b) has an
estimated length of 82 mm and a width of 59 mm ( W/L = 0-72). The umbonal region is not so clearly
isolated as in the specimen of /. cf. ovatus and not noticeably inclined forwards either. It merges

484

PALAEONTOLOGY, VOLUME 28

ventrally with the moderately inflated central region of the valve and there are gentle descents to the
posterior and posterodorsal margins. The latter is noticeably flatter and more pronounced than in the
previous species.

Both valves bear the impressions of two phases of concentric ornament. The larger ribs have widths
of 3-5 mm and low, rounded profiles. Regularly spaced and symmetrically curving over the central
part of the valve, they sweep strongly forwards in the posterodorsal region. Finer secondary ribs (up
to 2 mm in width) are superimposed on the primary ornament (text-fig. 4b).

Remarks. Even though there are similarities between this and the previous species (I. cf. ovatus ) in
style of ornament (see Thomson and Willey 1972, p. 9), it is apparent that there are significant
differences in their respective shell forms. This specimen, with its more prominent posterodorsal
region, is closer to broad forms in the /. ovatus group such as I. quatsinoensis and I. ellioti (Table 1 ). It
is judged to be marginally closer to the latter, although poor preservation means that it can be only
tentatively assigned to it.

Inoceramus heteropterus group
Inoceramus cf. heteropterus Pokhialainen, 1969

Text-fig. 5a, b

cf. 1969 Inoceramus heteropterus Pokhialainen, p. 141, pi. 14, figs. 2-4; pi. 17, fig. 4; pi. 20, fig. 3; pi. 21,
tig. 4.

Holotype. Inoceramus heteropterus Pokhialainen (1969, p. 141, pi. 14, fig. 4; pi. 17, fig. 4); specimen No. 46a/3,
Museum of the NE Complex Scientific Research Institute, Magadan; Upper Hauterivian, basin of the River
Veseloy, NW Kamchatka; by original designation.

Material. Five incomplete internal moulds of whole specimens bearing traces of shell material (M.l 196.3a-e);
approximately the 715 nr level in the Lower TulTMember, Annenkov Island (54° 29' S., 37° 03' W.; text-fig. 1 )
(Pettigrew 1981, fig. 5; Crame 1983a, figs. 6 and 7).

text-fig. 5. Inoceramus cf. heteropterus Pokhialainen from Annenkov Island, a, internal mould (with traces of a
thick prismatic shell layer) of an incomplete whole specimen (M.l 196.3c) viewed from the right side. 6, internal
mould (with traces of a thick prismatic shell layer) of an incomplete whole specimen (M.l 196.3 d) viewed from

the right. Both specimens x 1.

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

485

Occurrence. As for material. The 1. heteropterus group is Upper Hauterivian in the Far East of USSR (Table 1),
but this age may have to be extended down to at least the Valanginian if I. vallejoensis Anderson is a true member
of the group. Similarly, if certain younger species are subsequently referred to it, it may have to be extended to the
Albian or even Cenomanian (Table 1). The Annenkov Island Formation (which includes the lower TufY
Member) is imprecisely dated in the interval Neocomian- Aptian (Thomson et al. 1982).

Description. The specimens are conspicuous for their oval outlines, very reduced ornament, and
narrow, pointed umbones. The latter project strongly above the hinge and forwards, but are not
significantly curved inwards (text-fig. 5a , b). The mean shell length of the five specimens is 84-8 mm
(S.D. = 17-95; range = 71-113) and mean width is 62-4 mm (S.D. = 15-27; range = 49-86). There is
some evidence that the left valves may be slightly longer than the right valves but the specimens are
not well enough preserved for this to be accurately determined.

The best preserved right valve is that of specimen M.l 196.3c (text-fig. 5a). It is moderately convex,
with the maximum inflation occurring in the umbonal region and along the growth axis. There is a
gentle gradient from the centre of the valve to the ventral region and also to the posterodorsal region
where there is an indistinctly recessed wing. There is a significantly steeper descent from the anterior
margin of the valve to the plane of commissure. This occurs across a flat to gently concave shelf up to
13 mm wide (text-fig. 5a). At its dorsal end, directly beneath the umbo, the shelf assumes a more
strongly concave profile and could be described as a gutter. Its presence is related to the curve of the
umbo up and away from the hinge margin. The beak is missing, but by comparison with specimens
M.l 196.36 and cl, it can be judged to have been narrow, pointed, and prosogyrous. A similar, but
slightly shallower, concave gutter separates it from the hinge margin.

The left valves are not so well preserved as the right valves. In outline and general form they appear
to be close to the latter and there are some indications that they may be marginally narrower and
more evenly inflated. Specimen M.l 196. 3d has a very pronounced anterior shelf which again runs
steeply from the anterior margin to the plane of commissure and has an approximate width of 1 5 mm.
The net effect of the very pronounced anterior shelf on both valves must have been to produce a
projection with a distinctly ‘V’ shaped cross-section.

The anterior shelf of the left valve also narrows and deepens directly beneath the umbo (text-fig.
5a, b). The separation of the left umbo from the hingeline is best seen from the right side where a
smooth, narrow, triangular area can be clearly distinguished between the beak and the dorsal surface
of the ligamentat (text-fig. 5a, b). The latter feature, which is the surface on which the ligament pits
are arranged (Pokhialainen 1969), is partially displayed on specimens M.l 196.36-d; on M.l 196.3c
it has an approximate length of 30 mm and a depth of 5 mm in the centre. There is some evidence that
it may have had a convex upper (dorsal) surface but this could not be definitely confirmed. The largest
ligament pits certainly seem to be in the centre of the ligamentat where they are preserved as a rather
irregular row of barrel-like proturberances with convex sides (text-fig. 5a). Much of the surface detail
of the pits has been lost and it is uncertain whether they are simple or composed of composite
elements.

On both valves there are faint traces of broad, low concentric folds.

Remarks. These almost smooth, oval shells with prominent, pointed umbones can be readily linked to
the I. heteropterus group from the Soviet Far East. Unfortunately, this group is almost exclusively
known from a single taxonomic study (Pokhialainen 1969) and the relationships between the
component species are not always clear. The Annenkov Island specimens are judged to be closest to
I. heteropterus itself, which exhibits a very similar range of shell sizes and morphologies (cf.
Pokhialainen 1969, pi. 14, figs. 2 and 3; pi. 17, fig. 4; pi. 20, fig. 3; pi. 21, fig. 4). In particular, this
species has a steeply descending anterior margin on the right valve which assumes a deep concave
profile at its dorsal end, as well as a conspicuous ligamentat bearing large rounded-rectangular
ligament pits (Pokhialainen 1969, pi. 15, fig. 6; pi. 20, fig. 3). As the specimens may be slightly less
inequivalve than the Russian material, and have more prominent left anterior margins, they are only
tentatively referred to I. heteropterus.

I. solus Pokhialainen (1969, pi. 13, fig. 6; pi. 20, fig. 7) is very close to I. heteropterus but can be

486

PALAEONTOLOGY, VOLUME 28

distinguished by a slightly larger apical angle, distinctive ligamentat, and less prominent hollow
(gutter) on the anterior margin. I. peltiformis Pokhialainen (1969, pi. 12, fig. 2; pi. 13, figs. 1-5) is a
rounder and more convex form with traces of a distinct posterodorsal wing. The same feature is also
present on /. semicostatus Pokhialainen (1969, pi. 14, fig. 1; pi. 17, figs. 1-3) which is an elongate-
rounded form with prominent concentric ornament present on the smaller (?juvenile) specimens. It
too, has a distinctive ligamentat and lacks a deep concave hollow along the anterodorsal margin.

Inoceramus neocomiensis group
Inoceramus deltoides sp. nov.

Plate 54, figs. 1-7; Plate 55, figs. 1, 2

Type material. Holotype: KG. 1678.7 (RV?). Paratypes: KG.1677.2-4; KG.1678.5-7, 9; KG. 1701. 33, 34;
KG. 1730. 13- 16, 18, 20-22, 24; KG. 1735.4-9, 13-24, 26-30, 32-37; KG. 1743. 1-7, 9; KG. 1745. 15; KG.2800.9,
31, 33, 84-86, 88, 264, 281,282, 343, 344, 347, 348, 1113-1121, 1127-1130, 1 133, 1 136-1 139 (both int.m. + ext.
m.; RV’s, LV’s + indet. V's). From the following localities (see text-fig. 2): KG. 1677 -ridge between Mt. Lassell
and Mt. Phoebe (71° 45' 30" S„ 68° 49' W.); KG. 1678/1730— N face of Mt. Lassell, Neptune Glacier (71° 43' S„
68° 45' W.); KG. 1701 -Citadel Bastion (71° 59' S„ 68° 32' W.) (probably loose); KG. 1 735-nunatak 0-6 km W
of Mt. Lassell (71 43' 30" S„ 68 52' W.); KG. 1743/2800— Fossil Bluff (71° 19' S„ 68° 17' W.); KG. 1745-
locality U— 2 km NW of Fossil Bluff (71° 18' S., 68° 20' W.). At locality KG. 1743/2800 the specimens occur
between 96 and 133 m in a 426 m section (text-fig. 10); their occurrence at localities to the south of Fossil Bluff
(text-fig. 2) is due to tectonic repetition of the Fossil Bluff Formation.

Occurrence. As for the type material. Associated aconeceratid and heteromorph ammonites strongly suggest an
Aptian age, although there is a possibility that some of them may have Barremian affinities too (Thomson 1974,
1983). Co-occurring specimens belonging to the Aucellina andina-radiatostriata group suggest an Aptian-Albian
age (Crame 1 983a).

Derivation of name. From the distinctive delta-like, or triangular, outline.

Diagnosis. Medium to large, weakly inflated Inoceramus with a triangular outline; ornament of
narrow, regularly spaced concentric ribs that are slightly asymmetric about the growth axis;
ornament lacking in the ventral regions of the largest specimens; slightly inequivalve.

Description. First impressions of this species are those of a medium-sized, flat triangular shell that
varies from roughly equilateral in outline (e.g. KG. 1678.6 and 7; PI. 54, fig. 1 and PI. 55, fig. 2) to
slightly oblique and inequilateral (e.g. KG. 1677.4 and 1735.9; PI. 54, figs. 4 and 6). Although it is
often difficult to distinguish between right and left valves, it is apparent that, in all the asymmetric
specimens, it is the anterior region which becomes reduced; the extent of this reduction varies from
barely detectable to the pronounced state seen in two of the three articulated specimens (e.g.
KG. 1745. 15; PI. 54, fig. 2). The latter specimens are also of interest in that their left valves are slightly
larger and have more inflated umbonal regions than the right valves. There is also a general
impression from study of the single valves that at least some of the left valves have more inflated
umbones than the right valves (e.g. KG. 1677.4; PI. 54, fig. 6).

EXPLANATION OF PLATE 54

Figs. 1 -7. Inoceramus deltoides sp. nov. from the Fossil Bluff Formation of Alexander Island. 1, internal mould
of probable right valve (KG. 1678.6); Mt. Lassell. 2, internal mould of small whole specimen (KG. 1745. 15)
viewed from the right side; locality U near Fossil Bluff. 3, rubber peel from the external mould of
indeterminate valve (KG. 2800. 1 128); Fossil Bluff. 4, internal mould of probable left valve (KG. 1735. 9);
nunatak 0-6 km due W of Mt. Lassell. 5, rubber peel from external mould of probable right valve
(KG. 1678.9); Mt. Lassell. 6, internal mould of left valve (KG. 1677.4); Mt. Lassell Ml. Phoebe ridge.
7, rubber peel from external mould of probable right valve (KG. 1678.5); Mt. Lassell. All specimens x 1,
except for fig. 2 which is x 1 -5.

PLATE 54

CRAME, Inoceramus from Antarctica

488

PALAEONTOLOGY, VOLUME 28

Owing to the uncertainty of the orientation (i.e. left or right) of many valves the following
measurements are based on an undifferentiated sample. It should be borne in mind that the left valves
may be very slightly longer than the right valves, and also that, owing to the poor state of
preservation, most specimens are probably incomplete. The mean estimated shell length is 69- 1 1 mm
(S.D. = 22-65; range = 41-175; N = 44), mean width is 62-30 mm (S.D. = 31-48; range = 32-210;
N = 44), and mean ratio of width to length is 0-89 (S.D. = 0-23; range = 0-62-1.55; N = 44). Thus,
the average shell shape is roughly that of an isosceles triangle that is marginally taller than it is broad
(e.g. KG. 1678.7; PI. 55, fig. 2); common variants are narrower, oblique types (e.g. KG. 1677.4; PI. 54,
fig. 6) and squatter, broader forms with expanded ventral regions (e.g. KG.1735.14u; PI. 55, fig. 1).
Where preserved the anterior and posterior margins appear to be essentially straight features that
enclose an apical angle that approximates to a right angle (x = 91-16°; S.D. = 14-69; range = 73°-
132°; N = 45) (e.g. KG. 1678.6, 7, 1735.9, and 2800.1 128; PI. 54, figs. 1, 3, 4 and PI. 55, fig. 2). In those
valves with attenuated anterior regions, a short, straight anterodorsal margin passes into a gently to
strongly convex anteroventral margin (e.g. KG. 1677.4; PI. 54, fig. 6). The ventral margin always
appears to be well rounded.

The valves are weakly to moderately inflated with the maximum inflation occurring in the umbonal
region. As well as the single left valves with more inflated umbones (e.g. KG. 1677.4), there are some
large right valves in which the umbones rise steeply from a weakly inflated valve surface (e.g.
KG. 1677.2 and 3). A distinctive feature of this species is its narrow, regularly spaced concentric
ornament. This is best seen in the early stages of the shell where the ribs are typically 1-1-5 mm in
width, have acute to rounded profiles, and are separated by narrow interspaces (e.g. KG. 1678.7,
1735.9, 2800. 1 1 28; PI. 54, figs. 3, 4; PI. 55, fig. 2). Usually arranged slightly asymmetrically about the
growth axis, they occasionally anastomose or die out (e.g. KG. 1678.6, 1735.14a, 2800.1128; PI. 54,
figs. 1, 3; PI. 55, fig. 1); for the most part, however, they are remarkably uniform in their size and
distribution. On some of the larger specimens there is a tendency for the ribs to become coarser and
irregular in the posteroventral and ventral regions. Here they are bunched into thicker features
4-5 mm in width, which, when traced into the posterodorsal region, sweep strongly forwards and
diminish in intensity (e.g. KG. 1678.5 and 9; PI. 54, figs. 5 and 7). On the largest specimens the con-
centric ribs can be up to 2 mm in width and are separated by shallow, flat-floored interspaces of
similar, or even slightly greater, size. They are essentially restricted to the umbonal and central
regions of the valve and, when traced towards the anterior, posterior, and ventral margins, are seen to
die out rapidly (e.g. KG. 1677.2, 3 and 1735.8). They are replaced by areas of smooth, undulatory
shell and it can be concluded that this dimorphic ornament pattern is a consistent feature of all the
largest valves. A final characteristic to be noted about the ornament is the tendency for the fine
concentric ribs of some specimens to be grouped (and slightly raised) on low primary folds. This is
best seen in the umbonal region of the largest specimen (KG. 1677.3) but can also be detected on a
number of smaller valves. This style of ornament is close to Heinz's ( 1928c/) Anwachsringreifen.

Those specimens with the most strongly reduced anterior regions show obvious similarities in
outline to Anopaea (e.g. KG. 1 745. 1 5; PI. 54, fig. 2); these are enhanced by the presence on some valves
of a very shallow anteroventral sulcus (e.g. KG. 1735.4). Such material can be taken as further
evidence of possible morphological transitions between Inoceramus and Anopaea.

Remarks. The flat, triangular form and regular, closely spaced concentric ornament readily link I.
deltoides sp. nov. with either the /. neocomiensis or I. anglicus groups. Features which suggest a
greater affinity to the former of these include: a mean apical angle in the region of 90°, the absence of a
protruding anterior region, the slightly inequivalve nature of the shell (with the left valve being the
more convex), comparatively narrow ornament with acute to rounded summits, and the occasional
presence of Anwachsringreifen (Table 1). There are in fact considerable similarities, especially in style
of ornament, with I. neocomiensis itself, although this species is usually interpreted as a comparatively
small and more strongly inequivalve form (e.g. d’Orbigny 1846, pi. 403, figs. 1 and 2; Woods 191 1, pi.
45, figs. 1 and 2; Glazunova 1973, pi. 19, fig. la, b). I. subneocomiensis reaches larger sizes (> 100 mm
in length) and shows a distinct tendency to develop Anwachsringreifen (e.g. Glazunova 1973, pi. 16,

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

489

fig. 4 and pi. 17, fig. 1); however, on the early stages of the valve the ribbing is less distinct than on I.
deltoides sp. nov. and on some of the smaller specimens it is noticeably less regular (e.g. Glazunova
1973, pi. 17, fig. 2). Much clearer regular ornament is seen on both I. borealis and /. obtusus from the
Russian Platform (Glazunova 1973, pi. 13, figs. 1-3; pi. 14, figs. 1 and 2; pi. 15, figs. 1-3; pi. 16, figs. 1
and 2), but in both these species the interspace width exceeds that of the ribs and the overall ornament
style is closer to that of I. anglicus than I. neocomiensis. Large specimens of I. cadottensis McLearn
from the Albian of the Canadian west coast and Alaska also show a pronounced change from
narrow, regularly spaced ribs over the central regions of the valve to almost smooth ventral and
posterior margins (e.g. Imlay 1961, pi. 9, fig. 1; Jeletzky 1964, pi. 27, fig. 7). However, it is apparent
that this species typically has a longer hinge and more rounded-rectangular form than I. deltoides sp.
nov. The early ribbing is also more typical of the /. anglicus group.

Inoceramus liwerowskyae group
Inoceramus stoneleyi sp. nov.

Plate 55, figs. 3-12

Type material. Holotype: D.8212. 1 35 (int.m. WS with traces of outer shell layer). Paratypes: D. 8210. 25 (ext.m.,
WS); D. 8210. 50, 8212.103 (both int.m., LV); D.8212. 180, 181 (both ext.m., LV); D.8212. 101, 104, 136, 138, 139,
141-144, 182, 212, 213, 228 (all int.m., RV); D.8212. 73, 140, 177 (all int.m., indet. V). Locality D.8210 -N flank
of pale ridge 1 km NE of Stoneley Point (63° 51' 40" S., 58° 06' 20"W.); D.8212— floor of valley to S of pale ridge
(63° 51' 40" S., 58° 05' 20" W.) (text-fig. 3). These localities occur between 600 and 925 m in the combined
stratigraphic section measured on NW James Ross Island (text-fig. 11).

Occurrence. As for the type material. Aptian-Albian, with a possible downward extension into the Barremian
(Crame 1983a, b).

Some of the poorly preserved inoceramids recorded from the Crabeater Point region on the east coast of the
Antarctic Peninsula (text-fig. 1) (Thomson 1967) may also belong within this species. In particular, a series of
internal moulds bearing traces of thin, blade-like concentric ribs (BMNH specimens LL16068 and 16069) closely
resemble the smaller and smoother forms of I. stoneleyi sp. nov. (cf. PI. 55, fig. 9; PI. 55, figs. 6, 7, 10, 12).

Derivation of name. In recognition of the pioneer geological work in the area by Professor R. Stoneley.

Diagnosis. A small, erect Inoceramus ; moderately inflated in the central regions but with noticeably
flattened and extended valve margins; ornament weak and irregular in the centre of each valve and
almost absent on the margins; equivalve.

Description. This species is judged to have been equivalve, or very nearly so. The mean estimated shell
length is 25-43 mm (S.D. = 6 02; range =15-37; N = 14) and mean shell width is 16-86 mm
(S.D. = 4-47; range = 10-25; N = 14); (N.B. these measurements exclude two very small right valves
(D.8212. 101 and 228) with lengths of 10 and 12 mm and widths of 6 and 7 mm, respectively; these are
almost certainly juveniles). The right valve of the holotype has an erect, oval outline which is slightly
accentuated by the missing tip of the umbo (PI. 55, fig. 5). However, this feature is present on other
right valves and is clearly narrow and turned weakly forwards (e.g. D.8212. 104; PI. 55, fig. 7). It is
apparent that the apical region is always considerably narrower than the ventral and that the valves
are more asymmetric about an imaginary midline than the first impression gained by cursory
examination of the holotype. The maximum degree of inflation occurs in the centre of the valve and
usually follows the growth axis from the beak to a point close to the ventral margin. On the anterior,
ventral, and posterior margins there is an abrupt flattening to form a distinct rim (or shelf) between
4 and 6 mm in width; the junction between the main part of the valve and this rim is marked by a
shallow depression (PI. 55, figs. 3 and 5). Apparently restricted to just the largest specimens, there are
some indications that this rim may diminish in the anterodorsal region where there is usually the
steepest descent from the centre of the valve to the margin. On some specimens, such as D. 82 12.212
(which is a small, flattened right valve), there appears to have been a steep descent along the greater
part of the anterior margin (PI. 55, fig. 8), and thus the rim may have been essentially restricted to the
ventral and posterior regions.

490

PALAEONTOLOGY, VOLUME 28

The ornament on the right valve of the holotype, which is preserved in prismatic shell material,
consists of narrow ( < 1 mm wide) concentric ribs that are slightly asymmetric about the growth axis
(PI. 55, fig. 5). Their spacing is close but by no means regular, and, when traced towards the valve
margins, they show some tendency to anastomose or disappear. In any event, all the ornament
disappears at the annular depression and the marginal rim is smooth (PI. 55, figs. 3 and 5). The style of
ornament over the central region of the valve is characteristic of all the right valves in the collection,
although there is some variation in intensity of ribbing (PI. 55, figs. 5, 7, 8, 10-12). On specimen
D.8212.212, ribs of up to 2 mm in width are preserved in thick shell material in the posterodorsal
region; here they are closely intercalated with narrower, discontinuous ribs (PI. 55, fig. 8). Along the
dorsal margin of this specimen the same thick shell layer has preserved a ligamentat measuring
approximately 19 mm in length and 4-5 mm in height. This feature, which is set at an angle to the
plane of commissure, bears the remnants of three large and four small, intercalated, ligament pits.
The larger types have rectangular flask-shaped profiles, tapering gradually from the ventral to dorsal
margins of the ligamentat, whilst the smaller, narrower ones taper correspondingly in the opposite
direction. A large, oval, posterior adductor muscle scar is visible in the posterodorsal quadrant of the
holotype’s right valve (PI. 55, fig. 5).

Although less well preserved, the left valve of the holotype appears to be essentially similar in form
and style of ornament to the right. Left valves as a whole are less well represented in the collection but
those that are present suggest that they exhibit a range of variation very similar to that seen in the
rights (D. 8210.50, 8212.103, 135; PI. 55, figs. 3, 4, 6).

Remarks. Originally thought to be a small member of the I. neocomiensis group (Crame 1983c/, b) it is
now apparent that this species is better classified within the I. liwerowskyae group (Table 1). Perhaps
its closest resemblance is with I. spitzbergensis Stolley (1912, p. 20 and pi. 1, figs. 5 and 6) from the
Aplian/Albian of Spitzbergen. This species too, is small and erect, and shows a similar transition
from a well-inflated central region to much flatter margins. It has narrow, closely spaced concentric
ribs which occasionally become larger, irregular in their course and wedge-shaped in cross-section.
Nevertheless, despite these similarities, it would appear that I. spitzbergensis has a more quadrate
outline, due principally to the presence of a right-angled posterodorsal wing, and there is also the
possibility that it has a more angular ventral margin (Stolley 1912, pi. 1 , fig. 5). The narrower form of
I. liwerowskyae Saveliev (1962, p. 228 and pi. 5, figs. 2-8) from the Upper Albian of Mangishlak is
close to specimens such as D.8212.104 and 141 which lack an outer rim (PI. 55, figs. 7 and 12),
although it would seem that this species has consistently finer and closer spaced ornament. It also has
slightly more pointed umbones and there is some tendency towards the formation of a small postero-
dorsal wing (e.g. Saveliev 1962, pi. 5, fig. 2). Both I. kedroviensis Pergament (1965, p. 28 and pi. 9, figs.
3 and 4) from the Albian of north-west Kamchatka and 7. saratoviensis Glazunova (1973, p. 46, pi. 21 ,
figs. 3 and 4) from the late Aptian of the Russian Platform could be matched with some of the

EXPLANATION OF PLATE 55

Figs. 1 and 2. Inoceramus deltoides sp. nov. from the Fossil Bluff Formation of Alexander Island. 1, rubber peel
from an external mould of an indeterminate valve (KG. 1735. 14a); small nunatak 0-6 km due W of Mt. Lassell.

2, holotype (KG. 1678.7), an internal mould of a probable right valve; Mt. Lassell. Both specimens x 1.

Figs. 3-8, 10-12. Inoceramus stoneleyi sp. nov. from the Stoneley Point region, NW James Ross Island.

3, posterodorsal view of holotype (D.8212.135); the specimen, which is an internal mould bearing traces of a
prismatic shell layer, is joined at the hinge but gapes widely ventrally. 4, internal mould of left valve
(D. 8212. 103). 5, right valve of holotype. 6, internal mould of left valve (D. 8210. 50). 7, internal mould of right
valve (D.8212.104). 8, internal mould of right valve (D.8212.212) partly covered by a thick prismatic
shell layer. 10, internal mould of incomplete right valve (D. 8212. 138). 1 1, internal mould of right valve
(D. 82 12. 144). 12, internal mould of right valve (D. 8212. 141 ). All specimens x 1 - 5.

Fig. 9. Possible specimens of I. stoneleyi sp. nov. (BMNH LL. 16068) from Crabeater Point, Kenyon Peninsula.
Both specimens are internal moulds of right valves, x 1 -5.

PLATE 55

CRAME, Inoceramus from Antarctica

492

PALAEONTOLOGY, VOLUME 28

smoother forms of I. stoneleyi sp. nov. (e.g. PI. 55, figs. 7, 10, 12). However, there is some evidence to
suggest that the former of these Russian species is narrower and closer in outline to I. liwerowskyae,
whilst the latter has a much more rounded-triangular outline.

Inoceramus flemingi sp. nov.

Plate 56, figs. 1, 2

Type material. Holotype: KG. 1726. 10 (int.m., LV). Paratypes: KG. 1726.2, 3, 14 (all int.m., LV); KG. 1726.7
(ext. in., LV); KG. 1726.5, 8 (both int.m., RV); ?KG. 1682.36 (int.m., RV). Locality KG. 1726 is on the ridge
running SSW from Mt. Phoebe (71° 48' 30" S., 68° 48' 00" W.) and KG. 1682 is at a high level in Waitabit Cliff's,
Alexander Island (71° 30' S., 68° 48' W.; text-fig. 2). At both these localities the specimens occur in close
association with Birostrina ? cf. concentrica.

Occurrence. Albian (probably Middle-Upper Albian), from its close association with BP. cf. concentrica at both
localities KG. 1682 and 1726.

Derivation of name. In honour of the Rt. Revd. W. L. S. Fleming, geologist on the British Graham Land
Expedition ( 1934-1937) and the first person to collect fossils from Alexander Island.

Diagnosis. A small, erect Inoceramus with prominent concentric ornament, slightly prosogyrous
umbones, and a flange-like posterodorsal wing; prominent ligamentat bearing flask-shaped ligament
pits; equivalve, or very nearly so.

Description. This is a small species with a mean shell length of 27-38 mm (S.D. = 7-96; range = 18-38;
N = 8) and a mean width of 17-13 mm (S.D. = 5-64; range = 11-27; N = 8). Both valves have erect,
rounded-quadrate outlines and prosogyrous umbones that terminate in narrow, pointed beaks
(PI. 56, figs. 1 and 3). The latter reach up to the level of the long, straight hingeline which subtends
a mean apical angle of 75-88° (S.D. = 6-56; range = 68°-84°; N = 8) with the straight to gently convex
anterior margin. The ventral margin is well rounded but the posterior is much less steeply convex and
in some specimens comes to lie subparallel to the anterior. On the holotype a prominent ligamentat is
displayed (PI. 56, fig. 1), which, when complete, would have had dimensions of approximately
14x2 mm. The prominent ligament pits are distinctly flask-shaped, with well-rounded, bulbous
bases tapering up into narrow necks.

Both valves are moderately inflated, with the maximum degree of inflation occurring along the
growth axis. There are even descents from the apex of convexity to all the margins and in the
posterodorsal region there is a narrow, indistinctly recessed wing. This is best seen on the holotype
where it forms a smooth flange-like feature up to 6 mm in width (PI. 56, fig. 1). The ornament consists
of a simple pattern of clearly defined and regularly spaced concentric ribs that are symmetrically
arranged about the growth axis (e.g. KG. 1726.8, 10; PI. 56, figs. 1 and 3); in cross-section these ribs

EXPLANATION OF PLATE 56

Figs. 1 and 3. Inoceramus flemingi sp. nov. from the Fossil Bluff Formation of Alexander Island. 1, holotype, an
internal mould of a left valve (KG. 1726. 10) from the ridge running SSW from Mt. Phoebe. 3, internal mould
of right valve (KG. 1726.8) from the same locality. Both specimens x 1-5.

Fig. 2. Inoceramus urius Wellman. Plaster cast of the holotype (a left valve, NZGS TM21 16) from the Upper
Albian Motuan stage of New Zealand, x 1 .

Figs. 4, 5, and 7. Inoceramus cf. anglicus elongatus Pergament from the Fossil Bluff Formation of Alexander
Island. 4, internal mould of left valve (KG. 1680.72); Keystone Cliffs. 5, internal mould of left valve
(KG. 1680.2); same locality. 7, internal mould of left valve (KG. 106.3); same locality. All specimens x 1.

Figs. 6 and 8. Inoceramus sp. aff. comancheanus Cragin from the Fossil Bluff Formation of Alexander Island.
6, internal mould of right valve (KG. 1606.18); Hyperion Nunataks. 8, internal mould of left valve
(KG. 1 680.73); Keystone Cliffs. Both specimens x 1 .

Fig. 9. Inoceramus sp. aff. anglicus Woods. Rubber peel from an external mould of a large, incomplete right valve
(BR. 151.12); Welchness, western Dundee Island, x 1.

PLATE 56

CRAME, Inoceramus from Antarctica

494

PALAEONTOLOGY, VOLUME 28

typically have an acute profile, tapering from a broad base to a summit no more than 0-5 mm in width.
On most specimens the interspace width increases from approximately 0-5 mm over the umbo to
2-0-2-5 mm close to the ventral margin. Although there may be some difference in the size of the
posterodorsal wing between the left and right valves, they are in all other aspects very similar.

Remarks. This small, erect, equivalve (or nearly so) species probably has its closest links with the
I. liwerowskyae group (Table 1). However, it should be noted that the ornament seems to be con-
sistently stronger and more regular than almost all members of this group, with the possible exception
of I. liwerowskyae itself from the Upper Albian of Mangishlak. The holotype of this species does bear
moderately strong, regular concentric ribs (cf. Saveliev 1962, pi. 5, fig. 6; PI. 56, fig. 1) but all Saveliev’s
(1962, pi. 5, figs. 2-5, 7, 8) other illustrated material is characterized by much finer ribbing. The pro-
minent ligamentat suggests a possible link with I. wins Wellman from the Upper Albian of New
Zealand (see PI. 56, fig. 2), but this species has a slightly different outline and much less uniform
ornament. Although I.flemingi sp. nov. is similar in size to I. stoneleyi sp. nov. and some specimens
possess a flattened posterodorsal area, it can be readily distinguished by its prominent ligamentat and
strong, regular ornament.

Inoceramus anglicus group
Inoceramus sp. aff. anglicus Woods, 191 1

Plate 56, fig. 9

cf. 1911 Inoceramus anglicus Woods, p. 264, text-fig. 29.

cf. 1961 Inoceramus anglicus Woods; Imlay, p. 52, pi. 9, figs. 4 and 6; pi. 10, fig. 9.

cf. 1965 Inoceramus cf. anglicus Woods; Pergament, pi. 3, fig. 2.

cf. 1965 Inoceramus sp. aff. anglicus Woods; Pergament, pi. 6, fig. 4.

cf. 1973 Inoceramus anglicus Woods; Glazunova, p. 49, pi. 19, fig. 4; pi. 21, figs. 1 and 2.

Lectolype. Inoceramus anglicus Woods (191 1, pi. 45, fig. 8a, b ), Red Limestone (Albian), Hunstanton; designated
by Saveliev (1962, p. 223).

Material. Two large fragments (with traces of shell material): BR.151.12 (ext.m., RV); BR. 151.1 1 (int.m., LV).
From the large lateral moraine at Welchness, western Dundee Island (63° 29' S., 56° 15' W.) (text-fig. 1).

Occurrence. I. anglicus is Middle-Upper Albian, with possible extensions into both the Lower Albian and Lower
Cenomanian (Table 1). The predominantly Middle-Upper Albian species, B. concentrica , has also been
recorded from Dundee Island (Crame 1980).

Description and remarks. Whereas specimen BR.151.12 had an original shell length of approximately
160 mm and a width of 100 mm, specimen BR. 1 51.11 appears to have been significantly smaller, with
corresponding dimensions of approximately 105 and 75 mm. All the valve margins are incomplete,
but it can be judged, from the course of the ornament, that the form was sub-erect to erect. Both
specimens are only weakly inflated. The most characteristic feature of this material is its strong,
regular, and widely spaced concentric ornament; the component ribs, which are 1-2 mm in width and
have well-rounded cross-profiles, are separated by broad, flat interspaces that vary from 2-3 mm in
width on the early part of the shell to up to 7 mm towards the ventral margin (e.g. BR. 151 .12; PI. 56,
fig. 9). This style of ornament is reminiscent of that found on both large specimens of I. anglicus typica
(e.g. Woods 191 1, text-fig. 29; Imlay 1961, pi. 9, figs. 4 and 6, pi. 10, fig. 9; Glazunova 1973, pi. 19,
fig. 4, pi. 21, figs 1 and 2) and large fragments tentatively assigned to the I. anglicus group (e.g.
Pergament 1965, pi. 3, fig. 2, pi. 6, fig. 4).

Inoceramus cf. anglicus elongatus Pergament, 1965
Plate 56, figs. 4, 5, 7

cf. 1965 Inoceramus anglicus elongatus Pergament, p. 19, pi. 2, figs. 3 and 4; pi. 6, fig. 3.

1972 Inoceramus sp. /3; Thomson and Willey, p. 1 1 and fig. Id.

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

495

Holotype. Inoceramus anglicus elongatus Pergament; specimen 3/70« from the Albian of the River Kedrovoy
region, Kamchatka (Pergament 1965, p. 19, pi. 2, fig. 3); by original designation.

Material. KG. 106.3, 1680.2, 3, 30, 72, 75 (all int.m., LV); KG. 1680.59 (ext.m., RV); RV. 1680.5 (ext.m., ?RV);
KG. 1663.33, 1680.4 (both int.m., ?RV); KG. 1681. 5. (ext.m., indet.V). Localities KG. 106 and 1680 are in the
lowest 32 m of the section at Keystone Cliffs (71' 33' 00" S., 68 15' 30" W.) and KG. 1681 is from an equivalent
stratigraphic level at the top of Waitabit Cliffs (71 30' 00" S., 68 14' 30" W.). Locality KG. 1663 is on the W side
of Stephenson Nunatak (72° 08' 30" S., 69° 09' 00" W.; text-figs. 2 and 10).

Occurrence. On balance the upper levels of Waitabit Cliffs and lower levels of Keystone Cliffs have an Albian age
(Taylor et at. 1979); however, a few ammonites from these localities have older affinities (Thomson 1974, 1983)
and their presence has yet to be fully explained. I. anglicus elongatus is Middle-Upper Albian in Kamchatka
(Pergament 1965, fig. 6) and in England specimens provisionally identified as /. cf. anglicus elongatus occur in the
lower Cenomanian (Kauffman 19786, p. IV. 5).

Description. This is a small to medium-sized species with an erect to slightly obliquely elongated
outline; the mean shell length is 46-36 mm (S.D. = 9-00; range = 36-63; N = 1 1) and mean width is
30-73 mm (S.D. = 5-95; range — 22-40; N = 1 1 ). At first sight, it would appear to be significantly
narrower than the similarly ribbed /. deltoides sp. nov. (cf. PI. 54, figs. 1 -7; PI. 56,” figs. 5-7), but, when
the respective mean apical angles (cf. anglicus elongatus = 82-3°: deltoides sp. nov. = 91-1 6 ) are
compared, they are not found to be significantly different (Student’s /-test, p > 0 05). This is largely
due to the presence of a prominent hinge region on some specimens of /. cf. anglicus elongatus which
rises well above the posterodorsal margin of the valve (e.g. KG. 1 06.3 and 1 680.2; PI. 56, figs. 5 and 7).
There is no evidence of a corresponding feature of similar magnitude on I. deltoides sp. nov., although
it should be stressed that the hinge region of this species is never well preserved. The W/L ratio for I. cf.
anglicus elongatus is 0-66 and this is significantly less (Student’s /-test, /?< 0-001) than that for
I. deltoides sp. nov. (0-89).

The long, gently convex anterior margin forms an acute angle at the beak with an almost straight
posterodorsal margin (e.g. KG. 1680.72; PI. 56, fig. 4). The latter is sharply differentiated from a
prominent ligamentat which appears to have had a length up to two-thirds that of the total valve
length and a height of 1 -5—2-0 mm. It bears the remnants of rounded-quadrate to rectangular
ligament pits with a mean width in the region of 1 mm; these may be either closely spaced (e.g.
KG. 1680.30) or separated by interspaces of up to 1 mm. All the specimens are weakly inflated and
there are very gentle descents from the centre of the valve to all the margins. On some specimens (e.g.
KG. 106.3; PI. 56, fig. 7) the umbo is slightly differentiated from the main body of the valve, but in
others (e.g. KG. 1680.72, 1681 .5; PI. 56, fig. 4) it is barely discernible as a separate entity. In all cases it
terminates in a narrow, pointed beak which does not rise above the level of the hingeline.

The style of ornament is very similar to that seen in the early stages of I. deltoides sp. nov. Typically
less than 0-5 mm in width at their summits, the ribs are closely and regularly spaced, sub-symmetrical
about the growth axis and rounded in cross-section (PI. 56, figs. 4, 5, 7). There is a slight increase in
both rib and interspace width towards the ventral margin of the largest specimens and occasional ribs
anastomose or become irregular in their course. The irregularities in the early stage of specimen
KG. 1680. 72 (PI. 56, fig. 4) may be due to shell damage during growth. Although the right valves are
less well preserved than the left valves, it would appear that the species was equivalve.

Remarks. These comparatively small, finely ribbed individuals closely resemble a number of juvenile
and incomplete specimens that have been assigned to the I. anglicus group (e.g. Imlay 1961, pi. 9,
fig. 3; Saveliev 1962, pi. 1, figs. 1-5; Pergament 1965, pi. 1, fig. 3, pi. 4, fig. 3, pi. 5, fig. 2, pi. 9, fig. 6;
Glazunova 1973, pi. 19, fig. 5 a-c) (see Table 1). In particular, the broader, more erect forms (e.g.
KG. 1680.2; PI. 56, fig. 5) compare well with certain specimens of I. anglicus forma typica from
Mangishlak (e.g. Saveliev 1962, pi. 1, fig. 56) and the narrower, more oblique ones (e.g. KG. 1680.72;
PI. 56, fig. 4) with the subspecies I. anglicus elongatus from NW Kamchatka (Pergament 1965, p. 19,
pi. 2, figs. 3 and 4; pi. 6, fig. 3). Although lack of a hinge means that the true orientation of some of
the Alexander Island material is unknown, it is likely that the majority of specimens are at least
slightly obliquely elongated; as such, they are closer to anglicus elongatus than forma typica. For two

496

PALAEONTOLOGY, VOLUME 28

of his specimens, Pergament (1965, p. 19) gives lengths of 47 and 44 mm, widths of 32 and 37 mm, and
apical angles of 62 and 72°. One of these (Pergament 1965, pi. 2, fig. 3) clearly has a few bifurcating
ribs, and, were it not for the fact that no hinge regions are preserved, it would appear that the Russian
specimens closely resemble those from Alexander Island. I. volgensis from the Aptian of the Russian
Platform (Glazunova 1973, p. 43, pi. 12, figs. 1-5) is a finer ribbed species whilst I. borealis
(Glazunova 1973, p. 43, pi. 13, figs. 4 and 5; pi. 14, figs. 1 and 2; pi. 15, figs. 1-3; pi. 16, figs. 1 and 2),
even though it has a similar style of ornament, has a much wider apical angle.

Inoceramus sp. aff. bellvuensis Reeside, 1923
Text-fig. 6

cf. 1923 Inoceramus bellvuensis Reeside 1923, p. 203, pi. 46, fig. 1 .

Holotype. Inoceramus bellvuensis Reeside; USNM 325 1 4 ( L V ); N of Bellvue, Colorado; Dakota Formation (late
Albian-early Cenomanian?) (Reeside 1923, p. 203, pi. 46, fig. 1); by original designation.

Material. KG. 1675.3 (ext.m., LV), from Adams Nunatak, Neptune Glacier, Alexander Island (71° 44' 00" S.,
68 33' 30" W.); KG. 1606. 16 (ext.m. , ?LV) from near the top of a 600 m section on the north-westernmost

text-fig. 6. Inoceramus sp. alf. bellvuensis Reeside from the Fossil Bluff Formation
of Alexander Island; rubber peel from an external mould of a left valve (KG. 1 675.3)
from Adams Nunatak. x 0-75.

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

497

nunatak in the Hyperion Nunataks group (approximately 5 km south of Dione Nunataks) (71° 58' 30" S.,
68° 59' 00" W.; text-fig. 2).

Occurrence. I. bellvuensis is probably late Albian -early Cenomanian in the Western Interior USA (Scott 1970;
Kauffman et al. 1978). In the Far East of the Soviet Union, specimens referred to /. cf. bellvuensis by Pergament
(1965, p. 22, pi. 4, figs. 1 and 2) occur in association with members of the I. anglicus group in Middle-Upper
Albian strata.

Description. Specimen KG. 1675.3 (text-fig. 6) had an original shell length in excess of 190 mm and
a width of at least 140 mm. The outline is judged to have been roughly oval and the orientation slightly
oblique; although the hinge is only partially preserved and there has been a certain amount of post-
mortem distortion, it would appear that the blunt umbonal region was slightly opisthogyrous. The
apical angle of approximately 1 10° is formed between a comparatively short straight hinge and a
long, gently convex anterior margin. The ventral margin is incomplete but the posterior is a long
convex feature slightly more strongly curved than the anterior. One of the most striking features of
this specimen is the extensive crescent-shape posterior wing which can be traced from the hinge to the
ventral margin and has a maximum width in its central region of 50 mm (text-fig. 6). Whereas the
umbonal and central regions of the valve are slightly inflated and bear strong, regular concentric
ornament, the posterior wing is flat and bears very reduced ornament. The ribs of the former regions
are up to 2 mm in width, well rounded in cross-section and separated by flat-floored interspaces of
similar dimensions. As they pass on to the wing these ribs sweep strongly forwards and become very
much finer (text-fig. 6).

Specimen KG. 1606. 16 is a less complete left valve with an original shell length in the 90-100 mm
range. The ribbing over its umbonal region is similar to that of KG. 1675.3 but ventrally some of the
ribs fuse in the centre of the valve and have widths of up to 3 mm. There are again indications of a
flattened, crescentic posterior wing (up to 18 mm in width) bearing very fine, strongly recurved ribs.

Remarks. The North American species I. comancheanus Cragin and I. bellvuensis Reeside have
generally been regarded as less regularly ribbed members of the /. anglicus group (e.g. Reeside 1923;
Imlay 1961; Pergament 1 965). The latter species, which is the larger and more erect, is characterized
by an extensive, flattened, almost smooth posterior region; specimen KG. 1675.3 (text-fig. 6) is close
to the holotype (Reeside 1923, pi. 46, fig. 1) but has, if anything, a slightly more oval outline. It is also
apparent that at least some specimens of I. bellvuensis possess a terminal, pointed beak that rises
slightly above the hinge and is gently prosogyrous (e.g. Reeside 1923, pi. 46, figs. 1 and 2; Kauffman
et al. 1978, pi. 7, fig. 9).

Inoceramus sp. aff. comancheanus Cragin, 1895
PI. 56, figs. 6, 8

cf. 1923 Inoceramus comancheanus Cragin; Reeside, p. 202, pi. 45, fig. 2.

cf. 1965 Inoceramus cf. comancheanus Cragin; Pergament, p. 27, pi. 9, figs. 1 and 2.

cf. 1978 ‘ Inoceramus ’ comancheanus Cragin; Kauffman, Cobban and Eicher, p. XXIII. 29 and pi. 7, fig. 4.

Lectotype. Inoceramus comancheanus Cragin; USNM 32686 (RV); 2-3 miles NE of Denison, Texas; Duck Creek
Formation; designated by Kauffman et al. (1978, p. XXIII. 29 and pi. 7, fig. 4).

Material. KG.1606.18 (int.m., RV); KG. 1677.9, 1680.73 (both int.m., LV); KG. 1735. 11 (int.m. ?RV). Locality
KG. 1 606 — near the top of a 600 m section on the north-northwesternmost nunatak of the Hyperion Nunataks
group (71° 58' 30" S., 68° 59' 00" W.); KG. 1677 — ridge between Mt. Lassell and Mt. Phoebe (71° 45' 30" S„
68° 49' 00" W.); KG. 1680 lower levels. Keystone Cliffs (71° 33' 00" S„ 68° 15' 30" W.); KG. 1735-small
nunatak 0-6 km W of Mt. Lassell (71° 43' 30" S., 68° 52' 00" W.) (text-fig. 2).

Occurrence. Associated fossils indicate an undifferentiated Albian age. I. comancheanus is Upper Albian
Lower Cenomanian.

Description. These four poorly preserved specimens have lengths in the region of 60 mm and widths of
approximately 42 mm. They all have erect profiles, apical angles between 90° and 1 10 ”, and weak.

498

PALAEONTOLOGY, VOLUME 28

somewhat irregular concentric ornament. The anterior margin, which is long and comparatively
straight, leads into moderately well-rounded ventral and posterior margins (e.g. PI. 56, figs. 6 and 8).
There are traces of a short, straight hinge and on specimen KG. 1680.73 (PI. 56, fig. 8) four rounded-
rectangular ligament pits (measuring approximately 2-5 x 1-5 mm) are preserved. The valves are
weakly inflated in the umbonal and anterior regions and compressed posteriorly. Narrow ( < 1 mm),
fairly regularly spaced concentric ribs cover the former of these regions but they become noticeably
less distinct on the posterior (e.g. KG. 1 606. 1 8; PI. 56, fig. 6). In places the course of the ribs is irregular
and on specimen KG. 1680.73 (PI. 56, fig. 8) they are slightly asymmetric about the growth axis.

Remarks. The slightly irregular style of ornament and flattened, faintly ribbed posterior region links
these specimens with I. comancheanus Cragin. Although this species is usually regarded as an
obliquely elongated one (e.g. Reeside 1923, pi. 45, figs. 1,3,4, 6, 7; Eigenheerand Sornay 1974, pi. 1,
fig. a), it does include some more erect individuals (e.g. Reeside 1923, pi. 45, figs. 2 and 5; Kauffman
el al. 1978, pi. 7, fig. 4). The density of ribbing on the Antarctic specimens is perhaps not so high as is
usually encountered on I. comancheanus and there is also less of a tendency for individual ribs to split
(PI. 56, figs. 6 and 8). Nevertheless, Pergament (1965, p. 27, pi. 9, figs. 1 and 2) has illustrated two
specimens of /. cf. comancheanus from the Middle-Upper Albian of the Soviet Far East with regular,
more widely spaced ornament and these are comparable to the Alexander Island material.

Inoceramus carsoni group
Inoceramus carsoni M’Coy, 1865

Plate 57, figs. 1-3; Plate 58, fig. 2a, b; text-fig. 7

1865 Inoceramus carsoni M'Coy, p. 334.

1866 Inoceramus carsoni M'Coy; IVTCoy, p. 50.

1867 Inoceramus carsoni M'C'oy; M'C’oy, p. 196.

1872 Inoceramus pernoides Etheridge (non Goldfuss), p. 343, pi. 22, fig. 3.

1892 Inoceramus carsoni M'C’oy; Etheridge Jr., p. 463 ( non pi. 25, figs. 9 and 10 = I. sutherlandi).

1892 Inoceramus pernoides Etheridge; Etheridge Jr., p. 464, pi. 25, figs. 7, 8, 12.

1892 Inoceramus sp. indet.; Etheridge Jr., pi. 21, fig. 19.

1901 Inoceramus etheridgei Etheridge Jr. (non Woods), p. 22.

1905 Inoceramus etheridgei Etheridge Jr., p. 13, pi. 2, figs. 7-9.

1928ft Inoceramus pictus Sowerby; Heinz, p. 129 (pars).

1966 Inoceramus carsoni M’Coy; Ludbrook, p. 157, pi. 17, figs. 2 and 3.

Lectotype. P.2712 (RV) (National Museum of Victoria): base of Walker’s Table Mountain, W bank of Flinders
River, Queensland; Albian; designated by Ludbrook (1966, p. 157 and pi. 17, fig. 3).

Material. D. 8411. 11, 12ft, 13, 16, 8412.73, 8413.1, 8422.96c, 108, 114, 134-136, 8424.3 (all int.m., RV);
D. 8412. 57 (int.m., ?RV); D. 841 1.12c/, 14, 17, 37, 38, 58, 8412.59-61, 8422.66, 69, 74 ft, 75, 76, 80, 89, 95, 96 a, ft,
122-127, 131, 132, 137, 141 , 8424.4c/-c (all int.m., LV); D.8422.70 (ext.m., LV); D. 841 1.10, 15, 61, 66, 8412.62,
66, 68, 74, 75, 8413.3, 8422.64, 65, 71-74, 77, 78, 89, 97-99, 106, 115-117, 127-129 (probable juveniles).
Localities D.841 1 and 8413— reworked nodules in a breccio-conglomerate unit at 151 m in the composite 346 m
section (D.85 15-85 18) at N end of Tumbledown Cliffs, NW James Ross Island (64° 03' 50" S., 58° 26' 00" W.);
D.8412 — 67-121 m interval in the same section; D.8422 and 8424- uppermost 121.5 m of a 483.5 m section
(D. 8521-8523) in Lost Valley, NW James Ross Island (64 02' 20" S„ 58° 24' 10" W.) (text-figs. 3 and 11).

Occurrence. As for material; associated throughout both the Lost Valley and Tumbledown Cliffs sections with
Maccoyella and Aucellina (? A. hughendenensis (Etheridge)). An upper Albian age inferred for the James Ross

EXPLANATION OF PLATE 57

Figs. 1-3. Inoceramus carsoni M'Coy from NW James Ross Island. 1, internal mould of right valve (D.8422. 136);
Lost Valley, xO-75. 2, internal mould of left valve (D.8422. 123); same locality, xl. 3, internal mould of right
valve (D. 8413.1); northern Tumbledown Cliffs, x 1.

PLATE 57

CRAME, Inoceramus from Antarctica

500

PALAEONTOLOGY, VOLUME 28

Island material by analogy with Great Artesian Basin faunas (Ludbrook 1966; Day 1969), is compatible with the
occurrence of the ammonite Ptychoceras at approximately the 466 m level in the Lost Valley section and presence
of B. concentrica and a turrilitid ammonite in the upper levels of the Tumbledown Cliffs section (text-fig. 1 1).

text-fig. 7. Inoceramus carsoni M'Coy; internal mould of a large left
valve (D. 8422. 132) from Lost Valley, NW James Ross Island, x 0-75.

Description. Adult specimens have a mean shell length of 105-33 mm (S.D. = 39-36; range = 56-225;
N = 46) and a mean width of 58-26 mm (S.D. = 21-56; range = 28-127; N = 46). With their narrow,
pointed umbones and broader, rounded ventral regions, they have a distinctive mytiloid (or pernoid)
outline, although it is noticeable that the orientation (with respect to a horizontal hinge) varies from
sub-erect to strongly oblique (e.g. PI. 57, figs. 1-3; text-fig. 7). In the more erect forms the long
anterior margin varies from almost straight to a sigmoidal curve composed of an initial concave
portion beneath the umbonal region and a subsequent long, gently convex portion (e.g. D. 841 3.1;
PI. 57, fig. 3). The posterodorsal margin in these types is usually slightly convex and the
posteroventral and ventral margins well rounded. In the obliquely elongated forms the anterior
region is usually divisible into an early short sigmoidal section and a later, and much longer, gently
convex one (e.g. D.841 1.12, 8422. 1 32; PI. 58, fig. 2 a and text-fig. 7). The latter lies subparallel to the
long, feebly convex posterodorsal border and there is a narrow, very convex ventral region. As might
be expected the obliquely elongated specimens are consistently narrower than the more erect ones,
and have W/L values well below 0-5. Some of the erect types are narrow too, but others show a
pronounced trend towards ventral expansion (e.g. D.8422.136; PI. 57, fig. 1); overall, a mean W/L
value of 0-56 was obtained (S.D. = 9-35; range = 0-36-0-76; N = 46). The hinge, which is never well
preserved, is short and straight and sometimes forms a low, oblique-angled wing at its junction with
the posterodorsal margin (e.g. D.841 1.12; PI. 58, fig. 2a). The umbo rises sharply above the hinge,
curves moderately strongly forwards and inwards, and terminates in a narrow, acuminate beak. The
slim form of the whole umbonal region is reflected in the narrow apical angle (x = 62-85°;
S.D. = 10-61; range = 46-92°; N = 47).

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

501

From the umbonal region, which is moderately inflated, there is typically a steep descent to the
anterior margin and a slightly less sharp one to the posterodorsal border. There is also a smooth, even
gradient to the ventral region, which, in the largest specimens, is almost flat (e.g. D.8422.132, 136;
PI. 57, fig. 1 and text-fig. 7). The basic ornament pattern is one of simple, narrow ( < 1 mm) concentric
rings ( Anwachsringen of Fleinz 1928a) that are remarkably regular in their course and distribution;
very few tapering or anastomosing ribs are seen. The only significant variation in ornament style is in
the rib density, with some specimens having interspaces of only slightly greater dimensions than the
ribs (e.g. D.841 1.12; PI. 58, fig. 2b) and others exhibiting a spacing of 3-4 mm, especially in their
ventral regions (e.g. D.841 3.1; PI. 57, fig. 3). Very occasionally, interspaces up to 10 mm in width
developed, and on the largest specimen (D. 8412. 57), which has an estimated length of 215 mm, there
is evidence of extensive disruption of coarsely spaced ribs in the ventral region. Some of the closely
spaced ornament shows signs of being grouped on low primary folds into Anwachsringreifen (e.g.
D.841 1.12, 8422.132; PI. 58, fig. 2b and text-fig. 7). Although no adult whole specimens were found, it
is judged that this species was equivalve, or very nearly so.

Juvenile specimens tend to have more erect profiles than adults and less obviously protruding
umbonal regions. Nevertheless, the umbones are still consistently narrow and at least some terminate
in slender, pointed beaks. Many juveniles display closely spaced Anwachsringen ornament.

Remarks. The narrow, pointed and strongly projecting umbones of these specimens, together with
their sub-erect to obliquely elongate form and simple, regular concentric ribs, readily link them to
the Australian I. carsoni group (Table 1). In particular the comparatively low width to length
ratio, shallow sigmoidal curve of many of the anterior margins, and predominance of fine, closely
spaced ornament, suggest that they are closest to I. carsoni itself (cf. Etheridge 1872, pi. 22, fig. 3;
Etheridge Jr. 1892, pi. 25, figs. 7, 8, 12; Etheridge Jr. 1905, pi. 2, figs. 7-9; Ludbrook 1966, pi. 17,
figs. 2 and 3; Hill et al. (ed.) 1968, pi. K.V., fig. 1 1). However, it should be stressed here that the
principal distinction between I. carsoni and I. sutherlandi is one of size, with the latter being
broader, and usually longer, than the former (M'Coy 1865, 1866, 1867; Ludbrook 1966, p. 159).
It is not always easy to separate large carsoni from small sutherlandi and the possibility that some
of the Antarctic specimens may be at least transitional to sutherlandi should be born in mind.
Specimen D.841 1.12 (PI. 58, fig. 2a), for example, has a higher than average ventral expansion for
its length and approaches some small forms of sutherlandi (cf. Ludbrook 1966, pi. 17, figs. 4 and 6),
whilst one of the largest specimens, D.8422.136 (PI. 57, fig. 1), has a distinctive rounded-
triangular form that may be better accommodated in sutherlandi than carsoni (cf. Ludbrook 1966,
pi. 18, fig. 1). Re-examination of the extensive Australian material may yet show that these two
species can be combined.

Inoceramus cf. sutherlandi M'Coy, 1865
Plate 58, fig. 1

cf. 1865 Inoceramus sutherlandi M'Coy, p. 334.
cf. 1866 Inoceramus sutherlandi M'Coy, p. 50.
cf. 1867 Inoceramus sutherlandi M'Coy, p. 196.

cf. 1872 Inoceramus allied to /. problematicus d’Orbigny; Etheridge, p. 344, pi. 22, fig. 4.
cf. 1 889 Inoceramus maximus Lumholtz, p. 367, fig.

cf. 1892 Inoceramus carsoni M'Coy (pars); Etheridge Jr., p. 463, pi. 25, figs. 9 and 10.

cf. 1892 Inoceramus sutherlandi M'Coy; Etheridge Jr., p. 463.

cf. 1901 Inoceramus maximus Lumholtz; Etheridge Jr., p. 24.

cf. 1924 Inoceramus maximus Lumholtz; Whitehouse, p. 128, pi. 7, figs. 1 and 2a, h.

cf. 19286 Inoceramus sutherlandi M'Coy; Heinz, p. 144.

cf. 1966 Inoceramus sutherlandi M'Coy; Ludbrook, p. 157, pi. 18, fig. 1.

Holotype. Inoceramus sutherlandi M'Coy; P.2170 (RV) (National Museum of Victoria): base of
Walker’s Table Mountain, Flinders River, Queensland; Albian; illustrated by Ludbrook (1966, pi. 18,
fig. 1).

502

PALAEONTOLOGY, VOLUME 28

Material. D. 8403. 53-57: numerous large, incomplete valves contained within a series of mudstone concretions;
approximately the 200 m level in the section measured at Kotick Point (D.8403, 64° 00' S., 58° 21' W.), NW
James Ross Island (text-figs. 3 and 1 1).

Occurrence. As for material. Lithological correlations between the Kotick Point and Lost Valley sections
indicate that I. cf. sutherlandi occurs approximately 100-150 m beneath the first appearance of I. carsoni (text-
fig. 1 1). The specimens are associated with Aucellina and probable representatives of the ammonite Silesites.
Both these types are present at approximately the 900-1000 m level in the Brandy Bay-Whisky Bay area (text-
fig. 1 1 ) in Aptian -Albian beds that have also yielded I. stoneleyi sp. nov. and Anopaea sp. nov. j8. I. sutherlandi is
an Upper Albian species in Australia (Ludbrook 1966; Day 1969).

Description. There are indications that some of these specimens had lengths and widths in excess of
1 50 mm, and occasionally substantially more. The prismatic shell layer is in places up to 4 mm thick
and it is readily apparent that the original species was a broad, thick-shelled form. Some idea of the
original form can be obtained from three imperfect internal moulds of right valves (D.8403. 53a, (rand
57); these have estimated lengths of 144, 142, and 1 10 mm, corresponding widths of 105, 107, and
88 mm, and W/L values of 0-73, 0-75, and 0-80. Specimens D.8403. 53a and h seem to have erect,
rounded-triangular outlines, with narrow, pointed umbones and much broader, rounded ventral
regions. They are moderately and evenly inflated and exhibit smooth, gentle descents from the centre
of the valve to the ventral margins and slightly steeper ones from the umbonal region to the antero-
and posterodorsal margins. The umbones, which are not sharply differentiated from the valve
surface, curve gently forwards and inwards and appear to taper to a point.

Specimen D.8403. 57 (PI. 58, fig. 1) also has an erect profile but seems to have been considerably
broader than the previous two. In addition, it is more strongly inflated in the umbonal region and
there is an almost vertical descent from the latter to the anterodorsal margin. There is a shallower
gradient in the opposite direction to the posterodorsal region, which is considerably extended by the
presence of an extensive, flat, obtuse-angled wing. This feature is separated from the inflated umbo by
a well-marked radial groove and bordered dorsally by the remnants of a long, straight hinge. On all
three specimens there are only very faint traces of shallow, widely spaced, concentric folds.

Remarks. These broad, erect, almost smooth valves immediately invite comparison with medium-
sized and large forms of I. sutherlandi (e.g. /. maximus Lumholtz 1 889, fig. on p. 367; I. carsoni IVPCoy
in Etheridge Jr. 1892, pi. 25, fig. 9; I. maximus Lumholtz in Whitehouse 1924, p. 128, pi. 7, figs. 1
and 2; all of which = I. sutherlandi in Ludbrook 1966, p. 157). The holotype itself (Ludbrook 1966,
pi. 18, fig. 1) is very similar in style to these specimens, although of somewhat larger dimensions
(L = 187 mm, W = 130 mm). It should be noted, however, that the posterodorsal wing on specimen
D.8403. 57 (PI. 58, fig. 1) is rather more clearly defined than that normally seen on I. sutherlandi and
that the Antarctic specimens lack sigmoidally curved anterior margins. It is only possible, at present,
to suggest a tentative assignment to I. sutherlandi.

Inoceramus of uncertain group affinity
Inoceramus annenkovensis sp. nov.

Text-fig. 8a-c

1947 Inoceramus sp.; Wilckens, p. 37, pi. 5, figs. 2-4.

71982 Inoceramus sp.; Thomson, Tanner and Rex, p. 179, fig. 19.2g.

EXPLANATION OF PLATE 58

Fig. 1. Inoceramus cf. sutherlandi M'Coy. Internal mould of right valve (D.8403. 57); Kotick Point, NW James
Ross Island, x 1.

Fig. 2. Inoceramus carsoni M'Coy. a (left), internal mould of left valve (D.841 1 12a); b (right), internal mould of
right valve (with traces of shell material) (D.841 1 .126); specimens from northern Tumbledown Cliffs, James
Ross Island, x 1.

PLATE 58

CRAME, Inoceramus from Antarctica

504

PALAEONTOLOGY, VOLUME 28

Type material. Holotype: M. 1165. 30c. 1 (int.m., RV). Paratypes: M. 1 165.30c.2, 1165.32a (both int.m., RV);
M. 1 165.326 (ext.m., ?RV); M. 1 165.32c (int.m., indet. V). Locality M. 1 1 65 occurs at approximately the 518-5 m
level in the Lower Tuff Member, Annenkov Island (54° 29' S., 37° 03' W.; text-fig. 1) (Pettigrew 1981, fig. 5;
Crame 1983a, figs. 6 and 7).

Occurrence. As for material. The Annenkov Island Formation is imprecisely dated in the interval Neocomian-
Aptian (Thomson et al. 1982). The overlying species, /. cf. heteropterus, suggests a late Hauterivian age and the
closely related species, I. anomiaeformis , a Hauterivian-Barremian one.

Derivation of name. From the occurrence on Annenkov Island.

Diagnosis. A small, weakly inflated Inoceramus with a rounded, sub-symmetrical outline; irregular
and ill-defined ornament is mainly confined to the umbonal region; prominent ligamentat comprises
a sequence of complex ligament pits; equivalence of valves undetermined.

Description. The holotype has a length and width of 44 mm and specimens M.l 165.326 and c are of
similar dimensions; M.l 165.32a and M.l 165. 30c. 2, however, are significantly smaller, with lengths
and widths in the 20-25 mm range. The outline appears to be well rounded and the holotype is sub-
symmetrical about an axis joining the midpoints of the dorsal and ventral margins (text-fig. 8a). All
the valves are weakly inflated, with the only significant convexity occurring in the umbonal region. A
steep descent from the latter towards the anterodorsal margin leads into a narrow, concave gutter
which, on the holotype, has a length of approximately 10 mm and a maximum width of 3 mm. At its
deepest directly beneath the umbo, this feature progressively diminishes in strength when traced in a
ventral direction (text-fig. 8a).

The umbo is slightly prosogyrous and terminates in a narrow, pointed beak (text-fig. 8a). Although
the tip of the latter must have been approximately level with the dorsal surface of the hinge, it was
separated from it by another narrow, deep gutter (text-fig. 8a). This feature merges anteriorly with the
anterodorsal gutter and also diminishes in intensity when traced in a direction away from the umbo
(i.e. posteriorly). The ligamentat is well preserved on M.l 165. 30c. 2 (text-fig. 86) and partially
preserved on M.l 165. 30c. 1 and M.l 165.32a. On the former two of these specimens it can clearly be
seen to slope steeply inwards towards the plane of commissure between the valves; this indicates that
the ligament area had a ‘V’ shaped cross-section. The ligamentat varies from 1 5-20 mm in length and
2-3 mm in depth and comprises somewhere between seven and ten ligament pits. These pits are
complex features with essentially rounded-quadrate to rounded-rectangular outlines and at least two
structural elements. On specimen M.l 165. 30c. 2 these can be resolved into a simple, striated, square,
or rectangle which alternates with a more deeply impressed figure ‘J’ (text-fig. 86).

text-fig. 8. Inoceramus annenkovensis sp. nov. from Annenkov Island, a, holotype, internal mould of a right
valve (M.1165.30c.l). 6, internal mould of a right valve (M. 1 165.30c.2). c, latex peel from external mould of a
probable right valve (M.l 165.326). All specimens x 1.

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

505

The ornament on four of the specimens (M.l 165.30c. 1, 2; M.l 165.32a, c) is extremely weak and
irregular. Narrow, concentric ribs can be made out over the umbonal regions but these vary from
acute to well rounded in profile and are noticeably irregular in their distribution (text-fig. 8a, 6). On
the holotype (M.l 165.30c. 1) and specimen M. 1165. 32c, this style of ornament rapidly diminishes
away from the umbo and much of the ventral surface of the valve has an uneven, undulatory
appearance (text-fig. 8a). Stronger ornament is preserved on the external mould (M. 1 165.326),
which, at first sight, seems to be readily distinguishable from the other material by its slightly
posteriorly directed umbo (text-fig. 8c). Nevertheless, it is apparent that this specimen has been
crushed and it very probably is a right valve with a slightly displaced umbo. Narrow, irregular ribs
with acute summits on the earliest shell stages give way ventrally to closely spaced ones with
rounded summits and widths in the 1 -2 mm range. These ribs, which are slightly erratic in their course
and variable in thickness, sweep strongly forwards to fuse along the anterior margin (text-fig. 8c). On
the posterior and ventral margins there are indications of a marked reduction in rib intensity.

Specimen M.l 165.326 is of particular importance as it provides a possible link with a hitherto
unique specimen of Inoceramus previously described from the Lower Tuff Member of Annenkov
Island ( Inoceramus sp. of Thomson et al. 1982, p. 179, fig. 19.2 g). This external mould of a whole
specimen was collected loose from locality L on Lawther Knoll (Thomson et al. 1982, fig. 19.1) and
thus probably came from a stratigraphic level equivalent to either station M.l 165 or M.l 196
(Pettigrew 1981, fig. 2). It has clearly defined, narrow, and closely spaced concentric ornament that
shows some considerable similarities in style to that of specimen M.l 165.326 (cf. text-fig. 8c and
Thomson et al. 1982, fig. 19.2g). Although both valves appear to be mytiliform, it is apparent that this
specimen has been crushed and that the anterior margins may have been foreshortened.

Remarks. The three small, rather poorly preserved specimens of Inoceramus sp. described by
Wilckens (1947, p. 37, pi. 5, figs. 2-4) from the north-east coast of Annenkov Island most likely
belong within this new species. Their outlines are slightly more elongate than those of the valves just
described, but in degree of inflation and ornament pattern they agree closely. They were collected
from a series of localities that are very close to Pettigrew’s (1981, fig. 2) M.l 153 and locality J of
Thomson et al. (1982, table 19.1); thus, they probably originate from somewhere within the lowest
80 m of the Lower Tuff Member (Pettigrew 1981, fig. 5; Crame 1983a, figs. 6, 7). The only existing
species with which I. annenkovensis sp. nov. could be compared is I. anomiaeformis Feruglio (1936,
p. 29, pi. 2, figs. 1 and 2) from Tithonian-Lower Cretaceous strata of the Lago Argentino region of
Patagonia. This species was based on the internal moulds of two rounded, almost symmetrical, right
valves, the larger of which (Feruglio 1936, pi. 2, fig. I) was subsequently designated the lectotype
(Leanza 1967, p. 150). This specimen has a moderately inflated, centrally positioned umbo whose tip
is turned very slightly forwards. Either side of the umbo are two small, subequal ears that are
moderately well differentiated from the main disc of the valve. On both Feruglio’s illustrated
specimens there is an ornament of fine narrow concentric ribs that occasionally anastomose or
intergrade. This style of ornament is close to that of I. annenkovensis sp. nov., but the sub-
symmetrical form, distinct ears, and apparent lack of a prominent ligamentat probably serve to
distinguish it as a separate species. Riccardi (1977, p. 222, fig. 2a) has described a single incomplete
specimen of I. aff. anomiaeformis from the Springhill Formation of southern Patagonia which is
notable for the more pointed, prosogyrous form of its umbones as well as its indistinct ornament. It
may well provide a link with the Annenkov Island specimens but at present there is insufficient
material available for this to be firmly established. In Patagonia, I. anomiaeformis occurs in
association with the ammonite Favrella in beds that are generally assigned to the Hauterivian-
Barremian (Riccardi 1970, 1977; Riccardi et al. 1971).

Genus Birostrina J. Sowerby, 1821

1864 Actinoceramus Meek.

Type species. Inoceramus sulcatus Parkinson, 1819, from the Gault (middle-upper Albian) of Folkestone,
England; by subsequent designation (Cox 1969, p. N315).

506

PALAEONTOLOGY, VOLUME 28

Birostrina concentrica group
Birostrina concentrica Parkinson, 1819

Plate 59, figs. 1-11, 13

1819 Inoceramus concentricus Parkinson, p. 58, pi. 1, fig. 4.

1821 Inoceramus concentricus Parkinson; J. Sowerby, p. 183, pi. 305, figs. 1-6.

1846 Inoceramus concentricus J. Sowerby; d’Orbigny, p. 506, pi. 404, figs. 1 5.

1876 Inoceramus concentricus Parkinson; Whiteaves, p. 79.

1904 Inoceramus concentricus Parkinson; Airaghi, p. 183, fig. 2.

71907 Inoceramus volviumbonatus Etheridge, p. 73, pi. 2, figs. 1-6.

191 1 Inoceramus concentricus Parkinson; Woods, p. 265, pi. 45, fig. II; pi. 46, figs. 1-10; pi. 47, figs. 1
and 2.

71911 Inoceramus concentricus Parkinson; Schlagintweit, p. 94.

1917 Inoceramus concentricus Parkinson; Woods, p. 9, pi. 3, figs. 9 and 10.

71921 Inoceramus concentricus Parkinson; Bonarelli and Nagera, p. 22, pi. 2, fig. 9.

1930 Inoceramus concentricus Parkinson; Heinz, p. 683, fig. 1.

1933 Actinoceramus ( Taenioceramus ) concentricus Parkinson; Heinz, p. 245.

1936 Inoceramus concentricus brasiliensis (White); Maury, p. 107, pi. 8, figs. 9, 10, 13.

71939 Inoceramus concentricus Parkinson var. nipponicus Nagao and Matsumoto, p. 267, pi. 24, fig. 2;
pi. 25, figs. I -6.

71939 Inoceramus concentricus Parkinson var. costatus Nagao and Matsumoto, p. 270, pi. 24, figs. 1,4, 5;
pi. 27, fig. 2.

71960 Inoceramus (Actinoceramus) concentricus Parkinson; Jones, p. 157, pi. 29, figs. 1 and 2.

1962 Inoceramus concentricus Parkinson; Saveliev, p. 235, pi. 7, figs. 3-7 and pi. 8, figs. 1 and 2.

71966 Inoceramus cf. concentricus Parkinson; Pergament, p. 30, pi. I, figs. 1 -4.

71976 Inoceramus ( Inoceramus ) concentricus Parkinson; Chiplonkar and Badve, p. 199, pi. 1, fig. 5.
1978u Birostrina concentrica (Parkinson) ( sensu la to): Kauffman, p. IV. 2.

1978 b Birostrina concentrica (Parkinson) (sensu lato)\ Kauffman, p. XVII. 1 and pi. 1, figs. 1-3, 5-; 4,
16, 18.

1978 Birostrina concentrica (Parkinson) (sensu Iato)\ Wiedmann and Kauffman, p. III. 4, pi. 1, figs. 1-10
and 12-14.

1980 Inoceramus concentricus Parkinson; Crame, p. 283, fig. 2 a-c.

N.B. this synonymy comprises only those references that were useful in identifying the Antarctic material. No
attempt has been made here to fully revise all the European and Japanese specimens of/, concentricus. Woods

EXPLANATION OF PLATE 59

Figs. 1-11, 13. Biostrina concentrica (Parkinson) from James Ross and Dundee Islands. 1, anterior view of
internal mould of a whole specimen (D. 8228. 12); Brandy Bay. 2, posterodorsal view of internal mould of a
whole specimen (D. 8228. 5); umbonal region of the left valve slightly displaced across that of the right; same
locality. 3, internal mould of a whole specimen (D. 8413. 58) that bears traces of a thin prismatic shell layer;
viewed from the right side; northern Tumbledown Cliffs. 4, internal mould of a whole specimen (D. 3862. 6),
viewed from the right side; Welchness (Dundee Island). 5, internal mould ofleft valve (D. 8227. 5); Whisky Bay.
6, internal mould ofleft valve (D. 8214. 13); small gully approximately 2 km E of Stoneley Point. 7, internal
mould ofleft valve (D. 8214. 31); same locality. 8, left valve of specimen D. 8228. 5 viewed from the anterior.
9, internal mould ofleft valve (D. 841 3.42); northern end ofTumbledown Cliffs. 10, internal mould ofleft valve
(D. 841 3.38); same locality. 1 1, internal mould of right valve (D. 82 14.29); small gully approximately 2 km E of
Stoneley Point. 13, internal mould (with traces of prismatic shell) of right valve (D. 8215.1 1); gully approxi-
mately 3 km ENE of Stoneley Point. All the specimens are from N W James Ross Island, except for fig. 4 which
is from western Dundee Island. All x 1, except for fig. 1 which is x T5.

Figs. 12 and 14. Birostrina 7 cf. concentrica (Parkinson) from the Fossil Bluff Formation of Alexander Island.
12, rubber peel from external mould of right valve (KG. 1680.74); Keystone Cliffs. 14, internal mould
(with traces of shell material) of small right valve (KG. 2801 .250); northern end of Succession Cliffs. Both
specimens x 1.

PLATE 59

CRAME, Birostrina from Antarctica

508

PALAEONTOLOGY, VOLUME 28

(1911) and Saveliev (1962) contain references to a number of early European works not cited here and
Kauffmann (1977) gives a preliminary revision of the Japanese forms.

Type specimen. Inoceramus concentricus Parkinson ( 1 8 1 9, p. 58, pi. 1 , fig. 4); by monotypy.

Material. D.3862.6, 8215.3, 28, 8227.6, 8, 8228.5, 10, 12, 14, 20, 8413.58, 60, 8414.4, 5, 8431.97, 98, 8531.1 (all
int.m., WS); D. 8214.3, 5, 6, 8, 13, 30, 31, 37, 39, 40, 44, 45, 48, 50, 8215.1, 4, 9, 12-14, 16, 25, 8227.5, 10, 12,
8228.2, 3, 7, 13, 16, 19, 8413.35-42, 59, 8414.6, 8531.5-1 1, BR 151.2 (all int.m., LV); D. 8414.8, 8431.95 (int.m.,
?LV); D. 8214.4, 7,9, 12, 14 17,20,22,26-29, 32,33,42,46,51,52,8215.2, 7, II, 19,20,23,8227.11, 13,8228.6,

1 1 , 1 7, 1 8, 8344. 1 , 3, 84 1 3.43, 44, 843 1 .96, 853 1 .2-4, BR 151.1, 4, 5 a, b , 6, 10 (all int.m., RV); D. 8414.7 (int.m.,
?RV); D. 841 3.69 (ext.m., RV). Many of the internal moulds bear traces of a thin outer shell layer. All the
following localities (except for D.3862 and BR. 151) are on NW James Ross Island (text-fig. 3): D. 8214— small
gully approx. 2 km E of Stoneley Pt. (63° 52' S., 58° 04' 30" W.); D. 8215— small gully approx. 4 km ENE of
Stoneley Pt. (63° 51' 15" S„ 58° 03' 20" W.); D.8227-NE shore of Whisky Bay (63° 52' 40" S„ 58° 06' 55" W.);
D.8228-SW shore of Brandy Bay (63° 51' S., 58° 01' W.); D.8344 -2 km E of Bibby Pt. (63° 48' 20" S.,
57° 54' 30" W.); D.8413-N of Tumbledown Cliffs (64° 03' 00" S„ 58° 24' 30" W.); D.8431/8531 — S end of
Tumbledown Cliffs (64 05' 00" S., 58° 26' 40" W.); D.3862/BR. 151 —moraine ridge, Welchness, Dundee Island
(63° 29' S„ 56° 15' W.) (text-fig. 1).

Occurrence. As for material. On James Ross Island, B. concentrica is associated with an Albian-Cenomanian
turrilitid ammonite in the northern Tumbledown Cliffs assemblage (D.841 3) and at locality D. 8414 (text-fig. 3).
In the combined section for western James Ross Island (text-fig. 1 1) it occurs 125 m above the highest occurrence
of I. carsoni and approximately 400 m above the I. stoneleyi sp. nov.— Silesites assemblage of the Whisky Bay-
Brandy Bay area.

The Middle-Upper Albian age range established by Woods (191 1, 1912) for B. concentrica has been widely
accepted by other authors (e.g. Pergament 1981; Troger 1981). However, the possibility that this species both
extends into the top of the Lower Albian and the base of the Lower Cenomanian should not be discounted
(Kauffman 1978a, fig. 1; Sornay 1981). If forms such as I. concentricus nipponicus and 1. cf. concentricus
(Pergament 1966) prove to be true members of the B. concentrica group (see Table 1 ), then extension of the range
into well within the Cenomanian will be established. There is also a possibility of Cenomanian representatives of
7. concentricus ' in New Zealand (see below). Other definite Southern Hemisphere occurrences of B. concentrica
are in South Africa (Middle-Upper Albian of Zululand; Heinz 1930; Kauffman 19786) and Brazil
(?Middle-Upper Albian; Maury 1936), whilst there are further possible records from the Albian of Argentinian
Patagonia (Bonarelli and Nagera 1921 ) and the Albian of Madagascar (Heinz 1933).

Description. The James Ross Island specimens agree closely with those previously described from
Dundee Island (Crame 1980, p. 283). They are comparatively small, with the left valves having a
mean shell length of 34-48 mm (S.D. = 8-58; range = 16-56; N = 42) and the right valves 26-88 mm
(S.D. = 7-49; range = 1 1-50; N = 49). These respective mean shell lengths are significantly different
(Student’s /-test, p < 0-001 ) and attest to the strongly gryphaeoid form of the species. This is clearly
exhibited by the whole specimens, each of which has a left valve with an inflated umbonal region that
towers over that of the right (e.g. D.3862.6, 8228.5, 12, 8413.58; PI. 59, figs. 1-4). The larger volume of
material from James Ross Island enables the range of variation of this biostratigraphically important
species to be more fully assessed.

The outline of the left valve is variable and at least three main types can be made out: obliquely
elongated, elongate-pyriform, and rounded-quadrate. The first of these is the most strongly
asymmetric, possessing a comparatively narrow, pointed umbonal region that slopes gently forwards
(e.g. D.8214. 1 3 and 8227.5; PI. 59, figs. 5 and 6); the second is more upright and pear- (or tear-)shaped
(e.g. D.82 14.31, 8228.5, 12; PI. 59, figs. 1 , 7, 8); and the third is similar to the second but considerably
squatter (e.g. D.841 3. 38; PI. 59, fig. 10). No rigid morphologic or stratigraphic divisions can be placed
between these types and it is almost certain that they intergrade. In profile, all the left valves are
strongly convex (i.e. incurved) and in lateral view the maximum degree of inflation can be seen to lie in
the central regions of the valve and along the growth axis. The form of the central region varies from a
broad, shallow dome, with only moderate descents on either side (e.g. D. 8227. 5; PI. 59, fig. 5) to a
narrow, strongly convex ridge that is bounded laterally by extremely steep drop-offs (e.g. D. 8228. 5
and 8413.42; PI. 59, figs. 2, 8, 9). A particularly noticeable feature of all the left valves is the almost

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

509

vertical descent along the anterior margin; this forms a prominent flat, smooth shelf of anything up to
15 mm in height. There is a shallower descent to the ventral and posterior margins and at the
posterodorsal extremity a small flange-like wing is delimited by a sharp break in slope on one side and
the short, straight hinge on the other. The strongly incurved, slightly to strongly prosogyrous umbo
rises above the hingeline by anything up to 1 1 mm and frequently tapers to a beak of less than 2 mm
width. The apical angle generally lies in the 85°-95° range.

The smaller right valve of B. concentrica has an erect outline and is typically somewhat longer than
wide (xW/L = 0-71; S.D. =012; range = 0-48-1 -08; N = 61 ) (e.g. D.8215.1 1; PI. 59, fig. 13); there
are, however, both occasional narrow, elongate forms and squatter, rounded ones. The hinge
typically forms an apical angle with the anterior margin of between 90° and 1 10°. The latter feature is
straight to very slightly concave and usually measures at least half the total shell length. It leads into
well-rounded anteroventral, ventral, and posteroventral margins, but the posterior margin proper is
less steeply curved (PI. 59, fig. 13). The small, prosogyrous umbo, which is not nearly so prominent as
that on the left valve, does not rise above the hingeline and a narrow posterodorsal wing is indistinctly
recessed. The principal mode of variation in the right valve is in its degree of inflation. Normally, they
are moderately and evenly inflated, with smooth, regular descents occurring to the valve margins (e.g.
D. 8215. 11; PI. 59, fig. 13). Nevertheless, some specimens are more strongly convex, with the
maximum degree of inflation being concentrated along the growth axis (e.g. D. 3862. 6 and 8228.5;
PI. 59, figs. 2 and 4). There are even some indications of this variation being carried to the extreme
form that has a convex ridge extending from the beak to the ventral margin with very steep drop-offs
on either side (e.g. D. 82 14. 29; PI. 59, fig. 1 1 ). Such a form is similar to that described by Woods (1911,
p. 267, pi. 46, figs. 8-10) in some specimens from the Blackdown Greensand and Hunstanton Red
Limestone of England.

Owing to the poor state of preservation of many of the moulds the precise nature of the original
ornament is difficult to define. Whenever traces of concentric ribs are present, on either valve, they
always appear to be simple, narrow, regular, and closely spaced. Typically less than 1 mm wide and
separated by interspaces of slightly greater dimensions, they may have either sharp or rounded
summits (e.g. D. 3862. 6, 8214.31, 8227.5; PI. 59, figs. 4, 5, 7). On some valves slightly coarser
secondary ribs, with a 3-10 mm spacing, appear to be superimposed on the primary ones (e.g.
D. 8228. 5 and 8413.58; PI. 59, figs. 2, 3, 8) and there are some indications too, of both occasional
growth pauses and irregularities.

A small number of internal moulds bear traces of up to five radial riblets (e.g. D. 8214. 27, 821 5.25,
8228.13, 8413.38, 42; PI. 59, figs. 9, 10). These have a maximum width of 1 mm and seem to radiate
from the umbo to the ventral margin; however, it should be emphasized that they are nearly all weak
and discontinuous. Some of the interspaces between them are slightly concave and as such can be
described as sulci.

Remarks. Kauffman (1978a) suggested that this highly variable species may eventually be split into as
many as six subspecies, but these have yet to be formally described. The stratigraphic control on the
present collections does not permit such an approach here, and all the material is referred to a single
taxonomic category.

Small to medium-sized elongate-pyriform whole specimens (e.g. D. 3862.6, 8228.10, 20, 8413.58,
60; PI. 59, figs. 3 and 4) as well as more obliquely elongated forms (e.g. D. 8214. 13 and 8227.5; PI. 59,
figs. 5 and 6) compare well with the majority of European specimens illustrated by Parkinson (1819,
pi. 1, fig. 4), J. Sowerby (1821, pi. 305, figs. 1-6), Woods (191 1, pi. 46, figs. 1-10; pi. 47, figs. 1 and 2)
and Saveliev (1962, pi. 7, figs. 3-7; pi. 8, figs. 1 and 2). It is worth noting too that the latter two works
illustrate collections with a range of variation comparable in scale, if not in precise detail, with the
Antarctic material. The tendency towards a more pyriform outline of specimens such as D. 8214. 31
and 8228.5 (PI. 59, figs. 2, 7, 8) can be matched to that of Woods’s (1911, pi. 45, fig. 11) largest
illustrated specimen and, probably, to Maury’s (1936, pi. 8, figs. 9 and 10) subspecies I. concentricus
brasiliensis (White) too.

The squatter, more rounded-quadrate Antarctic left valves (e.g. D. 841 3.38, PI. 59, fig. 10) fit in less

510

PALAEONTOLOGY, VOLUME 28

well with the traditional European concepts of B. concentrica. Nevertheless, they appear to grade into
obliquely elongated forms and Kauffman (1978a, 6; Wiedmann and Kauffman 1978) has shown how
they can also be distinguished in both European and South African collections. Types such as
specimen D.84 13.38 closely resemble ‘B. concentrica n. subsp. 2’ and associated forms from Spain (cf.
PI. 59, fig. 10 and Wiedmann and Kauffman 1978, pi. 1, figs. 3-6, 8-10, 13), and B. concentrica
brasiliensis and B. concentrica n. subsp. C’ from South Africa (Kauffman 1978/?, pi. 1, figs. 11, 16,
18). Kauffman’s (1978a) concentrica n. subsp. A', which is characterized by an anterior sulcus and
a few coarse, widely spaced ribs, has no direct counterparts in the Antarctic collections. The
European and South Africa ‘ B. concentrica n. subsp. B’, with its slanting and strongly projecting
umbo (e.g. Kauffman 19786, pi. 1 , figs. 5 and 10), also seems to be missing, and it would appear that
the bulk of the Antarctic specimens fall within Kauffman’s (1978a, b) concepts of B. concentrica
concentrica and the more quadrate forms, B. concentrica brasiliensis and ‘ B . n. subsp. C’.

Birostrinal cf. concentrica Parkinson, 1819

Plate 59, figs. 12, 14; Plate 60, figs. 1-4; Plate 61, figs. 1-4

cf. 1846 Inoceramus concentricus J. Sowerby; d’Orbigny, p. 506, pi. 404, figs. 1 and 2.

cf. 191 1 Inoceramus concentricus Parkinson; Woods, p. 265, pi. 45, fig. 11.

cf. 1914 Inoceramus aff. concentricus Parkinson; Spengler, p. 235, pi. 15, fig. 18.

cf. 1917 Inoceramus concentricus Parkinson; Woods, p. 9, pi. 3, figs. 9 and 10.

cf. 1936 Inoceramus concentricus brasiliensis (White); Maury, p. 107, pi. 8, figs. 9 and 10.

1972 Inoceramus aff. concentricus Parkinson; Thomson and Willey, p. 13, figs. 9 a and 10.

Material. KG. 1674.7 (int.m., WS); KG. 103. 181, 1609.16, 1675.2, 1681.6, 1721.23, 1746.7,9, 1748.24, 2801.158,
1 82, 1 83, 206, 215,218, 239, 262 (all int.m. , LV); KG. 1 674.8 (?int.m., LV); KG. 1 663.35, 280 1 . 1 80, 223 (all ext.m.,
LV); KG. 103. 158, 1674.6, 9, 1677.1, 1746.8, 2801.185, 207, 250, 256 (all int.m., RV); KG.1677.8, 1726.96 (both
?int.m„ RV); KG.1677.3, 1674.5, 1680.74, 2801.150, 184, 214, 216 (all ext.m., RV). All the following localities
are within the Fossil Bluff Formation of Alexander Island (text-fig. 2): KG. 103/1681— upper levels, Waitabit
Cliffs (71° 30' 00" S., 68° 14' 30" W.); KG. 1609— westernmost nunatak, Hyperion Nunataks group (72° 02'
30" S., 68° 55' 00" W.); KG. 1 663— W side, Stephenson Nunatak (72° 08' 30" S„ 69° 09' 00" W.); KG. 1667-small
nunatak immediately to NW Tethys Nunatak (72° 08' 00" S., 68° 59' 30" W.); KG. 1674— small nunatak
approximately 2 km SW Adams Nunatak (71° 08' 00" S., 68° 38' 45" W.); KG. 1675 — Adams Nunatak, Neptune
Glacier (71° 44' 00" S„ 68° 33' 00" W.); KG.1677-ndge between Mt. Lassell and Mt. Phoebe (71° 45' 30" S.,
68° 49' 00" W.); KG. 1680-lower levels, Keystone Cliffs (71° 33' 00" S„ 68° 15' 30" W.); KG. 1721 -ridge running
E of Mt. Phoebe (71° 47' 00" S„ 68° 43' 45" W.); KG. 1726-ridge running SSW from Mt. Phoebe (71 ° 48' 30" S„
68° 48' 00" W.); KG. 1 746/ 1 748/280 1 -upper levels, North Succession Cliffs (71° 08' 30" S„ 68° 17' 30 W.).

Occurrence. As for material. The age in the upper Waitabit Cliffs and lower Keystone Cliffs is almost certainly
Albian (?Middle-Upper Albian) (text-fig. 10) (Willey 1972; Thomson 1974; Taylor et al. 1979); however, as
previously mentioned, the presence of some ammonites with older (?Barremian) affinities at Waitabit Cliffs
(Thomson 1983) has yet to be fully explained.

Description. These specimens show all the typical features of B. concentrica , except that they are
somewhat larger. The respective mean shell lengths of the left and right valves, for example, are
89-77 mm (S.D. = 12-64; range = 68-114; N = 13) and 84-80 mm (S.D. = 19-31; range = 54-118;
N = 1 0), and these values are significantly greater (Student’s /-test, p <0-001) than the corresponding

EXPLANATION OF PLATE 60

Figs. 1-4. Birostrinal cf. concentrica (Parkinson) from the Fossil Bluff Formation of Alexander Island.
1, anterior view of internal mould of left valve (KG. 1674.8); some prismatic shell material visible; small
nunatak approximately 2 km SW of Adams Nunatak. 2, anterior view of internal mould of incomplete left
valve (KG. 2801. 262); northern end of Succession Cliffs. 3, exterior view of the left valve of a whole specimen
(KG. 1674.7); small nunatak approximately 2 km SW of Adams Nunatak. 4, the same specimen, which is an
internal mould, viewed from the right. All specimens x 1.

PLATE 60

CRAME, Birostrinal from Antarctica

512

PALAEONTOLOGY, VOLUME 28

ones (34-48 and 26-88) from the James Ross Island specimens. The foregoing measurements suggest
that the degree of inequality between the valves may not be so great in the Alexander Island material
but it should be emphasized that they are nearly all incomplete single valves. The one articulated
specimen (KG. 1674.7; PI. 60, figs. 3 and 4) has a left valve of 100 mm length and a right of 76 mm and
it is likely that this scale of difference may be close to the true value for the species.

The left valves consistently have obliquely oval to pyriform outlines and prominent, strongly
enrolled umbones (e.g. KG. 103. 181, 1674.7, and 8, 2801.262; PI. 60, figs. 1-3 and PI. 61, fig. 2). They
were moderately to strongly inflated, with the maximum convexity occurring in the central region
along the growth axis. From the latter there was a very steep descent to the anterior margin and a
somewhat gentler one towards the posterior. One specimen, however, is noticeable for its very steep
descents to both the anterior and posterior margins; this gives it a markedly ‘humpback’ cross-profile
(KG. 168 1.6; PI. 61, figs. 3 and 4). The basic ornament pattern preserved on internal moulds of left
valves is one of narrow (generally < 1 mm in width) closely and regularly arranged concentric ribs
(e.g. KG. 1674.7 and 2801.262; PI. 60, figs. 2 and 3). On some specimens there is evidence of extremely
fine ribbing on the early stages (e.g. KG. 1 674.7; PI. 60, fig. 3), whilst on others somewhat coarser ones
develop towards the ventral margin (e.g. KG. 103. 181; PI. 61, fig. 2). There are also indications that
secondary stronger ribs developed in certain specimens (e.g. KG. 1674.8; PI. 60, fig. 1), in a manner
similar to that described for specimens D. 8228. 5 and 8413.58 (PI. 59, figs. 2, 3, 8) from James Ross
Island. These ribs have approximately 3-7 mm spacings and their sharper profiles rise significantly
higher than those of the primary ribs. A small number of the internal moulds are almost smooth, their
surfaces being broken only by occasional shallow concentric depressions which may indicate growth
pauses (e.g. KG. 1681. 6; PI. 61, figs. 3 and 4).

The right valve has an erect outline and a moderately to strongly prosogyrous umbo that scarcely
rises above the level of the hingeline (e.g. KG. 1674.7, 1680.74, 2801.250; PI. 59, figs. 12, 14 and PI. 60,
fig. 4). Where preserved the anterior margin is seen to be a nearly straight feature that is equal in
length to at least half the total valve length. It normally subtends an angle close to a right angle with
the hinge, although some compressed specimens have values considerably greater than this (e.g.
KG. 103. 158; PI. 61, fig. 1). The anterior margin leads into well-rounded ventral and posteroventral
margins but the posterodorsal border is somewhat straighter. The degree of inflation is considerably
less than that of the left valve, with most of it being concentrated in the umbonal and central regions;
the descents to the valve margins are correspondingly gentler, with only the anterior edge being
sharply defined. Specimen KG. 1674.7 (PI. 60, fig. 4) has a narrow, smooth flange running along the
anterior margin which tapers from approximately 9 mm in width at the ventral end to less than 3 mm
beneath the umbo. This feature could be interpreted as an anterior wing but it is also possible that it
represents a near vertical anterior edge that has collapsed on compression. The same specimen bears
traces of a narrow, tapering posterodorsal wing but this is not a prominent characteristic of the right
valve.

One right valve, KG. 1674.6, is considerably more inflated than all the others. This inflation is con-
centrated in the dorsal half of the valve, especially along the growth axis. There are steep descents to
both the antero- and posterodorsal margins and the overall form of the valve is similar to that of
specimen D. 8214. 29 (PI. 59, fig. 11) from James Ross Island. The fine regular ornament of the left
valve is generally repeated on the right (e.g. KG. 1680.74 and 2801 .250; PI. 59, figs. 12 and 14). There
are occasional interruptions to the basic pattern caused by growth pauses or superimposition of the
fine ribs on low primary folds (e.g. KG. 1680.74; PI. 59, fig. 12). Both rib width (which is generally

EXPLANATION OF PLATE 61

Figs. 1-4. Birostrinal cf. concentrica (Parkinson) from the Fossil Bluff Formation of Alexander Island.
1, internal mould of possible right valve (KG. 103. 1 58); Waitabit Cliffs. 2, anterodorsal view of internal mould
of incomplete left valve (KG. 103. 181); same locality. 3, exterior view of internal mould of large left valve
(KG. 168 1.6); same locality. 4, anterior view of same specimen. All specimens x 1.

PLATE 61

CRAME, Birostrina! from Antarctica

514

PALAEONTOLOGY, VOLUME 28

< 15 mm) and interspace width increase slightly towards the ventral margin. Notable variants
include types such as specimen KG. 103. 1 58 (PI. 61 , fig. 1), which has acute, more widely spaced ribs
that show some tendency to anastomose, and KG. 1674.6, which is an almost smooth form bearing
traces of fine radial striae. No obvious radial folds or sulci were observed on either valve.

Remarks. The strongly gryphaeoid form and regular concentric ornament link these specimens with
B. concentrica. However, as their mean dimensions are considerably greater than those normally
associated with this species, and their preservation rather poor, they are only tentatively assigned to
it. There is a fairly close correspondence with the few known large forms of B. concentrica from
Europe and Brazil (e.g. d’Orbigny 1846, pi. 404, figs. 1 and 2; Woods 191 1, pi. 45, fig. 11; Maury 1936,
pi. 8, figs. 9 and 10). The correspondence is less precise with the large specimens of 7. concentricus '
from New Zealand (Woods 1917, pi. 3, figs. 9 and 10) as the latter have somewhat broader and more
symmetrical left valves. Some specimens currently assigned to the New Zealand species, I. warakius
Wellman (see Speden 1977, figs. 10-16), may well be a better match for the Antarctic material, but
this taxon is currently in need of extensive revision. It certainly seems to possess a strongly
graphaeoid form and probably forms part of a lineage of 'I concentricus' -Uke forms which span the
Albian-Cenomanian Motuan and Ngaterian stages. A left valve described by Spengler (1914, pi. 15,
fig. 1 8) as I. aff. concentricus from the Lower Utatur Group of southern India shows some similarities
to the Antarctic specimens but is less strongly inflated. In view of their size and apparent lack of radial
ornament, all the foregoing large specimens can only be tentatively linked with Birostrina.

The affinity of specimen KG. 103. 158 (PI. 61, fig. 1) to BP. cf. concentrica must be held in some
doubt as it is both considerably broader (apical angle 128°) than any other right valve and bears
sharper, more irregular ornament. It could just be closer to certain members of the I. ang/icus group.
There are some doubts too about the allegiance of specimen KG. 1681. 6 (PI. 61, figs. 3 and 4), whose
strongly inflated profile and almost smooth surface could be matched to Cenomanian species such as
I. corpulentus from Canada (e.g. Warren and Stelck 1940, pi. 4, figs. 4-6) and I. reduncus from the
Soviet Far East (e.g. Pergament 1966, p. 40, pi. 16, fig. 1 a; pi. 18, figs. 1 a, b and 2a) (see Table 1).
However, no other Cenomanian fossils have yet been recognized in the Fossil Bluff Formation.

Genus Anopaea Eichwald, 1861

Type species. Inoceramus lobatus Auerbach and Frears (1846, p. 492, pi. 7, figs. 1-3) = I. brachovi Rouillier
(1849, p. 439), from the Upper Jurassic of the Ural Mts., USSR; by monotypy (Cox 1969, p. N317).

Anopaea trapezoidalis (Thomson and Willey, 1972)

Text-fig. 9 a

1972 'Inoceramus' trapezoidalis Thomson and Willey, p. 1 1, fig. 8 a-b.

1981 Anopaea trapezoidalis (Thomson and Willey); Crame, p. 215, pi. 2, figs. a-d.

Holotype. KG. 18.3 la (int.m., LV) from the Fossil Bluff Formation of Alexander Island; illustrated in Thomson
and Willey (1972, fig. 8a) and Crame (1981, pi. 2, fig. a); by original designation.

Additional material. KG. 2880.225, 293, 296 (all int.m., LV); KG. 1 745. 10 (ext. m„ LV); KG. 1 745. 1 1 (ext.m., RV).
Both localities are in the Fossil Bluff Formation of Alexander Island (text-fig. 2): KG. 1 745— locality U, 2 km
NW of Fossil Bluff (71° 18' S„ 68° 20' W.); KG. 2800-locality Q, Fossil Bluff (71° 19' S., 68° 17' W.). The
specimens at the latter locality occur between 88-5 and 1110 m in a 426 m section (text-fig. 10).

Occurrence. As for material. Previously regarded as Berriasian in age (Crame 1981, fig. 5), it is now apparent that
A. trapezoidalis may be, in part at least, as young as Aptian (Crame 1983a).

Description and remarks. Specimens KG. 1745. 10 and 11, with shell lengths (as measured from
anterior to posterior extremities) in the region of 1 5 mm, and specimen KG. 2800. 295, with a length of
approximately 32 mm, closely resemble the smallest members of this species previously described
from locality K (Thomson and Willey 1972, fig. 86; Crame 1981, pi. 2, figs. b-d). The two smallest
specimens bear traces of a distinct anterior sulcus and specimen KG. 2800. 293, although lacking this

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

515

text-fig. 9. Anopaea from the Antarctic Peninsula region, a, incomplete internal mould of a large left valve of
A. trapezoidalis (Thomson and Willey) (KG. 2800. 255); specimen from Fossil Bluff, Alexander Island.
b , incomplete internal mould of a left valve of A. sp. nov. aff. mandibula (Mordvilko) (KG. 1682.37); specimen
from Waitabit Cliffs, Alexander Island, c, internal mould of a right valve of A. sp. nov. f3 (D. 8212. 261) from a
locality approximately 1 -5 km ENE of Stoneley Point, NW James Ross Island. All specimens x I .

feature, has the remnants of a very deep anterior lunule. Specimen KG. 2800. 225 (text-fig. 9a) is
similar in general form to the holotype (KG. 18.31a; Thomson and Willey 1972, fig. 8a; Crame 1981,
pi. 2, fig. a), but slightly larger. It has a very high, rounded posterior region and much narrower
anterior, and, like the holotype, lacks a clearly defined anterior sulcus. This seems to be a
characteristic of the largest individuals of this species, as is the presence of coarse concentric
ornament. On specimen KG. 2800. 225 (text-fig. 9a) the ribs are in the region of 3 mm apart over the
centre of the valve and 5 mm apart towards the ventral margin. Specimen KG. 2800. 296 is a very
incomplete large left valve with a height in the posterior region of about 65 mm and an estimated
length of 80 mm. It is covered with closely set, coarse concentric ribs separated by deep interspaces.

Anopaea sp. nov. aff. mandibula (Mordvilko, 1949)

Text-fig. 9b

cf. 1962 Inoceramus mandibula Mordvilko; Saveliev, p. 230, pi. 6, figs. I 11.

Material. KG. 1682.37 (int.m., LV), from a high level in Waitabit Cliffs, Alexander Island (71° 30' 00" S.,
68° 14' 30" W.; text-figs. 2 and 10).

Occurrence. As for material. Albian (?Middle-Upper Albian), from its position at a high level in Waitabit Cliffs.

Description. The estimated length of this specimen is 35 mm and the maximum height (dorsal to
ventral margins) is 32 mm. Although most of the anterior region is missing it can be judged to have
been much narrower than the posterior with a height of approximately 17 mm. The well-rounded
posterior region can be traced forward into a short, straight hinge which is partially obscured by the
strongly prosogyrous umbo. The latter terminates in a narrow, pointed beak which overhangs a
deeply excavated anterodorsal lunule (text-fig. 96). The ventral margin is slightly sinuous due to the
presence of a broad, shallow sulcus of some 12 mm width in its mid-region. This sulcus can be traced
right up into the beak where it sweeps forwards and narrows to just under 2 mm in width (text-fig. 9b).
The weakly to moderately inflated valve surface is covered with narrow (predominantly < 1 mm in
width) concentric ribs that are regularly spaced but somewhat variable in their intensity; some tend to
fade across the sulcus and others in the posterodorsal region (text-fig. 9b). Slightly coarser ribs

516 PALAEONTOLOGY, VOLUME 28

approaching 1 mm in width and separated by interspaces of up to 2 mm occur close to the ventral
margin.

Remarks. The form of this specimen is too rounded and the ornament too fine for it to be related to
A. trapezoidalis. It would seem instead to be closer to A. mandibula (Mordvilko) from the Lower
Albian of Mangishlak (Saveliev 1962, p. 230, pi. 6, figs. 1-11). This is an erect, moderately inflated
and finely ribbed species whose left valve can be closely matched with the Alexander Island specimen.
However, there may be some differences between the two (especially in the form of the anterior
sulcus) and specific separation seems to be necessary. B. salomoni (d’Orbigny) also has a strongly
sulcate left valve, but this is typically more rounded-rectangular in outline and bears a broad,
orthogyrous umbo (e.g. Woods 191 1, pi. 45, figs. 3-7; Saveliev 1962, pi. 9, figs. 6-9).

Anopaea sp. nov. /3
Text-fig. 9c

Material. D. 8212. 261 (int.m., RV); D. 8212. 262 (int.m., LV). Locality D.8212 — valley floor approx. 1 -5 km ENE
of Stoneley Point, James Ross Island (63° 51' 40" S., 58° 05' 20" W.; text-fig. 3); the specimens originate from
approximately the 980 m level in the combined section measured on NW James Ross Island (text-fig. 11).

Occurrence. As for material. Aptian -Albian, from its association with I. stoneleyi sp. nov. and probable silesitid
ammonites.

Description. The right valve (D.8212. 261; text-fig. 9c) has an extremely accentuated rounded-wedge
shaped profile; over a length of 23 mm it tapers from a height of 23 mm in the posterior region to
approximately 8 mm at the anterior. The straight hinge has a length of 10 mm and subtends an angle
of 140 with the anterior margin. It leads posteriorly into what appear to have been well-rounded
posterior and posteroventral margins and these in turn pass into an anteroventral region whose
outline is interrupted by a broad, shallow sinus (text-fig. 9c). Although incomplete, it would appear
that the latter region was extremely narrow and pointed. A slender, tapering posterodorsal wing with
a height at its posterior end of 2 mm and a concave cross-section is sharply demarcated from the main
surface of the valve. Ornament consists of narrow ( < 1 mm), regular concentric ribs separated by flat
interspaces of 10-L5 mm width (text-fig. 9c).

Remarks. Although there are some similarities in style of ornament between these specimens and the
Tithonian species, A. stoliczkai (Holdhaus) (Crame 1981, pi. 1, figs, a-f ), they differ significantly in
shell form. In particular, specimen D.8212. 261 (text-fig. 9c) has a less deeply excavated lunule but
more prominent anterior sulcus. There are differences too in shell form from A. trapezoidalis, and in
style of ornament from A. sp. nov. aff. mandibula.

This material most likely represents a new species. The suffix /3 is used as a new but unnamed species
of Anopaea has already been described from probable Upper Jurassic strata within the Fossil Bluff
Formation (Crame 1981, p. 213).

STRATIGRAPHIC DISCUSSION

Alexander Island

Preliminary accounts of the distribution of Lower Cretaceous inoceramid bivalves through the Fossil
Bluff Formation have already been given (Crame 1983a, b). In these works a simplified stratigraphic
correlation scheme was presented for a series of major localities between Ablation Point and
Keystone Cliffs (text-fig. 2; Crame 1983a, fig. 3; 19836, fig. 2); these were thought to comprise the
greater part of the Fossil Bluff Formation and total approximately 3820 m in thickness. With the
more detailed taxonomic and stratigraphic information given in this study, it is necessary to briefly
re-examine these distributions and further assess their stratigraphic implications.

In the lower levels of Tombaugh Cliffs and highest parts of Callisto Cliffs (text-fig. 10),
Retroceramus everesti is of Berriasian age (Crame 1982). It is probably also present towards the top of

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

517

KEYSTONE

CLIFFS

(W)

fb-il

WAITABIT

-Z-Z-i

CLIFFS.

-zz-:

F-E-Eq

(T)

! B.7 cl.concentrica

i-z-B

f. mandibula

*

— -

FOSSIL BLUFF
(O&R)

/. deltoides sp.nov.

SUCCESSION CLIFFS
(C)

gsg (B)

(A)

Anopaea trapezoidalis
LOC.K

, Birostrina ? cl.concentrica

TOMBAUGH

CLIFFS

(Z)

ABLATION

VALLEY

-1000

Predominantly siltstone
Principal sandstone intervals
Conglomerate

CALLISTO

CLIFFS

I. sp.aff .ellioti.
Retroceramus everesti

text-fig. 10. Occurrence of Lower Cretaceous inoceramids in the Fossil Bluff Formation of Alexander Island.
Localities given in text-fig. 2. Correlations based in part on Taylor et al. 1979, fig. 6 and Crame 1982, text-fig. 9.

Vertical scale in metres (lowest 500 m omitted).

the main Ablation Valley section and it is thought that the uppermost limit of its range can be set at
1675 m. Thus, it overlaps with the occurrences of I. sp. aflf. ellioti (1480 m) and /. cf. ovatus (1675 m)
(text-fig. 10), both of which have also been well established as Berriasian (Taylor et ai 1979; Crame
1983n, b). Moving southwards, structural, lithological, and faunal considerations all suggest that the
top of the Tombaugh Cliffs section correlates with a level slightly beneath the base of the section at
locality K (text-fig. 10). The next inoceramid to appear in the sequence, A. trapezoidalis, commences
in the mid-levels of the section at locality K (2300 m) and continues upwards through locality D to the
2810 m level at localities Q and R (text-fig. 10). The extension of the range of this species into the base
of the Fossil Bluff sections means that it may be, in part at least, as young as Aptian in age; indeed, it is
even possible that it is also Aptian at locality K, as was originally suggested by Thomson and Willey
(1972). At present there is no palaeontological evidence for the age of the beds occurring between
I. cf. ovatus and A. trapezoidalis (text-fig. 10).

I. deltoides sp. nov. at the 2800-2820 m level (text-fig. 10) is most likely Aptian in age, although
previous comments about possible late Neocomian or Barremian affinities of certain heteromorph
ammonites occurring above it should be borne in mind. If various ammonite species of the genera
Aconeceras, Theganeceras, and Sanmartinoceras, and belemnites of the genera Peratobelus and
Neohibolites , are taken as Aptian (Taylor et al. 1979), then it would appear that this stage is well
represented in the sections at localities Q and R and in the lower levels of Waitabit Cliffs (text-fig. 10).
The transition up into the Albian seems to occur at the latter locality with specimens of BP. cf.
concentrica (3500-3700 m) and Eotetragonites being taken as indicators of the younger stage. Not-
withstanding the possible Barremian affinities of ammonites such as Antarcticoceras and Silesites

518

PALAEONTOLOGY, VOLUME 28

(Thomson 1983), both Anopaea sp. nov. aff. mandibula (3560 m) and /. cf. anglicus elongatus
(3620-3680 m) are also taken to be Albian (?Middle-Upper Albian) in age (text-fig. 10).

The gradual southerly younging of the Fossil Bluff Formation between Ablation Valley and Fossil
Bluff is interrupted by the abrupt occurrence of Albian strata at localities A, B, and C, Succession
Cliffs (text-figs. 2 and 1 0). These beds, which have yielded BP. cf. concentrica together with ammonites
such as Antarcticoceras antarcticum Thomson and Ptychoceras sp., are thought to have been
emplaced by a combination of thrusting and normal faulting (Taylor et al. 1 979). Further evidence of
stratigraphic repetition of the Fossil Bluff Formation by faulting is provided by the recurrence of
both I. deltoides sp. nov. and BP cf. concentrica at several localities between the Neptune Glacier and
Stephenson Nunatak (text-fig. 2). The exact stratigraphic positions of I.flemingi sp. nov., I. sp. aff.
bel/vuensis, and I. sp. aff. comancheanus are uncertain but it is assumed that they occupy a level at least
as high as that of BP cf. concentrica (text-fig. 10).

James Ross Island

The Lower Cretaceous biostratigraphy established for the Brandy Bay-Whisky Bay region (Crame
1 983a, b) can now be extended to the Gin Cove-Rum Cove area (text-fig. 3). This correlation is based
primarily on the occurrence of B. concentrica , which, although unrecorded in poorly exposed strata
at Kotick Point, is present in both the North and South Tumbledown Cliffs sections (text-figs. 3 and
1 1). It is enhanced by certain broad lithological comparisons between the two regions and by similar
levels of occurrence of the ammonite provisionally identified as Silesites.

I. stoneleyi sp. nov. (previously identified as a member of the I. neocomiensis group— Crame
1983a, b) occurs in the combined section between approximately 600 and 925 m (text-fig. 11).
Associated fossils, such as ‘ Ancyloceras' patagonicum, small aconeceratids and Aucellina , suggest an
Aptian-Albian age, as do close relatives within the /. liwerowskyae group; nevertheless, the
possibility that at least some ancyloceratids and aconeceratids may be Barremian (or even older) in
Alexander Island should be considered in any final age-determination of these beds. To date the only
fossils collected from the lowest 600 m of strata on James Ross Island are a series of juvenile buchiid/
oxytomid bivalves and a small gaudryceratid ammonite. If the beds containing I. stoneleyi sp. nov.
are confirmed as Aptian-Albian, then this lowest biostratigraphic unit (text-fig. 1 1 ) may well prove to
be Barremian or even earlier in age.

Anopaea sp. nov. /3 from the 980 m level at locality D. 8212 (text-fig. 1 1) is also judged to be Aptian-
Albian. The silesitid ammonites with which it is associated can be traced again at the 900- 1 000 m level
in the Kotick Point section (text-fig. 1 1 ). Other fossils occurring in the lower half of the Kotick Point
section include occasional phylloceratid and lytoceratid ammonites, Aucellina and I. cf. Sutherland /;
the latter species in particular has been used to assign an undifferentiated Aptian-Albian age to this
interval. Inoceramus carsoni, an Upper Albian species in Australia, occurs in the Lost Valley and
North Tumbledown Cliffs sections between 1 140 and 1250 m (text-fig. 1 1 ). Throughout this range it is
associated with a species of Aucellina close to A. hughendenensis and a so far unidentified species of
Maccoyella. The heteromorph Ptychoceras (Upper Aptian-Upper Albian) has been recorded from
the top of the Lost Valley section and a probable specimen of Beudanticeras (Lower-Upper Albian)
from an equivalent stratigraphic level nearby (M. R. A. Thomson, pers. comm.). Neither I. cf.
sutherlandi nor I. carsoni has yet been found in the Brandy Bay-Whisky Bay region, where the
1000-1300 m interval is predominantly composed of unfossiliferous conglomeratic beds.

B. concentrica occurs in the combined section at approximately the 1375 m level (text-fig. 1 1). The
Middle-Upper Albian age for this species established in other regions would seem to be confirmed by
its position above I. carsoni and its association at North Tumbledown Cliffs and locality D.8414
(text-figs. 3 and 1 1) with turrilitid ammonites (M. R. A. Thomson, pers. comm.). At South Tumble-
down Cliffs there appears to be a regular transition upwards from the B. concentrica beds into
Cenomanian strata characterized by acanthoceratid ammonites and probable members of the I. pictus
Sowerby group (M . R. A. Thomson, pers. comm.). However, at all other localities so far investigated
the Lower Upper Cretaceous boundary is obscured by poorly exposed conglomeratic strata and
local unconformities (Crame 19836).

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

519

BRANDY BAY

WHISKY BAY

B. concentrica

B. concentrica

l.stoneleyi sp.nov.

D.8209
& 8210

TUMBLEDOWN CLIFFS

KOTICK POINT

Birostrina concentrica

I. cl. sutherlandl .

LOST

VALLEY

gg

B concentrica

Siltstone

Sandstone & Pebbly Sandstone
Conglomerate

text-fig. 1 1. Occurrence of Lower Cretaceous inoceramids on the NW coast of James Ross Island. Localities
given in text-fig. 3. Correlations based in part on unpublished information kindly supplied by J. R. Ineson.

Vertical scale in metres.

Annenkov Island and South Georgia

Comparisons with similar forms suggest that the most likely age affinities of both I. annenkovensis sp.
nov. ( = /. cf. anomiaeformis of Crame 1983a) and I. cf. heteropterus from the Lower Tuff Member of
the Annenkov Island Formation are Hauterivian-Barremian. Nevertheless, it should be emphasized
that these comparisons are somewhat tenuous and the precise age of the sediments on this island
remains in some doubt. A slightly younger age (Barremian or Aptian) age is suggested by the poorly
preserved specimens of aconeceratid ammonites and Aucellina that occur in the upper levels of the
range of /. annenkovensis sp. nov. (Crame 1983a, figs. 6 and 7), but there are also a number of other
faunal elements in both the Lower Tuff and Upper Breccia Members whose significance is as yet
uncertain (Thomson et al. 1982). There is still considerable scope for clarification of the
biostratigraphy of Annenkov Island.

Only a single indeterminate specimen of Inoceramus has been recorded from the Cumberland Bay
Formation of South Georgia (Thomson et al. 1982, p. 178).

SYNTHESIS

Berriasian representatives of the I. ovatus group (I. cf. ovatus and I. sp. aff. ellioti) from Alexander
Island provide a possible means of correlation between the earliest Cretaceous strata of Antarctica
and those of the Pacific coast of North America and Siberia. Moving up through the succession of
Antarctic Lower Cretaceous inoceramids, it is apparent that there is then a pronounced stratigraphic
gap before the probable Aptian-Albian faunas of the upper Fossil Bluff Formation and the lower
part of the James Ross Island succession. There is very little evidence of Valanginian, Hauterivian, or
Barremian inoceramids. Of course this time interval may be at least partially filled in the Fossil Bluff
Formation when taxonomic revisions of certain ammonites (notably the heteromorphs) have been
undertaken and stratigraphic studies at localities AP and CC2 (text-fig. 2) completed. Nevertheless,
at present, there is very little firm palaeontological evidence for the Valanginian-Barremian stages in

520

PALAEONTOLOGY, VOLUME 28

Alexander Island. Anopaea trapezoidalis (text-fig. 10) may be late Neocomian or Barremian in the
early part of its range but this has yet to be confirmed.

Although the Cretaceous sedimentary succession in the James Ross Island area is thought to
commence at approximately the Aptian stage, there is evidence of lower stratigraphic horizons at at
least two localities on the east coast of the Antarctic Peninsula. Late Hauterivian-Barremian
dinoflagellates and coccoliths have been recovered from a 7 50- 1 000 m conglomeratic sequence on the
Sobral Peninsula (text-fig. 1), and at Pedersen Nunatak (text-fig. 1) a 142 m sequence of con-
glomerates and sandstones has yielded ammonite fragments referable to the South American
Hauterivian species, Favrella wilckensi (Favre) (Farquharson 1982; Thomson and Farquharson
1 984). Further north, /. annenkovensis sp. nov. and I. cf. heteropterus from the Lower T uff Member of
Annenkov Island have Hauterivian-Barremian affinities. The latter species in particular, through its
link with a distinctive North Pacific inoceramid group, suggests an Upper Hauterivian age, but
balanced against this is the presence of both aconeceratid ammonites and Aucellina in the same
stratigraphic unit. Perhaps the best compromise is to regard the age of this unit as Barremian. Early
Cretaceous marine sediments on Byers Peninsula (South Shetlands) are probably confined to the
Berriasian and Valanginian (Smellie et al. 1980) but isolated marine intervals within the Mesozoic
alluvial fan conglomerates of the South Orkney Islands (text-fig. 1) may range as high as the
Hauterivian (Thomson 1981).

The paucity of Valanginian-Barremian inoceramids from the Antarctic Peninsula region as a
whole enhances the impression of a general hiatus at this time gained from the study of other faunal
groups (notably the ammonites, e.g. Thomson 1974, 1982). This may well be linked with a significant
marine regression as the Antarctic Peninsula underwent a major phase of uplift and magmatic activity
(Farquharson 1 982). It has been suggested that the simplest explanation of this orogenic pulse was an
increased rate of subduction of the Pacific Aluk plate beneath the Antarctic Peninsula margin of
Gondwana. Such an event is thought to have been a likely precursor to the Valanginian opening of
the South Atlantic and Weddell Sea basins (Farquharson 1983).

I. stoneleyi sp. nov., the lowest species to occur in the James Ross Island succession, can be
matched with Aptian-Albian members of the I. liwerowskyae group from Spitsbergen, south-
western USSR, and the far-eastern USSR. Similarly, I. deltoides sp. nov. from the Fossil Bluff
Formation is close to specimens of I. subneocomiensis and I. neocomiensis known from a number
of Northern Hemisphere localities. The I. neocomiensis group is a particularly interesting one in
that it may be one of the earliest Cretaceous inoceramid groups with a cosmopolitan distribution.
It is succeeded by the B. concentrica group which is truly worldwide in its occurrence. The Middle-
Upper Albian age of B. concentrica on James Ross Island is supported by ammonites and this
species is generally regarded as one of the most useful for both local and regional correlations in
the Antarctic Lower Cretaceous. On Alexander Island the presence of BA cf. concentrica in
the Waitabit Cliffs and Keystone Cliffs sections suggests that a general correlation can be made
between the uppermost levels of the Fossil Bluff Formation and the 1375 m level on James Ross
Island (text-figs. 10 and 1 1).

Other high level inoceramids in the Fossil Bluff Formation include l.flemingi sp. nov., a probable
member of the I. liwerowskyae group, and three species (/. cf. anglicus elongatus , I. sp. aff. bellvuensis,
and I. sp. aff. comancheanus) that have their strongest affinities with the cosmopolitan I. anglicus
group. Both the latter three and I. sp. aff. anglicus from Dundee Island can be compared with
specimens from a wide range of Northern Hemisphere localities. Representatives of the I. carsoni
group in the James Ross Island succession provide a direct link between the Antarctic Peninsula and
Great Artesian Basin of Australia; in particular, I. carsoni indicates a correlation of the 1 1 40- 1 250 m
interval with stratigraphic units such as the Allaru Mudstone of Queensland (Day 1969). Anopaea sp.
nov. aff. mandibula may furnish another connection between the Aptian-Albian Fossil Bluff
Formation faunas and those of the south-western USSR.

The Fossil Bluff Formation apparently terminates in the Albian (Taylor et al. 1979), but the James
Ross Island succession passes up into Cenomanian and younger beds (Crame 1983fi).

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

521

Acknowledgements. I would like to thank Mr. C. P. Nuttall, Dr. N. J. Morris, and Mr. R. Cleevely for allowing
me ready access to the Mesozoic Mollusca collections of the Department of Palaeontology, British Museum
(Natural History). On a visit to Australia, Dr. N. H. Ludbrook kindly arranged for me to examine the collections
held in the South Australian Department of Mines and Energy and Dr. B. P. Webb (Director-General)
subsequently allowed me to borrow a number of specimens from them. Mr. P. J. G. Fleming facilitated my
examination of the Geological Survey of Queensland’s collections and I would also like to acknowledge use of a
set of inoceramid plaster casts supplied by the New Zealand Geological Survey.

My colleagues Dr. M. R. A. Thomson, Mr. J. R. Ineson, and Dr. G. W. Farquharson offered constructive
criticisms of a number of points raised in the text. Dr. B. Haigh (Cambridge) translated some of the cited Russian
material and Mr. C. Gilbert of the British Antarctic Survey took the photographs.

REFERENCES

airaghi, c. 1904. Inocerami del Vento. Boll. Soc. geol. ital. 20, 178-198.

anderson, f. m. 1938. Lower Cretaceous deposits in California and Oregon. Spec. Pap. geol. Soc. Am. 16,
1-339.

1945. Knoxville Series in the California Mesozoic. Bull. geol. Soc. Am. 56 , 909-1014.
auerbach, j. and frears, h. 1846. Notices sur quelques passages de l’ouvrage de MM. Murchison, E. de
Verneuil et la Comte A. de Keyserling: Geologie de la Russie d’Europe et des montagnes de l’Oural. Byull.
mosk. Obshch. Ispyt. Prir. 35 , 486-500.

bibby, j. s. 1966. The stratigraphy of part of north-east Graham Land and the James Ross Island group. Sclent.
Rep. Br. Antarct. Surv. 53 , 1-37.

bonarelli, g. and nagera, j. j. 1921. Observaciones geologicas en las inmediaciones del Lago San Martin
(Territorio de Santa Cruz). Boln Dir. gen. Minas Geol. Hidrol . , B. Aires, Ser. B. 27 , 1-39.
casey, r. 1961 . The stratigraphical palaeontology of the Lower Greensand. Palaeontology , 3 , 487-621.
chiplonkar, G. w. and badve, r. m. 1976. On some inoceramid species occurring in the Bagh Beds of Narmada
Valley. J. Univ. Poona , 48 , 195-209.

cox, l. r. 1969. Family Inoceramidae Giebel, 1852. In moore, r. c. (with teichert, c., mccormick, l. and
williams, R. B.) (eds. ). Treatise on invertebrate paleontology. Pt. N , Vol. 7, Mollusca 6. Bivalvia , N314-N321.
Geological Society of America and University of Kansas Press.
crame, j. a. 1980. The occurrence of the bivalve Inoceramus concentricus on Dundee Island, Joinville Island
group. Bull. Br. Antarct. Surv. 49 , 283-286.

1981. Upper Cretaceous inoceramids (Bivalvia) from the James Ross Island group and their stratigraphical
significance. Ibid. 53, 29-56.

1982. Late Jurassic inoceramid bivalves from the Antarctic Peninsula and their stratigraphic use.
Palaeontology , 25 , 555-603.

1983a. Lower Cretaceous bivalve biostratigraphy of Antarctica. Zitteliana, 10 , 399-406.

1983fi. Cretaceous inoceramid bivalves from Antarctica. In Oliver, r. l., james, p. r. and jago, j. b. (eds.).
Antarctic Earth Science , 298-302. Australian Academy of Science, Canberra.
day, r. w. 1969. The Lower Cretaceous of the Great Artesian Basin. In Campbell, k. s. w. (ed.). Stratigraphy
and palaeontology. Essays in honour of Dorothy Hill, 140 173. Australian National University Press,
Canberra.

eighenheer, r. and sornay, j. 1974. Sur une forme apparentee a Inoceramus comancheanus Cragin dans le
Cenomanien basal de Montlaux (Alpes-de-Provence). Bull. Mus. natn. Hist, nat., Paris, 229 , 141- 144.
etheridge, R. 1872. Description of the Palaeozoic and Mesozoic fossils of Queensland. In daintree, r. Notes on
the geology of the colony of Queensland. Q. Jl geol. Soc. Lond. 28 , 317-350.

— 1907. Cretaceous fossils of Natal collected by Mr William Anderson, Government Geologist. Part II. The
Umsinene River deposit. In anderson, W. Third and final report of the Geological Survey of Natal and
Zululand, 67-90.

etheridge, R. jr. 1892. In jack, R. L. and etheridge, R. JR. The geology and palaeontology of Queensland and New
Guinea , with sixty-eight plates and a geological map of Queensland. 2 vols. Dulau and Co., London.

— 1901. Additional notes on the palaeontology of Queensland (part 2). Bull. geol. Surv. Qld, 13, 1-37.

— 1905. Contributions to the palaeontology of South Australia. No. 14. Cretaceous fossils from Dalhousie
Springs. In Report on geological explorations in the west and north-west of South Australia, 13-17. C. E.
Bristow, Government Printer, Adelaide.

farquharson, g. w. 1982. Late Mesozoic sedimentation in the northern Antarctic Peninsula and its relationship
to the southern Andes. J. geol. Soc. Lond. 139 , 721-728.

522

PALAEONTOLOGY, VOLUME 28

farquharson, G. w. 1983. Sedimentation associated with the Late Mesozoic volcanic arc of the northern
Antarctic Peninsula. Unpubl. Ph.D. thesis, C.N.A.A.

feruglio, E. 1936. Palaeontolographica patagonica. Pt. 1. Memoire 1st. geol. miner. Univ. Padova , 11,
1 192.

gillet, s. 1924. Etudes sur les lamellibranches neocomiens. Mem. Soc. geol. Fr. ( N.S. ), 1, fasc. 3-4, 1 224.
Glazunova, A. e. 1973. Paleontological evidence of the stratigraphic separation of Cretaceous deposits in the Volga
region, 200 pp. Lower Cretaceous. ‘Nedra’, Moscow. [In Russian.]
hayami, i. 1960. Jurassic inoceramids in Japan. J. Fac. Sci. Tokyo Univ., ser. II, 12, 277-328.
heinz, r. 1928a. Uber die bisher wenig beachtete Skulptur der Inoceramen-Schale und ihre stratigraphische
Bedeutung. Beitrage zur Kenntnis der oberkretazischen Inoceramen IV. Mitt, miner. -geol. Stlnst. Hamh., 10,
3-39.

— 1928 b. Uber die Kreide-Inoceramen Australiens und ihre Beziehungen zu denen Europas und anderer
Gebiete. Beitrage zur Kenntnis der oberkretazischen Inoceramen VIII. Ibid. 131-147.

- 1930. Ueber Kreide-Inoceramen der siidafrikanischen Union. C.r. 15th hit. geol. Congr., Pretoria, 1929, 2,
681-687.

1933. Inoceramen von Madagaskar und ihre Bedeutung fur die Kreide-Stratigraphie. Z. dt. geol. Ges.
85,241-259.

hill, d., playford, g. and woods, J. t. (eds.). 1968. Cretaceous fossils of Queensland, 35 pp. Queensland
Palaeontographical Society, Brisbane.

imlay, r. w. 1960. Ammonites of early Cretaceous age (Valanginian and Hauterivian) from the Pacific Coast
states. Prof. Pap. U.S. geol. Surv. 334-F, 167-228.

— 1961. Characteristic Lower Cretaceous megafossils from northern Alaska. Ibid. 335, 74 pp.
jeletzky, j. a. 1964. Illustrations of Canadian fossils. Lower Cretaceous marine index fossils of the sedimentary

basins of western and Arctic Canada. Geol. Surv. Pap. Can. 64-11, 1-101.

- 1970. Cretaceous macrofaunas. In DOUGLAS, r. j. w. (ed.). Geology and economic minerals of Canada,
649-662. Department of Energy, Mines and Resources, Ottawa.

- 1977. Mid-Cretaceous (Aptian to Coniacian) history of Pacific slope of Canada. In kanmera, k. (ed.). Mid-
Cretaceous Events. Hokkaido Symposium, 1976. Spec. Pap. palaeontol. Soc. Japan, 21, 97-126.

jones, d. l. 1960. Lower Cretaceous (Albian) fossils from southwestern Oregon and their paleogeographic
significance. J. Paleont. 34, 152-160.

and gryc, G. 1960. Upper Cretaceous pelecypods of the genus Inoceramus from northern Alaska. Prof.
Pap. U.S. geol. Surv. 334E, 149-165.

kauffman, e. G. 1977. Systematic, biostratigraphic and biogeographic relationships between Middle Cretaceous
Euro-american and North Pacific Inoceramidae. In kanmera, k. (ed.). Mid-Cretaceous Events. Hokkaido
Symposium, 1976. Spec. Pap. palaeontol. Soc. Japan , 21, 169-212.

1978n. British Middle Cretaceous inoceramid biostratigraphy. In Evenements de la partie moyenne du
Cretace. Uppsala 197 5 -Nice 1976. Annls. Mus. Hist. nat. Nice, 4 (1976), IV. 1 -IV. 12.

19786. South African middle Cretaceous Inoceramidae. Ibid. XVII. 1 -XVII. 6.

1978c. Middle Cretaceous bivalve zones and stage implications in the Antillean subprovince, Caribbean
province. Ibid . XXX. 1 -XXX. 11.

— Cobban, w. a. and eicher, d. l. 1978. Albian through lower Coniacian strata. Biostratigraphy and principal
events in western interior states. Ibid. XXIII. 1 -XXIII. 52.
kelly, s. R. a. 1984. Bivalvia of the Spilsby Sandstone and Sandringham Sands (Late Jurassic-Early Cretaceous)
of eastern England. Part I . Palaeontogr. Soc. [ Monogr .], 94 pp., 20 pis.
leanza, a. f. 1967. Anotaciones sobre los losiles jurasico-cretacicos de Patagonia austral (coleccion Feruglio)
conservados en la Universidad de Bologna. Acta geol. lilloana, 9, 121-188.
ludbrook, n. h. 1966. Cretaceous biostratigraphy of the Great Artesian Basin in South Australia. Bull. geol.
Surv. S. Aust. 40, 1 223.

lumholtz, c. 1889. Among Cannibals, 395 pp. John Murray, London.

m‘coy, f. 1865. Note on the Cretaceous deposits of Australia. Ann. Mag. nat. Hist. 16, 333-334.

1866. On the discovery of Cretaceous fossils in Australia. Trans. Proc. R. Soc. Viet. 7, 49-51.

1867. On the Recent Zoology and palaeontology of Victoria. Ann. Mag. nat. Hist. 20, 175-202.
mclearn, f. h. 1919. New species of pelecypods from the Cretaceous of northern Alberta. Bull. geol. Surv. Can.

Mus. 29, 9-12.

1933. Pelecypods of the Lower Cretaceous Clearwater Formation, northern Alberta. Trans. R. Soc. Can.
27, 139-156.

1943. Trends in some Canadian species of Inoceramus. Can. Fid. Nat. 57, 36-46.

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

523

matsumoto, t., okada, h., hirano, h. and tanabe, k. 1978. Mid-Cretaceous zonation in Japan. In Evenements
de la partie moyenne du Cretace. Uppsala 1975-Nice 1976. Annls Mas. Hist. nat. Nice , 4 (1976), XXXIII. I
XXXIII. 6.

maury, c. 1936. O Cretaceo de Sergipe. Monografias Serv. Geol. Min. Brasil , 1 1, 1-25.

meek, f. b. 1864. Check list of the invertebrate fossils of North America. Cretaceous and Jurassic. Smithson, misc.
Col/ns, 7, 1 -40.

nagao, T. and matsumoto, T. 1939. A monograph of the Cretaceous Inoceramus of Japan. J. Fac. Sci. Hokkaido
Univ., 4, 241-299.

newton, r. b. 1909. Cretaceous Gastropoda and Pelecypoda from Zululand. Trans. R. Soc. S. Afr. 1,
1-106.

orbigny, a. d’. 1846. Paleontologie franqaise . Description des mollusques et rayonnes fossiles. Terrains Cre faces.
3. Lamellibranches, V. 449-520. Masson, Paris.

Parkinson, j. 1819. Remarks on the fossils collected by Mr Phillips near Dover and Folkestone. Trans, geol. Soc.
Lond. 5, 52-59.

pergament, m. a. 1965. Inocerams and Cretaceous stratigraphy of the Pacific region. Trudy Inst. geol. Nauk ,
Mosk. 118, 1 102. [In Russian.]

— 1966. Zonal stratigraphy and inocerams of the lowermost Upper Cretaceous on the Pacific coast of the
USSR. Ibid. 146, 1-83. [In Russian.]

— 1981. Pacific regions of the USSR. In reyment, r. a. and bengtson, p. (eds.). Aspects of Mid-Cretaceous
regional geology, 69-102. Academic Press, London.

Pettigrew, T. H. 1981. The geology of Annenkov Island. Bull. Br. Antarct. Surv. 53, 213-254.
pokhialainen, v. p. 1969. Neocomian inocerams of the Anadyrskiy-Koryakskiy folded region. Trudy sev.-vost.
kompl. nauchno-issled. Inst. 32, 1 18-123. [In Russian.]

— 1974. Osobennosti rasprostraneniya inotseramid neokoma Tikhookeanskoi oblasti (Spreading of the
Neocomian Pacific inoceramids). Trudy Inst. Geol. Geofiz. sih. Otd. 80, 174-187. [In Russian.]

reeside, j. b. 1923. The fauna of the so-called Dakota Formation of northern central Colorado and its equivalent
in southeastern Wyoming. Prof. Pap. U S. geol. Surv. 131 H, 199-212.
riccardi, a. c. 1970. Favrella R. Douville, 1909 (Ammonitina, Cretacico inferior): edad y distribucion.
Ameghiniana, 7, 1 19-138.

1977. Berriasian invertebrate fauna from the Springhill Formation of Southern Patagonia. Neues Jb.

Miner. Geol. Palaont. Abh. 155, 216-252.

— westermann, G. e. G. and levy, r. 1971. The Lower Cretaceous Ammonitina Olcostephanus, Leopoldia and
Favrella from west-central Argentina. Palae onto graphic a, 136, Abt. A, Lief. 1-6, 83-121.
rouillier, c. 1849. In rouillier, c. and vossinsky, a. Etudes progressives sur la geologie de Moscou; quatrieme
etude. Byull. Mosk. Abshch. Ispyt. Prir. 22, 237-355.

Saveliev, A. A. 1962. Albian inoceramids of Mangishlak. Trudy vses. nauchno-issled. geol.-razv. neft Inst. 196,
219-254. [In Russian. |

schlagintweit, o. 1911. Die fauna des Vracon undCenoman in Peru. In steinman, g. Beitragezur Geologie und
Palaontologie von Siidamerika. XVII. Neues Jb. Miner. Geol. Palaont. BeilBd, 33, 43-135.
scott, G. r. 1970. Paleoecology and paleontology of the Lower Cretaceous Kiowa Formation, Kansas. Paleont.
Contr. Univ. Kans. 52, 1 -94.

smellie, j. l., davies, R. e. s. and Thomson, m. r. a. 1980. Geology of a Mesozoic intra-arc sequence on Byers
Peninsula, Livingston Island, South Shetland Islands. Bull. Br. Antarct. Surv. 50, 55-76.
sornay, j. 1965. Les Inocerames du Cretace Inferieur en France. Mem. Bur. Rech. geol. minier. 34, 393-397.

— 1966. Idees actuelles sur les Inocerames d’apres divers travaux recents. Annls. Paleont., Invertebres , 52,
59-92.

— 1981. Inocerames. In Travail de synthese du groupe de travail sur les “ Evenements Biologiques" (Mid-
Cretaceous Events). Cretaceous Res. 2, 417-425.

sowerby, J. 1821. The Mineral Conchology of Great Britain, 3, 127-186, pis. 272-306. London.
speden, i. G. 1977. Taitai Series (Early Cretaceous) and the elimination of the Mokoiwian stage. N.Z. Jl Geol.
Geophys. 20, 537-562.

spengler, E. 1914. Nachtrage zur Oberkreidefauna des Trichinopoly-Distriktes in Sudindien. Beitr. Palaont.
Geol. Ost.-Ung. 26,213-239.

stanton, t. w. 1895. Contributions to the Cretaceous palaeontology of the Pacific coast. The fauna of the
Knoxville Beds. Bull. U.S. Geol. Surv. 133, 1-132.

stevens, g. r. and speden, i. g. 1978. New Zealand. In moullade, m. and nairn, a. e. m. (eds.). The phanerozoic
geology of the world, II. The Mesozoic, A. 251-328. Elsevier, Amsterdam, Oxford, and London.

524 PALAEONTOLOGY, VOLUME 28

stolley, E. 1912. Uber die Kreideformation und ihre fossilien auf Spitzbergen. K. svenska VetenskAkad. Hand!.
47, 1-29.

tanner, p. w. G. 1982. Geologic evolution of South Georgia. In craddock, C. (ed.). Antarctic geoscience ,
167-176. University of Wisconsin Press, Madison.

— storey, b. c. and Macdonald, d. i. m. 1981. Geology of an Upper Jurassic-Lower Cretaceous island-arc
assemblage in Hauge Reef, The Pickersgill Islands and adjoining areas of South Georgia. Bull. Br. Antarct.
Surv. 53, 77 - 1 1 7.

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.

Thomson, m. r. a. 1967. A probable Cretaceous invertebrate fauna from Crabeater Point, Bowman Coast,
Graham Land. Bull. Br. Antarct. Surv. 14, 1 14.

-1974. Ammonite faunas of the Lower Cretaceous of south-eastern Alexander Island. Scient. Rep. Br.
Antarct. Surv. 80, 1 -44.

1979. Upper Jurassic and Lower Cretaceous ammonite faunas of the Ablation Point area, Alexander
Island. Ibid. 97, I -37.

1981. Late Mesozoic stratigraphy and invertebrate paleontology of the South Orkney islands. Bull. Br.
Antarct. Surv. 54, 65-83.

— 1982. A comparison of the ammonite faunas of the Antarctic Peninsula and Magallanes Basin. J. geol. Soc.
Lond. 139, 763-770.

- 1983. ‘European’ ammonites in the Lower Cretaceous of Antarctica. Zitteliana , 10, 407-412.

and farquharson, g. w. 1984. Discovery and significance of the ammonite genus Favrella in the Antarctic
Peninsula area. Bull. Br. Antarct. Surv. 62, 7-14.

— and willey, L. E. 1972. Upper Jurassic and Lower Cretaceous Inoceramus (Bivalvia) from south-east
Alexander Island. Ibid. 29, 1 19.

— tanner, p. w. g. and rex, d. c. 1982. Fossil and radiometric evidence for ages of deposition and meta-
morphism of sedimentary sequences on South Georgia. In craddock, c. (ed.). Antarctic geoscience, 177-184.
University of Wisconsin Press, Madison.

trautschold, h. 1865. Der Inoceramen-Thon von Simbirsk. Bull. Soc. Nat. Moscou , 38, 1-24.
troger, k.-a. 1981. German Democratic Republic. In reyment, r. a. and bengtson, p. (eds.). Aspects of Mid-
Cretaceous regional geology, 1-28. Academic Press, London.
warren, p. s. and stelck, c. r. 1940. Cenomanian and Turonian faunas in the Pouce Coupe district, Alberta
and British Columbia. Trans. R. Soc. Can. 34, 143-152.
weaver, c. 1931. Palaeontology of the Jurassic and Cretaceous of west central Argentina. Mem. Univ. Wash. 1,
1-169.

wellman, h. w. 1959. Divisions of the New Zealand Cretaceous. Trans. R. Soc. N.Z. 87, 99-163.
whiteaves, j. f. 1876. On some invertebrates from the coal-bearing rocks of the Queen Charlotte Islands,
collected by Mr James Richardson in 1872. In Mesozoic fossils, 1, Pt. I, 1-92. Geological Survey of Canada,
Montreal.

— 1883. On the Lower Cretaceous rocks of British Columbia. Trans. R. Soc. Can. 1, 81-86.
whitehouse, f. w. 1924. The Queensland Inocerami collected by M. Lumholz in 1881. Proc. R. Soc. Qd , 35,
127-132.

wiedmann, j. and kauffman, e. g. 1978. Mid-Cretaceous biostratigraphy of northern Spain. In Evenements de la
partie moyenne du Cretace. Uppsala 1975-Nice 1976. Annls Mus. Hist. nat. Nice, 4 (1976), III. 1 -III. 34.
wilckens, o. 1947. Palaontologische und geologische Ergebnisse der Reise von Kohl-Larsen (1928-29) nach
Siid-Georgien. Ahh. senckenb. naturforsch. Ges. 474, 1-66.
willey, l. E. 1972. Belemnites from south-eastern Alexander Island: 1. The occurrence of the family
Dimitobelidae in the Lower Cretaceous. Bull. Br. Antarct. Surv. 28, 29-42.
wollemann, a. 1906. Die Bivalven und Gastropoden des nordeutschen Gaults (Aptiens und Albiens). Jb.
preuss. geol. Landesanst. 27, 259-300.

woods, h. 191 1. A monograph of the Cretaceous Lamellibranchia of England. Palaeontogr. Soc. [ Monogr .], 2,
Pt. 7, 261-284.

— 1912. The evolution of Inoceramus in the Cretaceous period. Q. Jl geol. Soc. Lond. 68, 1 -20.

1917. The Cretaceous faunas of the north-eastern part of the South Island of New Zealand. Palaeont. Bull.,
Wellington, 4, 1-41.

zakharov, v. A. 1 966. Late Jurassic and early Cretaceous bivalve molluscs of the north of Siberia and their ecology.
Part 1. Order Anisomyaria, 191 pp. ‘Nauka’, Moscow. [In Russian.]

— 1 968. Changes of complexes of bivalve species at the boundary between the Jurassic and Cretaceous

CRAME: CRETACEOUS INOCERAMID BIVALVES FROM ANTARCTICA

525

periods in the Boreal Arctic zoogeographic regions. Trudy Inst. Geol. Geofiz. sib. Otd. 48, 90-100. [In
Russian.]

and turbina, a. s. 1979. Early Neocomian inoceramids from northern Siberia and their role in benthic
assemblages. In saks, v. n. and zakharov, v. a. (eds.). Conditions of the existence of the Mesozoic marine
Boreal fauna. Ibid. 411 , 23 36. [In Russian.]

Manuscript received 23 May 1984

Revised manuscript received 28 September 1984

J. A. CRAME

British Antarctic Survey
Natural Environment Research Council
High Cross
Madingley Road
Cambridge CB3 OET

CORONATE ECHINODERMS FROM THE
LOWER PALAEOZOIC OF BRITAIN

by STEPHEN K. DONOVAN and CHRISTOPHER R. C. PAUL

Abstract. Coronates are pelmatozoan echinoderms with a functional stem, a bud-shaped theca, and erect,
biserial, pinnate arms. They evolved early in the Middle Ordovician, probably from the ‘eocrinoid’ Bockia ,
and gave rise to the blastoids sensu stricto. The subclass Coronata contains six genera. Mespilocystites (early
Caradoc-Ashgill) had geniculate radial furrows, while all later genera had planar or gently convex radii. Of these
Stephanoblastus (Caradoc-Wenlock) had a very narrow stem and triradiate keels at the base of the theca. All
other genera had triangular thecal bases. Tormoblastus (Ashgill) had a conical theca with a protruding base
bearing three flanges, Paracystis (Caradoc) had a bowl-shaped theca with a sunken base, Stephanocrinus
(AshgilULudlow) a tall, steeply conical theca, and Cupulocorona gen. nov. (Ashgill-Wenlock) a conical to
cup-shaped theca with a protruding base. The British coronate fauna includes five new species: S. ramsbottomi
(Hirnantian) characterized by a large, angular conical theca with low coronal processes, C. salopiae (early
Wenlock) characterized by a pyriform theca with low coronal processes and low ridges at the plate sutures,
C. rugosa (Cautleyan Rawtheyan) with a conical theca bearing very coarse ornament, C. digitalis (Cautleyan-
Rawtheyan) with a conical theca and very long coronal processes, and Stephanocrinus sensu lato sp. (Cautleyan)
which is poorly known but had very fine ribbing.

Coronates are a small but distinctive group of Lower Palaeozoic pelmatozoans with a functional
stem, a small theca with five interradial coronal processes ventrally and a fixed arrangement of thecal
plates very similar to that of blastoids, and erect, biserial, pinnate arms. Opinions as to their affinities
have varied in the past, but we believe they were most closely related to the blastoids. Although thecae
of coronate echinoderms are not uncommon in the Ordovician and Silurian of Britain (no columns or
arms have yet been discovered), the only previous references to their occurrence are in Bather (1900,
pp. 96, 145), who mentioned that Stephanocrinus was found in the Silurian of Britain, and King
and Wilcockson (1934, p. 17), in which Bather recognized a calyx from the Upper Ordovician
at Hunterstye as Stephanocrinus. The latter is described below as S. ramsbottomi sp. nov. In 1952
W. H. C. Ramsbottom prepared an unpublished manuscript on some coronates including a proposed
new species, S. salopiae , which is probably the species referred to by Bather (1900) and is redescribed
here as Cupulocorona salopiae. Otherwise British coronates have been ignored or referred to as ‘new
cystidean’ (Reed 1907, p. 537). Following the recent reassessment of Stephanocrinus ( Brett et al. 1983),
which elevated the Coronata to class status, it is timely to describe British representatives of the group.

GENERAL MORPHOLOGY OF THE CORONATES

Stem. Apparently homeomorphic, composed of small circular columnals with convex latera, narrow
(TA thecal diameter) or extremely narrow (A thecal diameter), circular in section and with a
circular lumen as far as is known. Brett et al. (1983, fig. Ik) illustrated a discoidal holdfast which they
associated with S. angulatus.

Theca. Elongate-conical through conical to cup- or bowl-shaped, base triangular or triradiate, top
with five distinct coronal processes. Constant plate arrangement (text-fig. 5a) consisting of three
basals of which the azygous basal occupies the AB interray; five radials all with cleft upper margins,
the two ‘wings’ each forming half of the aboral side of a coronal process; six deltoids, a sub- and
super-deltoid in the CD (posterior) interray adoral to the anus and one deltoid in each of the other

[Palaeontology, Vol. 28, Part 3, 1985, pp. 527-543, pis. 62-63-1

528

PALAEONTOLOGY, VOLUME 28

four interrays, deltoids triangular and raised to form the adoral portions of the coronal processes; five
primary ambulacral plates at the tip of each ambulacrum and in the radial furrows between the
coronal processes; five oral cover plates over the central mouth, the posterior of which is very slightly
the largest; ten elongate ambulacral cover plates, one pair per ambulacrum. The basals and radials are
usually ornamented with fine ridges or rows of granules which form rhombic patterns. In some genera
there are also major ridges on the theca diverging upwards from the basals and forming a V W pattern,
the base of the V being centred on the azygous basal.

Three orifices all in the oral surface (text-fig. 5b). A central, nearly circular mouth. An anus, almost
as large as the mouth, situated at the base of the CD coronal process on the adoral side, and covered
with an anal pyramid of three or four cover plates. Just adoral to the anus on the common suture of the
sub- and super-deltoid a small tubercle with a transverse slit is sometimes found. This has been
interpreted as a possible hydropore (Brett et al. 1983, p. 632). Both suture and orifice are frequently
cryptic.

Subvective system. The ambulacra commence as five broad main grooves in the radial furrows
between the coronal processes. In the floor of each main groove are two narrow food grooves which
may unite near the mouth, but in some species remain separate along the entire length of the main
groove. The main grooves terminate in a primary ambulacral plate, which bears two rounded facets of
which the left (as viewed from the mouth down the ambulacrum) is always the smaller. The free arms
are biserial and give rise to biserial lateral branches (brachioles) alternately. The arms are strongly

text-fig. 1 . Standard measurements of coronate thecae, a, b, lateral views of thecae, c, base of theca.
D, ambulacrum. H, height of theca, H DMAX , height of theca at maximum diameter, H ORAL , height of oral surface,
H BAS al’ height of basal circlet, D, maximum diameter of theca, D ORAL , diameter of oral surface, D BASE , diameter
of base of theca, KD, diameter of columnal facet, L AM , length of ambulacrum, L FG , length of food groove.

DONOVAN AND PAUL: CORONATE ECHINODERMS

529

coiled when at rest (like the proboscis of a butterfly) and must have been considerably longer than the
coronal processes when extended for feeding.

Measurements taken from specimens are explained in text-fig. 1 and summarized in Table 1.
Although many parameters can be defined, it is not easy to differentiate between species graphically
(text-fig. 2) due to the extreme similarity of thecae. Three bivariate plots are given; thecal height
against height of the oral surface (2A), ambulacral length against length of radial furrows (2B), and
thecal height against thecal diameter (2C). The best separation is given by text-fig. 2b, although the
graphs suffer from a lack of data as most of the available specimens are not sufficiently well preserved
to allow many measurements to be taken.

12

Hmm

L M mm

-AM * 12

H mm

11

A A

11

A

-

10

-

10

9

-

* 4 9

-

8

2

□

* 8

_

7

A

□

r\

<

<

□

6

■ □

□

6

.

5

□

■*o A

to

□

□ □□

_

4

°

a °Fdigitalis 1

A 4

-

3

- salopiae

3

2

° rugosa

2

-

1

‘ ramsbottomi

i i i i i i i i i —

1

I i

1 L

123456789
hL „ A1 mm

B

L F G mm

C

1 2 3 4 5 6 7

D mm

text-fig. 2. Bivariate plots of some of the measurements shown in text-fig. 1. A, height against height of oral
surface, b, length of ambulacrum against length of food groove, c, height against diameter.

CHARACTERS OF THE INCLUDED GENERA

Brett et al. (1983) provided a very detailed and accurate description of S. angulatus Conrad, 1842, the
first coronate to be described. They thus settled the basic characters of the type species of the type
genus of the only family recognized within the Coronata. However, the status and detailed
morphology of the other genera remain uncertain, as do the phylogenetic relationships of the group as
a whole. North American coronates, including S. angulatus , are exclusively Silurian, whereas in
Europe Ordovician representatives are definitely known in Britain (King and Wilcockson 1934, and
herein), Bohemia (Barrande 1887), France (Chauvel and Le Menn 1973), Spain (Chauvel and Le Menn
1979), and Sweden (Regnell 1945) and probably occur in Portugal as well (Delgado 1908). Coronates
form a small but very distinctive group of pelmatozoan echinoderms. As far as is known all species and
genera share an identical arrangement of thecal plates. Features which vary are general thecal shape,
relative proportions of the coronal processes, surface ornament, and relative size of the stem
(text-fig. 3), none of which is usually considered the type of character on which genera should be
defined. Indeed it is not difficult to argue the case for assigning all known species to one genus, for
which Stephanocrinus Conrad, 1842, is the oldest available name.

530

PALAEONTOLOGY, VOLUME 28

text-fig. 3. Thecal profiles in coronate genera, a, Stephanocrinus. b, Mespilocystites. c, Para-
cystis. D, Tormoblastus. E, Stephanoblastus. F, Cupulocorona gen. nov. All based on the type
species of the genera. An, anus,/, primary ambulacral plate.

Nevertheless, some of the variations within the Coronata seem to be of phylogenetic significance.
Mespilocystites bohemicus Barrande, 1887 (text-fig. 3b) from the Lower Caradoc of Trubsko, Czecho-
slovakia, is the oldest known species and a suitable starting point for comparisons (see Paul 1985, for
a full description). It is characterized by a relatively low, broadly conical cup with very prominent
coronal processes reaching half the total height which flare outwards so that their tips mark the widest
part of the theca. All later forms have coronal processes that curve inwards towards their tips and in
most the widest part of the theca is level with the oral surface. M. bohemicus also has strongly
geniculate radial furrows between the coronal processes. The ambulacral grooves radiate horizontally
from the mouth and terminate in facets for the arms which lie about half-way from the mouth to the
periphery of the oral surface. Beyond the arm facets the radial furrows slope steeply downwards at an
angle of about 40°. All other species, except M. tregarvanicus Le Menn from the Ashgill of Brittany,
have planar or gently convex radial furrows, although the position of the arm facet varies from species
to species. The base of the theca in M. bohemicus protrudes slightly and is triangular in cross-section.
Within the triangle is a narrow, circular stem facet about one-tenth the diameter of the oral surface.
The ornament of the main cup plates consists of obvious fine ridges which form rhombic patterns
across plate sutures.

Several changes to this basic morphology had already occurred in the Caradoc, from which two
Swedish species are known. Paracystis ostrogothicus Sjoberg, 1915 (text-fig. 3c), differs in having
planar radial furrows, much lower coronal processes, and a stem facet which is impressed into the
base of the theca. The corners of the basal triangle hang down slightly over the stem and the profile
of the cup is a regular bowl shape. The surface ornament is composed of much coarser ridges
than those of M. bohemicus and there are even more prominent ridges which form a V W pattern

DONOVAN AND PAUL: CORONATE ECHINODERMS

531

around the theca, with the base of the V centred on the azygous basal, i.e. in the AB interray. The
other Swedish Caradoc species, although discovered by F. A. Bather in 1907, has yet to be described.
It has a slightly angular, bud-shaped theca with a minute stem facet about one-twenty-fifth of the
diameter of the oral surface. From this facet three sharp ridges radiate to the points of the V and W
angles on the lateral surface of the theca, so that the base of the theca is triradiate, not triangular.
Otherwise cup plates are almost completely smooth and show only faint traces of concentric
growth lines. Stephanoblastus minis (Barrande 1887), from the Wenlock of Czechoslovakia is
the only other described species with a triradiate base (text-fig. 3e) and is similar to the Swedish
species, but it is more elongate and has very fine rhombic ridges on cup plates. S. minis was made
type species of Stephanoblastus by Jaekel (1918, p. 110) and it seems that this line became established
in the Caradoc.

Tormoblastus bodae Jaekel, 1927 (text-fig. 3d), is known from a unique type specimen which has a
conical theca with prominent VW ridges and a protruding triangular base which bears three
horizontal flanges. Each flange is formed by two adjacent basals. Most of the remaining species fall
into two distinct genera, Stephanocrinus and a new genus. S. angulatus Conrad, 1 842 (text-fig. 3a), type
species of Stephanocrinus, has a tall, steeply conical theca with tall coronal processes and a thecal
profile that is slightly concave at the level of the basal : radial sutures. It also has prominent ridges in a
V W pattern as well as finer ridges in rhombs. The second genus has a low, conical to cup-shaped theca
with a triangular basal prominence, no V W ridges but finer ridges in rhombic sets, and generally low
coronal processes (text-fig. 3f). It resembles M. bohemicus except that the coronal processes do not
flare, the radial furrows are usually planar, and the theca is generally cup-shaped rather than conical.
This genus lacks a suitable name and we propose Cupulocorona for it (see below). Both Stephanocrinus
and Cupulocorona were already present in the Ashgill. S. ramsbottomi occurs in the Hirnantian of
northern England, south-west Wales, and in the Boda Limestone (Ashgill) at Osmundsberget and
Skalberget, Dalarna, Sweden, while a species of Cupulocorona also occurs in the Boda Limestone at
Boda and Osmundsberget, and three new species from the Ashgill of Britain are described below. Our
view of the evolutionary relationships of these genera is shown in text-fig. 4.

text-fig. 4. Cladogram showing inferred relationships between
coronate genera. Synapomorphies (18) as follows: 1, loss of
geniculate radial furrows; 2, development of V W pattern of ridges
or angles; 3, very narrow stem and triradiate base to theca;

4, development of two additional ridges between V and W ridges;

5, tall conical theca with concave lateral surfaces; 6, ornament of
coarse ridges only; 7. basal flanges; 8, bowl-shaped theca.

CO

CD

co

o

o

5.

co

CD

s

CD

C

s

o

o

o

3

§

o

CO

3

-5

3

o

c

CX3

-C

Q.

Q5

CO

co

c

o

o

c

CTJ

-C

&

00

-5

-2

532

PALAEONTOLOGY, VOLUME 28

PHYLOGENETIC RELATIONSHIPS

The systematic position of coronates has given rise to considerable debate. Coronates have been
variously assigned to the crinoids (e.g. Wachsmuth and Springer 1886, p. 283; Bather 1900, p. 145; Fay
1962, p. 209; 1978, p. T575) or the blastoids (e.g. Etheridge and Carpenter 1886; Jaekel 1918, p. 109).
Sprinkle (1980) considered them to be ancestral to the blastoids, while Brett et al. (1983) erected a new
class, the Coronoidea, for them. Coronates share common features with blastoids and the enigmatic
Silurian genus Lysocystites, most recently described by Sprinkle (1973, p. 139). Blastoids, coronates,
and Lysocystites all have very similar arrangements of cup plates, including the following: three basals
with the azygous basal in the AB interray, five radials, four single deltoids and two or more anal
deltoids in the CD interray, and five radially positioned plates at the bases of the ambulacra. This
combination of characters is not found in any other group of pelmatozoans.

The radially positioned plates at the bases of the ambulacra demonstrate most clearly the close
affinities of blastoids, coronates, and Lysocystites. Blastoids differ from all other pelmatozoans in
having hydrospires and lancet plates beneath the ambulacra. Hydrospires are an autapomorphy of
the Blastoidea, being without homologue in other echinoderms; the lancet plate is not. Coronates
have a single plate at the base of each arm which lies between the radial and two adjacent deltoids
(text-fig. 5b). An erect arm with alternate biserial branches (brachioles) arises from this plate. As
Sprinkle ( 1 980) pointed out, to derive a blastoid ambulacrum from a coronate arm requires extension
of the primary ambulacral plate to become the lancet and the development of a recumbent
ambulacrum on top of it. Lysocystites (text-fig. 5c, d) has five elongate, radially positioned plates
between the deltoids at the adoral ends of which are facets for erect ambulacra of unknown structure.

text-fig. 5. Plate diagrams in coronates and Lysocystites. a, b.
Coronates. A, lateral view (broken lines indicate V and W ridges, dotted
lines additional ridges found in Stephanocrinus, Tormoblastus, and
Paracystis)\ B, oral surface, c, d, Lysocystites. c, lateral view (dotted lines
indicate outlines of triradiate thecal pore structures); d, oral surface.
A-E, Carpenter radii; An, anus; BB, basal plates; DD, deltoid plates;
G, gonopore; H, hydropore; M, mouth; RR, radial plates; supposed
homologues of blastoid lancet plates stippled. Note that these plates
bear the facets for erect arms and always lie in a radial position between a
radial plate and two deltoids. Note also that the mouth frame is
composed of five deltoids even in Lysocystites where they are elongate.

DONOVAN AND PAUL: CORONATE ECHINODERMS

533

These plates lie between two deltoids and are in contact with a radial. In our view they, too, are
homologues of the lancet plate in blastoids. Coronates, blastoids, and Lysocystites differ in their pore
structures. Lysocystites has unique pore structures running to the corners of the basals, radials, and
deltoids. Coronates have n-shaped canals within the coronal processes (see Brett et al. 1983, for a
thorough description), while blastoids have hydrospires. Brett et al. (1983) have argued that the
differences between the ambulacra and pore structures in the Blastoidea and Coronata warrant the
recognition of a separate class for the latter. Whether one accepts this new class or not, Lysocystites
has the same taxonomic status as coronates and blastoids. Since it is clearly possible to infer
relationships between blastoids, coronates, and Lysocystites (text-fig. 6), an alternative course of
action is to unite the three within an enlarged class Blastoidea. With the exception of the inclusion of
Lysocystites and the omission of the parablastoids, this is essentially the classification advanced by
Jaekel (1918, p. 107 et seq.).

Blastoidea

text-fig. 6. Cladogram showing inferred relationships between the
‘eocrinoid’ Bockia, the coronates, and blastoids. Synapomorphies (1-8) as
follows: 1, three basal plates; 2, erect, biserial, pinnate arms; 3, cup formed
by BB, RR, and DD only; 4, ambulacra on distinct base-plate ( = lancet in
eublastoids); 5, elongate ambulacral base-plate; 6, recumbent ambulacra;
7, hydrospires; 8, spiracles.

Eublastoidea

■Se

o

o

CQ

o

o

00

o

o

00

There remains the question of the ancestry of the Blastoidea, as modified and enlarged here. The
most likely candidate is the ‘eocrinoid’ Bockia, even though at first sight it differs dramatically from
blastoids. Bockia shares the following characters with blastoids sensu lato: a narrow circular stem,
a cup with three basal plates (unfortunately the position of the smallest basal with respect to
the Carpenter ambulacra remains unknown), and erect, biserial arms with alternating biserial
lateral branches (brachioles) which arise from an oral prominence that had four normal deltoids
and three posterior deltoids in the CD interray. The ambulacral structure is similar to that found
in coronates, although not absolutely identical (see Bockelie 1981, for a more complete description
of Bockia ). To transform Bockia into a coronate requires cessation of plate addition early in growth,
but after the basals, radials, and deltoids had formed, development of coronal processes, and the in-
corporation of the primary ambulacral plates into the theca. These last plates may have been derived
from the small plates in the oral prominence of Bockia or alternatively they may be enlarged
ambulacral plates. Regrettably the details of this part of the cup in Bockia are insufficiently known to

534 PALAEONTOLOGY, VOLUME 28

settle this point. The blastoid lineage was relatively successful, but ultimately became extinct in
the Permian.

SYSTEMATIC PALAEONTOLOGY

Subphylum blastozoa Sprinkle, 1973
Class blastoidea Say, 1825
Subclass coronata Jaekel, 1918
Family stephanocrinidae Wachsmuth and Springer, 1886

Definition. Pelmatozoans with a slender homeomorphic stem; theca formed of three basals (the
azygous in the AB interray), five radials, six deltoids (a sub- and super-deltoid in the CD interray and
one each in the other four interrays), five primary ambulacral plates each bearing two ambulacral
facets, five interradial oral cover plates of which the one in the CD interray is slightly the largest, ten
long, thin ambulacral cover plates (a pair in each ambulacrum); deltoids and radials produced into
erect coronal processes containing n-shaped canals; oral surface with three orifices, a large central
mouth, a moderately large anus in the CD coronal process and covered by a pyramid of three or four
anal cover plates, and a small slit-like hydropore on the sub-superdeltoid suture between the anus and
the mouth; primary ambulacral plates bearing erect, biserial, pinnate arms with biserial lateral
branches (brachioles), the two facets of the primary ambulacral plate bearing the main arm trunk and
the first lateral branch.

Genus stephanocrinus Conrad, 1842

Type species. By monotypy, Stephanocrinus angulatus Conrad, 1842, from the Middle Silurian, Rochester Shale,
of New York State.

Diagnosis. Narrow stem, large, elongate, conical theca with a triangular base, often with concave sides
and with low or high coronal processes, planar radial furrows, and ornament of very strong V W
ridges as well as finer ridges in rhombic sets.

Stephanocrinus ramsbottomi sp. nov.

Plate 62, figs. 1, 5, 6, 8; Plate 63, fig. 6; text-figs. 7a, 8b
vl907 New cystidean; Reed, p. 537.

vl934 Stephanocrinus sp.; Bather in King and Wilcockson, p. 17.

Derivation of species name. In honour of Dr. W. H. C. Ramsbottom.

EXPLANATION OF PLATE 62

Figs. 1, 5, 6, 8. Stephanocrinus ramsbottomi sp. nov. 1, 8, holotype, BMNH E25427. 1, lateral view of theca, x 8.
8, oral surface, x 8-3. 5, BMNH E25428. Anus, two ambulacra, and part of the mouth, x 22. 6, BMNH
E25426. Ambulacrum showing paired food grooves (mouth below), x 50.

Figs. 2, 3, 9, 10. Cupulocorona salopiae gen. et sp. nov. 2, BMNH E45531. Lateral view of the theca, x 15.
3, BMNH E45532. Lateral view of theca, x 1 1 . 9, BMNH E6393. Oblique view of the triangular base of theca.
The stem articulated on the depressed, circular facet, x 28. 10, BMNH E45530. Sutures between two radials
(left) and a basal, preserved in positive relief, x 24.

Fig. 4. Cupulocorona rugosa gen. et sp. nov. SM A3 1002a. Lateral view of theca, x 12.

Fig. 7. ? Stephanocrinus s. I. sp. BMNH E69218. Lateral view of theca (orientation unknown), x 13.

All scanning electron micrographs of latex casts from natural moulds, except figs. 2, 3, 9, 10.

PLATE 62

DONOVAN and PAUL, coronate echinoderms

536

PALAEONTOLOGY, VOLUME 28

Diagnosis. Low, triangular coronal processes, with ambulacral facets about half-way from the
mouth to the periphery of the oral surface, and with a prominent convex ridge aboral to the
facets.

Types. Holotype, British Museum, Natural History, (BMNH) E25427 from the Hirnantian at Hunterstye,
north-west Yorkshire. Eighteen paratypes in BMNH and Sedgwick Museum, Cambridge (SM).

Other material. Twenty-one thecae, all of which are external moulds unless otherwise stated. BMNH E25426,
E25427a, b , E25428a-d (a = internal mould); SM A31830, 1, 2a, b, 3 a, b, 4-7 (internal moulds), A32077, A39061,
2a, b, A41084a-d (a = internal mould), A109797/8 (part and counterpart), X.772-X.774.

Horizons and localities. This species is known from three British localities, two of which indicate a definite
Hirnantian age.

1. Ashgill Shales, old quarry on Hunterstye, Upper Crummackdale, Austwick district, north-west York-
shire, NGR SD 780716. King and Wilcockson (1934, p. 17) also identified Phacops cf. mucronata, Meristina
crassa, and Rafinesquina cf. Strophomena hirnantensis from this locality. Williams et al. (1972) regard the
Ashgill Shales in this area to be of Hirnantian age, based on assessment of the brachiopod fauna by Wright
(1968, table 3).

2. St. Martin’s Cemetery Beds, roadside 1 10 m west of St. Martin’s Cemetery, Haverfordwest, Dyfed, South
Wales (Reed 1907, Cantrill 1907). Hirnantian (Williams et al. 1972, fig. 5).

3. North side of the bend in the lane from Keisley Hamlet to the west quarry, Keisley, Westmorland, NGR NY
712238. The precise horizon is unknown, but may be Lower Llandovery in age (Wright 1982).

In addition, specimens from the Boda Limestone (Ashgill) at Osmundsberget and Skalberget, Dalarna,
Sweden, appear to belong to this species.

Description. Theca angular, elongate conical, with a pentagonal oral surface (PI. 62, fig. 8; text-fig. 8b) and
triangular base. The five angles of the oral surface joined to the three at the base by prominent ridges which form
a V (PI. 63, fig. 6) with its base in the AB interray, and a W with the tops in the C, D, and E rays. Faces of the theca
slightly concave (PI. 62, fig. 1; PI. 63, fig. 6; text-fig. 7a). Stem facet round, usually more than half the diameter of
the triangular base. Coronal processes triangular, varying from low with straight sides (PI. 63, fig. 6) to high with
incurved sides (PI. 62, fig. 1). Ornament of low ridges arranged in rhombic patterns (PI. 62, fig. 1; PI. 63, fig. 6).
Ambulacral grooves extend about half-way along the radial furrows (PI. 62, fig. 5). Ambulacral, oral, and anal

text-fig. 7. Thecal profiles in British coronates, a, Stephanocrinus
ramsbottomi, holotype, BMNH E25427. b, Cupulocorona salopiae,
holotype, BMNH E6390. c, e, C. rugosa. c, paratype, SM A31002n.
E, paratype, SM A3 12616. d, C. digitalis, holotype, SM A40229.

DONOVAN AND PAUL: CORONATE ECHINODERMS

537

cover plates unknown. Mouth central, circular. Anus circular, about half the diameter of the mouth (PI. 62, figs. 6,
8; text-fig. 8b). Oral surface slopes slightly away from the mouth.

Discussion. S. ramsbottomi is the largest British coronate and is also the earliest member of the genus
Stephanocrinus as we understand it. The earliest American species are from the Upper Llandovery
(C 5 ), Hopkington Formation (Brett et al. 1983, p. 632). The present species differs from S. angulatus in
having a plate ornament of ridges rather than rows of tubercles, lower coronal processes, the
ambulacral facets nearer the mouth, and a more angular theca. It differs from other British coronates
in not having a cup-like outline (text-fig. 7).

? Stephanocrinus sensu lato sp.

Plate 62, fig. 7; Plate 63, fig. 1

Material. A single, incomplete external mould, BMNH E69218, collected by C.R.C.P.

Horizon and locality. Sholeshook Limestone (Cautleyan; Price 1973, 1980), Sholeshook railway cutting, near
Haverfordwest, Dyfed, South Wales, NGR SM 968171.

Description. The specimen is an external mould, without counterpart, of part of a theca (PI. 62, fig. 7). Neither oral
surface nor base of the cup visible. Part of two sides of theca apparent, the junction between them angular,
suggesting a species of Stephanocrinus. Theca probably pentagonal in cross-section. On one side an ornament of
fine striae is preserved (PI. 63, fig. 1) which fan out and lie between 40° and 80° to the angle between the sides of the
theca. On the second side the striae are subparallel to this ridge. The specimen is 3-2 mm wide by 7-6 mm long,
individual striae about 007 mm wide.

Discussion. This specimen has a much finer ornament than other specimens from the Sholeshook
Limestone (i.e. Cupulocorona rugosa gen. et sp. nov.). The theca is more angular and, apparently,
proportionately more conical. The incomplete preservation precludes a more detailed description.
The specimen seems to represent a stephanocrinid with an angular theca, but without better material
a definite generic assignment cannot be made.

COELOMIC AMBULACRAL

CANALS COVER PLATES

PAIRED

FOOD

GROOVES

text-fig. 8. Oral surfaces of British coronates, a, Cupulocorona salopiae , holotype, BMNH E6390,
a, approximate position of anus, b, Stephanocrinus ramsbottomi, holotype, BMNH E25427.

538

PALAEONTOLOGY, VOLUME 28
Genus cupulocorona gen. nov.

Type species. Cupulocorona salopiae sp. nov.

Derivation of generic name. From the Latin cupula = a small cask, corona = a crown.

Diagnosis. Narrow stem, conical to cup-shaped theca with protruding triangular base and generally
low coronal processes, planar or gently convex radial furrows, ornament of fine to coarse ridges in
rhombic sets, but no prominent V W ridges.

Cupulocorona salopiae sp. nov.

Plate 62, figs. 2, 3, 9, 10; Plate 63, figs. 2, 3; text-figs. 7b, 8a
vl952 Stephanocrinus salopiae: Ramsbottom ms.

Derivation of species name. From the adjective Salopian = of Shropshire.

Diagnosis. Blunt coronal processes, pyriform theca with plate suture emphasized by low ridges.

Types. Holotype, BMNH E6390, paratypes, BMNH E6392-6394, E45530-45532.

Other material. Seven complete thecae, five of which are distorted.

Horizon ami locality. The specimen labels state: ‘Silurian. Lower Wenlock Shales, Buildwas Beds. E. bank of
River Severn, a short distance above Buildwas Bridge, Shropshire.’ Vine (1882, table, p. 48) records calyces of
echinoderms from Maw’s samples 36 and 37 in the Buildwas Beds, and the records probably refer to this species.
The Buildwas Beds are of early Wenlock age, from mid -centrifugus zone to mid -riccartonensis zone (Cocks et al.
1971). It is not known if Maw’s samples came from a single horizon or were gathered from throughout the
outcrop. The specimen labels seem to refer to locality 132 of Dean (1968).

Description. Theca pear-shaped (PI. 62, figs. 2 and 3). Base triangular with rounded corners (PI. 62, fig. 9), stem
cicatrix circular and depressed. Thecal cross-section changes from triangular to rounded pentagonal just above
the base. Coronal processes large and triangular, with blunt angles (text-fig. 7b). Sutures between adjacent plates
bear broad ridges (PI. 62, fig. 10). Aboral surface of theca ornamented with fine ridges in rhombic patterns.
Ambulacra bear paired food grooves, which slope gently away from the mouth and extend about 65 % of the way
to the periphery of the oral surface (PI. 63, figs. 2 and 3; text-fig. 8a). Food grooves concealed by paired ambulacral
cover plates. Five oral cover plates lie interradially over the mouth. The anus has not been identified with
certainty, although the approximate position can be predicted (text-fig. 8a). The hydropore has not been seen.
Two broken coronal processes, in the DE and EA interrays of the holotype, expose paired coelomic canals within.
Measurements are given in table 1.

Cupulocorona rugosa sp. nov.

Plate 62, fig. 4; Plate 63, figs. 4, 7-9; text-fig. 7c, E
Derivation of species name. From the Latin rugosa = rough, corrugated.

EXPLANATION OF PLATE 63

Fig. I. ? Stephanocrinus s. I. sp. BMNH E69218. Detail of ornament, x 55.

Figs. 2, 3. Cupulocorona salopiae gen. et sp. nov. Holotype, BMNH E6390. 2, oral cover plates (centre) and
five pairs of ambulacral cover plates (A ray above), x 30. 3, oral surface (anus below), x 12-3.

Figs. 4, 7-9. Cupulocorona rugosa gen. et sp. nov. 4, SM A3 1 8 1 3. Oral surface (anus below), x 12-3. 7, SM
A40187. Lateral view showing base of theca, x21. 8, holotype, SM A310046. Lateral view, x 15-5.
9, SM A53930a. Lateral view, x 13.

Fig. 5. Cupulocorona digitalis gen. et. sp. nov. Holotype, SM A40229, x 15.

Fig. 6. Stephanocrinus ramsbottomi sp. nov. SM A109798. Lateral view of crushed theca, x 8.

All scanning electron micrographs of latex casts from natural moulds, except figs. 2, 3.

PLATE 63

DONOVAN and PAUL, coronate echinoderms

540

PALAEONTOLOGY, VOLUME 28

table 1. Measurements in mm of British coronate echinoderms

Spec. No.

D

Dbase

Dqral H

Hdmax

HoRAL

Hbasal

KD

Lfg

Lam

Stephanocrinus

ramsbottomi

BM E25426

6-8

?

6-7

?

?

?

?

?

1-0

2-2

BM E25427

6-7

2-0

6-3

11-5

= H

90

?

0-7

1-3

2-4

BM E25428

7-0

?

7-0

10-0

= H?

7-7?

?

?

1-5

2-3

BM E30485

Internal mould

SM A3 1830

Poorly preserved base

of a theca

SM A3 1831

60

11

6-0

?

= H?

6-0

?

1-1

9

?

SM A3 1832

8-0

1-8

= D

12-0

9-2

9-2

?

1-3

0-7

2-4

SM A31833

8-1 +

2-0?

13-0?

= H

9-0?

?

9

1-3

3-0

SM A31834-7

All internal moulds

SM A32077

Deformed oral surface

only

SM A39061

7-3 +

?

7-3

?

?

?

?

?

1-5

2-7

SM A39062

6-2

1-0

5-7?

7-3

= H

6-5

9

0-8

1-0

2-0

SM A40184

Poorly preserved

SM A 109797/8

?

?

?

11-5

= H?

8-3

?

?

?

?

Cupulocorona salopiae

BM E6390

4-5

1-2

4-4

5-2

= H

4-0

2-0

0-6

1-2

1-8

BM E6392

?

1-6

?

6-7

= H

4-7

?

0-6

9

?

BM E6393

?

1-4

?

6-0

= H

4-2

3-2?

0-8

9

?

BM E6394

?

1-1

= D

4-4

= H

3-4

?

0-5

?

?

BM E45530

?

1-6

9

7-7

6-2

6-2

4-0

0-8

1-4

1-9

BM E45531

31

1-2

3-1

3-8

3-4

2-8

1-6?

0-65

0-8

1-0

BM E45532

?

1-2

?

5-1

3-8

3-4

2-4

0-6

9

?

Cupulocorona digitalus

BM E30484

?

10

4-1

6-5

?

4-2

?

9

?

?

SM A40229

3-0

0-6

= D

5-2

= H

31

?

?

?

?

Cupulocorona rugosus

SM A3 1001

5-6

?

5-4

4-6

= H

3-8

?

?

1-2

2-1

SM A3 1002

40

?

4-0

4-4

3-3

3-3

?

?

0-8

1-6

SM A31003

4-0?

?

3-8

5-4

4-2

4-2

?

?

?

1-7

SM A3 1004

4-2 +

?

4-2

9

?

?

?

?

0-8

1-4

SM A31259

5-6

?

5-6

6-4 +

5-2 +

5-2 +

?

?

0-9

1-8

SM A3 1260

4-2 +

0-7?

40 +

6-7 +

= H

5-5 +

?

0-7

?

?

SM A31261

4-3?

0-9

4-1

5-9

= H

5-2

?

0-6

0-7?

1-7

SM A3 1813

4-3

?

4-0

?

= H

?

?

?

0-7

1-4

SM A40185

Preserves part of the cup ornament

only

SM A40186

3-0 +

0-8

= D

4-5

= H

3-2

?

0-6

?

?

SM A40187

2-2

0-6

= D

3-2

= H

2-6

?

0-5

?

?

SM A40188

3-5

0-9

= D

5-0

= H

3-8

?

0-6

0-7

1-4

SM A40189

3-2?

?

3-2

?

?

?

?

?

0-7

1-25

SM A53929

5-4

16

50

4-2

= H

3-6

?

9

?

?

SM A53930

Poorly preserved

Diagnosis. Coarse-ribbed ornament and coronal processes which are lower than wide.

Types. Holotype, SM A3 1004, from the Upper Ordovician, Redhill Beds, at Prendergast Place near
Haverfordwest, Dyfed, South Wales, plus fourteen paratypes (SM).

Other material. Fifteen specimens in the Sedgwick Museum collections (all external moulds unless otherwise
indicated). A31001a, h (part and counterpart), A31002a, b (part and counterpart), A31003, 4a, b (a = internal

DONOVAN AND PAUL: CORONATE ECHINODERMS

541

mould), A31259u, b (part and counterpart), A31260, 1«, b (part and counterpart), A3 1813, A401 85-40187, 8 a, b
(part and counterpart), 9a, b(a = internal mould), A53929 a, b(b = artificial cast), A53930«, b (b = artificial cast).

Horizons and localities. This species is known from five localities and horizons, all of which are Ashgill
in age.

1, 2. The Sholeshook Limestone, both in the railway cutting (NGR SM 968171) and at Prendergast
Place (NGR SM 957166), near Haverfordwest, Dyfed, South Wales. Most of the Sholeshook Limestone is
Cautleyan in age, but the youngest rocks at Prendergast Place are Rawlheyan, Zone 5, according to Price
(1980).

3. The Redhill Beds at Prendergast Place which Price (1980, p. 486, table 1) considered to be confined to the
Rawtheyan, Zone 5.

4. The Lower Phillipsinella Beds at locality 1 of King ( 1923, p. 494, fig. 2), Aber Marchnant, SW Berwyn Hills,
Powys (NGR SJ 039194). These beds contain Kloucekia robertsi (Reed 1904) which Ingham (1977, p. 118)
regarded as an important index fossil for Rawtheyan, Zone 5.

5. The Dolhir Beds (mid-Cautleyan to Rawtheyan; Hiller 1981) at two localities in the Glyn Ceiriog district,
Clwyd. (a) Tram cutting ENE of Coed-y-Glyn-isaf’ (specimen label) and (b) on the road close to, and north of,
Gelli (NGR SJ 184367; Groom and Lake 1908).

Description. Thecal cross-section round (PI. 63, fig. 8) to pentagonal (PI. 62, fig. 4), with circular to pentagonal oral
surface (PI. 63, fig. 4) and triangular base (PI. 63, fig. 7). Stem facet round. Coronal processes triangular, lower than
wide. Ornament of coarse ridges in rhombic patterns (PI. 62, fig. 4; PI. 63, figs. 7-9). Ambulacral, oral, and anal
cover plates unknown. Ambulacral grooves about 50% of the length of the radial furrows, with paired food
grooves at their distal ends. Mouth circular to weakly pentagonal; anus circular.

Cupulocorona digitalis sp. nov.

Plate 63, fig. 5; text-fig. 7 d

Derivation of species name. From the Latin digitus = finger.

Diagnosis. Long finger-like coronal processes, ornament of coarse ridges, conical thecal profile, and
constricted base.

Types. Holotype, SM A40229, from the Ashgill, Phillipsinella Beds at Aber Marchnant, SW Berwyn Hills.
Paratype, BMNH E30484.

Other material. Two thecae, the holotype being an external mould with a counterpart.

Horizons and localities. 1, type locality (see C. rugosa ) and 2, 5 m above the Pusgillian Pen-y-Garnedd black
shales at Powys Arms Quarry, Pen-y-Garnedd, south-east of Llanfyllin, Powys, possibly Cautleyan (P. J.
Brenchley, pers. comm.).

Description. Theca low, conical, with a nipple-like triangular base. Stem facet not seen. Coronal processes
finger-like or triangular with curved sides (neither specimen shows both sides of the theca so the precise nature of
the coronal processes is not established). Ornament of coarse rhombic ridges. Oral surface not seen.

Acknowledgements. We thank Dr. D. Price, Sedgwick Museum, Cambridge, and Dr. A. B. Smith and Mr. D. N.
Lewis, British Museum, Natural History, for the loan of specimens. Dr. W. H. C. Ramsbottom kindly gave
us a copy of his unpublished manuscript on British and Swedish coronates. Scanning electron micrographs
were produced by Mr. C. J. Veltkamp of the Botany Department, Liverpool University, and Table 1 was
typed by Ms. Grainne Moloney. Some of the photomicrographs were printed by Mr. D. .1. McCabe. Finally,
part of this work was undertaken during the tenure of NERC research grant GR3/4732, which is gratefully
acknowledged.

REFERENCES

barrande, j. 1887. Systeme Silurien du centre de la Boheme. le partie: Recherches paleontologiques. v. 7. Classe
des Echinodermes. Ordre des Cystides, xix + 233 pp., 39 pis. Leipzig and Prague.
bather, F. a. 1900. The Echinoderma. In lankester, e. r. (ed.), A Treatise on Zoology, vol. 3, 344 pp. London.

542

PALAEONTOLOGY, VOLUME 28

bockelie, j. f. 1981. The Middle Ordovician of the Oslo Region, Norway. 30. The eocrinoid genera
Cryptocrinites, Rhipidocystis and Bockia. Norsk geol. Tidskr. 61, 123-147.
brett, c. e., frest, t. j., sprinkle, j. and clement, c. R. 1983. Coronoidea: a new class of blastozoan
echinoderms based on taxonomic reevaluation of Stephanocrinus. J. Paleont. 57, 627-651.
cantrill, t. c. 1907. Stratigraphical note. Geol. Mag. 44, 537-538.

chauvel, j. and le menn, j. 1973. Echinodermes de 1’Ordovicien Superieur de Coat-Carrec, Argol (Finistere).
Bull Soc. geol. miner. Bretagne Ser. c, 4, 39-61, pis. 1 3.

1979. Sur quelques echinodermes (Cystoi'des et Crinoi'des) de l’Ashgill d’Aragon (Espagne). Geobios,
12, 549-587, 3 pis.

cocks, l. r. m., Holland, c. h., rickards, r. b. and strachan, i. 1971. A correlation of Silurian rocks in the
British Isles. Quart. Jl geol. Soc. Lond. 127, 103-136.

conrad, t. a. 1842. Observations on the Silurian and Devonian systems in the United States, with descrip-
tions of new organic remains. J. Acad. nat. Sci. Philad. 8, 228-280, pis. 12-17.
dean, w. T. 1968. Geological itineraries in south Shropshire. Geol. Ass. Guide, 27, 48 pp.
delgado, J. F. N. 1908. Systeme Silurian du Portugal Etude de Stratigraphie Paleontologique, 247 pp„ 4 pis.
Lisbon.

etheridge, R. jn. and carpenter, p. H. 1886. Catalogue of the Blastoidea in the Geological Department of the
British Museum (Natural History), xv + 322 pp., 20 pis. London.
fay, r. o. 1962. Ventral structures of Stephanocrinus angulatus Conrad. J. Paleont. 36, 206-210, pi. 35.

1978. Order Coronata Jaekel, 1918. In moore, r. c. and teichert, c. (eds.). Treatise on Invertebrate

Paleo?itology. Part T. Echinodermata 2, pp. T574-T578. Geol. Soc. Am. and Univ. Kansas Press.
groom, t. and lake, p. 1908. The Bala and Llandovery rocks of Glyn Ceiriog (North Wales). Quart. Jl geol.
Soc. Lond. 64, 546-595.

hiller, n. 1981. The Ashgill rocks of the Glyn Ceiriog district. North Wales. Geol. J. 16, 181-200.
ingham, J. K. 1977. The Upper Ordovician trilobites from the Cautley and Dent districts of Westmorland and
Yorkshire. Palaeontogr. Soc. [Monogr.f 3, 89-121, pis. 19-27.
jaekel, o. 1918. Phylogenie und System der Pelmatozoen. Palao)it. Z. 3, 1-128.

— 1927. Uber Tormoblastus n.g., eine coronate Blastoidee, aus dem Ordovicium Schwedens. Ark. Zool.
19A (4), 6 pp., 1 pi.

king, w. b. r. 1923. The Upper Ordovician rocks of the southwestern Berwyn Hills. Quart. Jl geol. Soc. Lond. 79,
487-507, pi. 26.

— and wilcockson, h. w. 1934. The Lower Palaeozoic rocks of Austwick and Horton-in-Ribblesdale,
Yorkshire. Ibid. 90, 7-31.

paul, c. R. c. 1985. Ordovician and Silurian coronates from Czechoslovakia. Geol. J. 20, 21-29.
price, d. 1973. The age and stratigraphy of the Sholeshook Limestone of southwest Wales. Ibid. 8, 225-
246.

— 1980. A revised age for the topmost Sholeshook Limestone Formation (Ashgill) of South Wales. Geol. Mag.
117,485-489.

reed, f. r. c. 1904. New fossils from the Haverfordwest district. Ibid. 41, 106-109, pi. 5.

1907. The base of the Silurian near Haverfordwest. Ibid. 44, 535-537.

regnell, G. 1945. Non-crinoid Pelmatozoa from the Palaeozoic of Sweden: a taxonomic study. Meddn Lunds
geol-miner. Instn, 108, 255 pp., 15 pis.

say, t. 1825. On two genera and several species of Crinoidea. J. Acad. nat. Sci. Philad. 4, 289-296.
sjoberg, s. 1915. Paracystis ostrogothicus g. et sp. n., en egendomlig Pelmatozo fran Ostergotlands Chasmops-
kalk. Geol. For. Stockholm Forhandl. 37, 171-178, pis. 2-3.
sprinkle, j. 1973. Morphology and evolution of blastozoan echinoderms. Spec. Publ. Mus. comp. Zool. Harv.
284 pp., 43 pis. Cambridge, Mass.

— 1980. Origin of blastoids: new look at an old problem. Geol. Soc. Am. Abstr. Progr. 12, 528.

vine, g. r. 1882. Notes on the Polyzoa of the Wenlock Shales, Wenlock Limestone, and Shales over
Wenlock Limestone. From material supplied by G. Maw, Esq., F.L.S., F.G.S. Quart. Jl geol. Soc. Lond.
38, 44-68.

wachsmuth, c. and springer, f. 1886. Revision of the palaeocrinidae, pt. 3, sect. 2. Discussion of the classi-
fication and relations of the brachiate crinoids, and conclusion of the generic descriptions. Proc. Acad. nat.
Sci. Philad. (1885), 225-364, pis. 4-9.

WILLIAMS, a., STRACHAN, I., BASSETT, D. A., DEAN, W. T., INGHAM, J. K., WRIGHT, A. D. and WHITTINGTON, H. B. 1972.

A correlation of Ordovician rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 3, 74 pp.
wright, a. d. 1968. A westward extension of the Upper Ashgillian Hirnantia fauna. Lethaia, 1, 352-367.

DONOVAN AND PAUL: CORONATE ECHINODERMS

543

1982. The Ordovician -Silurian boundary at Keisley, northern England. In bruton, d. l. and williams,
s. h. (eds.). Abstracts for meetings 20, 21 and 23 August 1982, IVth International Symposium on the

Ordovician System. Palaeont. Contr. Oslo Univ. 280 , 60.

Typescript received 1 June 1984

Revised typescript received 29 November 1984

S. K. DONOVAN
Department of Geology
Trinity College
Dublin 2, Eire

Present address:

Natural Environment Research Council
Polaris House
North Star Avenue
Swindon SN2 1EU
England

C. R. C. PAUL

Department of Geology
Liverpool University
Liverpool L69 3BX
England

CORTICAL DEVELOPMENT IN CHALONERIA
CORMOSA (ISOETALES), AND THE BIOLOGICAL
DERIVATION OF COMPRESSED LYCOPHYTE
DECORTICATION T AX A

by KATHLEEN B. P I G G Cltul GAR W. ROTHWELL

Abstract. Several anatomically preserved stem fragments showing a wide range of surface features have been
discovered among specimens of the Upper Pennsylvanian isoetalean Chaloneria cormosa. A comparison of the
specimens demonstrates that stems produced a narrow zone of periderm, and that tissues external to the
periderm accounted for a moderate increase in stem circumference by two distinctive modes of cell divisions.
Depending on the presence or absence of secondary cortical tissues, on differential taphonomy, and on the level
at which the cortex is exposed, the surface of a specimen may be comparable to one of several distinctive
decortication morphotypes. The outer surface is similar to Bothrodendron and Cyclostigma, while specimens with
leaf bases removed are reminiscent of Stigmaria. When fractured through the periderm a Knorria surface is
produced, while secondary cortical features immediately external to the periderm conform to Asolanus.
Specimens reveal the anatomical bases for decortication morphotypes and demonstrate that such genera are
produced by members of Isoetales as well as Lepidodendrales.

Our current understanding of Carboniferous vegetation relies upon plant remains that are preserved
by several different modes, including compression/impression, cellular permineralization, and
mold/cast (Schopf 1975). Upper Carboniferous and Pennsylvanian fossils of varying preservational
types typically exhibit differential suites of characters, and traditionally have been studied
independently of one another. As a result, a great deal is known about fossils within each group, but
remains preserved by different modes seldom have been correlated with the precision necessary to
demonstrate whether they represent the same or different taxonomic species.

One group in which correlations of this type are of particular value is Carboniferous Lycopsida.
Large Palaeozoic lycophytes produced massive amounts of cortical tissue. Fractured at different
surface and subsurface levels, stems typically display characteristic, widely differing features (Renault
and Zeiller 1888; Weiss and Sterzel 1893; Thomas and Watson 1976). When preserved as
compressions, such morphotypes (DiMichele 1983) are given generic rank but their mode of
production, anatomical origin, and biological significance remain poorly understood. Among the
most prominent genera are Asolanus Wood (1860) and Knorria Sternberg (1825). Although Asolanus
has been described from numerous localities throughout North America, Europe, and Northern
Africa (White 1899; Janssen 1940; Crookall 1964; Daberand Kahlert 1970; Lejal-Nicol 1972; Boersma
1978), and is a common component of many compression floras from the Appalachian Basin (Darrah
1 969), it is understood only as a characteristic configuration on the rock surface. Through the years it
has been described as, or confused with, such diverse structural forms as Lepidodendron , Sigillaria,
Sigillarioides, Pseudosigillaria, and Stigmaria (Crookall 1964). It also has been interpreted as a
decortication layer of numerous Carboniferous compression genera (e.g. Daber and Kahlert 1970), or
as the outer surface of an otherwise unknown lycophyte stem (White 1899; Janssen 1940). In contrast,
Knorria is known to represent a decortication surface of lepidodendralean stem genera (Solms-
Laubach 1891; Thomas and Watson 1976).

Several additional taxa of Carboniferous lycophyte stem compressions exhibit features unlike those
of typical lepidodendrid taxa. The genus Bothrodendron Lindley and Hutton (1833) originally was

IPalacontology, Vol. 28, Part 3, 1985. pp. 545-553, pi. 64.|

546

PALAEONTOLOGY, VOLUME 28

described as a Carboniferous stem compression with tiny leaf scars and conspicuous halonial branch
scars, but it now includes several diverse forms which undoubtedly represent more than one genus of
plants. Morphotypes such as Bothrodendron , Pinakodendron (Weiss and Sterzel 1893), and some
species of Cyclostigma Haughton (1859) are delimited by overlapping taxonomic criteria. All are
characterized by rounded or lenticular leaf scars (or leaf bases) borne in a low helix and widely
separated on the stem surface (Crookall 1964). Despite detailed studies of some features (e.g. cuticle
and compression surfaces; Thomas 1967), anatomical features of these taxa are poorly known.
Consequently, the whole plants represented by Upper Carboniferous specimens of such taxa remain
poorly understood (Crookall 1964; Stubblefield and Rothwell 1981).

During the recent study of a new family of Pennsylvanian lycophytes, the Chaloneriaceae Pigg and
Rothwell (19836), several permineralized stems of Chaloneria cormosa were revealed on split coal-ball
surfaces. Some were exposed at the outer surface of the stem (figs. 21, 22, of Pigg and Rothwell 1983a)
with intact leaf bases, epidermis, and cuticle, while others were split at a variety of subsurface levels.
The specimens display a wide array of configurations that are very distinct from one another. When
compared to compression-impression taxa, they are similar to a variety of genera including
Cyclostigma , Pinakodendron , Bothrodendron , Stigmaria , Asolanus , and Knorria. Because the histology
and ontogeny of Chaloneria are well known, it is now possible to determine anatomical, taphonomic,
and ontogenetic bases for the surface configurations that delimit several morphotypes of
decortication compression taxa, and to relate more precisely these to the types of plants by which they
were produced.

METHODS AND MATERIALS

Specimens of C. cormosa are preserved by calcareous cellular permineralization in coal balls from the Duquesne
coal (Upper Pennsylvanian) where it outcrops in a roadcut on Ohio State Route 22, approximately 8 km west of
Steubenville, Ohio, USA (Rothwell 1976). Surfaces were etched in 5% HC1 for 30 sec. to increase contrast for
photography. Some coal balls were cut perpendicular to the stem surface and peeled to determine the exact level
at which the cortical tissue had split. Other specimens were peeled parallel to the stem surface to identify the
anatomical basis for the surface features. Peels were mounted on microscope slides for photography.

Compression specimens of A. camptotaenia and cf Bothrodendron photographed for comparative purposes
were collected from two Middle Pennsylvanian localities in eastern Ohio. The first is Dorr Run located
approximately L7 km northwest of Nelsonville (NE NE sec. 30, Wayne Twp„ Nelsonville 7-5 min.
Quadrangle), Athens Co., Ohio. Stratigraphically these specimens occur in the Snow Fork shale below the
Middle Kittanning coal. Middle Kittanning cyclothem, Allegheny Group. The second locality is a strip mine
operated by the James Bros. Coal Co., at Mineral City (South centre. Sec. 26, Rose Twp., 7-5 min. Quadrangle),
Carroll Co., Ohio. At this locality, fossils are preserved in the shale over the Lower Kittanning (Ohio no. 5 coal),
Lawrence-Lower Kittanning cyclothem, Allegheny Group. Pertinent specimens and slides are housed in the
Paleobotanical Herbarium, Department of Botany, Ohio University, where they bear acquisition numbers 584,
3375, 3825, 7640-7662.

DESCRIPTION

Surface Morphologies. C. cormosa is an Upper Pensylvanian member of Isoetales characterized by a
cormose base and an unbranched stem bearing helically arranged leaves (Pigg and Rothwell 1 983a, b).
Unlike typical members of Lepidodendrales that often have abutting leaf cushions prior to periderm
development, the leaf bases of Chaloneria are separated by wide areas of stem surface (PL 64, fig. 1).
Also unlike the leaf cushions of Lepidodendrales, Chaloneria leaf bases have irregularly shaped scars
where the leaves become detached.

Coal-ball specimens fractured at the outer surface of the stem are characterized by helically
arranged leaf bases, about 8-10 mm wide and 5 mm high separated from one another by wide areas of
stem surface (PI. 64, fig. I; text-fig. 1a). In this regard, Chaloneria is similar to several other taxa (e.g.
Bothrodendron , Cyclostigma ; Crookall, 1964) that typically lack abutting leaf cushions (PI. 64, figs. 6,
7; Jennings 1979). Leaf bases of Chaloneria have blunt lateral wings, a rounded keel, and either an

PIGG AND ROTHWELL. CHALONER1A (ISOETALES)

547

apical notch or a ligule pit (PI. 64, figs. 1, 2). They are also somewhat bulbous, bulging out from the
stem surface (PI. 64, fig. 2), and in this regard are similar to the leaf scars popularly attributed to
Asolamts (Crookall 1964; Darrah 1969). Distally the leaves of Chaloneria abruptly become much
narrower (Pigg and Rothwell 1983u). No distal portions of leaves were present on fractured surfaces.
When a leaf is detached at a single level the base displays either one or two vascular strands. More
frequently, leaves are broken off unevenly, resulting in irregular appearing bases ( PI. 64, fig. 1 ) in which
vascular strands are obscured. There is no evidence of an abscission zone that would provide for the
production of regular scars like those on the leaf cushions of Lepidodendrales. Dark, longitudinal
strips of tissue are often present on the interfoliar stem surface (PI. 64, fig. 1 ), and these represent strips
of epidermis and cuticle.

text-fig. 1a-d. Diagrams of cortical tissues in stems of Chaloneria cormosa that have significant
periderm; stippled area includes outer cortex in which randomly oriented internal cell divisions
occur; distorted lines indicate level at which Tso/unns-pattern is produced; straight lines indicate
periderm. Outer surface of a will appear as Chaloneria unless pattern of distorted cortex is impressed
on surface, in which case specimen will be recognized as Asolanus. Outer surface of b will resemble
Stigmaria unless pattern of distorted cortex is impressed on surface, in which case specimen will be
recognized as Asolanus. Outer surface of c has been split at level where regardless of whether leaf
bases are detached above (at left) or at interfoliar surface, specimen will exhibit Asolanus
configuration. Cortex in d has been split at level of periderm such that both inner and outer surfaces

will produce Knorria pattern.

In other specimens the epidermis is intact in interfoliar areas, but leaves have broken off near or
below the stem surface (PI. 64, fig. 5; text-fig. 1 b). In some specimens of this type the interfoliar surface
is smooth. On others there are vertically oriented, undulating striations (PI. 64, fig. 5 at arrows) that
conform to the subsurface primary cortical tissue as seen in tangential section ( PI. 64, fig. 3). The most
conspicuous features of many specimens are the prominent circular or elliptical parichnos strands
that accompany the leaf traces (PI. 64, fig. 3). They are similar to appendage scars on stigmarian axes,
with the rounded parichnos strands of Chaloneria ( PI. 64, fig. 5) corresponding to the aerenchymatous
middle cortex of stigmarian appendages. There is a further resemblance with Cyclostigma , which in
the past also has been confused with Stigmaria (Crookall 1964).

Two additional cortical patterns of Chaloneria are commonly represented on split coal-ball
surfaces. Several specimens are characterized by an anastomosing pattern of diagonal striations
(text-figs, lc, 2a, b). Of these, some exhibit bulbous leaf base scars (text-fig. 2b), while in others the leaf
base is broken off at a more proximal level, revealing axially elongate rather than rounded parichnos
strands (text-fig. 2a). This pattern conforms to the compression taxon Asolanus (Wood 1 860) based on
the characteristic interfoliar pattern (PI. 64, fig. 8; text-fig. 2a, b).

An additional, commonly preserved surface of Chaloneria cortex is characterized by densely
compact cells (text-fig. 2e, g) and lacks striations or other prominent interfoliar features (text-figs. 1 d,

548

PALAEONTOLOGY, VOLUME 28

2e). Specimens of that type each possess a vascular strand that is surrounded by an oval or vertically
elongate, lenticular parichnos strand (text-fig. 2c). This type of decortication pattern is comparable
to Knorria (Sternberg 1825).

Anatomical origin of surface features. In transverse section an immature specimen of Chaloneria
exhibits a medullated protostele surrounded by primary cortical tissues; this pattern is also present in
distal stem regions (Pigg and Rothwell 1983a). Most stems are flattened (PI. 64, fig. 4), lacking
preserved pith and inner cortex. In the inner zone of commonly preserved primary cortex the cells
possess differentially thickened walls and frequently broken tangential walls (PI. 64, fig. 4 at top). This
results in a resemblance to distorted radial rows of periderm (Pigg and Rothwell 1979). Tangential
sections through the zone (PI. 64, fig. 3) reveal the undulating, vertically oriented pattern that is
sometimes present on the epidermis (PI. 64, fig. 5 at arrows). This pattern is also characteristic of the
interfoliar region on the stems of many compressions, where it is interpreted as the result of distortion
during diagenesis (PI. 64, fig. 7; Rex and Chaloner 1983).

Older and more proximal stem segments exhibit secondary cortical tissues that are derived both
from a continuous cambium (viz. periderm) and from the internal subdivision of individual cortical
cells (PI. 64, fig. 2). Periderm is represented by a zone of radially aligned cells with differentially
thickened radial walls, and is present along the inside of the commonly preserved cortex (PI. 64, fig. 2).
In stems with little periderm (PI. 64, figs. 4, 5) the primary cortex and epidermis remain relatively
intact and unaltered. In stems with larger amounts of periderm the epidermis and cuticle are separated
into longitudinally oriented strips of tissue at the periphery of the stem (PI. 64, fig. 1). In these
specimens, cortex outside the periderm appears distorted (PI. 64, fig. 2 at sides; text-fig. 1a-c).
Tangential sections of the tissue reveal narrow, vertically elongated lenticular regions where the cells
are laterally expanded, and in which some internal cell divisions have occurred (text-fig. 2 f). These
regions alternate with areas of narrow primary cortical cells (text-fig. 2 d,f). The alternation of these
zones produces a pattern of discontinuous light and dark strips (text-fig. 2c) in specimens with little
periderm (e.g. PI. 64, fig. 4). Stems with a thicker zone of periderm (e.g. PI. 64, fig. 2) have a distinctive
pattern of diagonal lines in the interfoliar region. When seen on split surfaces this pattern produces the
Asolanus configuration (text-fig. 2a, b ). Cortical growth of this type undoubtedly could accommodate
only a moderate increase in circumference of the stem, and this is consistent with the periderm of
Chaloneria being up to only about 2 mm thick (PI. 64, fig. 2). Expansion of the subepidermal cortex is
accomplished by the randomly oriented subdivision of many cells (PI. 64, fig. 2 at arrow), another

EXPLANATION OF PLATE 64

Figs. 1 -5. Chaloneria cormosa. d, distorted cortex; p, parichnos; pc, primary cortex; pe, periderm; s, stele. 1, surface
view of stem in which secondary growth has split epidermis into longitudinally oriented strips, O.U.P.H.
no. 3825, x 1. 2, transverse section of cortex showing leaf base and interfoliar region after significant
secondary growth. Arrow indicates position of outer cortex in which randomly oriented internal cell divisions
have occurred, C.B. 1399a top no. 9, x 7. 3, tangential section near surface of stem with only primary growth.
Primary cortex exposed at centre and leaf bases sectioned at sides, C.B. 1 398d( 1 ) side no. 310, x 2-5.
4, transverse section of flattened stem with little periderm. Outer surface of cortex at top comparable to that in
fig. 5, outer surface of cortex at bottom comparable to that in text-fig. 2c. C.B. 2126a bottom x 7. 5, surface
view of stem with leaf bases broken off at level of interfoliar stem surface and intact epidermis. Note vertically
undulating, longitudinal pattern (at arrows) that conforms to cellular pattern of primary cortex in fig. 3, C.B.
2127a( 2), x 1 .

Fig. 6. Bothrodendron- type impression of stem surface with relatively smooth interfoliar surface. O.U.P.H.
no. 7642, x 2-5.

Fig. 7. Bothroderulron- type impression of stem surface with longitudinally oriented, vertical pattern similar to
that of primary cortex in Chaloneria , O.U.P.H. no. 7643, x 2 5.

Fig. 8. Asolanus impression of stem with leaf bases similar to those in figs. 6 and 7, and interfoliar pattern similar
to that of Chaloneria specimens in text-fig. 2a, b. O.U.P.H. no. 7644, x I.

PLATE 64

PIGG and ROTHWELL, Carboniferous Chaloneria ( Isoetales )

550

PALAEONTOLOGY, VOLUME 28

text-fig. 2a-g. Chaloneria cormosa cortex in surface views and in tangential sections; t, leaf trace, a, cortex of
specimen with well-developed periderm, split at level of ‘d’ in Plate 64, fig. 2, C.B. 2126, x 1. b, cortex of same
specimen as a , but with leaf bases split slightly further toward periphery, at level comparable to specimen in
Plate 64, fig. 5, C.B. 2126, x 1. c, surface of cortex at level similar to that in Plate 64, fig. I, but from specimen
with only limited periderm, comparable to outer surface of cortex at bottom of Plate 64, fig. 4, C.B. 2126, x 1.
d, tangential section through cortex of specimen with significant periderm, comparable to surface in b ,
C.B. 2 1 27a( 1 ) Side no. 6, x 2-5. e , tangential section through periderm of specimen in d showing anatomical
derivation of Knorria configuration, C.B. 2 1 27a( 1 ) side no. 9, x2-5 .f enlargement of anatomical detail in d,
C.B. 2127 no. 7, x 17. g , enlargement showing anatomical detail of tissue that compresses to produce Knorria

configuration, C.B. 1 398d( 1 ) side no. 310, x 17.

indication of only limited increase in girth due to secondary growth. The interfoliar pattern similar to
Asolanus commonly found among specimens of Bothrodendron, Cyclostigma , and Pinakodendron is
consistent with the interpretation that they also had limited periderm production, and a cortical zone
that became distorted by secondary growth similar to that in Chaloneria. If some species of these

PIGG AND ROTHWELL: CHALONERIA (ISOETALES)

551

genera represent small trees, as supposed by some authors, then the majority of the ultimate stem
diameter must have been produced by primary growth.

Tangential sections or fractures through the periderm of Chaloneria reveal homogeneous tissue
traversed by leaf traces and accompanying parichnos (text-figs. Id, 2e, g ), and produce a Knorria
configuration (text-fig. 2c). The occurrence of Knorria in association with many lepidodendrid
specimens probably also represents tangential sections through the periderm. Considering the
massive amounts of periderm they produced, it is not surprising that Knorria is of such common
occurrence among lepidodendrids. However, because smaller forms such as Chaloneria also produced
periderm, Knorria should be expected to occur among specimens of them as well.

DISCUSSION

Plant Size Estimation. Because the broad inner cortical region and pith are rarely present in
anatomically preserved stems of Chaloneria , specimens are almost always represented by a highly
flattened cylinder (DiMichele et al. 1979; Pigg and Rothwell 1 983a) that must have collapsed quite
early during diagenesis. Under these circumstances, instead of the width of the flattened Chaloneria
stems being a rough approximation of their diameter in life (as proposed in the compression model of
Walton 1936; Rex and Chaloner 1983), it was probably exaggerated significantly. A more realistic
approximation of the original diameter of the flattened stems is obtained by considering their width to
represent one-half the circumference. In such a case the original diameter can be calculated by
doubling the width of the stem, and considering that figure to represent the original circumference.
The stem diameter is then easily calculated by considering circumference (C) to be equal to 7rD, where
D = stem diameter. If this interpretation of the taphonomic alteration of stems is correct, then the size
of compressed plants with a similar mode of cortical growth (e.g. as in Asolanus ) may have been
overestimated by many previous authors.

Developmental Interpretation. Some stems of C. cormosa exhibit only primary growth, and in these
axes the preserved cortical tissues are relatively homogeneous (PI. 64, fig. 3 at centre). Variations
from specimen to specimen are due primarily to taphonomic processes including uneven leaf
detachment, differential preservation, dissimilar amounts of crushing, and the impression of
subsurface cortical patterns (PI. 64, fig. 3) on to the epidermis (PI. 64, fig. 5; Rex and Chaloner 1983).
The cortex in specimens of this type lacks regions of anastomosing cells and radially aligned cells.

By contrast the onset of secondary growth produced several distinctive layers of cortical tissues.
The most regular of these is the zone of radially aligned peridermal cells that are located at the inner
margin of the commonly preserved cortex. This is also the zone that initiated an increase in the girth of
the stem. In stems with prominent periderm, lenticular areas that are characterized by internally
septate cells developed to the periphery of the periderm ( PI. 64, fig. 2; text-fig. 2d ). Such areas represent
a mechanism for maintaining tissue continuity immediately outside the periderm during the increase
in girth. It is clear that many cells in the outermost cortex also remained meristematic at the onset of
periderm production, such that the longitudinal fissures on the stem surface are restricted primarily to
the epidermis. However, the outermost zone of cortical cells grew by more random patterns of internal
cell divisions (Pigg and Rothwell 1983a).

Systematic Implications. Documentation among specimens of C. cormosa of the anatomical and
taphonomic features that produce several morphotypes of lycophyte stem surface patterns provides a
basis for relating such morphotypes to a diverse assemblage of Palaeozoic plants. It also provides
evidence for the modes of cortical development in compression/impression forms that previously have
been known primarily as surface patterns. In taxa of compressed stems with distantly spaced leaf bases
such as Bothrodendron , Cyclostigma , and Pinakodendron, a relatively smooth interfoliar region (PI. 64,
fig. 6) probably displays a less distorted outer surface than specimens with characteristic interfoliar
patterns (PI. 64, figs. 7, 8; Rex and Chaloner 1983). Among the latter specimens, those with gently
undulating patterns (PI. 64, fig. 7) show features similar to those of the primary cortex in Chaloneria,
and were undoubtedly preserved prior to, or distal to, secondary cortical development. In contrast.

552

PALAEONTOLOGY, VOLUME 28

stems with the Asolcmus- type diagonal interfoliar pattern (PI. 64, fig. 8) reveal that significant
secondary cortical activity like that of Chaloneria had occurred prior to fossilization. Some specimens
of this type represent decortication external to the periderm (text-fig. 2 a, b ), while others have had the
subepidermal configuration impressed on the outer surface of the stem (text-fig. 2e). As emphasized by
Rex and Chaloner (1983) specimens of the latter type may show both surface and subsurface features
at the same level. This accounts for the occurrence of external leaf-base scars that are separated by a
subsurface Asolanus-type cortical pattern on some specimens (PI. 64, fig. 8). While isoetaleans like
Chaloneria were probably the most common source of such decortication surfaces, some lepidoden-
draleans with relatively moderate secondary development (e.g. Lepidodendron dicentricum , Eggert
1961) produced ‘phellem meshes’ that may appear similar to Asolanus when seen on a split surface.
The specimens of Asolanus interpreted as a Lepidodendron by Daber and Kahlert (1970) could reflect
this mode of growth.

It is now clear that the Knorria pattern is produced when the cortex is split through or at the surface
of the periderm in both Lepidodendrales and Isoetales, and that its more common occurrence with
lepidodendraleans is probably the result of the much more massive periderm produced by the group.
Likewise, the Asolanus configuration results from either decortication peripheral to the periderm or
from the imprinting of subsurface features on the epidermis in plants with the capacity for only limited
periderm production after much of the secondary cortical production had been completed.

Through continuing efforts to determine the biological and taphonomic origins of compression
features (e.g. Rex and Chaloner 1983) and the anatomical basis for morphological features, the
significance of many lycophyte morphotypes is becoming increasingly clear. In terms of both
taxonomic significance and ontogenetic potential the increased knowledge allows for a better
understanding of lycophyte evolution, and provides the basis for more accurate interpretations of
remains preserved as coalified compressions or impressions.

Acknowledgements. The specimens figured in Plate 64, figs. 7 and 8 were collected by Mr. L. J. Millhorn, and
provided by Dr. M. T. Sturgeon, Department of Geology, Ohio University. We wish to thank Dr. T. N. Taylor for
reading the manuscript. This study was supported in part by National Science Foundation grant BSR83 1-0576
(to G. W. R.).

REFERENCES

boersma, m. 1978. A survey of the fossil flora of the Tllinger Flozzone’ (‘Heusweiller Schichten’, Lower
Stephanian, Saar, German Federal Republic). Rev. Palaeobot. Palynol. 26 , 4 1 92.
crookall, R. 1964. Fossil plants of the Carboniferous rocks of Great Britain (2nd sect.), Mem. Geol. Surv. G.B.,
Palaeontol. 4, 217-354.

daber, R., and kahlert, E. 1970. Lepidodendron (Asolanus) camptotaenia (Wood, 1860) comb. n. Palaontol. Abh.,
B. Palaobot. 3 , 349-355.

darrah, w. c. 1969. A critical review of the Upper Pennsylvanian floras of eastern United States with notes on the
Mazon Creek flora of Illinois. Privately publ., 219 pp.

DiMiCHELE, w. A. 1983. Lepidodendron hickii and generic delimitation in Carboniferous lepidodendrid lycopods.
Syst. Bot. 8 , 317-333.

mahaffy, j. f. and Phillips, t. l. 1979. Lycopods of Pennsylvanian coals: Polysporia. Can. J. Bot. 57 ,
1740 1753.

eggert, D. a. 1961. The ontogeny of Carboniferous arborescent Lycopsida. Palaeontographica , Abt. B, 108 ,
43-92.

haughton, s. 1859. On Cyclostigma, a new genus of fossil plants from the Old Red Sandstone of Kiltorcan,
Co. Kilkenny; and on the general law of phyllotaxis in natural orders — Lycopodiaceae, Equisetaceae, Edlices,
etc. J. Roy. Dublin Soc. 2 , 407-420.

janssen, R. E. 1940. Some fossil plant types of Illinois. Sci. Papers Illinois State Museum, Springfield, I, I 123.
jennings, J. R. 1979. Lower Pennsylvanian plants of Illinois. III. Bothrodendron from the Drury Shale.
J. Paleontol. 53 , 519-523.

LEJAL-NICOL, a. 1972. Contribution a l’etude des lycophytes paleozoiques de Bassin de Fort-Polignac (Illizi).
Nord. Alger Bull. Soc. Hist. nat. Afr. 63, 49-79.

PIGG AND ROTHWELL: CHALON ERIA (ISOETALES)

553

lindley, J. and hutton, w. 1833. The fossil flora of Great Britain. Vol. 1, 218 pp. James Ridgway, London.
pigg, K. B. and rothwell, G. w. 1979. Stem-root transition of an Upper Pennsylvanian woody lycopsid. Am.
J. Bot. 66, 914-924.

— 1983a. Chaloneria gen. nov., heterosporous lycophytes from the Pennsylvanian of North America. Bot.
Gaz. 144, 132-147.

— 19836. Megagametophyte development in the Chaloneriaceae fam. nov., Paleozoic Isoetales
(Lycopsida). Ibid. 295-302.

Renault, b. and zeiller, r. 1888. Etudes sur le terrain houiller de Commentary— Flore fossile, Pt. 1. Soc.
Industrie mineral Si. Etienne Bull. 2, 1-366.

rex, G. m. and chaloner, w. g. 1983. The experimental formation of plant compression fossils. Palaeontolocjv ,
25, 231-252.

rothwell, g. w. 1976. Petrified Pennsylvanian age plants of eastern Ohio. Ohio J. Sci. 76, 128 132.
schopf, j. m. 1975. Modes of fossil preservation. Rev. Palaeobot. Palynol. 20, 27-53.
solms-laubach, h. 1891. Fossil Botany (English Translation), 401 pp. Clarendon Press, Oxford.

Sternberg, k. 1825. Versuch einer geognostisch botanischen Darstellung der Flora der Vorwelt , 1 (pt. 4), 1-48.
Leipzig.

Stubblefield, s. p., and rothwell, G. w. 1981. Embryogeny and reproductive biology of Bothrodendrostrobus
mundus (Lycopsida). Am. J. Bot. 68, 625-634.

thomas, b. a. 1967. The cuticle of two species of Bothrodendron (Lycopsida: Lepidodendrales). J. Nat. Hist.

I, 53-60.

and watson, i. 1976. A rediscovered 1 14-foot Lepidodendron from Bolton, Lancashire. Geol. J. 1 1, 15-20.
walton, j. 1936. On the factors which influence the external form of fossil plants; with descriptions of the foliage
of some species of the Palaeozoic Equisetalean genus Annularia Sternberg. Phil. Trans. R. Soc. Loud. B.
226, 219-237.

weiss, c. e. and sterzel, t. 1893. Die Sigillarien der preussischen Steinkohlen- und Rothliegenden-Gebiete.

II. Gruppe der Subsigillarien. Abhandl. konig. preuss. geol. Landesanst. n.f. 2, 1 -255.

white, d. 1899. Fossil flora of the Lower Coal Measures of Missouri. U.S. Geol. Survey Mon. 37, 467 pp.
wood, h. c. 1860-1861. Contributions to the Carboniferous flora of the United States. Proc. Acad. Nat. Sci.
Philadelphia for 1860 , 12, 236-240, 436-443, 519-522.

KATHLEEN B. PIGG

Department of Botany
The Ohio State University
Columbus, Ohio 43210
USA

GAR W. ROTHWELL
Department of Botany

„ . , . T Ohio University

Typescript received 4 June 1 984 Athens, Qhio 4570 ,

Revised typescript received 27 November 1984 USA

WEALDEN OCCURRENCE OF AN ISOLATED
BA R REM I AN DINOCYST FACIES

by n. f. hughes and i. c. harding

Abstract. Records of occurrence are presented in the form of two new taxa of peridinioid dinoflagellate cysts and
twenty-three comparison records from Upper Wealden strata of Barremian age in the British Geological
Survey’s Warlingham Borehole, Surrey. Because these well-preserved dinocysts are not accompanied by any
other dinocyst palynomorphs, their presence is interpreted as a short-lived and local record of very low palaeo-
salinity. Other occurrences at different levels in the Wealden usually contain three or four dinocysts of known
marine taxa, and appear to represent more widespread and perhaps stronger marine incursions. Palynofacies,
ostracods, and clay minerals are briefly discussed in connection with the possibility that these new dinocysts may
represent the earliest known occurrences of non-marine dinoflagellates to produce fossilisable cysts.

The purpose of this paper is to illustrate and to describe the circumstances of occurrence of a new
dinoflagellate cyst which has been found in abundance in some palynomorph-bearing Wealden
samples. No other dinocysts have been found in these samples which are from the upper part of the
Weald Clay in the British Geological Survey (B.G.S.) Warlingham Borehole in Surrey (Worssam and
Ivimey-Cook 1971) and appear to be of late Barremian age. The new dinocysts, first illustrated by
Hughes ( 1 980), have no close parallels in the literature and appear to be relatively simple peridinioids;
their occurrence without other dinocysts may indicate a palaeosalinity less than the marine salinity of
the time, and if so the beginning of a trend towards the fresh-water dinocysts of today. It may therefore
be necessary to qualify the generally accepted view (e.g. Hughes and Moody-Stuart 1967) that through
the Mesozoic any dinocyst occurrence indicates a marine environment.

PREPARATION METHOD

Samples have been treated by a standard palynologic extraction method for light microscope study: dilute HC1,
HF, cone. HN0 3 oxidation (time stated with record sample description), dilute NH 4 OH (to avoid using KOH
which causes swelling), and zinc bromide (S.G.2.0) as heavy liquid. Although residues are now stored in
glycerine/water, some early preparations twenty-five years ago were stored in glycerine jelly and have been
successfully brought into use with hot water and/or HC1/HF.

Slides for light microscopy were made by a standard Hydramount/Depex method; some of the older
preparations mentioned below for comparison records were mounted in Clearcol or even in glycerine jelly.
Micrographs were taken on Kodak Technical Pan Film 2415 (rated 100 ASA) using a Nikon FX35A camera on
Nikon Labophot microscope 01819.

SEM stubs were prepared with strew residue over a Cambridge Mark 2 nickel grid (see Hughes et al. 1979).
Micrographs were taken on Ilford 70 mm film FP4, using a Phillips 501B Scanning Electron Microscope. All
specimens and preparations are lodged in the Sedgwick Museum.

RECORD OF SPECIMENS

All specimens observed are here listed as belonging either to a biorecord (unchangeable reference
taxon) or to a graded comparison record (see Hughes 1976, p. 26). To distinguish this separate
treatment in which neither emendation of a taxon definition nor unqualified attribution of specimens
to a taxon are admissible, the taxa are not arranged in any formal hierarchy nor are the names
latinized.

[Palaeontology, Vol. 28, Part 3, 1985, pp. 555-565, pis. 65-67. |

556

PALAEONTOLOGY, VOLUME 28

Qualified attribution in the form of graded comparison records is set out as follows:

cfA = variation in the comparison record agrees with that described for the reference taxon in all qualitative
aspects, but minor quantitative differences may be included. This grade (cfA) is taken to indicate such close
resemblance as would indicate stratigraphic correspondence.

cfB = variation in the comparison record may differ from that of the reference taxon in one specified
qualitative aspect. This comparison is taken to indicate less close stratigraphic correspondence. The grade cfB
occurrences would normally stratigraphically enclose those of cfA.

cfC = adequate resemblance to the reference taxon to be useful in discussion, but normally used for
occurrences which would be made into new biorecords if such were justified by geological necessity. Any one
occurrence may be compared cfC with more than one reference taxon.

Any of these records may be placed by others in attribution to a Linnean taxon if that procedure
appears to be rewarding, but the records themselves are designed to remain unchanged and separately
retrievable for any purpose.

Group DINOCYST PERIDINIOID CRETACEOUS
Genusbox cincturo-

Diagnosis. Peridinioid dinocyst with ambitus subcircular to ovoidal, without pericoels or horns. Apex
usually exhibits slight asymmetry. Periphragm thickness < 05 /<m; sculpture rugulae and/or
verrucae. Paratabulation obscure except where indicated by archaeopyle, believed to be (31) or (A 3 3I).
Pericingulum distinctly sculptured; perisulcus devoid of sculpture.

Comparison. Cyclopsiella Drugg and Loeblich is unsculptured.

Archaeopyle references. Evitt (1967) and Norris (1978, p. 303).

Biorecord cincturo-judith
Plate 65, figs. 1-9; Plate 66, figs. 1-5; text-fig. 1

Record sample. B.G.S. Warlingham Borehole, depth (ft) 1258/2-4 (in.); greenish-grey compacted mudstone
(blocky with weak lamination) with pyrite as streaks, burrow infillings and coatings on small fossils; small
red-brown oxidation patches. Fauna: ostracods, small black fish spines, fish vertebrae, bone fragments, and
brown spines. Preparation CH053/1(W120): oxidation 10 min cold cone. HN0 3 , centrifuged, heavy liquid
mineral separation. Palynologic facies (200 counted): palynomorphs compressed, little damage, little corrosion;
48-5% Cincturo-Judith, 1% tectate-columellates, 34% saccates; 15% Classopollis, 1 % Eucommiidites, 2-5%
Cicatricosisporites , 11-5% other triletes.

Dimensions (100 specimens). Length (31 4) 37 0 (45 7) pi n; breadth (25 7) 331 (42 8) pm.

Description. Shape and structure. Pericyst: ambitus ovoidal or sub-circular, elongated anterior to posterior,
widest immediately post-cingulum; slight dorsoventral flattening. Epicyst and hypocyst of approximately equal
size. Apical region rounded; apex asymmetrically located in 40% of observed specimens; 24% displayed a short
broad apical papilla. Antapex rounded. Periphragm thickness 0 4 /mi (2 specimens; SEM photos); sculpture of
rugulae (max. width 0-75 /mi) and verrucae (max. diameter 0-6 /an). The specimens with apical papilla have 8 to 10

EXPLANATION OF PLATE 65

Figs. 1-9. Cretaceous dinoflagellate. Biorecord Cincturo-judith: Barremian age; Upper Wealden, Warlingham
Borehole, depth 1258' 2-4"; stub IC 105, preparation CH053, sample WM 1258' 2 4". Films HSF62 and 67.
1, 2, 4, 5, 7-9, x 1500; 3, x 3000; 6, x 6500. Scanning electron micrographs. 1, ref. 213789, dorsal view
(HSF67/07); 2, ref. 360823, dorsal view (HSF62/08); 3, detail of same, sculpture in pericingular region
(HSF62/09); 4, ref. 238786, ventral view showing broad perisulcus (HSF62/06); 5, ref. 246892, oblique dorsal
view, with prominent pericingulum (HSF62/07); 6, ref. 358875, endocyst exposed where pericyst is torn
(HSF62/10); 7, ref. 315915. dorsal view (HSF62/11); 8, ref. 345910, slightly oblique ventral view showing
perisulcal fold (HSF67/01); 9, ref. 337796; oblique view (HSF67/03).

PLATE 65

HUGHES and HARDING, Cretaceous dinoflagellates

558

PALAEONTOLOGY, VOLUME 28

text-fig. I. Generalized reconstruction of Cincturo-Judith, designed to show important morphological
characters. Left, ventral view; right, dorsal view, x 1750.

rugulae radiating irregularly from the papilla. Endocyst: distinguishable in specimens with antapical ‘pericoef
(4%); closely adpressed to pericyst in most other specimens but separation of two walls seen when cyst is
ruptured. Endophragm laevigate, 0-5 /mi thick but considerably thinner in apical third of epicyst. Paratabulation.
Pericingulum: planar, always prominent; anterior and posterior parasutures shown as continuous ridges 0-3 /mi
wide set in smooth pandasutural zone 0 8 /mi wide; a row of papillate verrucae borders the pericingulum above
and below, while the main sculpture of parallel cross striae (rugulae) fills the whole 3 /mi width effectively
obscuring any paratabulation. Perisulcus: strongly developed, mainly on hypocyst; depressed area of much
reduced sculpture, bounded (in 59 % of specimens) by prominent parallel sulcal folds 4 -6 /mi apart. Archaeopyle:
(31) or ( A 3 3I) type, recognizable (4 % of specimens) by broadly triangular opening extending almost to apex; 59 %
of cysts showed no excystment feature, but 35 % were torn or ruptured in the apical region perhaps as a result of
archaeopyle formation.

EXPLANATION OF PLATE 66

Figs. 1-12. Cretaceous dinoflagellates. Barremian age; Upper Wealden, Warlingham Borehole. Films FlOF 13
and 14. 1-5, depth 1258' 2-4", slide CH053/1, biorecord Cincturo-Judith; 6-8, depth 1209' 9", slide W124/1,
cfB Cincturo-Judith; 9-12, depth 1 198' 2', slide BP556/3, biorecord Cincturo-Domed. All figures x 1000, light
micrographs. 2, 3, 10, phase contrast. Microscope reference SM306(OR3)Feitz646388. 1, 2, ref. 52.9/122.3,
specimen focussed to show archaeopyle (HOF13/32, 13/35); 3, ref. 35.3/120.4, specimen showing asymmetry of
apex (HOF13/12); 4, ref. 44.0/116.6, subcircular specimen with ?archaeopyle (HOF13/24); 5, ref. 38.3/124.6,
focussed on pericingulum (HOF13/17); 6, ref. 29.5/124.1, slightly oblate specimen (HOF14/16); 7, ref. 28.1/
126.2, specimen with reduced sculpture and thin apical region (HOF14/18); 8, ref. 50.4/123.1, showing
prominent pericingular folds ( HOF' 1 4/1 4); 9, ref. 43.5/119.9, specimen showing ‘domed’ epicyst and perisulcal
fold (HOF14/19); 10, 1 1, ref. 49.7/1 1 3.2, showing general morphology (HOF14/33, 14/34); 12, ref. 59.2/1 18.3,
displaying broad perisulcus (HOF 14/32).

PLATE 66

6

9

HUGHES and HARDING, Cretaceous dinoflagellates

560

PALAEONTOLOGY, VOLUME 28

Distinction. The single unnamed peridinacean cyst (Batten 1982, fig. 8.11/) with a prominent
sculptured pericingulum appears to have a (4A3I) archaeopyle and dissimilar sculpture.

Comparison. Taxa in the genusbox Cincturo- are readily distinguishable from those in such genera as
Palaeoperidinium (Deflandre) Sarjeant and Subtilisphaera (Jain and Millepied) Lentin and Williams,
by their lack of apical and antapical horns. They can also be distinguished from the Tertiary genus
Saeptodinium Harris by the strongly developed pericingulum, and from Geiselodinium Krutzsch by the
lack of pericoels. The overall morphology bears some resemblance to Cyclopsiella mura Duxbury,
although this is an unsculptured species of a genus which Stover and Evitt (1978) regarded as
acritarchs.

Name. The first observations were made by Mrs. Judith C. Moody-Stuart in 1966.

COMPARISON RECORDS

cfA Cincturo-Judith: Warlingham Borehole; lithology as for WM1258/2-4, except for differences
noted. Specimen counts represent whole content of one slide in each case.

WM 1254/1-3
WM 1254/8
WM 1255/3-6
WM 1255/7-8
WM 1256/3-4
WM 1258/2-4
WM 1258/9
WM 1259/0
WM 1259
WM 1259/6-9
WM 1260/ —
WM 1261/10-1
WM 1262/0-1
WM 1262/2-3
WM 1262/6
WM 1277/9
WM 1278/8
WM 1279/2
WM 1279/3
WM 1297/10

(18 specimens):
(56 specimens):
(309 specimens):
(34 specimens):
(95 specimens):
(428 specimens):
(401 specimens):
(63 specimens):
(2 specimens):
(78 specimens):
(210 specimens):
1 (588 specimens):
(294 specimens):
(490 specimens):
(126 specimens):
(21 specimens):
(7 specimens):

(2 specimens):

(5 specimens):
(11 specimens):

plus bivalves, Euestheria. Prep. W067/1.
darker and more brownish colour. Prep. K 127/1.

Prep. W 100/1.

plus bivalves, gastropods. Prep. W 121/1.
darker and more brownish colour. Prep. M68/1.

Biorecord (see above).

Prep. Z212/3.

plus Euestheria. Prep. Z 182/3.

Prep. CH009/2.

Prep. Z207/1.

Prep. W 1 19/2.

plus Euestheria. Prep. W068/1.
plus Euestheria. Prep. W069/1.

Prep. K098/2.

plus Euestheria. Prep. W108/1.

brownish; flaser bedding, pale brown; rare ostracods. Prep. Z216/3.
brownish; no ostracods. Prep. BP558/1.
brownish; no ostracods. Prep. W118/1.
brownish; no ostracods. Prep. Z217/3.

brownish; flaser bedding, pale brown; no ostracods. Prep. Y563/4.

cfB. CINCTURO-JUDITH
Plate 66, figs. 6-8; Plate 67, figs. 1-3

Record sample. B.G.S. Warlingham Borehole, depth (ft) 1209/9 (in); olive green compacted mudstone with
brownish tint (blocky with very weak lamination); few pyrite patches. Fauna: numerous fish fragments, black

EXPLANATION OF PLATE 67

Figs. 1-8. Cretaceous dinoflagellates. Barremian age; Upper Wealden, Warlingham Borehole. 1-3, stub IC112,
preparation W124, sample WM1209' 9", cfB Cincturo-Judith; 4-8, stub IC 161, preparation BP556, sample
WM1 187' 2", biorecord Cincturo-Domed. Films HSF62, 63, and 66. 1, 4, 5, 7, 8, x 1500; 3, 6, x 3000. 1, 3,
ref. 232754, 1, oblique ventral view showing perisulcal fold (HSF62/15); 3, detail of sculpture (HSF62/16);
2, ref. 236784, showing reduced size of epicyst (HSF62/14); 4, ref. 363771, dorsal view showing domed epicyst
(HSF63/09); 5, ref. 361775, oblique ventral view showing epicystal archaeopyle (HSF63/08); 6, ref. 365760,
detail of sculpture (HSF66/02); 7, ref. 350761, dorsal view (HSF63/13); 8, ref. 351762, oblique apico-ventral
view (HSF63/12).

PLATE 67

HUGHES and HARDING, Cretaceous dinoflagellates

562

PALAEONTOLOGY, VOLUME 28

bones, vertebrae, spines. Preparation W124/1: oxidation 10 min cold cone. HNO a , centrifuged, heavy liquid
mineral separation. Palynologic facies: 31 % cfB. Cincturo-Judith, 1 % tectate-columellates and monosulcates,
28-5% saccates, 10% Classopollis, 3% Eucommiidites . 3-5% Cicatricosisporites , 22% other triletes, 1% fungal
debris.

Dimensions (50 specimens). Length (45 6) 38-5 (34-2) pm; breadth (45 6) 38-3 (31-4) pm.

Distinction from biorecord. Pericyst shape: high proportion (42%) broader than long; hypocyst larger,
about six-tenths of whole. Periphragm: thickness 0-5 pm (one SEM photo); sculpture less prominent
as verrucae (max. diameter 06 pm) or short rugulae (max. length 3 pm). Paratabulation: pericingulum
wider, up to 4 pm. but often reduced by secondary folding; cingular sculpture of less distinct uneven
vertical corrugations; pandasutural zone bordered by discrete verrucae; parasulcus emphasized by
folds (70% of specimens); archaeopyle not seen, but apical rupture frequent.

Other cfB records

WM 1180/6 (36 specimens): Prep. K345/1.

WM 1181/2-10 (85 specimens): same lithology as biorecord, more fish remains. Prep. M679/1.

WM 1184/0-2 (40 specimens): more fish remains. Prep. K 125/1.

WM 1209/9 (672 specimens): (see above).

Paly nofacies. All of the cfA and cfB records occur in palynofacies rich in amorphous finely com-
minuted organic matter (‘mush’). This material may be a partial case of the greenish colouration of the
samples. Botryococcus has rarely been observed in these preparations and without it there is no
reason to regard the ‘mush’ as indicative of an anoxic lacustrine environment in the sense of Sladen
and Batten (1984).

Biorecord cincturo-domed
Plate 66, figs. 9-12; Plate 67, figs. 4-8

Record sample. B.G.S. Warlingham Borehole, depth (ft) 1187/2 (in.); khaki greenish compacted blocky mud-
stone, almost conchoidal fracture, ‘greasy’ to touch, weak lamination, small pyrite patches. Fauna: fish
vertebrae and spines (brown), Euestheria. Preparation BP556/3; oxidation 60 min cold cone. HN0 3 , centri-
fuged; mounted in glycerine jelly. Palynologic facies (200 counted): 51-5% Cincturo-Domed, 4-5% tectate-
columellates, 15-5% saccates, 10% Classopollis. 1 % Eucommiidites. 3% Cicatricosisporites. 3% other triletes,
6-5 % Celyphus rallus.

Dimensions. (50 specimens). Length (48-5) 42-3 (31 4) pm, breadth (48-5) 39-6 (28-5) pm.

Description. Shape and structure. Pericyst: ambitus pear-shaped, elongated anterior to posterior; epicyst
blunt-ended concave-sided; hypocyst rounded, of comparable overall dimensions with epicyst; greatest width
immediately post-cingular; slight dorsoventral flattening. Periphragm: thickness 0-3-0 4 pm (observed on two
SEM photos); sculpture more pronounced on hypocyst, short rugulae (up to 5 pm long) and/or verrucae; epicyst
rugulae of low relief, continuous, radiating from blunt apex. Endocyst: laevigate, 0-4-0- 5 pm thick (SEM photos),
thinning rapidly in apical third of epicyst. Paratabulation. Pericingulum: planar, narrow, distinct; width
(3-4 pm), sculpture uneven vertical corrugations and striae obscuring tabulation; width reduced by secondary
compression; pandasutural zones without sculpture. Perisulcus: mainly on hypocyst, 4 pm wide; bounded by
parallel ridges, secondarily developed into folds. Archaeopyle: not clearly observed, but frequent dorsal
rupturing of cyst wall in ‘intercalary’ position, affecting up to one-third of epicyst.

Distinction. Only known from this sample (1258 specimens); differs in shape and exeystment feature
from Cyclopsiella mura Duxbury 1983. Accompanied by very rare Veryhachium-type acritarchs.

STRATIGRAPHIC POSITION OF SAMPLES

Succession. The main occurrences of Cincturo-Judith are in the strata immediately below Topley’s
Bed 6 ‘Large Paludina’ at 1252 ft (Worssam and Ivimey-Cook 1971, p. 22). The occurrence of
Cincturo-Domed and most of the cfB records of Cincturo-Judith are from just above Topley’s

HUGHES AND HARDING: BARREMIAN DINOFLAGELL ATES

563

Bed 7 ‘Sandstone’ at 1210-1238 ft. The base of the Lower Greensand (Aptian) is taken at 1047 ft,
making all these samples Barremian or earlier.

Ostracods. F. W. Anderson (in Worssam and Ivimey-Cook 1971) placed all these beds in his Cypridea
bogdenensis beds of his C. clavata Zone, and further placed his Gillmans Cycle of ‘C’ and ‘S’ phases
between 1248 and 1263 ft. Our samples (1254-1263 ft), which were taken from the cores after those
used by Anderson, included identifications of C. clavata (spinose), C. valdensis (large, punctate),
C. rotundata (medium size, punctate), C. spinigera (unispinose), Theriosynoecum fittoni (especially
from WM 1258/9 in great numbers), Darwinula leguminella , and D. oblonga. Less certain identifica-
tions were of rare Mantelliana mantelli and Wabanella boloniensis (both from WM 1258/9) and
IMiocytheridea henfieldensis (from WM 1255/7-8). Of these all the Cypridea , Darwinula , and
Miocytheridea species form part of Anderson’s ‘C’ phase of fresh/brackish water; Theriosynoecum is
found in association with Cypridea and is probably from fresh-water (Kilenyi and Allen 1968, and our
data); the others relate to Anderson’s ‘S' phase, which he regarded as ‘marine’. The age indication by
Anderson (1973) and Kilenyi and Neale (1978) for these samples would be Late Hauterivian to Early
Barremian.

Angiospermid pollen. Various tectate-columellate pollen from lower levels in the Warlingham
Borehole down to 1415 feet (Hughes et al. 1979) are believed from correlation with similar occurrences
from north of the London-Brabant uplift to be of early Barremian age.

PALYNOMORPHS AND WEALDEN PALAEOSALINITIES

Hitherto the presence of dinocysts in Mesozoic palynomorph assemblages in which reworking was
considered unlikely, has normally been taken as straightforward indication of marine conditions.

Recently it has been pointed out (Hughes 1980; Batten 1982; Batten and Eaton 1980) that several
horizons through the otherwise fresh-water Wealden provide restricted dinocyst assemblages;
individual horizons may contain up to four dinocyst species normally found in marine rocks of
comparable age. These horizons have been interpreted as resulting from marine incursions; but
presumably these incursions were both incomplete and short-lived because stenohaline marine
megafossils have not been found except in one case at the very top of the Weald Clay in Kent, just
below the main Aptian incursion (Casey 1961).

The occurrences described in this paper differ in each featuring only one dinocyst taxon that is in
abundance up to 75% and that is not known to exist elsewhere in any fully marine circumstances.

Finally the great majority of Wealden palynomorph assemblages include no dinocysts or acritarchs
at all and are consequently interpreted as of fresh-water origin.

It is tempting to regard these occurrences as a gradation from fully marine, through major
fluctuations say down to one half salinity (with three or four dinocyst taxa) to minor fluctuations with
say some such low level as one quarter salinity (with one unique taxon) to fresh water. No salinity
percentage figures could be suggested because the Cretaceous general marine salinity was probably
not 34-4% 0 . In addition it is perhaps more likely that incursive water-masses bearing living dinocysts
were both local and transient, and that no sector of the Wealden ‘lake’ settled down to a distinct
geographical salinity gradient as in the present Baltic Sea with steady conditions and organism
representation for any length of time.

OTHER INDICATIONS OF WEALDEN PALAEOSALINITY

It is important that in the Wealden of Southern England from beginning Berriasian to end Barremian
no cephalopod, echinoderm, coral, foraminiferan, or brachiopod has been recorded and there were
thus no fully marine situations.

Anderson (1973) claimed that his ‘S’ phase ostracods represented marine conditions, alternating
with his more usual ‘C’ phase ostracods representing fresh/brackish water. All of our samples from

564

PALAEONTOLOGY, VOLUME 28

Warlingham 1254 to 1263 contained ostracods in quantity and of these only rare specimens of
IMantelliana and IFabanella at WM 1258/9 could be taken to represent ‘S’ phase; all others were
‘C’ phase ostracods. Anderson ( 1971) regarded the triple ‘S' phase, styled by him, Gillman’s cycle, as a
distinct incursion; our ostracod fossils from fifteen samples in these 10 ft of strata do not indicate
anything dramatic. Possibly there was a minor fluctation as indicated by the dinocysts, which is also
in keeping with the findings of Kilenyi and N. W. Allen (1968) but scarcely with those of P. Allen and
Keith (1965).

Dr. C. V. Jeans has kindly examined by X-ray diffraction the clay mineral assemblages (< 2/im
e.s.d.) from a series of eight samples spanning the horizons of occurrence of these dinocysts (Jeans
1978, p. 625 for method). Identification was carried out at the level of clay mineral groups and he
concluded that no systematic variations within these assemblages could be related to these particular
horizons. Such a view does not conflict with our interpretation that these consistently greenish
samples bearing ostracods, fish fragments, and pyrite, in addition to abundance of a unique dinocyst,
represent a minor salinity fluctuation. We further do not accept the suggestion of Sladen and Batten
(1984, p. 161) that these salinity changes could reflect evaporation cycles in the basin.

Clearly, chemical investigations should have been and have been attempted (Allen et al. 1973) but
the results are cautious and appear inconclusive in the context of our present observations. The
bivalve Filosina referred to by Allen et al. (1973) as marine, has not been observed in the borehole
sector we are studying.

CONCLUSIONS

Our evidence and that of previous papers suggests that in general early Cretaceous dinocysts were
confined to marine conditions which their presence may be taken to indicate.

On the other hand the dinocysts Cincturo-Judith and Cincturo-Domed as described above appear
to be genuine dinocysts which occur in a facies which is distinct from the rocks above and below and
represented perhaps some slight degree of salinity. The occurrence of these dinocysts in abundance,
without any others present, suggests the beginning of a trend that may have led in late Tertiary and
Recent time to the freshwater armoured dinoflagellate thecae.

Acknowledgements. We are grateful to assistance from Mrs. Audrey McDougall with sample preparation, to Mr.
David Newling with Scanning Electron Microscopy, and to Mr. R. Lee and Mr. R. Hill with light microscopy.
I. C. H. acknowledges a NERC studentship GT4/82/GS/122 and N. F. H. a NERC Research Grant 3/4048.

REFERENCES

allen, p. and keith, m. l. 1965. Carbon isotope ratios and palaeosalinities of Purbeck- Wealden carbonates.
Nature , Loud. 208 , 1278-1280.

tan, F. c. and deines, P. 1973. Isotopic ratios and Wealden environments. Palaeontology, 16 , 607-621.
anderson, F. w. 1971. The sequence of ostracod faunas in the Wealden and Purbeck of the Warlingham bore-
hole. Bull. Geol. Surv. Gt. Br. 36 , 122-138.

1973. The Jurassic-Cretaceous transition: the non-marine ostracod faunas. In casey, r. and rawson, p. f.
(eds.). The Boreal Lower Cretaceous , 101-110. Geol. Jl, Spec. Issue No. 5. Liverpool.
batten, d. J. 1982. Palynofacies and salinity in the Purbeck and Wealden of southern England. In banner, f. t.
and lord, a. r. (eds.). 278-295. Allen & Unwin.

and eaton, g. l. 1980. Dinoflagellates and salinity variations in the Wealden (Lower Cretaceous) of
southern England, 32. In Abstracts 5th Int. Palynological Conf. Cambridge 1980.
casey, R. 1961. The stratigraphical palaeontology of the Lower Greensand. Palaeontology, 3 , 487-621.
duxbury, s. 1983. A study of dinoflagellate cysts and acritarchs from the Lower Greensand (Aptian to Lower
Albian) of the Isle of Wight, Southern England. Palaeontogr. Abt. B. 186 , 18-80.
evitt, w. r. 1967. Dinoflagellate studies II, the archaeopyle. Stanford Univ. Publ. Geol. Sci., 10 ( 3 ), 1-83.
hughes, N. F. 1976. Palaeobiology of angiosperm origins, pp. 24-31. Cambridge University Press, Cambridge.

1980. Palynological correlation of Early Cretaceous freshwater to marine strata. Proc. IV Int. Palynol.
conf. Lucknow (1976-1977), 2, 497-499.

HUGHES AND HARDING: BARREMIAN DINOFLAGELLATES

565

drewry, G. E. and laing, J. F. 1979. Barremian earliest angiosperm pollen. Palaeontology , 22, 513-525.
and moody-stuart, j. c. 1967. Palynological facies and correlation in the English Wealden. Rev. Palaeobot.
Palynol. 1, 259-268.

jeans, c. v. 1978. The origin of the Triassic clay assemblage of Europe with special reference to the Keuper
Marl and Rhaetic of parts of England. Phil. Trans. Roy. Soc. A 289, 549-639.
kilenyi, t. j. and allen, n. w. 1968. Marine-brackish bands and their microfauna from the lower part of the
Weald Clay of Sussex and Surrey. Palaeontology , 11, 141-162.

and neale, j. w. 1978. The Purbeck/Wealden. In bate, r. h. and robinson, e. (eds.). A stratigraphical
index of British Ostracoda , 299-324. Geol. J1 Spec. Issue No. 8. Liverpool.
norris, G. 1978. Phylogeny and a revised supra-generic classification for Triassic-Quaternary organic-walled
dinoflagellate cysts (Pyrrophyta). Part I, Cyst terminology and assessment of previous classifications. Neues
Jahrb. fur Geol. und Paldont. Abhandl. 155, 300-317.

sladen, c. p. and batten, d. j. 1 984. Source-area environments of Late Jurassic and Early Cretaceous sediments
in Southeast England. Proc. Geol. Assoc. 95 (2), 149 164.
stover, l. e. and evitt, w. r. 1978. Analyses of pre-Pleistocene organic-walled dinoflagellates. Stanford Univ.
Publ. Geol. Sci. 15, 1-300.

worssam, b. c. and ivimey-cook, h. c. 1971. The stratigraphy of the Geological Survey Borehole at Warling-
ham, Surrey, Bull. Geol. Surv. Gt. Br. 36, 1-146.

N. F. HUGHES
I. C. HARDING

Typescript received 10 July 1984

Revised typescript received 13 November 1984

Department of Earth Sciences
University of Cambridge
Downing Street
Cambridge CB2 3EQ

A NEW GENUS OF C A R BO N I F E RO U S SPIRIFERED
BRACHIOPOD FROM SCOTFAND

by MARIE LEGRAND-BLAIN

Abstract. The microsculpture and vascular markings of the classical species ‘ Spirifef trigonalis are described,
mainly from Scottish Brigantian (late Visean) specimens. The genus Angiospirifer and subfamily Angiospiri-
ferinae are created. Angiospirifer is related to the genus Brachythyrina, which became widespread during the
upper Carboniferous.

Carboniferous spiriferids present unusual systematic difficulties, especially when specimens have
a simple external shape and ornamentation, and when the microsculpture and internal features
become taxonomically important. Among upper Palaeozoic spiriferids, in addition to the dental
plates, the mantle canals are important taxonomic and evolutionary features (Ivanova 1960, 1971;
Lazarev and Poletaev 1982). However, good internal surfaces or moulds are needed for such
observations: in consequence, mantle canals are unknown on many of the long-established Dinantian
species that were collected from limestone facies.

‘S.’ trigonalis (Martin, 1809) sensu Muir-Wood 1956, a species often cited from the Eurasian and
North African Carboniferous, has been studied in detail by Dunlop (1961), so its biometry and shell
structure are well established. This paper adds to Dunlop’s study by researching the microsculpture
and mantle canals of ‘S.’ trigonalis : a species previously attributed to a variety of genera, but here
assigned to a new genus, Angiospirifer.

SYSTEMATIC PALAEONTOLOGY
Genus angiospirifer nov.

Type species. Spirifer trigonalis (Martin, 1809) sensu Muir-Wood 1956; Brigantian, Scotland.

Derivatio nominis. From the Greek word aggeion (= vessel): a reference to the development of the vascular
markings.

Diagnosis. Shell of trigonal shape, megathyrid to slightly brachythyrid; lateral ribs conspicuous, rarely
divided, median sinus and fold costate; sub-imbricated micro-lamellae bearing faint radial tubercles;
adminicula short, intra-sinal; elongated apical callosity located between apex and muscle scar; genital
markings surrounding the apical-muscular area; vascular markings roughly reticulated, developed
on adult interior lateral ventral region.

Description of the type species: S. trigonalis

The type specimen of the species trigonalis is a neotype, proposed by Muir-Wood to the International
Commission on Zoological Nomenclature (1956, p. 1 12). It bears British Museum (Natural History)
number BB 7340; and is figured by Davidson (1858, pi. 5, fig. 33; 1863, pi. 50, fig. 4) and Dunlop (1961,
pi. 64, figs. 1-3). The locality ‘Cousland, near Dalkeith, Midlothian’, is a well-known old quarry in the
North Greens Limestone and underlying shales (Wilson 1974, locality 32; P. Brand, pers. comm.), of
upper Brigantian age (George et al. 1976, fig. 14j). The lost original specimens described by Martin
came from a different locality: ‘Derbyshire’, with an imprecise age of Asbian to Brigantian. So, the
neotype choice may have alterated the original sense of the species trigonalis.

I Palaeontology, Vol. 28, Part 3, 1985, pp. 567-575.|

568

PALAEONTOLOGY, VOLUME 28

LEGR AND-BLAIN: NEW CARBONIFEROUS BRACHIOPOD

569

Newly collected topotypes would have provided the firmest basis for this study, but they proved
impossible to obtain. P. Brand (British Geological Survey, Edinburgh) recently visited Cousland
quarry and found it partly filled. Some specimens collected from Cousland have been examined in
the British Geological Survey, Edinburgh (EV 517-519, 524, B 1735°, B 1787°, T 3277), but they are
poorly preserved.

On the other hand, D'Arcy quarry, 4 km S-SW of Cousland in the North Greens Limestone
(Wilson 1 974, locality 3 1 ), has yielded good specimens ( British Geological Survey, Edinburgh, T 3278;
text-fig. lc, m). In addition the S. trigonalis material described by Dunlop (1961) is from another
Scottish locality: Brockley, near Coalburn, Lanarkshire. This locality is 65 km SW from Cousland, in
the Douglas Main Limestone, at the base of the lower Limestone group, and a little below the stratum
typicum (George et al. 1976, fig. 14b). Dunlop (1961) has established statistically that the Brockley
spiriferid population belongs to the species trigonalis. I have collected extra specimens from this
locality (text-fig. 1b, d). Therefore, in the absence of good trigonalis topotypes, D’Arcy and Brockley
specimens may be used. In both localities the shells are preserved in shaly limestones, those from
D’Arcy being rather fragile and distorted. Internal moulds have been prepared by partly removing the
shell material with needles (text-fig. 2a, d, f).

Other specimens are natural internal moulds from outside Scotland: British Geological Survey
no. RV 733, figured in Burgess and Holliday (1979, pi. 6, fig. 14) and herein (text-fig. 1a), from the
Pendleian of Cumbria; British Museum (Natural History), Gilbertson Collection no. B 244, figured by
Waterhouse (1970, pi. 2, figs, g-j), from an unknown locality, probably Bolland, Yorkshire;
B.M.(N.H.) nos. BD 2062-2063 (text-fig. 1e, f, g), from the late Dinantian of Alnwick, Northumber-
land. These latter specimens show apical interiors.

Morphological terms. Most morphological terms are in current use, as in Williams and Rowell (1965). Some less
common terms not cited by Dunlop (1961) are listed in Table 1.

Biometry. The Brockley population has been studied statistically by Dunlop (1961, p. 480, text-
fig. 1, tables 1, 2). Some specimens from D’Arcy attain 42 mm in width, whereas the Brockley shells
never exceed 39-3 mm wide. The external variability of A. trigonalis is important, and more
so than previously recognized. The neotype is decidedly narrower than the largest members
of these populations (maximum width 27-8 mm; hinge width 24-9 mm; length 24 mm; thickness
18-2 mm).

text-fig. 1 A-G, M., Angiospirifer trigonalis (Martin, sensu Muir-Wood), a, British Geological Survey, Keyworth,
RV 733; natural internal mould of adult specimen, formerly figured by Pattison in Burgess and Holliday (1979,
pi. 6 fig. 14). Knucton Shell Beds, Pendleian, Coldberry Gutter (Cumbria), x 2. b, British Museum (Nat. Hist.)
BD 1700, prepared internal mould of adult pedicle valve collected by the author. Douglas Main Limestone of
Lower Limestone Group, Brigantian, Brockley (Lanarkshire), x 2. c, British Geological Survey, Edinburgh,
T 3278-1, prepared internal mould of adult pedicle valve. North Greens Limestone of Lower Limestone Group,
Brigantian, D’Arcy quarry (Midlothian), x 2. D, B.M.(N.H.) BD 1701, prepared internal mould of young pedicle
valve collected by the author. Same locality as b, x 2. e, f, B.M.(N.H.) BD 2062, adult pedicle valve;
Carboniferous Limestone (Brigantian), Alnwick (Northumberland), e, interior, showing delthyrium, elongated
apical callosity, muscle scar, x 2. f, fragments of stegidial plates, x 5. G, B.M.(N.H.) BD 2063, pedicle valve
interior. Same locality as e and f, x 2. m, B.G.S., Edinburgh, T 3278-2, external microsculpture. Same locality as in
c, x 20. H and K, Anthracospirifer pellaensis (Weller); h, B.M.(N.H.) BD 1703, prepared internal mould of adult
specimen given to the author by Dr. G. A. Cooper; Pella beds, Meramecian, 2 miles S of Pella (Iowa, USA), xl.K,
B.M.(N.H.) BD 1704, external microsculpture of specimen given to the author by Dr. A. S. Horowitz: Pella
Formation, Meramecian, Mahaska quarry (Iowa, USA), x 10. i, J, l, ‘ Parachoristes ’ (sensu Lazarev and Poletaev
1982). i, j, Cayton Gill beds, Kinderscoutian, Ripley, Harrogate (N. Yorkshire), i, B.G.S. Keyworth, 50261,
natural internal mould of adult pedicle valve, formerly figured as ‘ Spirifer a’ by George (1932, text-fig. 8), x 1.
J, B.M.(N.H.) BD 1702, natural mould of rather young pedicle valve collected by the author, x 1. l, B.M.(N.H.)
BD 1705, external microsculpture of specimen collected by the author. Oued el Hamar Formation, upper
Bashkirian, Oued Tagnana section, loc. ML 290, 15 km W-NW. of Bechar (Algeria), x 10.

570

PALAEONTOLOGY, VOLUME 28

table 1. Morphological terms not cited in Dunlop (1961) or differently interpreted, and their equivalents in

the literature

Terms and their authors
(when post Treatise , 1965)

Definitions

Equivalents in:

(1) Dunlop 1961;

(2) Vandercammen 1959, 1962;

(3) Lazarev and Poletaev 1982

Adminicula

Ventral parts of dental plates, joining to the
bottom of the pedicle valve

(1) dental plates

Apical callosity
Delthyrial plate

Secondary thickening inside the umbo of
the pedicle valve

Transverse plate connecting the two dental
plates in their apical part, below the inter-
area level

(1) ‘ventral septum’

Genital markings

Radial ridges or pits on inside of shell within
genital areas

(2) gonoglyphe

(3) alveolate vascular system

Stegidial plates:

Pairs of laminate plates making up the del-

(1) a part of the ‘delthyrial plate’

Cowen 1968; Grant 1976

thyrial cover, forming a convex arch
above the interarea level; they are fragile
and rarely entirely preserved

(2) plaques deltidiales

Tabellae:
Waterhouse 1968

Dorsal supporting plates, called crural
plates by some authors

(1) crural plates

Vascular markings

Impressions of mantle canals on shell
interior

(2) angioglyphe

(3) pinnate, ramified, reticulate
vascular system

Microsculpture. The neotype exhibits fine but worn concentric lamellae. Davidson ( 1 863, pi. 50, fig. 9a)
figured the microsculpture of a specimen from Barrhead, Renfrewshire, Scotland. This specimen is not
preserved in the British Museum collections. However, Davidson’s observation is confirmed on
specimens from D’Arcy (text-fig. 1m). The concentric lamellae are variously spaced (4-10 per mm) and
are slightly raised and imbricated, like a tiled roof; their anterior borders display a festoon aspect, due
to fine tubercles being arranged regularly in radial rows. Sometimes, traces of these tubercles form
very fine radial lirae.

Delthyrium and stegidial plates. Dunlop (1961) described the area and delthyrial structures. None of
the Scottish specimens displays externally preserved stegidial plates, but fragments of these structures
are seen on specimens from Alnwick (text-fig. If) and on B.M.(N.H.) B 244 (Waterhouse 1970, fig. 2j).
Stegidial structures are found on apical sections of Brockley specimens, as described below.

Internal moulds and surfaces (text-figs. 1a-g, 2). On adult internal moulds the adminicula appear as
extremely short apical incisions, close to each other and within the sulcus. A narrow median apical
callosity forms a depression that runs from the inter-adminicular region to the posterior end of the
muscle scar; its length and shape are very variable, from short rhombic form (text-fig. 2d) to an
elongate septum-like shape (text-figs. 1g, 2a). The muscle scar is moderately to slightly depressed into
the shell substance, pyriform, and more or less separated from the apex, occurring either just in front
of the adminicular distal extremities (text-fig. 2d) or at some distance from them (text-fig. 2a). The
genital markings surround the region of the muscle scar and apical callosity, extending up on to the
inside of the interarea on both sides of the delthyrium (text-fig. 2c). This area bears small pits which
may be arranged in an indistinct radial pattern. The vascular markings, around the genital area,
extend over the whole posterolateral interior surface, including the lateral regions of the interarea.
Two distinct canals run from the tops of lateral apical cavities towards the lateral extremities. They
widen, divide into transverse canals, and form a rough reticulation upon the insides of the interarea

LEGRAND-BLAIN: NEW CARBONIFEROUS BRACHIOPOD

571

text-fig. 2. Angiospirifer trigonalis (Martin, sensu Muir-Wood), internal moulds, a, British Museum (Nat. Hist.)
BD 1700, adult pedicle valve. Douglas Main Limestone, Brigantian, Brockley (Lanarkshire), b, c, British
Geological Survey, Keyworth, RV 733, adult specimen, broken in the ventral muscular area. Knucton Shell Beds,
Pendleien, Coldberry Gutter (Cumbria). B, ventral view, c, dorsal view. D, E, British Geological Survey,
Edinburgh, T 3278-1, adult pedicle valve. North Greens Limestone, Brigantian, D’Arcy Quarry (Midlothian), d,
ventral view, e, dorsal view, f, B.M.(N.H.) BD1 701, young pedicle valve. Same locality as in a. ad = adminicula or
their internal moulds; ap = apical callosity; g = genital markings; m = muscle scar; r = external ribs on their

internal moulds; v = vascular markings.

and ears. Other radial canals extend from the lateral parts of the genital area, to branch and
anastomose together (text-fig. 2a, b, d). The shape of the external ribs is seen internally only
anteriorly. Brachial valve internal moulds (text-fig. 2c) display short thick tabellae; the cardinal
process and muscle scar are poorly preserved; reduced genital markings appear on both sides of the
muscle area.

A young pedicle valve, 25 mm wide, with a shell thickness of 04 mm, shows internally the external
ribs pattern within 7 mm of the apex (text-figs. Id, 2f). The adminicula are represented by two slender
incisions, 2 mm apart and 24 mm long; their apical parts are extra-sinal, whereas their distal
extremities coincide with the sinal borders. There is no apical callosity. The muscle scar is hardly
distinguishable, but posteriorly it projects to between the adminicula, whereas its anterior border

572

PALAEONTOLOGY, VOLUME 28

seems to extend beyond them. The genital and vascular markings are not differentiated. Lateral to the
adminicula, thin radiating canals run in different directions from the external ribs.

On another young pedicle valve, 41 mm wide and 1 mm thick (T 3278-3, not figured), the external
ribs are seen internally after one third of the length of the valve. Already the muscle scar is located
anterior of the apex, from which it is separated by a median callosity 4 mm long. The genital and
vascular markings are again poorly differentiated.

The characteristic adult pattern of genital and vascular markings appears on the pedicle valves
when the shell, in the posterior region, is at least 2 mm thick. Thus the ontogeny of A. trigonalis is
characterized by a rapid forward migration of the muscle scar; the development of the vascular
markings is linked with a thickening of the shell in the whole posterolateral ventral region.

Apical sections of pedicle valves. Apical sections have been described by Dunlop (1961, text-figs. 4 e-f,
le, 12 a-e; pi. 65, fig. 7), and again here (text-fig. 3a-c). The dental plates are well developed, made of
two distinctive parts: 1, short converging dental ‘flanges’ bordering the delthyrium; 2, long parallel or
slightly diverging adminicula. Both parts are made of tertiary columnar shell, on both sides of a
slender median fibrous layer (Vandercammen and Plodowski 1967). The ‘median septum’ of Dunlop
(1961, text-figs. 4 </, 12 a,d,e) corresponds mainly to the median apical callosity, posterior to the muscle
scar. The growth surfaces (as seen in sections, e.g. Dunlop 1961, text-fig. 12e) show that there is no
fibrous layer such as occurs in true septa. On the other hand, in the umbo infilling, the ‘V’ structure
called a septum by Dunlop (1961, text-fig. 12a), resembles a recrystallization figure (‘faux coussinet
septal’ of Vandercammen and Plodowski 1967). The ‘delthyrial plate’ described by Dunlop poses a
problem: its anterior part is convex, protruding above the level of the interarea. In fact, it is pro parte a
stegidial cover, as is strongly suggested by sections exhibiting paired but unequally developed plates
(text-fig. 3b, c). A true single internal delthyrial plate is restricted to the top of the delthyrial cavity, but
on most adult specimens it is obliterated by the umbonal infilling and recrystalization.

Remarks. The species trigonalis has been attributed in the literature to several spiriferid genera: to
Spirifer by most authors in the past, and by Dunlop (1961); to Fusella by Buckman (1906), Beznosova
(1959), and (with some uncertainty) Brunton and Rissone (1976); to 1 Anthracosp irifer by Thomas
(1971); and to VJnispirifer by Afanas'eva (1975).

Spirifer Sowerby 1816 attains a greater size than Angio spirifer, and has frequently divided radial
ribs and a microsculpture of radial lirae; vascular markings are absent or reduced to fine and short
canal traces. Fusella M‘Coy 1844 has dental plates comparable to those of Angio spirifer, being

ad

ao

1 mm

A

st

text-fig. 3. Angiospirifer trigonalis (Martin,
sensu Muir-Wood), transverse sections of
pedicle valves showing structures reinter-
preted from Dunlop (1961). a, Dunlop’s text-
fig. 12c. B, Dunlop’s text-fig. 11c. C, Dunlop's
text-fig. 11 g. Douglas Main Limestone,
Brigantian, Brockley (Lanarkshire), ad =
adminicula or their internal moulds; ap =

B

C

apical callosity; st = stegidial plates.

LEGRAND-BLAIN: NEW CARBONIFEROUS BRACHIOPOD

573

positioned within the sinus (Brunton and Rissone 1976, fig. 2a). The muscle scars are indistinguish-
able. the median callosity is restricted to the tip of the umbo, and the vascular markings are reduced. A
further important difference is the virtual absence of ribbing in the ventral sulcus of Fusella.
Unispirifer Campbell 1957 has an ornamentation of finer and more numerous ribs than on
Angiospirifer and a microsculpture of radial lirae. Its adminicula diverge and surround the posterior
part of the muscle scar and the vascular markings are not differentiated. Anthracospirifer Lane 1963
has the same external shape and ornamentation as Angiospirifer , but it differs in: a, its microsculpture,
with radial lirae as prominent as the concentric lamellae, forming a characteristic reticulation (text-
fig. Ik; Sutherland and Harlow 1973, pi. 16, fig. 10); 6, the position of the ventral muscle scar, which on
adult internal moulds creates a strong apical relief (text-fig. 1h; Sutherland and Harlow 1973, pi. 16,
fig. 4); c, the adminicula, which diverge around the posterior part of the muscle scar, in an extra-sinal
position; d, the vascular markings, which seem to be absent. In the pedicle valve of Anthracospirifer the
posteriorly thickened shell thins rapidly towards the margins, so that the internal surface shows the
external ribbing. Prochoristitella Legrand-Blain 1968 resembles Angiospirifer by the position of
the adminicula, apical callosity, and muscle scar; however, the microsculpture differs from that of
the latter genus by better development of the radial lirae; mantle cavities are reduced. Brachythyrina
Fredericks 1929 has the same shape and ornamentation as Angiospirifer. Its apical callosity, muscle
scar, and mantle canals are comparable, but Angiospirifer is distinguished by possessing adminicula.
Choristitids, like Angiospirifer , are provided with slightly diverging or parallel adminicula, and
reticulate vascular markings (Lazarev and Poletaev 1982, pi. 1, figs. 4, 5): this pattern is well exhibited
on British specimens from the Kinderscoutian Cayton Gill beds (text-fig. li, j). The choristitid
microsculpture, seldom figured in Russian literature, is not preserved on the Cayton Gill
material. On Algerian Bashkirian specimens (text-fig. 1l) the microsculpture is rather similar to that
of Angiospirifer , although the concentric lamellae are more closely spaced. The differences between
choristitids and Angiospirifer are: a , the radial ribs are generally more slender and bifurcating in
choristitids; 6, the adminicula are longer and the apical callosity reduced or absent; c, the vascular
markings are more extensive, especially in the anterior part of the pedicle valve, and their reticulation
is more delicate than in Angiospirifer. Finally, the genus Subspirifer Shan and Zhao 1981, which
resembles Angiospirifer by shape and costation, differs by its smooth dorsal fold. Its microsculpture is
unknown and its internal structure inadequately figured.

Generic composition

It is difficult to assign with certainty species of the broad trigonalis group to Angiospirifer. Some of the
‘ trigonalis ’ specimens widely cited in the Eurasian and North African Carboniferous may not belong
to Angiospirifer. For example, ‘ Spirifer trigonalis ’ from Silesia (Zakowa 1958, pi. 3, fig. 9; 1966, pi. 15,
figs. 3, 6) of probable Asbian age, displays a prominent apical muscle scar and pinnate mantle canals; it
is probably not congeneric. The species bisulcatus Sowerby, sometimes considered synonymous with
trigonalis, does not belong to Angiospirifer, since the syntypes in the Sowerby Collection, examined in
the British Museum (Natural History), have a lirate microsculpture.

Some Serpukhovian Russian species, described by Semikhatova (1941) as ‘ trigonalis group’, may
be true Angiospirifer: their adminicula show a tendency towards reduction and the muscle scar is
located far from the apex (S. gamma Semikhatova, 1941, pi. 5, figs. 146, 166). The microsculpture of the
species parabisulcatus Semikhatova is identical to that of A. trigonalis (Ivanova 1971, text fig. 4; pi. 1,
fig. 4). The assignment of these species to Angiospirifer should be checked by observations of the
vascular markings.

In addition the present author is currently investigating several undescribed Angiospirifer species
in the upper Visean and Serpukhovian of the northern Algerian Sahara.

Phyletic and systematic position of the genus Angiospirifer

The evidence of reticulate vascular markings as early as the upper Visean is a significant feature for the
phyletic trends of Spiriferidae. Ivanova (1972) pointed that the complex choristitid canal system
evolved from a spiriferid stock. In the diagram of Lazarev and Poletaev (1982), such a development

574

PALAEONTOLOGY, VOLUME 28

was recorded only from Bashkirian choristitids, and regarded as an important criterion for
distinguishing lower and upper Carboniferous Spiriferids. There is now evidence for the existence of
reticulate vascular markings as early as the end of lower Carboniferous, both in Angiospirifer and in
Brachythyrina ( Anthracothyrina ) Legrand-Blain, 1984. These taxa are related to each other,
displaying a reduction in their dental plates (studied in some Russian species of the trigonalis group by
Yanichewsky (1935), Semikhatova (1941), and observed also in Algerian species). On the other hand,
Angiospirifer is a possible ancestor of choristitids. The origin of Angiospirifer should be sought in
pre-Brigantian specimens, probably among the genera Unispirifer or ProchoristiteUa.

Placing Angiospirifer within its correct position and amongst the spiriferids is currently impossible.
Ivanova (1972) attaches importance to the canal system which characterizes the families Brachy-
thyrididae and Choristitidae. The existence in Angiospirifer of a denticulate interarea separates it from
the Brachythyrididae. The Choristitidae of Ivanova (1972) have well-developed adminicula, so it is
impossible to include both Angiospirifer and Brachythyrina in that family.

Carter (1974) distinguished many subfamilies in the Spiriferidae, but without considering the
mantle canals. He doubtfully placed Brachythyrina in the Prospirinae (the microsculpture of which is
lirate). His sub-family Choristitinae includes several genera characterized by numerous divided ribs.

Brunton and Rissone (1974), followed by Waterhouse (1981), placed Fusella and Brachythyrina in
the subfamily Strophopleuridae, which they considered as belonging to the family Mucrospiriferidae,
and not Spiriferidae. In my opinion the sinal costation of Angiospirifer is strongly different from the
ornamentation of Mucrospiriferids.

Thus, since Angiospirifer does not fit into existing families, I propose a new sub-family, belonging to
the family Spiriferidae King 1846 sensu Carter 1974.

Subfamily angiospiriferinae nov.

Diagnosis. Transverse shape; ribs not numerous, sometimes dividing; microsculpture sub-imbricated;
adminicula short or lacking; apical callosity and vascular markings more or less developed.

Genera included. Angiospirifer gen. nov.; Brachythyrina Fredericks, 1929; ProchoristiteUa Legrand-Blain, 1968;
Kinghiria Litvinovitch, 1969; IQuizhouspirifer Xian, 1979; ISubspirifer Shan and Zhao, 1981.

Stratigraphic range. Lower Carboniferous (Visean) to lower Permian.

Acknowledgements. The author thanks Dr. C. H. C. Brunton for his advice, supervision of the manuscript, and
loan of British Museum (Natural History) specimens; Dr. P. Brand for information about Scottish localities and
loan of Edinburgh Geological Survey specimens; Drs. W. H. C. Ramsbottom and J. Pattison for demonstrating
the Cayton Gill locality and for the loan of British Geological Survey specimens; Drs. G. A. Cooper and A. S.
Horowitz for the gift of American Anthracospirifer.

REFERENCES

afanas'yeva, g. a. 1975. The genus Unispirifer (Brachiopoda) in the Soviet Union. Paleont. J. 1975 , 412-413.
beznosova, G. a. 1959. Lower Carboniferous Brachiopods of the Kuznetsk basin (families Cyrtospiriferidae and
Spiriferidae). Tr. Pal. Inst. A.N. SSSR. 75 , 1-136. [In Russian.]
brunton, c. H. c. and rissone, a. 1976. Fusella M'Coy 1844, a problematic brachiopod genus from the lower
Carboniferous. Bull. Br. Mus. Nat. Hist. ( Geo! .) 27 , 275-284.
buckman, s. s. 1906. Brachiopod nomenclature. Ann. Mag. Nat. Hist. 58 , 321-327.

burgess, i. C. and holliday, D. w. 1979. Geology of the country around Brough-under-Stainmore. Mem. Geol.
Surv. G.B. sheet 31.

carter, J. L. 1974. New genera of Spiriferid and Brachythyridid brachiopods. J. Paleont. 48 , 674-696.
cowen, R. 1968. A new type of delthyrial cover in the Devonian brachiopod Mucrospirifer. Palaeontology , 11 ,
317-327.

davidson, t. 1858 1863. A monograph of British Carboniferous brachiopods. Palaeont. Soc. [ Monogr.~\ 2 , 5,
1-280.

dunlop, G. m. 1961. Shell development in Spirifer trigonalis from the Carboniferous of Scotland. Palaeontology,
4, 477-506.

LEGRAND-BLAIN: NEW CARBONIFEROUS BRACHIOPOD

575

GEORGE, T. N. 1932. Brachiopoda from the Cayton Gill beds. Trans. Leeds Ass. 5, 37-48.

- JOHNSON, G. A. L., MITCHELL, M., PRENTICE, J. E., RAMSBOTTOM, W. H. C., SEVASTOPULO, G. D. and WILSON, R. B.
1976. A correlation of Dinantian rocks in the British Isles. Geol. Soc. London Spec. Rep. 7, 1-87.
grant, R. E. 1976. Permian brachiopods from Southern Thailand. Mem. Paleontol. Soc. 9, 1-269.
ivanova, E. A. 1960. Order Spiriferida. In orlov, y. a. (ed.). Principles of Palaeontology , (7) Brvozoa and
Brachiopoda , 264-280. Mosk. A.N. SSSR.

1971. Introduction to the study of Spiriferids. Tr. Pal. Inst. A.N. SSSR , 126 , 1-104. [In Russian.]

1972. Main features of Spiriferid evolution (Brachiopoda). Paleont. J. 1972 , 309-320.

lazarev, s. s. and poletaev, v. i. 1982. The development of the brachiopod mantle canal system at the early-
middle Carboniferous boundary. In: ramsbottom, w. h. c. et al. (eds.). Biostratigraphic data for a mid-
Carboniferous boundary, 89-94. Subcommittee on Carboniferous Stratigraphy, Institute of Geological
Sciences, Leeds.

legrand-blain, m. 1968. Spiriferacea carboniferes et permiens d'Afghanistan central. Notes et Mem.
Moyen-Orient , Museum d'Hist. nat. Paris, 9 , 187-253.

— delvolve, j. j. and perret, m. f. 1984. Les brachiopodes carboniferes des Pyrenees Centrales frampaises.
2 — Etude des Orthida et des Spiriferida. Biostratigraphie, Paleoecologie, Paleobiogeographie. Geobios, 17 ,
297-325.

muir-wood, h. m. 1956. Request for the substitution of neotype as the standard of reference for six nominal
species belonging to the Class Articulata (Phylum Brachiopoda), the names published for which by Martin
( W.) in 1 809 have been validated by the International Commission on Zoological Nomenclature, in place of the
figures previously proposed for adoption as such standards. In: Opinion 419. Opin. and Decl. Intern. Comm.
Zool. Nomencl. 14 , 107-119.

semikhatova, s. v. 1941. ‘The group of Spirifer trigonalis Martin’ from the lower Carboniferous supra-
coalbearing beds of Moscow basin. Tr. Pal. Inst. A.N. SSSR, 12 , 1-175. [In Russian.]

Sutherland, p. k. and harlow, f. h. 1973. Pennsylvanian brachiopods and biostratigraphy in Southern
Sangre de Cristo Mountains, New Mexico. Mem. New Mexico Bur. Mines Miner. Res. 27 , 1 173.
thomas, G. a. 1971. Carboniferous and early Permian Brachiopods from Western and Northern Australia.
Bull. Austr. Bur. Min. Res. Geol. Geophys. 56 , 1-276.

vandercammen, A. 1959. Essai d’etude statistique des Cyrtospirifer du Frasnien de la Belgique. Mem. Inst. roy.
Sci. nat. Belg. 145 , 1-175.

— and lambiotte, m. 1962. Observations sur les sarcoglyphes dans Atrypa reticularis (C. Linne, 1767). Bull.
Inst. roy. Sci. nat. Belg. 38, 1-15.

— and plodowski, G. 1967. La question du genre Spirifer s. str. et des genres voisins. Ibid. 43, 1 11.
Waterhouse, j. b. 1968. The classification and description of Permian Spiriferida (Brachiopoda) from New

Zealand. Palaeontographica, A, 129 , 1 94.

— 1970. The lower Carboniferous brachiopod genus Fusella M‘Coy 1844. Occ. Pap. Roy. Ontario Mus. Life
Sci. 15 , 1-12.

— 1981. The Permian Stratigraphy and Palaeontology of Southern Thailand. Mem. Thailand Geol. Surv. 4 ,

I- 213.

williams, a. and rowell, a. j. 1965. Morphological terms applied to Brachiopods. In: moore, r. c. (ed.) Treatise
on Invertebrate Paleontology, Part H, Brachiopoda 1, H139-H155. Geological Society of America and
University of Kansas Press, New York and Lawrence.

wilson, r. b. 1974. A study of the Dinantian marine faunas of South-East Scotland. Bull. Geol. Surv. G.B. 46 ,
35-65.

yanichewsky, m. e. 1935. On some peculiarities of the shells of Spiriferida. Annuaire Soc. Paleont. Russie, 10 ,

II- 27.

zakowa, H. 1958. Biostratigraphy of the lower Carboniferous marine deposits of the area of Walbrzych
Miasto (Lower Silesia). Inst. Geol. Prace, 19 , 1-211.

1966. Zone Goniatites crenistria Phill. in the vicinity of Sokolec and Jugow, at the foot of the Sowie Gory
Mountains (Central Sudetes). Ibid. 43, 1-197.

Typescript received 19 July 1984
Revised typescript received 2 January 1985

marie legrand-blain

Laboratoire de Geodynamique des Bassins Sedimentaires
Universite de Pau et des Pays de l’Adour
Avenue de l’Universite
64000 Pau
France

OSTRACODES ACROSS THE IAPETUS OCEAN

by ROGER E. L. SCHALLREUTER and DAVID J. SIVETER

Abstract. A pilot study comparing ostracode faunas from the North American and European plates dispels the
notion that Ordovician ostracodes show strict endemicity. Problems addressed include: how some ostracodes
managed to cross the Ordovician Iapetus Ocean estimated at 3,000 4- km wide; and why some of their ostracode
counterparts in the Silurian show provinciality? It is concluded that, from the point of view of ostracode
dispersal, opposing Ordovician plates may have been in closer effective geographical proximity than hitherto
supposed. Ostracodes recovered only slowly and with general provincial aspect from global sea level changes at
the Ordovician-Silurian boundary.

Ostracode distributions support other biogeographic evidence that Tornquist’s Sea, the ocean between the
Gondwanan plate (containing southern Britain) and Baltica, had contracted by mid to late Ordovician and that
the Rheic Ocean, separating northern and southern Europe, was developing during the Silurian. Much more
work on Lower Palaeozoic ostracodes is needed to test these findings further.

Most of present day Europe and North America were separated by an ocean during the early
Palaeozoic. Wilson’s (1966) idea that the ‘proto-Atlantic’ had opened and closed again during the
Cambrian to Silurian has been amplified in many accounts which document sedimentological,
structural, and tectonic evidence for what has been termed the Iapetus Ocean (e.g. Dewey 1969;
Phillips et al. 1976; Mitchell 1984). Palaeontologists have independently tested for the existence and
distance apart of the plates by comparing Lower Palaeozoic faunas and floras from both sides of the
presumed ocean, and assessing their relative provincial or cosmopolitan nature (e.g. Spjeldnaes
1978). Faunal evidence relating to continental separation has been elaborated mainly using trilobites
(e.g. Whittington and Hughes 1972; Fortey 1975), brachiopods (Williams 1973, 1976), and
graptolites (Skevington 1978). Some microfossil groups (e.g. conodonts; Fortey and Barnes 1977)
have been similarly employed but ostracodes have been neglected in this respect.

There is no recognized Lower Palaeozoic equivalent of the ostracode fauna of the later
(Tertiary-Recent) deep-sea psychrosphere. Thus, no Lower Palaeozoic (benthic) ostracodes can be
regarded as diagnostically ‘oceanic’. However, ostracode endemicity can be used as a good
independent test for the presence of an ocean recognized from other geological and faunal evidence.
Ordovician ostracodes have previously been assumed to be endemic across the Iapetus Ocean: in a
paper which provided the standard reference model for progressive faunal migration across the
Iapetus Ocean, McKerrow and Cocks (1976, fig. 1) charted ostracodes as the last major animal
group, apart from freshwater fish, to migrate (in the middle Silurian) across the former oceanic
barrier.

The present paper represents a pilot study to assess the possible occurrence of generic or species
level links between pertinent ostracode faunas of the North American and European plates,
particularly during Ordovician times, and also to assess what implications such possible links may
have for palaeogeography and for the palaeozoology of Lower Palaeozoic Ostracoda. Supporting
the model of development of the Iapetus Ocean based on structural and allied evidence (Phillips et al.
1976), distributional patterns of trilobites and brachiopods indicate that the discrete lower
Ordovician faunal provinces, which help define each plate, had by Caradoc-Ashgill times broken
down as the ocean narrowed. Ostracode faunal links between the two plates are herein shown to
occur and increase throughout the Ordovician. It is concluded that, bearing in mind the possible
dispersal capabilities of Ordovician ostracodes, opposing Ordovician plates may have been in closer
effective geographical proximity than hitherto supposed.

[Palaeontology, Vol. 28, Part 3, 1985, pp. 577 598, pis. 68-70.]

578

PALAEONTOLOGY, VOLUME 28

According to recent palaeogeographic reconstructions based on facies and faunal analysis (Cocks
and Fortey 1982, figs. 2, 3; see text-fig. 1 herein), during the Arenig-early Llanvirn a North American
continent straddled the equator, while Baltic and Gondwanan continents were positioned at
southerly, relatively temperate and high latitudes respectively. North America included Britain north
of the Lake District, Spitzbergen, Greenland, and western Newfoundland, but excluded parts of the
present day American Atlantic seaboard. The Baltic continent extended from Scandinavia eastwards
to parts of the Russian platform. Gondwana included southern Britain, Africa, South America,
Iberia, Bohemia, and eastern Newfoundland. Tornquist’s Sea, the proposed ocean (Cocks and
Fortey 1982) separating the Baltic and Gondwanan continents, is thought to have closed during mid
to late Ordovician times, thus creating a latitudinally more continuous southern continental margin
to the Iapetus Ocean. Cocks and Fortey (1982) also document mid to late Silurian faunal evidence
supporting the presence of a widening Rheic Ocean between the latter southern continental area and
Gondwana-southern Europe (including France, Iberia, and Bohemia) further south.

SILURIAN OSTRACODE FAUNAS OF THE NORTH ATLANTIC REGION

Unusually for Silurian invertebrates, a barrier to geographical dispersal of many characteristic
ostracodes apparently still existed along the site of the remnant Iapetus Ocean. Late Silurian
(Ludlow-Pridoli) ostracode faunas of the European plate indicate widespread faunal links,
particularly in the distribution of kloedeniine beyrichiaceans (Martinsson 1963, 1965, 1967, 1970,
1977; Siveter 19786). Combinations of the genera Kloedenia, Londinia , and Frostiella, and
concomitant beyrichiine and amphitoxotidine taxa define a late Silurian faunal region incorporating
parts of Maritime Canada and New England (eastern Maine, southern New Brunswick, eastern
Nova Scotia), southern Britain including the Lake District, Scania, subsurface rocks in the Baltic and
Poland and extending into Podolia, Ukrainian SSR. Corresponding faunas on the North American
plate (Lundin 1971; Berdan 1983) are mostly different, being rare in ostracodes with strong ‘Baltic’
affinities. Thus, a parallel ‘Appalachian belt’ is traceable from westernmost Virginia to Gaspe and the
Silurian faunas of Anticosti Island, Canada (Berdan 1970; Copeland 1977a; Copeland and Berdan
1977). This belt has an essentially endemic beyrichiacean fauna (Copeland 1980) characterized, for
example, by early Silurian zygobolbines and a later Silurian-early Devonian ‘false Kloedenia ’ fauna.
Silurian faunas from central Europe (Czechoslovakia), dominated by non-palaeocopes (see Boucek
1936), are largely different again from those of northern Europe and eastern North America, and
thus could support the notion of a second more southerly barrier (the Rheic Ocean) hindering the
dispersal of invertebrates.

These three ostracode regions are broadly defined, largely on the occurrence and type of
beyrichiaceans present (a distinctive group of shallow-water benthic ostracodes), and are con-
spicuous during the late Silurian. Concurring with the faunal migrational patterns proposed by
McKerrow and Cocks (1976), Cocks and Fortey (1982, p. 474) concluded that in the Silurian, ‘The
assumption, from faunal grounds alone, that there was still a Iapetus Ocean is based on continued
provinciality of the thelodont fish and ostracods’. This seemed to confirm the migrational pattern one
might expect of ostracodes from a knowledge of the life history of extant forms, which lack pelagic
larvae.

The regions inevitably have some ostracode genera in common. For example, Beyrichia,
Craspedobolbina , and Aechmina are also known from the early Silurian sequence of Anticosti
(Copeland 1974a, 1982a). In Britain, where remnants of the North American and European plates
are sutured, the rare ostracode species known from the Silurian of Scotland are endemic, although
some of the genera (e.g. Beyrichia , Craspedobolbina) are also known from the prolific faunas of
southern Britain (Siveter 19786). Further afield, other ‘European’-‘North American’ Silurian
ostracode contacts are more difficult to explain palaeogeographically, such as those few (pelagic?)
podocopid genera from Bohemia and Soviet Asia which are also recorded from the mid to late
Silurian of north-western Canada (Copeland 19776). This part of Canada belongs to the late
Silurian -Devonian Cordilleran ostracode province extending from Nevada to Alaska (Berdan 1983)

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 579

H:\i-mcj. 1. Palaeogeography and chief sedimentary facies of the area surrounding Britain in Arenig

times (from Cocks and Fortey 1982).

580

PALAEONTOLOGY, VOLUME 28

which, although on the opposite side of the North American plate to the western margin of the
Iapetus Ocean, was nevertheless capable of recruiting a few north ‘European’ type beyrichiacean
genera in addition to central European podocopids.

OSTRACODES AS INDICATORS OF ENVIRONMENT

Studies of the life cycle and ecology of modern species (for example Elofson’s (1941) classic paper on
marine forms from Sweden) indicate that benthic ostracodes have no known pelagic larval stage and
that ontogeny normally proceeds within a single biotope. Unaided dispersal across deep oceanic
barriers would logically be a formidable task for benthic ostracodes, which generally live on, or
sometimes burrow into, the substrate. Although some ostracode groups such as the Myodocopida
are pelagic, the overwhelming number of both living and fossil marine forms have, or are thought to
have had, a benthic mode of life. The vast majority of extant marine ostracodes live in the typically
high diversity littoral and shallow marine zones. By contrast, the present day deep-ocean
(psychrospheric) benthic ostracode faunas (those living at depths of about 500-5000 m) are global
and cosmopolitan in distribution but relatively poorly diverse (Benson and Sylvester-Bradley 1971;
Benson 1979). The junction between the psychrosphere and the shallower, warmer ‘thermosphere’
represents a fundamental biological frontier (Hessler and Saunders 1 967)— an environmental barrier
which, if breached, induces amongst ostracodes skeletal and diversity changes in response to the new
stresses imposed (Benson 1975c, 1981).

Ostracodes can be good palaeoenvironmental indicators, particularly in post-Palaeozoic rocks
where closer taxonomic correlation with modern groups facilitates easier ecological interpretation
than for Palaeozoic forms. Ostracodes are opportunistic colonizers, occurring in marine, brackish,
freshwater, and even terrestrial environments. Unlike the plankton, extant benthic ostracodes are
affected by such environmental factors as the nature of the substrate (Elofson 1941). However,
salinity and depth-related factors, such as bottom-water temperatures, exert the most significant
control on their overall distribution. As Benson elegantly demonstrated in his studies of Tethys and
its Messinian salinity crisis (1972, 1973, 1975a, 1976), and of the early Tertiary origin of the
psychrosphere (19756), the development of a sedimentary basin or the evolution of oceanic systems
can be independently charted using ostracode faunal evidence. Ostracodes— often facies con-
trolled— have the capacity to crawl into or out of basins as changing geological and environmental
controls dictate. The history of Lower Palaeozoic ostracodes is presumably controlled in like fashion
and would similarly betray major events such as the relative movements of plates.

OCCURRENCE OF EUROPEAN AND NORTH AMERICAN
ORDOVICIAN OSTRACODES

The occurrence of described Ordovician ostracode faunas is shown in text-figs. 2 and 3. Ostracodes
are known from much of Scandinavia, with important faunas from the lower Ordovician of the Siljan
district (Hessland 1949) and the Oslo region (Henningsmoen 1954a); from the middle Ordovician of
Jamtland and Tvaren (Thorslund 1940), central Sweden (Jaanusson 1957), and the Oslo region
(Henningsmoen 1953); and from the upper Ordovician of Vastergotland (Henningsmoen 1948), the
Oslo region (Henningsmoen 19546), and Scania (Troedsson 1918; Schallreuter 1980). Sarv (e.g. 1959,
1963) has described Estonian Ordovician ostracodes and Schallreuter (1983; full bibliography) has
extensively documented faunas from Baltoscandian erratic boulders. Faunas described from eastern
Europe embrace the middle and upper Ordovician of Latvia and Lithuania (GailTte in Ulster a/. 1982;
Sidaraviciene 1971, 1975) together with records from the Leningrad region (Mannil 1963), Volhyn
(Krandievsky 1975), and Podolia (Krandievsky 1969). In central Europe ostracodes are recorded
from nearly all Ordovician stages in Bohemia, the late Ordovician of Thuringia, and a borehole in
Pomerania (Schmidt 1941; Blumenstengel 1965; Kniipfer 1968; Pribyl 1979).

Ostracodes occur in many Ordovician areas in Britain; Siveter (1978a) has reviewed all known

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 581

text-fig. 2. Occurrence and key papers of the main Ordovician ostracode faunas from Europe: 1, Oslo Region
(Henningsmoen 1953, 1 954c/, 6; Qvale 1980); 2, Jamtiand (Thorslund 1940); 3, Siljan district (Hessland 1949;
Jaanusson 1957); 4, South Bothnian area (Jaanusson 1957); 5, Vastergotland (Henningsmoen 1948; Jaanusson
1957); 6, Ostergotland (Jaanusson 1957); 7, Tvaren area (Thorslund 1940; Jaanusson 1957); 8, Oland (Jaanusson
1957; Schallreuter 1977c); 9, Scania (Troedsson 1918; Schallreuter 1980); 10, Bornholm (Poulson 1978); 11,
Gotland and Baltic erratics in general (e.g. see Schallreuter 1983, and references therein); 12, Gotska Sandon
(Jaanusson 1966); 13, Nyland (Martinsson 1956); 14, Leningrad district (Mannil 1963); 15, Estonia (Sarv 1959,
1963); 16, Latvia (GailTte in Ulst et al. 1982); 17, Lithuania (Sidaraviciene 1971, 1975); 18, Volyn (Krandievsky
1975); 19, Podolia (Krandievsky 1969); 20, Pskov district (Neckaja 1973); 21, Moscow syneclise (Prokofiev and
Kuznetzov 1982); 22, Pecorskian Urals (Zenkova 1977); 23, Central Urals (Persina et al. 1971); 24, Pomerania
(Bednarczyk 1974); 25, Thuringia (Kntipfer 1968); 26, Bohemia (Pribyl 1979); 27, Armoricain Massif (Vannier
1984); 28, Portugal (Vannier 1984); 29, Cantabrian Mountains (Bassler and Kellett 1934); 30, Almaden (Bassler
and Kellett 1934); 31, Sardinia (Bassler and Kellett 1934); 32, Central Urals (Varganov et al. 1970); 33, Great
Britain (Siveter 1978a, 19826, c, 1983; Jones and Siveter 1983; Schallreuter and Siveter 1983a).

faunal associations and revised key taxa. Vannier (1984) has recently provided comprehensive faunal
records from the Armoricain Massif, France, and a comparison with Ordovician ostracodes from
Iberia.

The better known faunas from the North American plate include those from Oklahoma (Harris
1957), Kentucky (Warshauer and Berdan 1982), Virginia (Kraft 1962), Michigan (Kesling 1960;
Kesling et al. 1962), Missouri (Keenan 1951), Iowa and Ontario (Kay 1934, 1940), Minnesota and
Iowa (Swain et al. 1961; Burr and Swain 1965), and various eastern states (e.g. Swain 1957, 1962).
Many faunas, particularly of middle and upper Ordovician age, have been described from Canada by
Copeland (e.g. 1962, 1965, 1966, 1970, 1973, 19746, 1977a, b , c, 1978, 19826, 1983). Unrevised faunas

582

PALAEONTOLOGY, VOLUME 28

text-fig. 3. Occurrence and key papers of the main Ordovician ostracode faunas from United States, Canada,
and Greenland. 1, Greenland (Teichert 1937a, 6); 2, Ellesmere Island (Teichert 1937a, 6); 3-6, Melville Peninsula
and Baffin Island (Copeland 1977c); 7, District of McKenzie (Copeland 19746, 19776, 19826); 8, Lake
Timiskaming and vicinity (Copeland 1965); 9, Ottawa Valley (Copeland in Steele and Sinclair 1971); 10,
Napanee, Ontario (Copeland 1962); 11, Anticosti Island (Copeland 1970, 1973); 12, Newfoundland (Copeland
and Bolton 1977; Whittington and Kindle 1963); 13, Yukon (Copeland 1966, 19776, 1978); 14, 14a,
Pennsylvania and New York (Swain 1957, 1962); 15, Vermont (Creath and Shaw 1966); 16, 17, New Jersey and
Maryland (Bassler and Kellett 1934); 18, Virginia (Kraft 1962); 19, Ohio (Warshauer 1975; Guber 1971); 20,
Michigan (Kesling 1960; Kesling et al. 1962); 21, Indiana (Guber 1971); 22, Illinois (Bassler and Kellett 1934);
23, Wisconsin (Kay 1940); 24, Minnesota (Burr and Swain 1965; Swain et al. 1961); 25, Iowa (Burr and Swain
1965); 26, Missouri (Keenan 1951); 27, Oklahoma (Elarris 1957); 28, Arkansas (Harris 1957); 29, Kentucky
(Warshauer and Berdan 1982); 30, 31, Tennessee and Alabama (Bassler and Kellett 1934); 32, 33, Texas and
Montana (Harris 1957); 34, Wyoming (Guber and Jaanusson 1964; Berdan 1976); 35, 36, Utah and Nevada

(Berdan 1976); 37, Alaska (Copeland 1983).

occur in Greenland (Teichert 19376) and Scotland (see Siveter 1 978a), and ostracodes are also known
to occur in the Arenig of Spitsbergen (Fortey 1975).

AFFINITIES BETWEEN EUROPEAN AND NORTH AMERICAN
ORDOVICIAN OSTRACODES

Ordovician ostracodes have previously been assumed to be endemic across the North Atlantic
region (McKerrow and Cocks 1976; Cocks and Fortey 1982), but the evidence indicates that this is
not the case.

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 583

Rader (1965), Copeland (1977a, p. 5; 1978, p. 97; 1981, p. 185; 1983), Copeland and Berdan (1977,
p. 22), Swain (1977, p. 39), Siveter (1978a), and Schallreuter and Siveter (1982, 19836) indicate some
Ordovician genera common to Europe and North America but no comprehensive comparison of
Ordovician faunas from opposite sides of the Iapetus Ocean has been made, partly because of
inherent difficulties. Most Ordovician faunas are incompletely known and taxonomic revision is
required for several major faunas, especially those described from North America. In general, early
Ordovician ostracode faunas are poorly represented or are less well documented than those of the
middle and late Ordovician. In addition, within the Palaeocopa, homeomorphy, misidentification,
and classificatory difficulties are widespread. Nevertheless, as Ordovician ostracodes are documented
from virtually all North American states and relevant western European countries, a faunal
comparison is possible.

Plates 68-70 illustrate just a few of the Ordovician ostracode genera common to two or more
relevant continents. They embrace a wide variety of both palaeocopes (Pis. 68, 70) and non-
palaeocopes such as Eridostraca (PI. 69, figs. 1 , 2), Metacopa (PI. 69, figs. 3, 4), Leiocopa (Pi. 69, figs.
5, 6), Binodicopa (PI. 69, figs. 7, 8), and Leperditiocopa (PI. 69, figs. 9, 10). Taxonomic revision and
improved documentation (of, for example, British faunas) will undoubtedly produce an extended list
of cosmopolitan genera (and possibly some conspecific taxa) from opposite sides of the Iapetus
Ocean.

Earliest contacts

Ostracode faunas of the North American and European plates reveal elements in common from at
least the later part of early Ordovician times. Eobromidella provides an early link, occurring in the
lower Ordovician of Sweden (Hessland 1949) and the Tulip Creek Formation (early Champlanian),
Simpson Group of Oklahoma (Harris 1957); differences between the two forms appear minimal (PI.
68, figs. 5, 6). This early ostracode contact across the Iapetus Ocean confirms the migrational pattern
defined from palaeogeography: an equatorial North America opposed a southerly, approaching
Baltica, with the Gondwanan continent further to the south beyond a contracting Tornquist Sea.

Other initial links would possibly be recognized if early Ordovician faunas were better known,
particularly from North America. Moreover, compared with the Scandinavian limestone-and-shale
successions in which Ordovician palaeocopes flourished from Arenig to Ashgill, known British
faunas are notably deficient in the Arenig and early Llanvirn. For instance, the oldest known
ostracode faunas of significance from Wales and the Welsh Borderland are from the Llandeilo of
South Wales (Siveter 1978a). In the strongly contrasted facies of the British Ordovician,
representative ostracode assemblages belong mostly to the shelly associations and not to the
alternative basin facies; it seems a valid generalization to group Ordovician ostracodes with trilobites
and brachiopods as shelf-sea benthos. It may be significant, therefore, that British ostracode faunas
largely come from post-Llanvirn sequences when the contrast between shelly and graptolitic facies
was strongest, and lithologies similar to ‘Llandeilo Flags’ prevailed.

Middle Ordovician

Most of the genera with supposed European affinities mentioned by Swain (1962, 1977) we consider
misidentified. Nevertheless, in the middle Ordovician clearly increased affinities exist between
opposing Iapetus Ocean ostracode faunas, presumably reflecting a closer proximity of plates than in
earlier times. Many middle Ordovician ostracode genera are common to both North American and
European plates; there is ample evidence also of contemporaneous ostracode contact between Baltica
and part of the old Gondwanan continent, as represented by southern Britain. The notion that
Gondwana moved north, so that Tornquist’s Sea had virtually ceased to exist by mid to late
Ordovician (Cocks and Fortey 1982), is corroborated in as much as any remnant of this ocean
evidently failed to provide an effective barrier to ostracode dispersal. Whittington and Hughes (1972)
and Williams (1973, 1976) showed that, by the late Caradoc, certain trilobite and brachiopod faunas
of southern Britain and Scandinavia respectively were very similar; generic links between the
ostracode faunas of the two areas, and between Britain and North America, can be traced somewhat

584

PALAEONTOLOGY, VOLUME 28

earlier to the Llandeilo Series of Wales. Additionally, the presence in the Caradoc and possibly the
Ashgill of southern Britain (Siveter 1978a, pi. 3, fig. 5 and p. 46 respectively) of the Bohemian genus
Crescentilla suggests that, unless the ostracode was pelagic, the postulated developing Rheic Ocean
was not yet a formidable obstacle to migration. Obviously many middle Ordovician ostracode genera
are endemic to one particular plate, but the fact that ostracode faunal links between the plates were
firmly established by mid Ordovician is also clear and requires explanation.

Homeokiesowia , Vittella , Klimphores, Pseudulrichia , and Tallinnellina are amongst many ostracode
genera occurring in the Llandeilo Series of South Wales and the middle Ordovician of Baltoscandia
(Siveter 1978a, 19826, 1983; Schallreuter and Siveter 1983a). For example, the Llandeilo material
includes a form close to Pseudulrichia bucera , a species recorded from middle Ordovician
Backsteinkalk boulders of North Germany and Gotland and from the higher part of the Idavere
Stage (C 3 ) ( = early Caradoc) in the Pskov district, Russia (Siveter 1978a). Cryptophyllus (PI. 69, figs.
1, 2) is one of the rare ostracodes to show moult retention and occurs as far afield as the Llandeilo
Series of South Wales (Siveter 1978a), the Ojlemyrflint erratics of the Baltic (Schallreuter 1977a), the
early to mid Ordovician of Oklahoma (Flarris 1957), and the Whiterockian Sunblood Formation of
north-western Canada (Copeland 1978). Ceratopsis is equally as distinctive morphologically and
characteristic of mid to late Ordovician sequences in the United States (e.g. Kentucky: Warshauer
and Berdan 1982; Ohio: Warshauer 1975) and north-western and eastern Canada (Copeland 19746,
1977a); it is represented by C. britannica , C. duftonensis , and undescribed species variously from the

EXPLANATION OF PLATE 68

Palaeocopes from the lower and middle Ordovician of both sides of the Iapetus Ocean {left North
America, right Europe).

Fig. 1. Platybolbina (Rimabolbina) omphalota Kesling, 1960, middle Ordovician (Bony Falls Limestone,
Blackriverian), Bony Falls on the Escanaba River, Deltay County, Michigan, U.S.A.; right valve
(holotype, UMMP 37355), x 34 (from Kesling 1960, pi. 8, fig. 4 left).

Fig. 2. P. (Rimabolbina) rima Schallreuter, 1964, middle Ordovician (Skagen Formation, Upper Viruan),
Backsteinkalk erratic boulder (no. Stal, 1B1 type), Staberhuk, Isle of Fehmarn (Baltic Sea), Germany;
right valve (GPIMH 2717), x 47.

Fig. 3. Hesperidella michiganensis Kesling, Hall and Melik, 1962, same bed and locality as fig. 1;
immature female right valve (paratype. UMMP 37225), x 87 (from Kesling et al. 1962, pi. 2, fig. 2 left).

Fig. 4. H. esthonica (Bonnema, 1909), same boulder as fig. 2; tecnomorphic right valve (GPIMH 2718),
x 48.

Fig. 5. Laccochilina [ Eobromidella ] eurychilinoides (Harris, 1957), middle Ordovician (Tulip Creek horizon,
Blackriverian), Sycamore Creek Simpson section, Oklahoma, U.S.A.; left valve (holotype, Museum of
Comparative Zoology, Harvard University, Boston, no. 4631), x 29 (from Harris 1957, pi. 8, fig. la).

Fig. 6. L. dorsoplicata Hessland, 1949, lower Ordovician (Upper Oelandian, Kundan), Silverberg II,
Eastern Siljan District, Dalecarlia, Sweden; right valve (holotype. Museum of the Palaeontological
Institute, University of Uppsala, no. ar.os.398), x42 (from Hessland 1949, pi. 6, fig. 6).

Fig. 7. Hithis colonus Schallreuter and Siveter, 1982, middle Ordovician (lower part of Edinburg
Formation), section in field on south side of road, 0-2 km south-east of Strasburg Junction, just west of
Strasburg, Shenandoah County, Virginia, U.S.A.; female right valve (paratype, GPIMH 2675; valve
tilted 10°), x45.

Fig. 8. H. hithis Schallreuter, 1964, middle Ordovician (Skagen Formation, Upper Viruan), Back-
steinkalk erratic boulder (no. 1 B 1 6, IB1 type), beach at Dornbusch, Isle of Hiddensee (Baltic Sea),
Germany; female right valve (holotype, SGWG 3/3; Schallreuter 1964b, pi. 12, fig. 2), x 80.

Fig. 9. Hippula (Hippida) varicata (Harris, 1957), middle Ordovician (Lower Esbataottine Formation,
Chazyan), Sunblood Mountain, southwestern District of Mackenzie, Canada; female right valve (GSC
49378), x 67 (from Copeland 19826, pi. 1, fig. 2).

Fig. 10. H. ( H .) latonoda (Schallreuter, 1964), middle Ordovician (Skagen Formation, Upper Viruan),
Backsteinkalk erratic boulder (no. 1B4, 1B1 type), beach at Dornbusch, Isle of Hiddensee (Baltic Sea),
Germany; female right valve (holotype, SGWG 2/3; Schallreuter 1964a, pi. 10, fig. 1), x 60.

PLATE 68

SCHALLREUTER and SIVETER, palaeocopes from the lower and middle Ordovician of
North America ( left ) and Europe (right )

586

PALAEONTOLOGY, VOLUME 28

Llandeilo, Caradoc, and Ashgill of South Wales, the Welsh Borders, and Cross Fell in southern
Britain (Siveter 1978a). Hithis (PI. 68, figs. 7, 8) is present in the mid and late Ordovician of
Baltoscandia and the Edinburg Formation of Virginia (Schallreuter and Siveter 1982), while
Hippula (PI. 68, figs. 9, 10) occurs in the mid Ordovician of Baltoscandia, Thuringia, Bohemia,
Siberia, and Oklahoma (Schallreuter and Kruta 1980), and in the District of Mackenzie, Canada
(= Oecematobolbina of Copeland 19826).

Later Ordovician

Many ostracode genera from the late Ordovician of the North American continent are typical of the
fairly homogeneous coeval faunas of the Baltic regions (see PI. 70) and also occur to a varying extent
in Scandinavian and southern British faunas. On Anticosti Island alone, the late Ashgill Ellis Bay
and underlying Vaureal formations (Copeland 1970, 1973) have the following genera in common
with the Baltic: Moeckowia , Anticostiella , Caprabolbina, Platybolbina, Tetradel/a, Eoaquapulex ,
Pseudulrichia , Antiaechmina, Cryptophyllus, Eographiodactylus, Hemeaschmidtella, Triangu/o-
schmidtella , Warthinia , Byrsolopsina , Monotiopleura, Brevibolbina , Pseudohippula (PI. 70, figs. 3, 4),
Rectella , and Brevidorsa. Distobolbina warthini Copeland, 1977, from the late Mohawkian of Baffin
Island, Canada, and D. grekoffi Schallreuter, 1977, from upper Ordovician Baltic erratics, may be
conspecific (PI. 70, figs. 1,2). Furthermore, Copeland (1983) has recently reported the typical circum-
Baltic species Steusloffina cuneata from the latest Ordovician/earliest Silurian of Anticosti Island.
The ostracode fauna of the Ashgill Brachiopodskiffer of Scania (Troedsson 1918) includes the
characteristic American genus Quadrijugator , whilst that of the late Ordovician Maquoketa Shale of
Missouri (Keenan 1951) contains many genera also found in the Baltic, exemplified inter alia by the
species Eoaquapulex barbatus, Byrsolopsina irregularis , Deefgella ? septinoda, Spinaechmina taura ,
and Antiaechmina maquoketensis.

EXPLANATION OF PLATE 69

Non-palaeocopes from the middle Ordovician of both sides of the Iapetus Ocean {left North America,

right Europe).

Figs. 1, 2. Eridostraca. 1, Cryptophyllus niagnus (Elarris, 1931), middle Ordovician (Sunblood Forma-
tion, Whiterockian, Chazyan), Esbataottine Mountain, southwestern District of Mackenzie, Canada;
left valve (GSC 38417), x 36 (from Copeland 1978, pi. 1, fig. 7). 2, Cryptophyllus ? sp. of Siveter 1978,
middle Ordovician (‘Llandeilo Flags'), old quarry 300 m north of Big House, Lampeter Velfrey, east of
Narberth Dyfed, Wales; right valve (BM(NH) no. OS 6676), x 76 (from Siveter 1978a, pi. 1, fig. 1).

Figs. 3, 4. Metacopa. 3, Balticella deckeri (Harris, 1931), middle Ordovician (lower part of Edinburg
Formation, Blackriverian), Strasburg Junction, Virginia, U.S.A.; left valve (GPIMH 2719), x45. 4,

B. binodis (Krause, 1897) (= B. oblonga Thorslund, 1940), middle Ordovician [Idavere- (C 3 ) or Johvi-
Stage (Dj), Upper Viruan], Backsteinkalk erratic boulder (type and no. 14B2), Teschenhagen near
Stralsund, Pomerania, Germany; juvenile left valve (SGWG 28/9; Schallreuter 1968, fig. 10.4), x 64.

Figs. 5, 6. Leiocopa. 5, Brevidorsa fimbriata (Ulrich, 1892), from same bed and locality as fig. 3; left valve
(GPIMH 2720), x 32. 6. B. crassispinosa (Schallreuter, 1973), from same bed and locality as Plate 68,
fig. 2; left valve (GPIMH 2721), x 67.

Figs. 7, 8. Binodicopa. 7, Pedomphalella intermedia Swain and Cornell in Swain et al. 1961, middle
Ordovician (Decorah Shale), 3 miles east of Rochester, Minnesota, U.S.A.; imperfect left valve
(paratype, Univ. Minnesota), x 115 (from Swain et al. 1961, pi. 48, fig. 7a). 8, P. jonesii (Krause, 1897)
(= Schmidtella egregia Sarv, 1963), same boulder as fig. 4; left valve (GPIMH 2722), x 99.

Figs. 9, 10. Leperditiocopa. 9, Bivia duncanae Berdan, 1976, middle Ordovician (Llanvirn-Llandeilo;
upper part of Kanosh Shale, Pogonip Group, Chazyan), Crystal Peak section. Ibex Area, Millard
County, Western Utah, U.S.A.; right valve (paratype, USNM 235540), x 8 (from Berdan 1976, pi. 5,
fig. 4). 10. BP. ordoviciana (Kummerow, 1924), middle Ordovician light grey limestone erratic boulder

corresponding in age to the Echinospharitenkalk (Lower Viruan), Voigtsdorf, Mecklenburg, Germany;
right valve (lectotype. Museum fur Naturkunde der Humboldt-Universitat Berlin, MB. 0.65), x4 (from
Kummerow 1924, pi. 20 [numbered 21], fig. 1).

PLATE 69

SCHALLREUTER and SIVETER, non-palaeocopes from the middle Ordovician of
North America (left) and Europe (right)

588

PALAEONTOLOGY, VOLUME 28

By contrast the faunas of the Maquoketa Shale in Iowa (Burr and Swain 1965) bear more
resemblance to those from Scania. Ostracoda in common with the upper Viru (Caradoc) Sularp Shale
fauna (Schallreuter 1980) include Pariconchoprimitia , Vogdesella , Orechina, Conchoprimitiella, and
Klimphores ; the last two genera also occur in the mid to late Ordovician of southern Britain.

While the appearance of ostracode genera in common is often broadly contemporaneous between
continents, the origins of many ostracode groups apparently lie on the North American plate with
subsequent migration to ‘Europe’. The Leperditiocopa (e.g. PI. 69, figs. 9, 1 0), for example, occurs in
North America throughout the Ordovician whereas in Europe only four species are known, including
one from the middle Ordovician of Norway. Other groups, such as the Lomatopisthiidae, the
Nodambichilinae, and genera such as Eoaquapulex (PI. 70, figs. 9, 10), Tetradella (PI. 70, figs. 5, 6),
and Eographiodactylus (PI. 70, figs. 7, 8) occur in North America by the middle Ordovician but not in
Baltoscandia until the end of the Ordovician.

In summary, ostracodes display progressively increasing faunal connections between the North
American and Baltica/Gondwanan continents throughout the Ordovician. The earliest known
connections occur in the late early Ordovician; there are many genera in common in the middle
Ordovician; and by the late Ordovician the ostracode faunas of the two areas show even greater
similarity, including possibly conspecific material. There are correspondingly increasing ostracode
links between the Gondwanan and Baltic continents as the intervening ocean waned. The pattern of
ostracode distribution across the Iapetus Ocean seems to mirror that shown by Ordovician
brachiopods and trilobites, with progressively increasing contacts associated with the approaching
continents.

EXPLANATION OF PLATE 70

Palaeocopes from the middle and upper Ordovician of both sides of the Iapetus Ocean ( left North America, right
Europe).

Fig. 1. Distobolbina warthini Copeland, 1977, middle Ordovician (late Mohawkian), Silliman’s Fossil
Mount, southern Baffin Island, Canada; female left valve (paratype, USNM 216133), x 82 (from
Copeland 1977c, pi. 6, fig. 9).

Fig. 2. D. grekoffi Schallreuter, 1977, upper Ordovician (Upper Harjuan), Ojlemyrflint erratic boulder
(no. G8), beach at Gnisvards, Isle of Gotland (Baltic Sea), Sweden; female left valve (holotype, GP1MH
1926; Schallreuter 19776, pi. 4.26, fig. 1 a), x 63.

Fig. 3. Pseudohippula castorensis (Copeland, 1970), upper Ordovician (Vaureal Formation), Anticosti
Island, Quebec, Canada; posteriorly incomplete right valve (paratype, GSC 24004) with spines broken
away, x 68 (from Copeland 1970, pi. 4, fig. 21).

Fig. 4. P. pseudopokornina Schallreuter, 1975, upper Ordovician (Upper Harjuan), Ojlemyrflint erratic
boulder (no. Wie2), Wielen, west of Uelsen, Emsland, Niedersachsen, Germany; right valve (GPIMH
2723), x 97.

Fig. 5. Tetradella scotti Guber, 1971, upper Ordovician (Elkhorn Formation), Fairhaven, Preble County,
Ohio, U.S.A.; female left valve (GPIMH 2724, valve tilted 20°), x 70.

Fig. 6. T. separata Sidaraviciene, 1971, upper Ordovician (Upper Harjuan), Ojlemyrflint erratic boulder
(no. G16), beach north of Lickershamn, Isle of Gotland (Baltic Sea), Sweden; female left valve (GPIMH
1992, valve tilted 20°), x 59.

Fig. 7. Eographiodactylus eos Kraft, 1962, middle Ordovician (Edinburg Formation, Blackriverian),
Tumbling Run section, Virginia, U.S.A.; left valve (paratype, USNM 136637), x 71 (from Kraft 1962,
pi. 16, fig. 9).

Fig. 8. E. sulcatus Schallreuter, 1975, same boulder as fig. 4; left valve (GPIMH 2725), x 100.

Fig. 9. Eoaquapulex sp. (= Oepikella frequens of Kraft, 1962), same bed and locality as Plate 69, fig. 3;
female right valve (GPIMH 2726), x 30.

Fig. 10. E. frequens (SteuslotT, 1895), upper Ordovician (Upper Harjuan), Leptaenakalk erratic boulder,
Neu-Brandenburg, Mecklenburg, Germany; female right valve (lectotype, SGWG 1 14/37), x 27.

PLATE 70

SCHALLREUTER and SIVETER, palaeocopes from the middle and upper Ordovician of
North America (left) and Europe (right)

590

PALAEONTOLOGY, VOLUME 28

SIGNIFICANCE OF THE PATTERNS OF OSTRACODE DISTRIBUTIONS

How did many ostracode genera manage to cross the Iapetus Ocean in the Ordovician and why were
some corresponding ostracodes still provincial during the Silurian?

Width of the Iapetus Ocean

Estimated widths of the Iapetus Ocean pose problems in explaining the pattern of ostracode
migration. Based on structural and stratigraphic palinspastic restoration of Iapetus, Williams (1980)
estimated a minimum width of 2000 km at its maximum development. He concluded (1980, p. 435)
that it was unlikely that such a geographically continuous orogen (10,000 km) ‘which included
equally continuous tectonic-stratigraphic facies belts related to ancient continental margins and an
ocean basin, could have been produced by the opening of a narrow rift, a marginal ocean, or even a
narrow major ocean’. Analysis of structural telescoping and palinspastic restoration of coeval
sedimentary facies across the western margin of Iapetus implies a continental slope/rise of at least 200
km width; modern continental margins of comparable width and length border major oceans
(Williams 1980, p. 421).

Cocks and Fortey (1982, p. 474) noted that ostracodes would have been unable to cross an ocean
barrier, even one quite narrow. Cocks et al. (1980) estimated an ocean 900 km wide by the early
Silurian. By analogy with modern subduction rates and with the survival periods and dispersal rates
of modern pelagic larvae, McKerrow and Cocks (1976) suggested a minimal 2000-3000 km width,
even during late Ashgill, with the first closure occurring in the north-east by the Silurian collision
of Greenland and the Baltic. They believed that a mid Silurian trans-Iapetus migration of benthic
ostracodes demonstrated closure of the ocean along part of its length.

Pelagic larvae?

One can speculate that ostracodes crossed the Ordovician Iapetus Ocean because, unlike modern
forms, they had pelagic larvae. This is improbable, a conclusion supported by the fact that many
Ordovician ostracodes apparently brooded their young (dimorphism believed to be associated with
egg/brood care dominates palaeocope morphology) and thus would be unlikely to also have had
widely dispersed larvae.

Benthic migration?

Fortey (1975) and Fortey and Owens (1978) have demonstrated shelf to slope related Iapetus trilobite
communities showing increasing cosmopolitanism with depth. Could the trans-Iapetus ostracodes
possibly represent deeper water assemblages, or did they perhaps migrate stepwise down shelf and
across the ocean? Alternatively, is the trans-Iapetus ostracode mixing due in part to transgressive
events in the manner recently modelled and persuasively tested by Fortey (1984) when assessing
biological effects of Ordovician eustatic events? The biological implications of a transgressive pulse
include not only generation of endemic taxa on flooded cratonic areas but also migration shelf-wards
of previously extra-cratonic, deep-water faunas, thereby giving a false impression of provincial
breakdown in response to tectonic events (Fortey 1984, p. 39). Thus, the pronounced Llandeilo-
Caradoc transgression augments tectonic explanations for the late Ordovician trilobite mixing across
Iapetus (Fortey 1984, pp. 46-47, fig. 4).

That some ostracode faunal mixing may be explained either by trans-oceanic migration or by
cratonic migration during transgressions undoubtedly justifies detailed investigation, but this lies
beyond the scope of the present study. As noted above, however, depth barriers are generally
formidable for ostracodes and, even though community and detailed facies related analyses of
Ordovician ostracode faunas have not yet been attempted, in broad terms they belong to the
relatively shallow-water shelf/platform environment and not to the outer slope (e.g. Copeland 19826).
The existence of a Tower Palaeozoic pyschrospheric (deep-sea) ostracode fauna is not documented,
but a controlling inflow of the cold polar waters necessary to create a more universal pyschrosphere
(Benson 1979) within a two-layer ocean model (Bruun 1957), similar to that determining global

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 591

distribution of extant oceanic ostracodes, was apparently a possibility during the Ordovician. Late
Ordovician glaciation is amply attested and Fortey (1984, p. 48, fig. 10) has suggested that a possible
early Ordovician South Polar ice sheet was twice as large as Antarctica today. We need to search the
right kind of deposits for possible Ordovician pyschrospheric ostracodes; the oceanic deposits of the
Ordovician of the Southern Uplands of Scotland (Leggett 1979) are an example, but from which
ostracodes are so far unknown.

Chance dispersal?

Dispersal of larvae (or even eggs or adults) by the passive transport of ocean currents is another
possibility, as with many other marine invertebrates (Scheltema 1977). The ostracode dispersal
patterns noted above suggest that, if current action was a factor, the significant current direction may
have been from the North American to the European plate. Rafting on ‘floats’ (of, for example,
marine algae) is yet another viable method of dispersal of modern marine animals, but is less effective
(Scheltema 1977) because most benthic invertebrates are not well-adapted to survive long periods as
epiplankton in the open sea.

The eggs of freshwater ostracodes are quite hardy; they are able to tolerate cold (Sohn and
Kornicker 1979) and, in the case of some (though not all: McKenzie and Hussainy 1968), to
withstand decades of desiccation (Van Morkhoven 1962, p. 1 39). Freshwater ostracode eggs can also
possibly survive transport by high altitude winds (Sohn and Kornicker 1979) and have been found
viable after ingestion and defecation by aquatic animals such as fish (Kornicker and Sohn 1971 ). The
same has not been demonstrated for marine ostracodes, but they themselves form part of the diet of
marine fish (Kornicker and Sohn 1971) and their eggs, when laid (outside the carapace), are usually
deposited as clusters on bottom sediment or vegetation— an obvious potential food for many marine
invertebrates.

Even though their dispersal mechanism is not in many cases understood, some modern ostracode
faunas (from such island sites as Hawaii, for example) combine genera from distant regions separated
by deep and in some cases almost ‘barren’ ocean floors (R. H. Benson, pers. comm.). It is thus
possible that Ordovician ostracode dispersal was the result of nothing more than chance rafting or
passive transport by other animals. Against this argument, however, is the fact that the ostracode
migrational pattern copies that of trilobites and brachiopods, indicating that it reflects something
more tangible than chance. Moreover, as noted above, there is reason to believe that in palaeocope
ostracodes the well-being of possibly eggs and young depends to some degree on parental care.
Potential Ordovician conveyors would also have been limited to cephalopods and possibly early fish,
whereas modern ostracodes have many other passive dispersal mechanisms, including carriage by
birds (e.g. De Deckker 1977) and man (Van Morkhoven 1962, p. 139).

Pelagic ostracodes?

Another solution is to reinterpret the palaeoecology of palaeocopes, but their shape, centre of gravity,
and ventral and adventral structures favour the notion that most were benthic crawlers (or maybe
bottom swimmers) and not pelagic; swimmers possibly included, for example, those forms with
ventrally open (unprotected) antra, like Foramenella (Henningsmoen 1965). Pelagic ostracodes are
characteristically planktonic forms and free-swimmers belonging to the Order Myodocopida; they
are typically poorly calcified with a limited potential for preservation and have a scant fossil record.
The origins of a free-swimming mode of life in Ostracoda is an untackled question but we know of no
convincing pelagic myodocopid earlier than the Silurian. The bolbozoids and associates, known from
basinal facies of the Ludlow Series of the Welsh Borderland and from coeval horizons of France,
Sardinia, and Czechoslovakia, are possibly amongst the pioneer ostracodes adapted to a life off the
bottom (Siveter 1984).

Closer geographical proximity?

Unless Ordovician ostracodes had some elusive mechanism of dispersal, their crossing of the Iapetus
Ocean may have been aided by geographical proximity of opposing plates, which were possibly much

592

PALAEONTOLOGY, VOLUME 28

closer together and the ocean shallower at some point along the 1 0 000 km length of the Appalachian-
Caledonides tract than estimates and palaeogeographical reconstructions have indicated. Bearing in
mind the contacts demonstrated above (for example, by the ostracode fauna of South Wales as early
as the early Llandeilo), the separation distance of 60° latitude between Britain and North America
proposed by Cocks and Fortey (1982) (text-fig. 1 herein) appears to be an overestimate. This
argument supports the tectonic model of Phillips et al. ( 1976) of oblique collision as early as the late
Ordovician. Likewise the occurrence of only major-sized plates is possibly too simplistic; geo-
graphically intermediate ‘islands’ (microplates?) (e.g. see Neuman 1972) would obviously have
provided easier pathways for the long distance oceanic migration of ostracodes. A number of extra-
cratonic island faunas occur, for example, in the Llandeilo of the mobile belt of Newfoundland, in
sites which concur with both regressive conditions and models, and which might engender speciation
events imprinted on later transgressions (Fortey 1984, pp. 40, 45, 47-48). The Llandeilo marks
approximately the initial, extensive mixing of trans-lapetus ostracodes, though no ostracode faunas
are as yet documented from supposed island sites.

Effects of late Ordovician eustatic changes

Animals on both the American and European plates, as typified by benthic brachiopods and planktic
graptolites, suffered a marked decline during the latest Ordovician— a time of eustatic regression
associated with the Gondwanan glaciation centred on North Africa (McKerrow 1979; Leggett et al.
1981 ). Ostracodes are no exception: the late Ordovician in eastern North America marks a period of
extinction of, for example, many hollinomorphs, followed by the re-establishment of new ostracode
faunas (Copeland 1977a); similar changes can be recognized in Europe (e.g. Siveter 1982a). In one of
the best known and conformable Ordovician-Silurian ostracode successions, that from Anticosti
Island, Quebec, about 15 m of strata across the boundary are barren of diagnostic ostracode faunas
(Copeland 1970, 1973, 1981). In the Welsh basin of Britain, latest Ordovician and particularly earliest
Silurian (early Llandovery) ostracodes are notably few. New Silurian stock in many of these regions
is characterized by the occurrence of pioneer representatives of beyrichiacean palaeocopes. The faunal
differences between Silurian beyrichiacean ostracodes across the Iapetus Ocean (see above) resulted
possibly from the Llandovery, post-glacial transgression which isolated the shelf-living ostracodes
into two stocks (Copeland 1977a; Copeland and Berdan 1977). Fortey (1984) has indeed predicted
that one of the biological implications of a transgressive episode is the generation of high diversity
and endemism on separated cratonic areas. The recovery of the shelf benthos was apparently slow
(Leggett et al. 1981) but, ecological conditions being generally similar on both sides of the remnant
Iapetus Ocean, this does not explain why some ostracodes display an endemicity which is particularly
marked during the Ludlovian, and at the level of the Ludlow-Pridoli transition (and which is atypical
of Silurian animals in general; Holland 1971).

CONCLUSIONS

1. Evidence from ostracode distributions can be used to test independently other faunal and
geological data which indicate the site of oceans. As no psychrospheric-type ostracodes are yet
documented from the Lower Palaeozoic, ostracodes alone cannot be used to diagnose the presence of
oceans.

2. A pilot study comparing ostracode faunas from the North American and European plates
dispels the notion of McKerrow and Cocks (1976) that Ordovician ostracodes show strict
endemicity. Many middle and late Ordovician ostracode genera are common to both sides of an
ocean which is recognized from the weight of geological and other faunal evidence.

3. In order for ostracodes to migrate across a mid Ordovician Iapetus Ocean which is estimated,
from independent geological evidence, to have been at least several thousands of kilometres wide, it is
concluded that opposing plates were at some point possibly in much closer proximity, or the ocean
shallower, than hitherto supposed. Alternatively, the existence of intermediately placed islands or
microplates may have aided animal dispersal.

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 593

4. The provincial aspect of some Silurian ostracodes may be related back to global changes in sea
level at the Ordovician-Silurian boundary.

5. Ostracode distributions support other biogeographic evidence for the contraction of Torn-
quist’s Sea by mid to late Ordovician, and for the development of the Rheic Ocean during the
Silurian.

6. The use of ostracodes to help solve Lower Palaeozoic biogeographic problems is at a rather
crude stage compared with equivalent trilobite or brachiopod studies. The data base is nowhere as
comprehensive and there has been no sophisticated analyses of, for example, facies and any asso-
ciated ostracode ‘communities’. A few specialists are fully engaged documenting the ostracode
faunas. Much remains to be done, for example, on the taxonomy of North American Ordovician
ostracodes. Detailed studies on the autecology of palaeocope species also need to be undertaken to
elucidate modes of life. Nevertheless, previous assumptions concerning the timing of ostracode
contacts across the Iapetus Ocean must now change and their implications be reappraised. The more
we examine the record, the more contacts emerge, even during the Silurian. Endemism is often a
reflection of the current state of knowledge.

Plates 68-70. The plates are designed to demonstrate congeneric ostracode occurrence across the Iapetus Ocean
in the middle and upper Ordovician, using examples from a variety of major taxa.The figures include scanning
electron micrographs of our own material and pertinent illustrations after previous authors. Restrictions on the
availability of several relevant specimens have prevented their illustration by scanning electron microscopy.
Abbreviations used in plate explanations: GPIMH, Geologisch-Palaontologisches Institut und Museum der
Universitat Hamburg, Federal Republic of Germany; SGWG, Sektion Geologische Wissenschaften der
Universitat Greifswald, German Democratic Republic; USNM, United States National Museum, Washington,
D.C., U.S.A.; UMMP, University of Michigan Museum of Paleontology, Ann Arbor, Michigan, U.S.A.; GSC,
Geological Survey of Canada, Ottawa, Canada.

Acknowledgements. We thank Dr. R. A. Fortey and Professor R. F. Lundin for kindly suggesting improvements
to the original manuscript, and Drs. R. H. Benson and J. M. Berdan for their helpful comments. D. J. S.
gratefully acknowledges support for this work from N.E.R.C., the Royal Society, the Research Board of the
University of Leicester, and the Deutsche Forschungsgemeinschaft in combination with the Geologisch-
Palaontologisches Institut, Universitat Hamburg. Text-fig. 1 is reproduced by kind permission of the Geological
Society of London and Drs. L. R. M. Cocks and R. A. Fortey.

REFERENCES

bassler, r. s. and kellett, B. 1934. Bibliographic index of Paleozoic Ostracoda. Spec. Pap. geol. Soc. Am. 1,
xiii + 500 pp.

bednarczyk, w. 1974. The Ordovician in the Koszalin-Chojnice region (Western Pomerania). Acta geol. pol.
24, 581-600, 2 pis.

benson, r. h. 1972. Ostracodes as indicators of threshold depth in the Mediterranean during the Pliocene. In
Stanley, D. J. (ed. ). The Mediterranean Sea , 63 73. Dowden, Hutchinson & Ross, Stroudsburg, Pennsylvania.

1973. An ostracodal view of the Messinian salinity crisis. In drooger, c. w. (ed.). Messinian events in the
Mediterranean , 235-242. North Holland Publ. Co., Amsterdam.

1975a. Ostracodes and Neogene history. Micropaleont. spec. Pubis , 1, 41-48.

1975fi. The origin of the psychrosphere as recorded in changes of deep-sea ostracode assemblages. Lethaia ,
8, 69-83.

1975c. Morphologic stability in Ostracoda. Bull. Am. Pa/eont. 65, 13-46.

— 1976. Changes in the ostracodes of the Mediterranean with the Messinian salinity crisis. Palaeogeogr.
Palaeoclimat. Palaeoecol. 20, 147-170.

— 1979. In search of lost oceans; a paradox in discovery. In gray, j. and boucot, a. j. (eds.). Historical
biogeography , plate tectonics and the changing environment , 379-389. Oregon State University Press.

— 1981. Form, function and architecture of ostracode shells. A. Rev. Earth planet. Sci. 9, 59-80.

— and Sylvester- bradley, p. c. 1971. Deep-sea ostracodes and the transformation of Ocean to Sea in the
Tethys. In oertli, h. j. (ed.). Paleoecologie Ostracodes. Bud. Cent. Rech. Pau , 5, 63-91.

594

PALAEONTOLOGY, VOLUME 28

berdan, j. m. 1970. American ostracode zonation. In berry, w. b. n. and boucot, a. j. (eds.). Correlation of the
North American Silurian rocks. Spec. Pap. geol. Soc. Am. 102, 39-40.

1976. Middle Ordovician leperditicopid ostracodes from the Ibex area, Millard County, western Utah.
Geology Stud. Brigham Young Univ. 23 (3), 37-65, 9 pis.

1983. Biostratigraphy of Upper Silurian and Lower Devonian ostracodes in the United States. In
maddocks, R. f. (ed.). Applications of Ostracoda, 313-337. University of Houston, Texas.
blumenstengel, h. 1965. Zur Ostracodenfauna eines Kalkgerolls aus dem Thiiringer Lederschiefer (Ordo-
vizium). Freiberger ForschHft. C182, 63-78, 2 pis.

bonnema, j. h. 1909. Beitrag zur Kenntnis der Ostrakoden der Kuckerschen Schicht (C 2 ). Mitt, miner. -geol. Inst.
Reichsuniv. Groningen , 2, 1-84, pis. 1-8.

bouLek, b. 1936. Die Ostracoden des bohmischen Ludlows (Stufe e jS). Neues Jb. Miner. 76, 31-98, pis. 2-6.
bruun, a. 1957. Deep sea and abyssal depths. In hedgpeth, j. w. (ed.). Treatise on marine ecology and
paleoecology , 1, 641-672. Geological Society of America.
burr, j. h. and swain, f. m. 1965. Ostracoda of the Dubuque and Maquoketa formations of Minnesota and
northern Iowa. Spec. Pubis Minn. geol. Surv. 3, 40 pp., 6 pis.
cocks, l. r. m. and fortey, r. a. 1982. Faunal evidence for oceanic separations in the Palaeozoic of Britain. J.
geol. Soc. 139, 465-478.

— mckerrow, w. s. and i.eggett, j. k. 1980. Silurian palaeogeography on the margins of the Iapetus Ocean in
the British Isles. Mem. Dept. geol. Sci. Virginia Polyt. Inst. 2, 49-55.

copeland, m. j. 1962. Two new Ordovician Ostracoda from southern Ontario. In Canadian fossil Ostracoda,
Conchostraca, Eurypterida and Phyllocarida. Bull. geol. Surv. Can. 91, 55-57, pi. 12.

— 1965. Ordovician Ostracoda from Lake Timiskaming, Ontario. Ibid. 127, 1-52, 1 1 pis.

— 1966. Some Leperditiid Ostracoda from the Richardson Mountains, northern Yukon Territory. Proc.
geol. Ass. Can. 17, 43-51, 1 pi.

1970. Ostracoda from the Vaureal Formation (Upper Ordovician) of Anticosti Island, Quebec. Bull. geol.

Surv. Can. 187, 15-29, pis. 4, 5.

— 1973. Ostracoda from the Ellis Bay Formation (Ordovician), Anticosti Island, Quebec. Geol. Surv. Pap.
Can. 72-43, 1 49, 9 pis.

— 1974a. Silurian Ostracoda from Anticosti Island, Quebec. Bull. geol. Surv. Can. 117, 1-13, 1 pi.

— 19746. Middle Ordovician Ostracoda from southwestern District of Mackenzie. Ibid. 244, 1-55, pis. 1 -9.

— 1977a. Early Paleozoic Ostracoda of eastern Canada. In swain, f. m. (ed.). Stratigraphic micropaleontology
of Atlantic Basin and Borderlands , 1 1 7, 3 pis. Elsevier, Amsterdam.

— 1 9776. Early Paleozoic Ostracoda from southwestern District of Mackenzie and Yukon Territory. Bull,
geol. Surv. Can. 275, 1-88, 16 pis.

1977c. Ordovician Ostracoda, southeastern District of Franklin. Ibid. 269, 77-97, 6 pis.

1978. Early Paleozoic ostracode assemblages, northwestern Canada. Spec. Pap. geol. Ass. Can. 18, 93-111,
3 pis.

1980. Innuitbeyrichia , a new Silurian ostracode genus from the Canadian arctic. Geol. Surv. Pap. Can.
80- IB, 29-37.

1981. Latest Ordovician and Silurian ostracode faunas from Anticosti Island, Quebec. In lesperance, p. j.
(ed.). IUGS Subcommission on Silurian Stratigraphy, Ordovician-Silurian Boundary working group. Field
meeting , Anticosti, Gaspe, Quebec, II: Stratigraphy and paleontology, 185-195.

— 1982a. An occurrence of the Silurian ostracode Beyrichia ( Beyrichia ) from Anticosti Island, Quebec. Geol.
Surv. Pap. Can. 82- IB, 223-224.

— 1 9826. Bathymetry of early middle Ordovician (Chazy) ostracodes. Lower Esbataottine Formation,
District of Mackenzie. Bull. geol. Surv. Can. 347, 1 -39, 9 pis.

1983. Steusloffina cuneata (Steusloff), 1895, from Anticosti Island, Quebec. Geol. Surv. Pap. Can. 83-1B,
201-204.

— and berdan, J. M. 1977. Silurian and early Devonian beyrichiacean ostracode provincialism in northeastern
North America. Ibid. 77-1B, 15-24, 3 pis.

— and bolton, t. e. 1977. Additional paleontological observations bearing on the age of the Lourdes
Formation (Ordovician), Port au Port Peninsula, Western Newfoundland. Ibid. 77-1B, 1 13, 4 pis.

creath, w. b. and shaw, a. b. 1966. Paleontology of northwestern Vermont. XIII. Isochilina from the
Ordovician Highgate Formation. J. Paleont. 40, 1331-1334, pi. 164.
de deckker, p. 1977. The distribution of the 'giant' ostracods (Family Cyprididae Baird, 1845) endemic to
Australia. In loffler, h. and danielopol, d. (eds.). Aspects of ecology and zoogeography of recent and fossil
Ostracoda, 285-294. W. Junk, The Hague.

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 595

dewey, J. F. 1969. Evolution of the Appalachian/Caledonian Orogen. Nature, Lond. 222, 124-129.
elofson, o. 1941. Zur Kenntnis der marinen Ostracoden Schwedens, mit besonderer Beriicksichtigung des
Skagerraks. Zoo/. Bidr. Upps. 19, 215-534.

fortey, R. A. 1975. Early Ordovician trilobite communities. Fossils Strata , 4, 339-360.

— 1984. Global earlier Ordovician transgressions and regressions and their biological implications. Palaeont.
Contr. Univ. Oslo, 295, 37-50.

and barnes, c. r. 1977. Early Ordovician conodont and trilobite communities of Spitsbergen: influence on

biogeography. Alcheringa, 1, 297-309.

— and owens, R. m. 1978. Early Ordovician (Arenig) stratigraphy and faunas of the Carmarthen district,
south-west Wales. Bull. Br. Mus. nat. Hist. ( Geol .) 30, 225-294.

guber, a. l. 1971. Problems of sexual dimorphism, population structure and taxonomy of the Ordovician genus
Tetradella (Ostracoda). J. Paleont. 45, 6-22, pis. 1-4.

-and jaanusson, v. 1964. Ordovician ostracodes with posterior domiciliar dimorphism. Bull. geol. Instn
Univ. Uppsala, 43 (1/3) 2, 43 pp., 6 pis.

Harris, r. w. 1931. Descriptions and illustrations of ostracodes and conodonts. In decker, c. e. and merritt,
c. a. The stratigraphy and physical characteristics of the Simpson Group. Bull. Okla. geol. Surv. 55,
112 pp., 15 pis.

1957. Ostracoda of the Simpson Group of Oklahoma. Ibid. 75, vi + 333 pp., 10 pis.

henningsmoen, g. 1948. The Tretaspis Series of the Kullatorp core. Bull. geol. Instn Univ. Uppsala , 42, 374-432,
pis. 23-25.

— 1953. The Middle Ordovician of the Oslo region, Norway. 4. Ostracoda. Norsk geol. Tidsskr. 32, 35-56,
5 pis.

— 1954a. Lower Ordovician ostracods from the Oslo region, Norway. Ibid. 33, 41-68, 2 pis.

1954/j. Upper Ordovician ostracods from the Oslo region, Norway. Ibid. 69-108, 6 pis.

— 1965. On certain features of palaeocope ostracodes. Geol. For. Stockh. Fork. 86, 329-334.

hessland, i. 1949. Investigations of the Lower Ordovician of the Siljan district, Sweden. I. Lower Ordovician
ostracods of the Siljan district, Sweden. Bull. geol. Instn Univ. Uppsala , 33, 97-408, 26 pis.
hessler, r. r. and saunders, h. l. 1967. Faunal diversity in the Deep Sea. Deep-Sea Res. 14, 56-78.

Holland, c. h. 1971. Silurian faunal provinces? Geol. J. spec. Issue, 4, 61-76.

jaanusson, v. 1957. Middle Ordovician ostracodes of central and southern Sweden. Bull. geol. Instn Univ.
Uppsala, 37, 173-442, 15 pis.

1966. Ordovician ostracodes with supravelar antra. Ibid. 43, (6/8) 7, 30 pp., 3 pis.

jones, c. R. and siveter, d. j. 1983. On Harperopsis scripta (Harper). Stereo-Atlas Ostracod Shells, 10, 5-12, 4 pis.
kay, G. m. 1934. Mohawkian ostracoda: species common to Trenton faunules from the Hull and Decorah
formations. /. Paleont. 8, 328-343, pis. 44-46.

— 1940. Ordovician Mohawkian Ostracoda: Lower Trenton Decorah fauna. Ibid. 14, 234-269, pis. 29-34.
keenan, j. e. 1951. Ostracodes from the Maquoketa Shale of Missouri. Ibid. 25, 561 574, pis. 78, 79.
kesling, r. v. 1960. Middle Ordovician Black River ostracods from Michigan. Part II. Levisulculus and

Eurychilina. Contr. Mus. Paleont. Univ. Mich. 15, 349-363, 4 pis.

-hall, d. d. and melik, J. c. 1962. Middle Ordovician Black River ostracods from Michigan. Part IV.
Species of Colacchilina (new genus), Laccochilina, and Hesperidella. Ibid. 17, 205-213, 2 pis.
knupfer, j. 1968. Ostracoden aus dem Oberen Ordovizium Thiiringens. Freiberger ForschHft (C), 234,
5-29, 6 pis.

kornicker, l. s. and sohn, i. g. 1971. Viability of ostracode eggs egested by fish and effect of digestive fluids on
ostracode shells— ecologic and paleoecologic implications. Bull. Cent. Rech. Pau , 5, 125-135, 1 pi.
kraft, j. c. 1962. Morphologic and systematic relationships of some Middle Ordovician Ostracoda. Mem. geol.
Soc. Am. 86, viii + 104 pp., 19 pis.

krandievsky, v. s. 1969. Stratigraphic distribution of Ostracoda in the Ordovician deposits of the Volyn-
Podolye. Trans. Akad. Sci. Ukrain. SSR (b) 1969, 870-874. [In Ukrainian.]

— 1975. First occurrences of the Lower Ordovician ostracods in western Volyn. Ibid. 1975, 691-694. [In
Ukrainian.]

krause, a. 1 897. Ueber die Ostrakodenfauna eines hollandischen Silurgeschiebes. Z. dt. geol. Ges. 48, 932-939,
pi. 25.

kummerow, E. 1924. Beitrage zur Kenntnis der Ostracoden und Phyllocariden aus nordischen Diluvial-
geschieben. Jb. preuss. geol. Landesanst . BergAkad. 44, 405-448, pis. 20, 21.
leggett, j. K. 1979. Oceanic sediments from the Ordovician of the Southern Uplands. Spec. Pubis geol. Soc.
Lond. 8, 495-498.

596

PALAEONTOLOGY, VOLUME 28

leggett, j. k., mckerrow, w. s., cocks, l. r. m. and rickards, r. b. 1981. Periodicity in the early Palaeozoic
marine realm. J. geol. Soc. 138 , 167-176.

lundin, r. f. 1971 . Possible paleoecological significance of Silurian and early Devonian ostracode faunas from
midcontinental and northeastern North America. Bull. Cent. Reck. Pau , 5 , 853-868.
mckenzie, a. h. G. and hussainy, s. u. 1968. Relevance of a freshwater cytherid (Crustacea, Ostracoda) to the
continental drift hypothesis. Nature, Lond. 220 , 806-808.
mckerrow, w. s. 1979. Ordovician and Silurian changes in sea level. J. geol. Soc. 136 , 137-145.

and cocks, l. r. m. 1976. Progressive faunal migration across the Iapetus Ocean. Nature, Lond. 263 ,
304-306.

mannil, r. m. 1963. On the correlation of the Ordovician strata of Estonia and Leningrad region. Eesti NSV
Tead. Akad. Geol. Inst. Uurim. 13 , 3-40. [In Russian.]

martinsson, a. 1956. Neue Funde kambrischer und ordovizischer Geschiebe im siidwestlichen Finnland. Bull,
geol. Instn Univ. Uppsala, 36, 79-105, 2 pis.

- 1963. Kloedenia and related ostracode genera from the Silurian and Devonian of the Baltic area and
Britain. Ibid. 42 ( 1 /6) 2, 63 pp.

- 1965. The Siluro- Devonian ostracode genus Nodibeyrichia and faunally associated kloedeniines. Geol. For.
Stockh. Fork. 87 , 109-138.

1967. The succession and correlation of ostracode faunas in the Silurian of Gotland. Ibid. 89 , 350-386.
1970. Ostracodes: correlation with Europe. Spec. Pap. geol. Soc. Am. 102 , 41-44.

1977. Palaeocope ostracodes. Int. Un. geol. Sci. (Ser. A) 5 , 327-332.

Mitchell, a. h. G. 1984. The British Caledonides: interpretations from Cenozoic analogues. Geol. Mag. 121 ,
35-46.

neckaja, a. i. 1973. Ostrakody ordovika i silura SSSR. Trudy vses. neft. nauchno-issled. geol.-razy. Inst. 324 ,
104 pp. [In Russian.]

neuman, r. b. 1972. Brachiopods of early Ordovician volcanic islands. Int. Geol. Congr., sec. 7, 297-302.
persina, a. i. et al. 1971. Biostratigrafija silurijskich i devonskich otlozenij Pecorskogo Urala, 131 pp. Nauka,
Leningrad. [In Russian ]

philips, w. e. a., stillman, c. J. and murphy, T. 1976. A Caledonian plate tectonic model. J. geol. Soc. 132 ,
579-610.

poulsen, v. 1978. Dalmanitina beds (late Ordovician) on Bornholm. Damn. geol. Unders 1976 , 53-87.
pribyl, a. 1979. Ostracoden der Sarka- bis Kraluv Dvur-Schichten-gruppe des bohmischen Ordoviziums. Sb.
nar. mus. Praze (B), 33, 53-145, 8 pis.

prokofiev, v. a. and kuznetzov, a. G. 1982. Fauna and certain stratigraphic problems in Ordovician deposits of
Moscow syneclise. Bull. Mosc. Soc. Naturalists (Geol. Ser.) 57, 67-82.
qvale, g. 1980. New Caradocian ostracodes from the Oslo-Asker district, Norway. Norsk geol. Tidsskr. 60,
93-116.

rader, e. k. 1965. First discovery of the hollinacean Ostracoda Pseudorakverella from the United States
(Trenton Member, Martinsburg Formation, Virginia). J. Paleont. 39 , 159 161.

SARV, L. i. 1959. Ordovician ostracods in the Estonian S.S.R. Eesti NSV Tead. Akad. Geol. Inst. Uurim. 4 ,
21 1 pp., 32 pis.

- 1963. New ostracods from the Ordovician of East Baltic. Ibid. 13 , 161-188, 7 pis.

schallreuter, R. E. L. 1964«. Neue Ostrakoden der Gattungen Platybolbina, Brevibolbina und Oecematobolbina
aus mittelordovizischen Backsteinkalkgeschieben. Ber. geol. Ges. D.D.R. 9, 381-383, 422-425.

1 9646. Hithinae— eine neue Unterfamilie ordovizischer Ostrakoden. Ibid. 385-387, 426-429.

1968. Ordovizische Ostracoden mit geradem SchloBrand und konkavem Ventralrand. Wiss. Z. Ernst
Moritz Arndt-Univ. Greifswald, 17 , 127-152.

1973. Die Ostracodengattung Hyperchilarina und das Aparchites-Prob\em. Geol. For. Stockh. Fork. 95 ,
37-49.

1975. Ostrakoden aus Ojlemyrgeschieben (Ordoviz) II. Neues Jb. Geol. Palaont. Abb. 150 , 270-293.

1977a. On Cryptophyllus gutta Schallreuter. Stereo- Atlas Ostracod Shells, 4 , 1-8.

19776. On Distobolbina grekoffi Schallreuter sp. nov. Ibid. 25-28.

1977c. Eine neue Art der Ostrakodengattung Pelecybolbina aus dem Ordoviz von Oland. Geol. For. Stockh.
Forh. 99 , 409-41 1.

1980. Ostrakoden aus dem Sularpschiefer (Mitlelordoviz) von Schonen (Schweden). Palaeontographica,
A, 169 , 1-27, pis. 1-9.

1983. Glossomorphitinae und Sylthinae (Tetradellidae, Palaeocopa, Ostracoda) aus Backsteinkalk-
Geschieben ( Mittelordoviz) Norddeutschlands. Ibid. 180 , 126-191, pis. 15-29.

SCHALLREUTER AND SIVETER: OSTRACODES ACROSS THE IAPETUS OCEAN 597

— and kruta, m. 1980. Taxonomy and nomenclature of the Ordovician ostracode genus Hippula. Neues Jb.
Geol. Palaont. Mh. 1980, 505-512.

— and siveter, d. j. 1982. On Hithis colonus Schallreuter & Siveter sp. nov. Stereo- Atlas Ostracod Shells , 9,
85-88.

1983a. On Tallinnellina dissita Schallreuter & Siveter sp. nov. Ibid. 10, 1-4.

19836. Ostrakoden beiderseits des Iapetus-Ozeans. Palaont. Ges. 53. Jahresversammlung Mainz , 66.
scheltema, r. e. 1977. Dispersal of marine invertebrate organisms: paleobiogeographic and biostratigraphic
implications. In kauffman, e. g. and hazel, j. e. (eds.). Concepts and methods in biostratigraphy, 73-108.
Dowden, Hutchinson & Ross, Stroudsberg, Pennsylvania.

Schmidt, e. a. 1941. Studien im bohmischen Caradoc (Zahofan-Stufe). 1. Ostrakoden aus den Bohdalec-
Schichten und iiber die Taxonomie der Beyrichiacea. Abh. senckenb. naturforsch. Ges. 454.
sidaraviciene, n. 1971 . New Ostracoda from Middle and Upper Ordovician of Lithuania. In Palaeontology and
stratigraphy of the Baltic and Byelorussia , 3, 23-36.

— 1975. New Ordovician Ostracoda of the South Baltic Area. In grigelis, a. (ed. ). The fauna and stratigraphy
of Paleozoic and Mesozoic of Baltic and Byelorussia , 21-43.

siveter, d. j. 1978a. The Ordovician. Geol. J. Spec. Issue , 8, 41-56.

1978/7. The Silurian. Ibid. 57- 100.

— 1982a. Late Ordovician and early Silurian ostracodes from the Oslo region, Norway. In worsley, d. (ed.).
IUGS Subcommission on Silurian Stratigraphy. Field Meeting, Oslo Region, 1982 , 121 -127.

19826. On Homeokiesowia epicopa Siveter sp. nov. Stereo- Atlas Ostracod Shells, 9, 89-92.

1982c. On Schallreuteria superciliata (Reed). Ibid. 93-100.

— 1983. On Vittella fecunda Siveter sp. nov. Ibid. 10, 13 16.

— 1984. Habitats and modes of life of Silurian ostracodes. In bassett, m. g. (ed.). Autecology of Silurian
organisms. Spec. Pap. Palaeont. 32, 71-85, 2 pis.

skevington, d. 1978. Latitudinal surface water temperature gradients and Ordovician faunal provinces.
Alcheringa, 2, 21-26.

sohn, i. g. and kornicker, l. s. 1 979. Viability of freeze-dried eggs of the freshwater Heterocypris incongruens. In
krstic, n. (ed.). Taxonomy, biostratigraphy and distribution of ostracodes, 1-4, 1 pi. Geological Society of
Serbia, Belgrade.

spjeldnaes, n. 1978. Faunal provinces and the Proto-Atlantic. Geol. J. Spec. Issue, 10, 139-150.
steele, h. m. and Sinclair, g. w. 197 1. A Middle Ordovician fauna from Braeside, Ottawa Valley, Ontario. Bull,
geol. Surv. Can. 21 1, 97 pp.

steusloff, a. 1895. Neue Ostrakoden aus Diluvialgeschieben von Neu-Brandenburg. Z. dt. geol. Ges. 46,
775-787, pi. 58.

swain, f. m. 1957. Early Middle Ordovician Ostracoda of the eastern United States. Part I. Stratigraphic
data and description of Leperditiidae, Aparchitidae and Leperditellidae. J. Pa/eont. 31, 528-570, pis.
59-62.

— 1962. Early Middle Ordovician Ostracoda of the eastern United States. Part II. Leperditellacea (part),
Hollinacea, Kloedenellacea, Bairdiacea and Superfamily Uncertain. Ibid. 36, 719-744, pis. 109-1 1 1.

1977. Early Paleozoic Ostracoda of the Atlantic margin, other than eastern Canada. In swain, f. m. (ed.).
Stratigraphic micropaleontology of Atlantic Basin and Borderlands , 19-48. Elsevier, Amsterdam.

— Cornell, j. r. and hansen, d. l. 1961. Ostracoda of the families Aparchitidae, Aechminidae,
Leperditellidae, Drepanellidae, Eurychilinidae and Punctaparchitidae from the Decorah Shale of Minnesota.
J. Paleont. 35, 345-372, pis. 46-50.

teichert, c. 1937a. Ordovician and Silurian faunas from Arctic Canada. Rep. Fifth Thule Exped. 1921-1924, 1
(5), 169 pp., 24 pis.

— 19376. A new Ordovician fauna from Washington Land, North Greenland. Meddr Gronland, 119 (1),
65 pp., 7 pis.

thorslund, p. 1940. On the Chasmops Series of Jemtland and Sodermanland (Tvaren). Sver. geol. Unders. Afh.
(Ser. C), 436, 191 pp., 15 pis.

troedsson, g. t. 1918. Om Skanes Brachiopodskiffer. Acta Univ. land. (n.f. 2), 15 (3), 1 10 pp., 2 pis.
ulrich, e. o. 1892. New lower Silurian Ostracoda, no. 1. Am. Geol. 10, 263-270, pi. 9.

ulst, r. z., gailTte, l. k. and jakovleva, v. i. 1982. Ordovik Latvii, 295 pp., 8 pis. Zinatne Riga. [In Russian.]
van morkhoven, f. p. c. m. 1962. Post-Palaeozoic Ostracoda; their morphology, taxonomy and economic use. 1,
204 pp. Elsevier, Amsterdam.

vannier, j. 1984. Ostracodes ordoviciens du Massif Armoricain. Unpublished thesis. University of Rennes,
158 pp., 26 pis.

598 PALAEONTOLOGY, VOLUME 28

varganov, v. g. et al. 1970. Ordovik central'noj casti Ufimskogo amfiteatra. Sov. Geol. 1970 (6), 44-57. [In
Russian.]

warshauer, s. M. 1975. The sexual dimorphism and ontogeny of Ceratopsis chambers i (Miller) (Ostracoda,
Palaeocopida) from the Upper Ordovician of southwestern Ohio and northern Kentucky. Bull. Am. Paleont.
67, 443-456, 3 pis.

- and berdan, j. m. 1982. Palaeocopid and Podocopid Ostracoda from the Lexington Limestone and Clays
Ferry Formation (Middle and Upper Ordovician) of Central Kentucky. Prof. Pap. U.S. geol. Surv. 1066 (H),
80 pp., 19 pis.

Whittington, h. b. and hughes, c. p. 1972. Ordovician geography and faunal provinces deduced from trilobite
distribution. Phil. Trans. R. Soc., B, 263, 235-278.

— and kindle, c. h. 1963. Middle Ordovician Table Head Formation, western Newfoundland. Bull. geol. Soc.
Am. 74, 745-758, 2 pis.

williams, a. 1973. Distribution of brachiopod assemblages in relation to Ordovician palaeogeography. Spec.
Pap. Palaeont. 12, 241-269.

- 1976. Plate tectonics and biofacies evolution as factors in Ordovician correlation. In bassett, m. g. (ed.).
The Ordovician System , 29-66. University of Wales Press and National Museum of Wales, Cardiff.

williams, h. 1980. Structural telescoping across the Appalachian Orogen and the minimum width of the
Iapetus Ocean. Spec. Pap. geol. Ass. Can. 20, 422-440.
wilson, J. T. 1966. Did the Atlantic close and then re-open? Nature , Lond. 210, 678-681 .

Zenkova, G. G. 1977. Novye vidy ostrakod ordovika zapadnogo sklona Srednego Urala. Novye vidy drevnich
rastenij i bespozvonocnych SSSR , 4, 74-75, 87, 180, 203. [In Russian.]

ROGER E. L. SCHALLREUTER

Geologisch-Palaontologisches Institut
Universitat Hamburg
Bundesstrasse 55, D 2000 Hamburg 1 3
Federal Republic of Germany

Typescript received 20 July 1984
Revised typescript received 31 January 1985

DAVID J. SIVETER

Department of Geology
University of Leicester
Leicester LEI 7RH, U K.

TWO NEW SPECIES OF HERBACEOUS LYCOPODS
FROM THE DEVONIAN OF VENEZUELA WITH
COMMENTS ON THEIR TAPHONOMY

by d. edwards and j. l. benedetto

Abstract. Two new species of herbaceous lycopods, Haskinsia sagittata and Colpodexylon cachiriense ,
are described from Devonian strata in the Cano Grande-Rio Cachiri region (north central Sierra de Perija)
of Venezuela. The recently discovered diverse assemblage of megafossils also contains the first record
of progymnosperms in South America as well as possible cladoxylaleans. The lycopods exhibit a number of
preservation forms, and these are related to their putative preservational history. The assemblage is of
Laurentian aspect, showing greatest similarity to those from New York State. Its biogeographical significance is
briefly discussed in relation to the distribution of continents in Devonian times.

Most of our knowledge of terrestrial vegetation in Devonian times derives from assemblages
collected in North America, Europe, and Asia (Chaloner and Sheerin 1979; Banks 1980; Gensel and
Andrews 1984). With the exception of Australia, records from southern continents are based on
fragmentary, usually poorly preserved megafossils which often lack independent faunal or
palynological evidence for their age. In the course of a biostratigraphical study of Devonian outcrops
in northern Venezuela, one of us (J. L. B.) and Dr. P. Rachebouef (University of Brest) discovered
a completely new fossil plant assemblage, more diverse and better preserved than hitherto recorded
from the Devonian of South America.

STRATIGRAPHY AND LOCALITIES

The Devonian of the Colombian-Venezuelan basin in the northern Andes comprises a number of
outcrops extending from the eastern cordillera in Colombia to the Sierra de Perija in Venezuela (text-
fig. In). The strata exposed in the Cano Colorado-Rio Cachiri area, c. 70 km west of Maracaibo, are
the most northerly (text-fig. 1 b). The geology and biostratigraphy of the area have been summarized
by Benedetto (1980, 1984). The fossil plants occur at two exposures of the same horizon (Co 8 )
towards the base of the Campo Chico formation in the Cano Colorado section (text-fig. lc). There
were no animal fossils present, but a rich brachiopod fauna collected about 50 m below the plant beds
(Co 6 horizon) is of Givetian age.

MATERIAL AND METHODS

The fossil plants occur in laminated micaceous siltstones ranging in colour from a light olive grey to
whitish or pinkish buff depending on the degree of weathering. The lycopod stems are strongly
compressed and usually incompletely preserved. Many comprise a thin fragmentary layer of fine-
grained sediment sandwiched between two coalified layers. Others are iron-stained impressions often
with a dusting of coaly matter. The complete range of preservation forms is discussed more fully
below.

All specimens were photographed under uniform and unilateral lighting. The most useful
technique for this extremely soft matrix was degagement (Leclercq 1 960), tungsten wire sharpened in
molten sodium nitrite being used to loosen the grains of sediment. Areas of cuticle were removed on

I Palaeontology, Vol. 28, Part 3, 1985, pp. 599-618, pis. 71-74.)

text-fig. 1. a , locality map indicating Devonian outcrops in the northern part of the Colombian
Venezuelan basin of South America, b , Devonian geology in the Cano-Colorado Rio Cachiri region.
c, Stratigraphical log of the section along the road to the Rio Socuy from the Cano Colorado. Plant
horizon is indicated by a star. Co,_ 8 refer to horizons with faunal assemblages recorded in detail by

Benedetto (1984).

EDWARDS AND BENEDETTO: DEVONIAN LYCOPODS

601

cellulose nitrate film pulls. Latex replicas (Dunlop) taken from impressions with cellular imprints
were coated with gold and examined by SEM. Palynological samples were prepared by standard
techniques.

The specimens and preparations are housed in the Palaeontology Laboratory, University of Brest,
France (prefix LPB).

SYSTEMATIC PALAEONTOLOGY

Class LYCOPSIDA
Order protolepidodendrales
Family haskinsiaceae Grierson and Banks, 1983
Genus haskinsia Grierson and Banks, 1983

Type species'. Drepanophycus colophyllus Grierson and Banks, 1963 (transferred to H. colophylla by
Grierson and Banks 1983) from the Kiskatom Formation, Tioughniogan Stage, Erian Series, Middle
Devonian (=Givetian) in New York State, U.S.A.; also recorded from Lower Frasnian strata of New
York State.

Haskinsia sagittata sp. nov.

Plate 7 1 , figs. 1 -8; Plate 72, figs. 1 1 3; Plate 73, figs. 1 -8; text-fig. 2
Derivation of name. Latin sagitta , arrow, referring to the arrowheaded shape of the leaf.

Holotype. LPB 16046 (PI. 71, fig. 6; PI. 72, figs. 1, 2, and 12).

Locality. Exposures on the road from the town. Villa del Rosario, to the Socuy River in the Colorado valley,
some 950 and 1500 m NNW of the farm ‘Alemania’, c. 70 km west of Maracaibo, north-western Venezuela.

Horizon. Carbonaceous lutites at base of the Campo Chico Formation; Devonian (?Givetian-Frasnian).

Diagnosis. Herbaceous lycopod with occasional dichotomous branching. Axes at least 10-5 cm long,
4-8-6-8 mm wide (x = 5-7 mm, n = 34) exclusive of leaves. Longitudinal lines separate leaf bases.
Leaves spirally arranged, sometimes pseudowhorled; 7-9 leaves per gyre (usually 7). Lateral margins
of adjacent leaf blades touch or overlap: apices of leaves of one gyre at same level as bases of leaves in
that above. Leaves simple and upright, comprising petiole 0-45-1-2 mm wide (x = 0-7 mm, n = 35)
widening into conspicuous lamina, at maximum width 1-13—1 -92 mm (x = 1 -48 mm, n = 30) with two
downwardly directed lateral projections and tapering into an acuminate tip. Leaf length, including
petiole, at least 2 47 mm.

Description. Seven small slabs, almost all of slightly differing lithology, contain some forty sterile leafy stems of
variable appearance. The variability is due to different forms of preservation and planes of fracture. With one
exception ( LPB 1 605 1 , PI. 71, fig. 2) the possession of distinctive sagittate leaves indicates that the stems belong
to a single taxon.

Stem morphology; Except in the two branching specimens the stems show little change in diameter along the
short lengths available for study. Branching is dichotomous (PI. 71, fig. 1). Depending on the preservation/
fracture state the stem surface displays elevations or depressions, sometimes both, marking the attachment of
leaves, the surface between the leaf bases being smooth (PI. 7 1 , fig. 2). Leaves borne on the straight margins of the
stem are falcate in profile. Further conspicuous features of a few stems are longitudinal dark lines (PI. 71, fig. 1)
which, on any one surface, pass between but not across leaf bases: their number is thus related to the number of
orthostichies. They are either parallel or curve slightly outwards around a leaf base. In a few cases much finer,
closely spaced, longitudinally orientated striations occur on a thin brown layer, presumed to be cuticular, but
when this is removed on a film pull no further detail is visible. A similar patterning is present on a few
impressions, but again scanning electron micrographs of latex casts of these stems failed to reveal any precise
details of surface cells (PI. 73, fig. 8).

Leaf morphology and arrangement: Leaves are most commonly seen in profile. A typical example before
uncovering (PL 72, fig. 5) shows a prominent base tapering upwards into a simple linear blade, which in this case
is parallel to the edge of the axis, but may occasionally be directed away from it (PI. 72, fig. 6). Removal of

602

PALAEONTOLOGY, VOLUME 28

sediment between the leaf and axis, and sometimes removal of the axis itself, reveals an approximately parallel-
sided petiole and approximately half of a lamina with a downwardly directed lobe (PI. 72, fig. 8). The tapering
apex of the lamina often extends down into the matrix, so that the length of the leaf is actually greater than that
visible in profile. At its base the petiole is vertically extended in cross-section, becoming circular and then
flattened distally in the same plane as the lamina. The presence of a thick layer of coalified material suggests that
the petiole possessed abundant, probably peripheral, strengthening tissues. Except for one fortuitously fractured
specimen (PI. 71, fig. 4) leaf shape in face view was elucidated by degagement. Text-fig. 2, where all the leaves are
drawn to the same scale, summarizes the range in shape. The leaf blade is roughly triangular with entire margin,
and widest point more or less at the level of the junction of the petiole (PI. 72, figs. 9-11). Two pronounced
downwardly directed projections produce the sagittate form (PI. 72, figs. 9 11) but very occasionally these flare
outwards so that the overall shape is hastate (PI. 72, fig. 1 2). A midrib is absent. The parallel-sided petiole widens
slightly below the blade, but the actual junction is impossible to define. We are not sure of the exact shape of the
extremities of the leaf blades since removal of a few grains of sediment can radically alter a tapering outline. At
the leaf apex the blade frequently fades away as if it were thinner distally (PI. 71, figs. 6 and 8). Evidence from
isolated tips and iron-stained impressions in addition to uncovered specimens suggests that the apices were
originally acuminate (PI. 71, fig. 7; PI. 72, fig. 11). The measurements for leaf lengths (blade plus petiole =
2-47-4-7 mm, x = 3-2, n = 19) are therefore too low. The downward projections were probably pointed, but
ended somewhat more abruptly than the apices. Any cuticle remaining on blades or petioles is featureless.

The positions of leaf bases on the stems indicate that the leaves were borne in a low spiral, sometimes
approaching a pseudowhorled condition, with 7 to 9 leaves in a gyre, the former being the more usual number.
The external appearance of leafy shoots in life must have been similar to that of the fossils figured in Plate 71,
fig. 4, where the plane of fracture passes through the laminae of the leaves leaving a dusting of coalified material
on part and counterpart. The width of this specimen is explained by a change in level which more or less bisects
the leafy surface longitudinally and indicates that there were probably two axes lying side by side. The fortuitous
and very informative fracture is thus even more remarkable. The subtending stems, originally completely
obscured by leaves and sediment, were revealed by degagement on the larger block. Adjacent leaves in a single
gyre meet or just overlap at their lateral extremities and the attenuated leaf apex just extends beyond the base
of the petiole in the gyre above. From the attitudes of the leaves it seems that the blades were rigid in life but
gently curved inwards and tangentially so that they formed a continuous sheath around the stem.

Comparisons. Grierson and Banks (1983) review Palaeozoic herbaceous lycopods, including the
families Drepanophycaceae, Protolepidodendraceae, Archaeosigillariaceae, Eleutherophyllaceae,
and Lycopodiaceae. The Venezuelan specimens most closely resemble their new taxon, H. colophylla.
They originally (1963) placed the leafy remains from New York State in Drepanophycus, but new
information on the leaves, originally considered falcate, and on stem anatomy, led to the generic
reassignment. The leaves of the American specimens are not as markedly sagittate and the vertical
files of strengthening cells in the outer cortex are more sinuous. On the basis of such differences, and
because the Venezuelan plants lack anatomy, we regard them as a new species.

EXPLANATION OF PLATE 71

Figs. 1-8. Haskinsia sagittata sp. nov. 1, LBD16050, branching, leafy stem with traces of coalified material,
x 1-7. 2, LBD16051, leafless stem, probably attributable to H. sagittata , seen at base as a cleavage

compression (sensu Chaloner and Collinson 1975) and as a cleavage impression distally. Note persistence
of coalified material at points of departure of leaves and on truncated petioles attached to sides of stem,
unilateral illumination, x2-8. 3-5, LBD16052. 3, coalified stem with sedimentary infill (s) and one un-
covered leaf (arrowed), x 2-8. 4, leaf lamina cleaved compressions of two aligned stems, x 3-7. 5, base of
stem illustrated in fig. 3, unilateral illumination, x 10. 6, LBD 16046, uncovered leaf on holotype. It was
attached (‘scar’ is arrowed) to lower surface of the stem and shows the typical lateral displacement of the
petiole, x 11. 7, LBD16054(/), acuminate apex of leaf, probably incomplete at tip. Note that it ends at
approximately same level as base of leaf above. 8, LBD 16046, petioles and lamina bases of three leaves
uncovered above fractured end of ridged stem. Arrowed is the elongate raised area in the oval depression
marking attachment of petiole on lower surface, x 10. (Photographs under uniform illumination except
where otherwise stated.)

PLATE 71

EDWARDS and BENEDETTO, Haskinsia sagittata

604

PALAEONTOLOGY, VOLUME 28

There remain a number of very fragmentary fossils with similarly shaped leaves, which on further
investigation may be assigned to Haskinsia. Grindley et al. (1980) described some stems from the
Middle Devonian of Marie Byrd Land, western Antarctica, with similar dimensions and numbers of
leaves per gyre. The leaves themselves are slightly smaller but rarely completely preserved, and are
sickle-shaped in profile. Face view drawings of the leaves show that their shape is variable: the one
just below no. 3 in their fig. 10 has a downwardly projecting tip on the left hand side reminiscent of
H. sagittata. A major difference is the presence of a central line presumably representing a vein or
midrib in the Antarctic specimens. The latter were placed in a new species, D. schopfi , but on the
basis of leaf shape they clearly do not belong to that genus (Grierson and Banks 1983).

Far more problematic are the short lengths of lycopod stems, usually lacking leaves but with leaf
bases or scars, that are superficially similar to some of the preservation forms seen in the Venezuelan
material. These include Haplostigma fwquei from the Middle Devonian of Argentina (Frenguelli
1954), H. irregularis. Middle Devonian of Brazil (Krause! 1960) and the Falkland Islands (Seward
and Walton 1923) as well as from the type area in South Africa, and Palaeostigma sewardi first
described from the Middle Devonian of Brazil (Krausel and Dolianiti 1957). The most recent
assessment of these two genera is in a preliminary survey of South African Palaeozoic floras by
Anderson and Anderson (in press). They note that Palaeostigma is atypical among South African
lycopods in that it invariably occurs as impressions with the leaves or granulae of the lower surface of
the stem showing through to the upper. We have described this preservation state in Haskinsia
sagittata , but the taxa differ in that the Andersons interpret the projections and depressions on the
stem as leaves, describing them as ‘rudimentary, rounded to oval granulate projections about as high
as wide, well spaced, irregularly aligned into imperfect low angle spiral’. Haplostigma , originally
based on South African material, has a more regular appearance and the leaves although rarely
preserved are described as ‘squat conical, acute tipped, length around 1 to 2 times basal width with
gentle distal inclination’. Anderson and Anderson placed H.furquei in synonomy with H. irregularis.
We have not had the opportunity to examine specimens of either genus, but from photographs it
seems that their relatively featureless stems more closely resemble leafless Haskinsia sagittata than
those lycopods with a characteristic surface patterning such as Archaeosigillaria. The latter, repre-
sented by A. picosensis (placed in synonomy with A. caespitosa by Anderson and Anderson (in press))
occurs with Protolepidodendron kegeli at the same Middle Devonian horizon as Palaeostigma sewardi
in Brazil. It is easily separated from H. sagittata (as is A. conferta from Argentina (Menendez 1965a))
on the grounds of leaf base characteristics (Krausel and Dolianiti 1957).

Finally a predominantly lycopod flora from the late Silurian-early Devonian of Libya (Klitzsch et
al. 1973; Boucot and Gray 1982) contains a number of fragmentary stems with leaf attachment sites,
but not leaves, visible. Their preservation states appear to parallel some of those we have described.
In particular, the illustrations of Protolepidodendron he/leri (Klitzsch et al. 1973, figs. 1. 12, 13) and
Precyclostigma tadrartense (ibid., figs. 2, 6, 9, 10) resemble H. sagittata in size, leaf arrangement, and

EXPLANATION OF PLATE 72

Figs. 1-13. Haskinsia sagittata sp. nov. 1, 2, LBD16046, holotype. I, stem with some uncovered leaves. Those
near the base are attached to the upper surface of the stem, x2-4. 2, unilateral lighting to show surface
features such as leaf bases and ridges, x 2-6. 3, 4, LBD 16047, pink stem impressions with dusting of coaly
powder. 3, xl -4. 4, unilateral illumination, x 1-7. 5-7, leaves in profile before uncovering. 5, LBD 16052,
x9-4. 6, LBD 16047, x 10. 7, LBD16049, x 7. 8, LBD 16047, half of leaf revealed on removal of overlying
stem and matrix, x9. 9-11, face views of uncovered leaves. 9, LBD16047; 10, 11, LBD16052 all x 10.
12, LBD 16046, hastate leaf exposed when sediment was removed from above holotype; probably attached
to upper surface of stem, x 12. 13, LBD 16052, stem on left has more or less continuous covering of coalified
material (leaf attachment sites are more obvious under unilateral illumination) and shows numerous leaves
in profile, successive gyres being close together, x 1 -6. (Photographs under uniform illumination except where
otherwise stated.)

PLATE 72

EDWARDS and BENEDETTO, Haskinsia sagittate!

606

PALAEONTOLOGY, VOLUME 28

the presence of longitudinal grooves between leaf bases, although there are differences in the
orientation of the leaf bases. In contrast the much larger leafy lycopod axes recorded from the
overlying Lower Devonian Tadrart Formation (Lejal-Nicol 1975) are completely different and more
comparable to arborescent forms reported from later in the Devonian elsewhere (Chaloner and
Sheerin 1979).

Order protolepidodendrales

Family protolepidodendraceae Krausel and Weyland, 1949
Genus colpodexylon Banks, 1944

Type species. Colpodexylon deatsii Banks, 1944 from the Delaware River Flags, equivalent to the marine Upper
Ithaca or Enfield formation. Finger Lakes Stage, Senecan Series, lower Upper Devonian ( = Frasnian) of New
York State, U.S.A.

Colpodexylon cachiriense sp. nov.

Plate 73, fig. 9; Plate 74, figs. 1-9

Derivation of name. From the Rio Cachiri, after which the most northern outcrop of Devonian rocks in
Venezuela is named.

Holotype. LPB 16045 (PI. 74, figs. 1 and 2).

Locality. Roadside exposures on the road from the town. Villa del Rosario, to the Socuy River in the Colorado
valley, some 950 to 1 500 m NNW of the farm ‘Alemania’, c. 70 km west of Maracaibo, north-western Venezuela.

Horizon. Carbonaceous Unites at the base of the Campo Chico Formation, Devonian (?Givetian-Fresnian).

Diagnosis. Vegetative stems at least 10 cm long, 5-3 to 9-5 mm wide (x = 7-7 mm, n = 10) with
persistent leaves borne spirally or in pseudowhorls. Leaf bases not contiguous with 7-9 leaves per
gyre (9 being the commonest number); successive gyres also widely spaced. Stem surface between leaf
bases probably smooth. Divaricate leaves with broad, flat, or abaxially curved bases, three forked;
0-62-0-96 mm (x = 0-76 mm, n = 14) wide in basal unbranched region, trifurcating 4-0 to 7-5 mm
(x = 5-6 mm, n = 11) from attachment. The central fork (0-5-0-72 mm) is approximately twice as
wide as the laterals (0-26-0-3 mm) and at least twice as long. All tips acuminate. Leaves at least
14-7 mm long.

EXPLANATION OF PLATE 73

Figs. 1-8. Haskinsia sagittata sp. nov. 1, LBD 16050, short length of stem with coalified covering, prominent
ridges, and leaf bases with sedimentary infill (text-fig. 3d, e), x 3. 2-6, LBD16046. 2, part of lower surface
of rock bearing holotype showing predominantly impression fossils, unilateral illumination. Lustre surface
produces brighter areas. Note range in form of projection marking bases of leaves attached to upper surface,
x 2-4. 3, area of left-hand stem in fig. 2 magnified to show detail of such projections and the infilled bases
of leaves (arrowed) attached on lower surface of stem, x 6. 4, two stems on same surface as holotype: latex
cast made from one on right, unilateral illumination, x 31. 5, part of stem on same surface as holotype,
lighter areas comprise fine-grained sedimentary infill; darker areas comprise granular coalified material.
Inverted Us mark sites of attachment of leaves on lower surface, x9-2. 6, oval area marking attachment
of leaf passing down into matrix; powdery coalified material has been removed to reveal sediment filling
base of petiole, x 20. 7, scanning electron micrograph of indeterminate spores recovered on surface of latex
cast (stub no. 620), x 650. 8, scanning electron micrograph of latex cast of stem impression illustrated in fig. 4
(stub no. 628), x 16.

Fig. 9, Colpodexylon cachiriense sp. nov. LBD16052(i), coalified compression fossil with very fine-grained
sedimentary infill (s) sometimes iron-stained on surface (r) and near base an uncovered trifurcating leaf
extending from a crescentic depression on lower surface. Arrow indicates possible vascular strand, x 29.
(Photographs under uniform illumination except where otherwise stated.)

PLATE 73

EDWARDS and BENEDETTO, Haskinsia , Colpodexylon

608

PALAEONTOLOGY, VOLUME 28

Description. The diagnosis is based on short lengths of unbranched stems showing a variety of preservation
forms united in the possession of trifurcating leaves. Apart from two isolated impressions, all the leaves were
uncovered. As they are small in number and generally incomplete or distorted, we anticipate that the dimensions
recorded in the diagnosis will be emended on the discovery of further specimens.

Stem characteristics: The parallel-sided stems are almost featureless except for the presence of leaf bases or
attachment sites marked by inverted crescentic (PI. 74, tig. 7) or more strongly arched lines (PI. 74, fig. 1)
enclosing depressions and, less frequently, by inverted crescentic ridges. The persistent leaves vary in their
arrangement from pseudowhorled to a more steeply pitched spiral condition, but in all cases successive gyres are
widely spaced, as are individual leaves in a gyre. The number of leaves per gyre ranges from 7 in the narrower
stems to 9 or possibly more in the wider, uncertainty arising from the fact that usually only one surface of a stem
is preserved.

Some evidence of a vascular strand comes from a stem 9-5 mm wide, with a longitudinal strip of sedimentary
infill and coalified material 1 mm wide, bearing coarse longitudinal striations (arrowed in PI. 73, fig. 9). The
coalified material was macerated in Schulze's solution, but no tracheids were recovered. Latex replicas taken
from rock surfaces show faint impressions of cells, but failed to reveal further detail under the SEM.

Leaf characteristics: The most complete leaves are long, slender, and trifurcating, showing no distinction
between petiole and blade. (PI. 74, figs. 2, 4, 5). The proximal unbranched region is parallel-sided and strongly
flattened in the plane of the trifurcation, widening slightly before branching. The crescentic attachment sites and
bases of leaves uncovered below such sites (PI. 74, fig. 8) suggest that the basal region was abaxially curved, but
this is not apparent on the leaves attached to stems fortuitously fractured transversely (PI. 74, fig. 7). Nor do they
show the broad longitudinal grooves visible on the proximal regions of some of the more complete leaves
(PI. 74, figs. 2 and 3).

Almost all the branching leaves were uncovered. The lateral segments are usually slightly more than half the
width of the central one, the three separated basally by U-shaped embayments of sediment where all are
preserved in the same plane. All three taper, the lateral ones being much shorter than the central. We doubt that
any of the extreme tips remained intact after uncovering as the sediment grains were wider than the distal parts of
the leaves, which were very easily dislodged. Vague outlines of cells are visible as impressions of the adaxial
surface near the base, but no further anatomical details of the leaf were obtained. The leaves were persistent, but
in the majority of cases only the divaricate basal parts are preserved showing tapering (PI. 74, fig. 7), truncated
(PI. 74, fig. 9), or more rarely, hooked tips (PI. 74, fig. 3). Such incompleteness suggests that the branches were
more delicate and also explains the change in orientation sometimes observed between the two regions. Whether
or not the entire leaf in life was a stiff spreading organ with all parts in the same plane, or was more flexuous, as
seen on the holotype, remains conjectural.

All the leaves were sterile. However, we noted two isolated possible sporangia preserved near typical axes. One
is elliptical and featureless; the other, 2-5 mm long and 1 -8 mm wide, has a notch at one end which is extended
into a central longitudinal depression (PI. 74, fig. 6). Neither structure yielded spores and we have no evidence,
apart from association, that they belonged to C. cachiriense.

Comparisons. The diagnosis of Colpodexylon is based on sterile and fertile leafy stems from the
Devonian of New York State (Banks 1944). Most are compression fossils, but some have pyritized

EXPLANATION TO PLATE 74

Figs. 1-9. Colpodexylon cachiriense sp. nov. 1, LBD 16045, holotype, impression fossil, unilateral light showing
attachment sites of leaves borne on the lower surface of the stem, x 1-8. 2, enlargement of part of holotype to
show the single attached trifurcating leaf, x 3-6. 3, LBD16048, impression fossil with incomplete attached
leaves with hooked tips, unilateral illumination, x 3-3. 4, LBD16049, iron-stained impression of leaf with
traces of coalified material, central segment of trifurcation missing, x 4. 5-8, LBD160148. 5, detached leaf
with incomplete tips, x 3-7. 6, possible isolated sporangium, x 10. 7, coalified stem with leaf bases in
profile. Arrow indicates another stem fractured transversely showing abaxial or adaxial surfaces of leaf bases,
x 2. 8, distal part of stem figured in fig. 7, preserved as impression fossil. Leaf scars have been developed
to expose adaxial surfaces of two leaves (arrowed) originally projecting down into the matrix below the stem.
Isolated leaf shows two of the three distal segments and typical curvature just below the trifurcation, x 7-4.
9, LBD 16045, stem impression on same surface as holotype. Incomplete leaves show fracture at point of
attachment (arrowed), x 5. (Photographs under uniform illumination except where otherwise stated.)

PLATE 74

EDWARDS and BENEDETTO, Colpodexylon cachiriense

610

PALAEONTOLOGY, VOLUME 28

anatomy. The Venezuelan plants, although far more fragmentary and lacking both reproductive and
anatomical details, are assignable to Colpodexylon because they possess persistent trifurcating leaves
which Grierson and Banks (1963) considered The chief diagnostic character’ (p. 253) of the genus.
Banks originally distinguished two species, the type C. deatsii from Frasnian strata and the
anatomically less completely known C. trifurcation from the Eifelian. The two are readily separated
from the South American plant in that their stems bear a greater number of leaves per gyre with leaf
bases both laterally and vertically contiguous. The preservation of the Venezuelan plants is such that
usually only the inside of leaf bases are visible, but these are so widely spaced that considerable areas
of stem surface must have been exposed between leaf bases. There are however certain similarities in
the structure of the leaves. When describing the basal segment of the leaf of C. deatsii , Banks (1944)
noted that it was necessary to remove matrix from ‘alongside the leaf margin in order to expose the
true widths’ (p. 654) and concluded that this was evidence for a slightly revolute margin. He also
recorded two longitudinal furrows on the leaf surface. In describing isolated sporangia he mentioned
that a few showed an apical notch sometimes extended into a median line. These were not illustrated
but may be similar to the isolated structure we figure in Plate 74, fig. 6.

Lemoigne and Ishchenko (1980) described a putative third species from the Famennian of the
Ukraine. In the leaves, only the lateral branches beyond the trifurcation are visible (a similar
preservation to that illustrated in PI. 74, fig. 4), the authors suggesting that the central one is present
in a different plane. The stem surface looks quite different from the other species including C.
cachiriense because each leaf base has an adaxial depression. The leaves are considered ligulate.

Of the herbaceous lycopods reported from South America the only taxon with distinctly divided
leaves is Protolepidodendron kcge// (Krausel and Dolianiti 1957) from the Maranhao Basin of Brazil.
The leaves as illustrated differ from those in C. cachiriense in that they resemble tuning forks, being at
least 1 5 mm long with a bifurcation about 2 mm from the base, although there is always the possibility
that further uncovering will show them to be more complexly branched.

TAPHONOMY

Several recent investigations demonstrate that both the type of preservation and subsequent planes of
fracture are responsible for the diversity of appearance in the stems of herbaceous and arborescent
lycopods (Chaloner and Collinson 1975; Thomas and Purdy 1982; Grierson and Banks 1983; Rex
and Chaloner 1983). Such studies greatly facilitate interpretations and descriptions of our fossils and,
in particular, the leaf bases of H. sagittata. Indeed, the leafy specimen illustrated in Plate 71, fig. 4
allows a leaf lamina cleaved compression plane to be added to the fracture plane terminology
summarized by Grierson and Banks (1983). However it has sometimes been difficult to apply their
terminology, based mainly on coalified compression fossils, to the heavily weathered Venezuelan
examples, where at best the original stem is replaced by a thin layer of sedimentary infill coated by a
film of coalified material, and where counterparts are missing. Our fossils exhibit a range of
appearances and preservation states (compare, for example, PI. 71, figs. 1-5) and the entombing
sediments vary in colour and hardness, a reflection of the degree of weathering. Some of the more
perplexing structures occur in those fossils where little or no organic material remains. The range of
types of preservation present permits the poorly preserved forms to be interpreted by reference to the
more complete and well-preserved stems. Such an assessment of different preservation types is
necessary as the identification and description of lycopods from the Devonian of South America,
South Africa, and Antarctica have been based on similar fragmentary impression fossils.

Types of preservation

a. Fossils with layer of sediment sandwiched between two coalilied layers that presumably represent the outer
tissues of the stem. The amounts of coalified material, sedimentary infill, and compression vary between
specimens.

H. sagittata. The least compressed specimen (PI. 71, fig. 2) lacks the diagnostic leaf blades, but the shape,
orientation, and arrangement (7 per gyre) of leaf bases and its size permit assignation to H. sagittata. The

EDWARDS AND BENEDETTO: DEVONIAN LYCOPODS

611

coalified stem, a cleavage compression sensu Chaloner and Collmson, exposed by a fracture plane running
between the upper surface and the matrix is present only at the base of the specimen. A fractured example of the
poorly defined elongate projections, the bases of the petioles, show them to be filled with sediment. Leaf
arrangement is much clearer where the fossil has dropped out revealing an impression of its lower surface,
a cleavage impression. Fine striations are visible under a dissecting microscope.

The stem illustrated in Plate 71, figs. 3 and 5 is similar in composition, but looks quite different as a result of
greater compression and irregular fracture removing parts of the upper surface. It comprises a papery layer of
very fine-grained whitish sediment(s), stained orange on the outside, and sandwiched between two sporadic
powdery coaly layers. Both impression and infill have areas of indistinct cell outlines. Removal of the rock distal
to a depression reveals the characteristically shaped leaf of H. sagittata (arrowed in PI. 71, fig. 3). In some cases
the depressions are filled with sediment and some additional coalified material producing mounds with irregular
topography, but in addition to these, smooth mounds of relatively low relief may be seen on the lower impression
surface as well as less regular ones on the sedimentary infill (see PI . 71, fig. 4 where the stem is illuminated
unilaterally). The latter are the incomplete leaf base infills of the leaves attached to the upper surface (i.e. that
closer to the observer) and the former show that the stem has been so compressed that its lower surface is pressed
into the bases of these leaves producing these ‘false leaf bases' (text-fig. 4 <?, /). Their positions indicate that they
do not represent the inward collapse of the stem between leaf bases on the lower surface.

C. cachiriense. The most complete fossils are of this preservation form. The unbranched stem (40 mm long) in
PI. 73, fig. 9, has a variable appearance due to uneven fracture. The extremely fine-grained infill is silvery
internally (s) but iron-stained on the surface (r). It lies above a thick coalified layer, but a similar layer
representing the organic remains of the stem closer to the observer is rarely preserved above the sedimentary
infill. The impression fossil on the rock below is also iron-stained and while some areas bear faint traces of cells,
others are distinctly warty. The sediment infill does not extend to the edges of the stem which are marked by a
strip of coaly material of variable width (0-3-0 4 mm) although this has sometimes broken. The bases of leaves
attached to the sides of the stem, represented by tapering, needle-like projections before developing, are heavily
coalified compressions continuous with this marginal band. The attachment sites of the widely spaced superficial
leaves are not conspicuous. On the lower impression surface they are shallow depressions limited at one end
by a crescentic coalified line. The depression is sometimes filled with sediment and additional coalified material.
On the cleaved sedimentary layer their positions are marked by short transverse furrows. Attachment sites
on the upper surface of the infill are marked by crescentic ridges, abruptly truncated on the convex edge and
tapering gradually on the other. They are most conspicuous where the surface stain has rubbed off.

Another example (PI. 74, fig. 7) has a more untidy appearance because the layers of coalified material and
sediment are not continuous over wide areas. They are easily dislodged; the impression below has been
developed (PI. 74, fig. 8) to demonstrate the broad abaxial curved junctions between leaves and stem.

b. Strongly compressed fossils comprising a granular mixture of inorganic and coalified material, only rarely
showing distinct layering.

H. sagittata. The holotype and about twenty other axes occurring at different levels in the rock, but most
orientated in the same direction, show this preservation type. The matrix was the least weathered of all the slabs
examined. The specimen provided an opportunity to study part and counterpart as during preparation one stem
was split. Its plane of fracture was irregular but more fossil remained on one surface (the part) than the other
(counterpart) thus paralleling the situation on the two major faces of the rock. The surface equivalent to the part
bears a number of stems of variable colour (ranging from grey to brown to rust) and appearance. Depending on
the plane of fracture and amount of weathering or damage since collection, granular sediment, sporadic coaly or
rusty material or stained rock below may be exposed. In some areas the impression beneath has a metallic sheen,
the entombing sediment being much finer grained than the general matrix and displaying longitudinally aligned
evenly spaced depressions presumably representing the outlines of epidermal cells. SEM examination of latex
casts of such stems failed to show the precise shape of individual cells (PI. 73, fig. 8). The predominantly
impressions on the reverse of the block (equivalent to the counterpart) are similar in appearance ranging from
those with a striking gold-lustre surface (PI. 73, fig. 2) to those stained brown and visible only in unilateral
lighting.

Before developing only a few leaves were visible in profile, but removal of sediment above and adjacent to
stems showed that they were ensheathed in leaves, slightly smaller than the mean, but identical in shape (see base
of holotype; PI. 72, figs. 1, 2 and PI. 71, fig. 8).

The most conspicuous macroscopic features of these axes, enhanced by unilateral illumination, are the
depressions and elevations, the latter sometimes standing over a millimetre above the stem surface marking the
positions of leaves. Here again the positions of the bases of leaves of both sides of the stem may be visible on one

612

PALAEONTOLOGY. VOLUME 28

text-fig. 2. Outlines of uncovered leaves showing range in
shape and size, drawn from specimens LPB 16046. 16047, 16049,
16052, and 16054. All x 24.

exposed surface. The depressions marking the leaves passing down into the matrix are most clearly defined and,
on the parts, have three different appearances. In the complete absence of the compression fossil they appear as
inverted Us (PI. 73, fig. 5), often delimited by a line of coalified material (the remains of the peripheral tissues of
the leaf base), enclosing a shallow depressed area which sometimes bears a central, longitudinally aligned fold
(PI. 71, fig. 8; text-fig. 2). More usually the attachment site is marked by a circular, elliptical, or ovate area of fine-
grained sediment frequently covered by a thin film of coalified matter (PI. 73, figs. 3, 4, 6). Although this may
represent the peripheral tissues of the side of the stem closer to the observer, in a few rare cases, it is further
obscured by the infill layer. The plug of sediment may be flat and almost flush with the stem surface or possess a
central, barely visible, fold. Finally, less regular mounds of coarser sediment mark the leaf bases.

The elevations are more variable and may be truncated with unevenly contoured tips or more usually are
smooth attenuated mounds (PI. 73, fig. 4). Little sediment or coalified material persists on the flanks of the
projections, but occurs sporadically on the surface of the stem particularly in the vicinity of the margins.

The projections are far more distinct on the counterparts accentuated to some extent by the lack of
compression fossil. The tips of the pegs are either rounded, irregularly fractured, or have a central crater-like
depression surrounded by a rim (PI. 73, figs. 2 and 3). Depressions or circular to elliptical areas of sediment
(sometimes covered by a thin dark layer) mark the attachment ofleaves beneath the stem. These may alternate
regularly with the pegs or almost coincide with them (PI. 73, fig. 3).

A few of the stems (e.g. the holotype) bear prominent elongate ridges or depressions of varying lengths (PI. 72,
fig. 2; PI. 73, figs. 1 and 4). On the specimens equivalent to parts, longitudinally aligned ridges occur between, but
never traverse, the leaf base depressions and are sometimes interrupted by the mounds. Between the vertically
superimposed depressions they are parallel, 06 to 0-9 mm apart, but may widen slightly to enclose the

EDWARDS AND BENEDETTO: DEVONIAN LYCOPODS

613

depressions. Where the body fossil is missing on this surface, ridges also occur on the impression below. A latex
replica of such a surface, a cast of the stem (PI. 73, fig. 8), suggests that the leaves of a single orthostichy were
borne on a broad probably slightly raised band isolated from each adjacent one by a groove. On the stems
( = counterparts) exposed on the reverse of the slab, longitudinally orientated grooves pass between the
projections and across some of the depressions, but there are no ridges.

c. Coalified compressions

One short length of H. sagittate! stem (PI. 73, tig. 1 ) comprises an almost continuous thin layer of coaly material
interrupted by approximately circular areas of sediments, thought to be trapped within the bases of leaves
attached to the lower surface. Also present are longitudinal ridges separating the leaf bases, and occasional
indistinct mounds on the surface of the coalified layer. The latter may mark the portions of the leaves attached to
the upper surface of the stem, i.e. that closer to the observer. A very thin film of coaly material has been noted on
two of the circular areas of sediment. This may represent the remains of the tissues of the upper surface of the
stem. A drawing of a transverse section through this stem is given in text-fig. 3d. Unfortunately its counterpart
is missing and no further information was gained on removing the sediment overlying one end.

In most fossils of this type the coalified layer is more powdery and rarely continuous, frequently revealing an
iron-stained impression on the rock beneath. Such a stem, figured on the left in Plate 72, fig. 13, is typical in that
the attachment sites of the leaves on the lower surface are marked by shallow depressions and on the upper by
elongate smooth mounds. Leaf bases in a single orthostichy are widely spaced. The stem surface bears numerous
faint longitudinal ridges and grooves that are not continuous over any great distance. The stem on the right
looks quite different: it has more pronounced relief and less coaly material. The leaves in profile are more
numerous. Although approximately the same width, it has at least nine leaves per gyre and successive gyres are
more closely spaced.

The most striking fossils in the assemblage are those where the coalified material is replaced by a granular
reddish brown layer.

d. Impressions. Many of the fossils available for study were of this type. Some may have been produced by
tropical weathering of originally rather fragmentary coalified compressions or coalified material plus
sedimentary infill, but others may be artifacts of collecting in that only the counterpart was originally retained
and the body fossil may have been dislodged and lost on subsequent transport.

H. sagittatev. The weathered specimens obviously intergrade with some of the coalified compressions just
described (PI. 72, fig. 13). All occur in a very soft, buff to pink matrix, one most useful in deducing leaf
morphology by uncovering. The stems illustrated in Plate 72, figs. 3 and 4 are pink iron-stained impressions with
traces of black powder associated with leaf bases. Also present on parts of the stem and certain leaves is a brown
film which may represent the remains of a cuticle, but this lacks cellular imprints.

The leafy stem figured in Plate 71, fig. 1 superficially resembles the cleavage compression illustrated by
Grierson and Banks ( 1 983, fig. 4). The most obvious common features are the circular and oval areas of sediment
(0-96 mm long and 0-83 mm wide) marking the attachment sites of leaves. In the Venezuelan fossils, just above
and slightly to one side of each leaf base is a slight mound, a false leaf base, marking the position of the leaf on the
surface nearer the observer. The stems have a high relief but very little organic material persists. Some of this is in
the form of a thin layer of cuticle with traces of cellular markings and more prominent widely spaced
longitudinally orientated dark lines delimiting orthostichies. These are sometimes replaced by less regular ridges
or furrows. Film pulls showed no further cellular detail.

C. cachiriense: The holotype (PI. 74, figs. 1 and 2) is predominantly an impression fossil, stained dull red towards
one end; leaves are preserved as compressions. Their attachment sites are marked by shallow inverted U-shaped
depressions containing oval or eliptical mounds of fine-grained sediment plus some coalified material. The
curves of the U extend as shallow ridges, almost to the level of the gyre below, giving the striated appearance best
seen in unilateral light. The strongly curved junction between leaf base and stem is represented by a pronounced
strip of coalified material much greater in thickness than the coalified film, presumed to be the remains of the
outer tissues of the stem, exposed on removing the sedimentary infill. Some very low mbunds perhaps marking
the sites of leaves on the upper surface are just visible between the depressions on the surface of the impression
towards its base, but are sometimes superimposed on the sedimentary infill distally.

A representative of the commonest preservation type of C. cachiriense is shown in Plate 74, fig. 3. Its surface
has a yellow, silky, almost bone-like, texture. Excavations around the ends of similar small fragments indicate
that originally present was a crumbling mixture of small irregular sheets of coalified material and sedimentary
infill. Leaf attachment sites on the impression resemble those on the holotype, but when sediment and coalified
material are removed, the surface of the stem impression is also convex, suggesting that the stem had collapsed

614

PALAEONTOLOGY, VOLUME 28

inwards just below a leaf. That this was a topographic feature of the stem in life is considered unlikely: similar but
more pronounced examples are seen on stems showing distortion from twisting and stretching.

Preservational history. The sequence of events that produced these preservation forms is broadly
similar for the stems of both species. Biodegradation of internal tissues of stems and leaves would
have resulted in collapse of these organs, but as emphasized by Rex and Chaloner ( 1 983) exactly when
this occurred in relation to burial, possible infiltration by sediment and then further compaction
under a mass of sediment, is difficult to determine. The situation is further complicated by the
differing rates of decay of soft and hard (strengthening) tissues. Here major problems in
interpretation relate to the resilient leaf bases and other strengthening tissues in Haskinsia and to the
compression of the flat divaricate leaves of Colpodexylon.

Impressions of epidermal cells in the extremely fine-grained matrix immediately around some of
the fossils suggest that an iron-rich layer quickly encrusted the plant surface after immersion (Spicer
1977). Decay of soft tissues continued during immersion and subsequent burial, but as the stems
began to collapse the bases of leaves retained their shape because of the presence of decay-delaying
peripheral strengthening tissues. In some instances sediment began to accumulate within the
increasingly hollow stems and may well have become trapped in the leaf bases early on, thus further
reinforcing them (text-fig. 3). As overlying sediment increasingly compressed the stems perhaps
squeezing more grains into the leaf bases, a cast of the inside of the stem surrounded by a very thin
layer of disintegrating tissues would have resulted.

Further compression of the infill would have resulted in considerable distortion (Rex and Chaloner
1983) with parts of stem and emtombing matrix pressed into leaf bases (text-fig. 3c) thus producing
the false leaf bases already noted (PI. 71, fig. 5). A similarly undulating cross-section would be seen in
fossils lacking any sedimentary infill (e.g. PI. 72, figs. 3, 4, 13). As is often the case features of one
surface appear less compressed than the other (text-fig. 3).

Rex and Chaloner (1983) investigating the behaviour of spines in a plant like Sawdonia during
compression used a hollow cylindrical foam rubber model, which in section looks remarkably like

text-fig. 3 a-c. Diagrams showing postulated
behaviour of stem of H. sagittata on compression
and subsequent fracture. Coarse stippling repre-
sents matrix, fine stippling sedimentary infill. For
ease of drawing a whorled arrangement of leaves
is shown with those at the margin both directed
downwards (a). The fossil above the fracture is
equivalent to the counterpart, that below, the part.

b, fossil with well-defined coalified layers and
sedimentary infill with fracture plane passing be-
tween matrix and upper surface of stem and across
(b') or around sediment infill of leaf bases ( b

c, highly compressed fossil with fracture plane
passing over upper surface and across leaf bases
(<:')• c" shows a weathered example with surface
of matrix below the compression exposed. In
even more compressed forms, sometimes lacking
infill, fracture is less regular across stem surface.

d, e , postulated section through the stem illustrated
in Plate 73, fig. I where fracture occurs across
?upper surface of coalified compression and across
bases of leaves attached to lower surface, e is en-
larged to show detail of the longitudinal ridges.

EDWARDS AND BENEDETTO: DEVONIAN LYCOPODS

615

our steins. On compression the ‘spines’ on the upper surface became much shorter, those on the lower
surface shortened but less so. Thus the rock surface with stems showing the most pronounced
depressions and the most compression material (the parts) was probably the original surface of the
sediment.

The longitudinal ridges and grooves on some of the preservation forms are more difficult to
interpret. The presence of dark lines in similar positions on certain stems (PI. 71, fig. 1) suggest that
they were indeed regions of thick-walled strengthening tissues equivalent to those described for
H. colophylla by Grierson and Banks (1983) who considered them hypodermal In many of the
Venezuelan examples (type 6, PI. 71, fig. 8, PI. 72, fig. 2; typec, PI. 73, fig. 1) they seem to be part of the
stem surface, where they are associated with the leaf bases on the lower surface of the stem. They thus
appear as ridges on the impression fossil below that surface, ridges on the body fossil (part), and
grooves on the impression counterpart. The latex cast of the lower surface of such a stem (PI. 73,
fig. 8) shows narrow grooves separating broad ridges on which the vertical files of leaves are borne.

The sequence of events that produced such an arrangement remains problematic. Inwardly
projecting rods of more persistent hypodermal tissues could have produced longitudinal grooves
on the surface of the sediment filling the hollow (decomposed) centre of the stem in a sequence
analogous to the production of grooves on pith casts of Catamites (text-fig. 3 b). During further
diagenesis the organic matter may have almost disappeared so that the entombing matrix was pressed
against the surface of the cast. The resulting impression in the surrounding rock would have
longitudinal ridges but these have nothing to do with the original surface of the plant (also seen in
Catamites). However, in these Venezuelan examples, as there is little or no sedimentary infill, it seems
likely that the longitudinal strands themselves projected into the matrix and subsequently became
coalified. While preferential preservation may account for their presence on one side only of the stem,
it is also possible that the positions of the strands on the two surfaces were coincident on
compression, thus reinforcing each other to produce the observed feature. Indeed on stems which are
twisted or where leaf bases on the two surfaces do not regularly alternate, the longitudinal features, if
present, are more closely spaced, cannot be traced over long distances, and occasionally occur as
furrows as well as ridges on the same surface.

Also puzzling is the presence of a raised area at the centre of certain depressions marking leaf bases
(text-fig. 2). Those illustrated in Plate 71, fig. 8 occur on a stem (type b preservation) adjacent to the
holotype: removal of the stem distal to the leaf attachment sites has revealed the leaves themselves.
The raised area on the left is elongate, that on the right more circular. The anatomical feature
responsible for such a topography remains unknown.

Considering Colpodexylon , different appearances of the leaves may be related to the behaviour on
compression of dorsoventrally flattened structures with extended horizontal connections with the
stem. Seen sideways-on, such leaves appear as linear structures with slightly expanded bases
(horizontal stem on PI. 74, fig. 7) produced by the lateral compression of the unbranched proximal
regions of the leaf. However, we encountered some such specimens in which removal of matrix
immediately adjacent to the straight edge, believed to represent the abaxial limits of the leaf, revealed
a further strip of more delicate coalified material below. This change in level and thickness suggests
that the originally gently abaxially curved leaf had been twisted slightly on compression so that part
of the dorsal (abaxial) surface has become visible at the lower level. A similar explanation fits the
zoned appearance of certain quite broad leaf bases in which the adaxial quarter or one third is more
heavily coalified or stained than the remainder below. Twisting or fracture of the extended flat or
crescentic junction may account for the constrictions sometimes visible at leaf junctions (PI. 74,
fig. 9). As a result of twisting in the basal region of the leaf attached to the holotype (PI. 74, fig. 2) the
abaxial or adaxial surface is preserved in the same plane as the flattened stem.

PALAEO BIOGEOGRAPHY

Lack of information from southern continents has, more than any other factor, frustrated attempts
to detect provincialism in Devonian floras (Edwards 1973; Edwards and Fanning 1985). The situation

616

PALAEONTOLOGY, VOLUME 28

in South America is typical, the plants (e.g. Haplostigma and Palaeostigma ) are often poorly
understood and independent dating of assemblages absent. Even for this relatively well-preserved
assemblage from Venezuela the age determination is equivocal in that animal fossils are lacking in the
beds containing the plants (Co 8 horizon in the Campo Chico Formation): an assemblage of
brachiopods and pelecypods (Co 6 horizon) suggests a Givetian age (probably middle to late
Givetian: Benedetto 1980, 1984) for the Co 6 horizon some 50 m below. Brachiopods and pelecypods
from 300 m above the plant horizons are late Mississippian-early Pennsylvanian. Bowen (1972)
mentions an assemblage of Middle to Upper Devonian palynomorphs (presumably Frasnian) from
the upper part of the Campo Chico Formation but we have been unable to correlate this horizon with
the plant beds. We have failed to isolate a spore assemblage from the matrix and spores found
adpressed to several of the axes are not sufficiently well preserved to permit an age determination
(PI. 73, fig. 7). The plants described here support a Givetian-Frasnian age (Grierson and Banks
1963), although C. trifurcation and Haskinsia colophylla are also recorded from Eifelian sediments in
New York State. Of the assemblages from Argentina originally described by Frenguelli (1954) and
reinvestigated by Menendez (1965a, b ) only one, that containing Haplostigma furquei, is indepen-
dently dated and considered Middle Devonian from associated trilobites. The far more diverse
assemblage from the Quebrada de la Chavela including Drepanophycus eximius (Menendez 19656)
and Archaeosigillaria confer t a (Menendez 1 965a) as well as Furqueia angladae , Charnelia dichotoma,
Hyenia argentina , and Adiantites devonica was originally thought to be Devonian but Cuerda et al.
(1968) consider it younger (Carboniferous) after re-examination and re-identification of some of
the plants.

Considering Brazil the assemblage from the Upper Porta Grossa Beds with a number of species of
Spongiophyton and Haplostigma (Krausel 1960) was originally described as Lower Devonian.
Chaloner et al. (1974) reviewed recent palynological and micropalaeontological (acritarchs and
chitinozoans) studies and concluded that the sediments are probably Givetian. Associated palyno-
morphs suggest that the lycopod assemblage ( Palaeostigma sewardi, Protolepidodendron kegeli , and
Archaeosigillaria picosensis ) in the Picos Formation of the Maranhao Basin (Krausel and Dolianiti
1957) is Middle Devonian (Bar and Riegel 1974) although age determinations of other workers range
from Emsian to Frasnian (Sampaio and Northfleet 1973; Brito 1967). Finally, strata on the Falkland
Islands containing H. irregularis are considered Middle Devonian on the basis of the plants them-
selves (Seward and Walton 1923). These are being reinvestigated by Dr. Sergio Archangelsky.

Thus on current evidence, vascular plant assemblages from South America are of Middle
Devonian age, and are dominated by lycopod genera, most of which are known from the northern
hemisphere. The recent taxonomic and stratigraphic review of African lycopods (Anderson and
Anderson (in press)) includes some revision of American lycopods, which reinforces similarities
between the floras of the two southern continents. Further support for a uniform world-wide flora in
mid-Devonian times comes from the description of Leclercqia complexa from Queensland (Fairon-
Demaret 1974), and late Devonian sediments of New South Wales contain species of the northern
hemisphere genera Leptophloeum , Barinophyton , and Archaeopteris (Gould 1975).

Our studies provide evidence for a uniform vegetation between a part of north Gondwana and the
Old Red Continent (Laurentia) in mid- to late-Devonian times, but in isolation do not provide
compelling evidence for global uniformity during that period. The similarities may simply reflect the
palaeogeographic proximity of the localities or their occurrence in the same climatic zone
encompassing more than one palaeocontinent. As far as is known the Venezuelan plants were
homosporous with the potential for colonizing very wide areas (Chaloner and Sheerin 1 979) and thus
provide inconclusive evidence for the past distribution of continents. Many Devonian palaeo-
geographic reconstructions show an east-west (Rheic) ocean separating north and southern land
masses (see e.g. Smith et al. (1981) where continent positions are based on palaeomagnetic data;
Scotese et al. (1979); and Scotese (1984) late Devonian (B) map). Some recent reconstructions have
closed or narrowed this ocean. That of Heckel and Witzke (1979), based on the distribution of
palaeoclimatically significant rock types, e.g. carbonates, show northern parts of Africa and South
America close to Laurentia (Old Red Continent), while Scotese (1984) presents an alternative

EDWARDS AND BENEDETTO: DEVONIAN LYCOPODS

617

reconstruction for the late Devonian (map A) based on new palaeomagnetic data which positions
Venezuela closer to north-east America, but Australia is at a much higher latitude and a considerable
distance from the Old Red Continent. Such a juxtapositioning of Venezuela and eastern North
America appears to complement our conclusions based on floras and receives further support from
similarities in marine faunas comprising brachiopods, pelecypods, and corals ( Benedetto 1 980, 1 984).

Finally, considering climatic zones, Barrett (in press) has predicted global climates for four
reconstructions of continents in early- to mid-Devonian times including those of Smith et al. (1981)
and Scotese et al. (1979). In all the models, he finds that northern South America (Amazon-
Colombian) is likely to have been cool, the land vegetation growing under wet temperate conditions
with eastern North America (Appalachian Basin) at lower latitudes in a tropical to subtropical wet
climate. This suggests that these early pteridophytes with small or divided leaves and an abundance of
strengthening tissues may have been tolerant of a range of climates and thus may provide an
explanation for the apparent uniformity of vegetation in Devonian times. Alternatively it may merely
indicate that such plants possessed a higher fossilization potential.

Acknowledgements. We thank Dr. Patrick Rachebouef for his assistance and advice throughout the Venezuelan
project. Dr. J. B. Richardson for palynological help, Ms. Jan Cawley for photographic expertise, and Dr.
S. Barrett for access to unpublished data.

REFERENCES

anderson, J. and anderson, H. (in press). Prodromus of South African Megafloras , Devonian to Lower
Cretaceous , 400 pp. A. A. Balkema, Rotterdam.

banks, h. p. 1944. A new Devonian lycopod genus from southern New York. Am. J1 Bot. 31 , 649-659.

- 1980. Floral assemblage zones in the Siluro-Devonian. In dilcher, d. and taylor, t. n. (eds.). Bio-
stratigraphy of fossil plants: successional and pa/eoecological analysis , 1-24. Dowden, Hutchinson and Ross,
Stroudsburg, Pennsylvania.

bar, p. and riegel, w. 1974. Microfloras of the Palaeozoic of Ghana and their palaeofloristic relations. Sci.
Geol. Bull. 27 , 39-58.

barrett, s. f. (in press). Early Devonian continental positions and climate: a framework for paleophyto-
geography. In tiffney, b. (ed.). Geologic factors and the evolution of plants. Yale University Press.
benedetto, j.-l. 1980. Sintesis biostratigrafica del Paleozoico tardio de la Sierra de Perija, Venezuela. An.
Acad. Brasil Cienc. 52 , 827-839.

- 1984. Les Brachiopodes devonicus de la Sierra de Perija (Venezuela). Biostratigraphie du Paleozoique
(Brest), 1, 1-191.

boucot, a. j. and gray, j. 1982. Geologic correlates of early land plant evolution. Proc. Third N. Am. Paleontol.
Conv. 1, 61-66.

bowen, j. m. 1972. Estratigrafia del Precretaceo en la parte norte de la Sierra de Perija. Mem. V. Congr. Geol.
Venezuela , 2 , 729-761.

brito, i. m. 1967. Silurian and Devonian Acritarcha from Maranhao basin, Brazil. Micropalaeontology , 13 ,
473-482.

chaloner, w. g. and collinson, m. e. 1975. Application of SEM to a sigillarian impression fossil. Rev.
Palaeobot. Palynol. 20 , 85-101.

- mensah, m. k. and crane, M. D. 1974. Non-vascular land plants from the Devonian of Ghana.
Palaeontology, 17 , 925-947.

- and sheerin, A. 1979. Devonian macrofloras. In house, m. r., scrutton, c. t. and bassett, m. g. (eds.).
The Devonian System. Spec. Pap. Palaeont. 23, 145-161.

cuerda, a. j., wagner, r. h. and arrondo, o. g. 1968. Observationes sombre algunas floras del Carbonifero
Argentino. Ameghiniana, 5 , 265-269.

Edwards, d. 1973. Devonian floras. In hallam, a. (ed.). Atlas of Palaeobiogeography, 105-115. Elsevier,
Amsterdam.

- and fanning, u. 1985. Evolution and environment in the late Silurian-early Devonian: the rise of the
pteridophytes. Phil. Trans. R. Soc. Land. B, 309 , 147 165.

fairon-demaret, m. 1974. Nouveaux specimens du genre Leclercqia Banks, H. P., Bonamo P. M. et Grier-
son, J. D. 1972, du Givetien (?) du Queensland (Australie). Bull Inst. r. Sci. nat. Belg. 50, I 4.

618

PALAEONTOLOGY, VOLUME 28

fairon-demaret, m. and banks, h, p. 1978. Leaves of Archaeosigillaria vanuxemii, a Devonian lycopod from
New York. Am. Jl Bot. 65, 246-249.

frenguelli, j. 1954. Plantas Devonicas de la Quebrada de la Charnela en la Precordillera de San Juan. Not.
Mus. La Plata. 17, 359-376.

gbnsel, p. g. and Andrews, h. n. 1984. Plant life in the Devonian, 381 pp. Praeger, New York.

could, r. e. 1975. The succession of Australian pre-Tertiary megafossil floras. Bot. Rev. 41, 453-483.

Grierson, J. D. and banks, H. p. 1963. Lycopods of the Devonian of New York State. Palaeontogr. Am. 4,
220-295.

- 1983. A new genus of lycopods from the Devonian of New York State. Bot. J! Linn. Soc. 86, 81-101.

grindley, G. w., mildenhall, d. c. and schopf, j. m. 1980. A mid late Devonian flora from the Ruppert Coast,
Marie Byrd Land, West Antarctica. Jl Roy. Soc. New Zealand , 10, 271-285.

heckel, p. h. and witzke, b. j. 1979. Devonian world palaeography determined from distribution of carbonates
and related lithic palaeoclimatic indicators. In house, m. r., scrutton, c. t. and bassett, m. g. (eds.). The
Devonian System. Spec. Pap. Pa/aeont. 23, 99 123.

klitzsch, e., lejal-nicol, a. and massa d. 1973. Le Siluro-Devonien a psilophytes et lycophytes du bassin de
Mourzouk (Libye). C.r. Iiehd. Seanc. Acad. Sci., Paris. Serie D , 277, 2465-2467.

krausel, r. 1960. Spongiophyton nov. gen. (Thallophyta) e Haplostigma Seward (Pteridophyta) no Devoniano
Inferior do Parana. Depart. Nac. Prod. Miner., Div. Geol. Miner., Monogr. 15, 1-41.

and dolianiti, e. 1957. Restos vegetais das Camadas Picos, Devoniano Inferior do Piaui. Minest. Agr.
Dep. Nac. Prod. Miner. Div. Geol. Miner. Bull. 173, 7-19.

and weyland, w. 1949. Pflanzenreste aus dem Devon. XIV. Gilboaphyton und die Protolepidophytales.
Senckenbergiana, 30, 129-152.

leclercq, S. 1 960. Refendage d’une roche fossilifere et degagement de ses fossiles sous binoculaire. Senckenberg
leth. 41,483-487.

lejal-nicol, a. 1975. Sur une nouvelle flore a lycophytes du Devonien inferieur de la Libye. Palaeontographica ,
151 B, 52-96.

lemoigne, y. and Ishchenko, t. 1980. Deux lycophytes avec structures convervees du Devonien Superieur
d'Ukraine (U.R.S.S.). Geobios, 13, 671-681.

menendez, c. A. 1965a. Archaeosigillaria conferta (Frenguelli) nov. comb, del Devonico de la Quebrada de la
Chavela, San Juan. Ameghiniana , 4, 67-68.

- 19656. Drepanophycus eximius (Frenguelli) nov. comb, del Devonico de la Quebrada de la Chavela, San
Juan. Ibid. 139-140.

rex, G. m. and chaloner, w. g. 1983. The experimental formation of plant compression fossils. Palaeontology,
26, 231-252.

sampaio, a. v. and northfleet, a. 1973. Estratigrafia e correlacao das bacias sedimentares Brasileiras (abs.).
Congr. Bras. Geol. 27, 148.

scotese, c. r. 1984. Palaeozoic palaeomagnetism and the assembly of Pangaea. In van der voo, r., scotese,
c. r. and bonhommet, n. (eds.). Plate reconstruction from Palaeozoic palaeomagnetism. Geodynamics series,
12, 1-10. American Geophysical Union, Washington, D.C.

bambach, r. k., barton, c., van der voo, r. and ziegler, a. m. 1979. Paleozoic base maps. Jl geol. 87,
217-277.

seward, a. c. and walton, j. 1923. On fossil plants from the Falkland Islands. Q. J! geol. Soc. Lond. 79,
313-333.

smith, A. G., hurley, A. M. and briden, J. c. 1981. Plianerozoic palaeocontinental world maps, 102 pp. Cambridge
University Press, Cambridge.

Spicer, r. a. 1977. The pre-depositional formation of some leaf impressions. Palaeontology, 20, 908-912.

thomas, b. a. and purdy, h. m. 1982. Additional fossil plants from the Drybrook Sandstone, Forest of Dean,
Gloucestershire. Bull. Br. Mus. Nat. Hist. (Geol.) 36, 131 142.

D. EDWARDS

Department of Plant Science
University College
Cardiff

Typescript received 3 December 1984
Revised typescript received 16 January 1985

J. L. BENEDETTO
Departamento de Geologia
Universidad Nacional de Cordoba y conicet
Argentina

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. D. E. G. Briggs, Department of Geology, Goldsmiths'
College, University of London, Creek Road, London SE8 3BU, England, who will supply detailed instructions for authors on
request (these were published in Palaeontology 1977, 20, pp. 92U929).

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 1985 is £33 (U.S. $53) for Institutional
Members, and £16-50 (U.S. S27) 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 1LB, 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. c. morris. 155 pp. 99 text-figs., 30 plates. Price £16 (U.S. $24).

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 and G. L. williams. 100 pp., 24 text-figs., 22 plates. Price £15 (U.S. $23).

25. (for 1980): Stereom Microstructure of the Echinoid 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, 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.
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 1 984): 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).

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. c. morris. 31 pp., 24 plates. Price £20 (U.S. $44).

© The Palaeontological Association, 1985

Palaeontology

VOLUME 28 ■ PART 3

CONTENTS

The origins and aerodynamics of flight in extinct vertebrates
[Palaeontology Review]

KEVIN PAD1AN 413

The microstructure of tooth enamel in multituberculate
mammals

G. FOSSE, Z. KIELAN-J A WOROWSK A and S. G. SKAALE 435

Micropalaeontology of the Late Proterozoic Veteranen Group,
Spitsbergen

ANDREW H. KNOLL and KEENE SWETT 451

Lower Cretaceous inoceramid bivalves from the Antarctic
Peninsula region

j. a. crame 475

Coronate echinoderms from the Lower Palaeozoic of Britain
STEPHEN K. DONOVAN and CHRISTOPHER R. C. PAUL 527

Cortical development in Chaloneria cormosa (Isoetales), and
the biological derivation of compressed lycophyte decortica-
tion taxa

KATHLEEN B. P I G G and G A R W. ROTHWELL 545

Wealden occurrence of an isolated Barremian dinocyst facies
N. F. HUGHES and I. C. HARDING 555

A new genus of Carboniferous spiriferid brachiopod from
Scotland

MARIE LEGRAND-BLAIN 567

Ostracodes across the lapetus Ocean

ROGER E. L. SCFIALLREUTER and DAVID J. SIVETER 577

Two new species of herbaceous lycopods from the Devonian

of Venezuela with comments on their taphonomy

D. EDWARDS and J. L. BENEDETTO 599

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

issn 0031 -0239

Palaeontology

VOLUME 28 • PART4 NOVEMBER 1985

Published by

The Palaeontological Association * London

Price £21 -50

THE PALAEONTOLOGICAL ASSOCIATION

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

COUNCIL 1985-1986

President : Professor C. Downie, Department of Geology, University of Sheffield, Sheffield SI 3JD

Vice-Presidents : Dr. M. G. Bassett, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Dr. R. Riding, Department of Geology, University College, Cardiff CF1 1XL
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. R. Lord, Department of Geology, University College, London WC1E 6BT
Secretary: Dr. P. W. Skelton, Department of Earth Sciences, Open University, Milton Keynes MK7 6AA
Circular Reporter: Dr. D. J. Siveter, Department of Geology, University of Hull, Hull HU6 7RX
Marketing Manager: Dr. R. J. Aldridge, Department of Geology, University of Nottingham, Nottingham NG7 2RD

Editors

Dr. D. E. G. Briggs, Department of Geology, University of Bristol, Bristol BS8 1 RJ
Dr. P. R. Crowther, Leicestershire Museums Service, Leicester LEI 6TD
Dr. D. Edwards, Department of Plant Science, University College, Cardiff CF1 1XL
Dr. L. B. Halstead, Department of Geology, University of Reading, Reading RG6 2AB
Dr. R. Harland, British Geological Survey, Keyworth, Nottingham NG12 5GG
Dr. T. J. Palmer, Department of Geology, University College of Wales, Aberystwyth SY23 2AX

Other Members

Dr. M. J. Benton, Belfast Dr. C. R. C. Paul, Liverpool

Dr. M. E. Collinson, London Dr. A. B. Smith, London

Dr. P. L. Forey, London Professor T. N. Taylor, Columbus

Dr. A. W. Owen, Dundee

Overseas Representatives

Australia: Professor B. D. Webby, Department of Geology, The University, Sydney, N.S.W., 2006
Canada: Dr. B. S. Norford, Institute of Sedimentary 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. Cuffey, Department of Geology, Pennsylvania State University, Pennsylvania
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 1985 are:

Institutional membership
Ordinary membership
Student membership
Retired membership

£45 00 (U.S. $68)
£21 00 (U.S. $32)
£11-50 (U.S. $18)
£10-50 (U.S. $16)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr. A. R. Lord,
Department of Geology, University College, Gower Street, London WC1E 6BT, England. 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 1985 will receive Palaeontology, Volume 28, Parts 1-4.
All back numbers are still in print and may be ordered from Marston Book Services, P.O. Box 87, Oxford OX4 1LB, England,
at £21-50 (U.S. $33) per part (post free).

Cover: The dinoflagellate cyst Impagidinium patulum (Wall) Stover and Evitt from bottom sediments in the North Atlantic
Ocean. British Geological Survey specimen MPK 4234. x 2500.

XIPHOSURID BURROWS FROM THE LOWER
COAL MEASURES (WESTPHALIAN A)

OF WEST YORKSHIRE

by J . I. CHISHOLM

Abstract. Bilobate burrows lying parallel to bedding are described from a shallow-water sandstone. The
burrows are linked to structures produced by oblique movement through the sediment, the outlines of these
suggesting that the animal responsible was a xiphosurid, perhaps Bellinurus. The internal structure of the
burrows shows transverse markings on either side of a median furrow but the external surfaces are poorly
preserved and it is uncertain to which of the existing bilobate ichnogenera the trace fossil belongs. It is
provisionally referred to Aulichnites Fenton and Fenton, 1937 as Aulichnites ? bradfordensis ichnosp. nov.

Trace fossils attributed to limulid movements are known from many parts of the geological column
but the majority of those described are walking (or half-swimming) traces ( Kouphichnium Nopsca; see
Hantzschel 1975, p. W75), or are resting traces (Limulicubichnus Miller, 1982). Records of these in the
Silesian rocks of the central Pennines are summarized by Eagar et a/. (1985). Limulid burrowing
traces have been described less commonly, which is surprising in view of the known burrowing habits
of Recent Limulus; the structures described here are thought to have been produced by this process.

GEOLOGICAL SETTING

The trace fossils were found at Bolton Wood Quarries (British National Grid Reference SE 1 62 366)
in Bradford, West Yorkshire. Some of the burrows were found in place, distributed sparsely through
some 1-2 m of fine-grained sandstone exposed in the north-western part of the quarry but the best-
preserved material, including the holotype, was found in slabs lying loose at the foot of this face.
The following section was visible in 1983:

Top of section at SE 1 627 3660 thickness

(metres)

8. Coal, interlaminated with black mudstone. 0-0 05

7. Alternations of grey striped siltstone and ripple-laminated sandstone; rooty top. 4-80

6. Sandstone, fine and very fine-grained, mainly trough cross-bedded in sets up to
about 1 0 m thick, with palaeocurrent flow to east-south-east; some mudstone clasts
and tree trunks; some ripple- and parallel-laminated bands. Uneven erosive base. 13-50

5. Siltstone, striped, draped over eroded surface of bed below. 0-0-20

4. Sandstone, fine and very fine-grained, with low-angle wedge-bedding; Arenicolites

carbonarius (Binney) sparsely present in top few cm of some beds. 1 -70

3. Siltstone, striped, argillaceous. 3-90

2. Sandstone, fine and very fine-grained, with low-angle wedge bedding. 1 -80

1. Sandstone, fine and very fine-grained; ripple lamination and parallel lamination
alternating in bands 0-06 to 2-26 m thick; parting lineation and rib and furrow
structures indicate consistent palaeocurrent flow to west; Aulichnites ? bradfordensis
ichnosp. nov. sparsely distributed through the top 1 -20 m. 7-00

Base of section at SE 1605 3643.

(Palaeontology, Vol. 28, Part 4, 1985, pp. 619-628, pis. 75-76-1

620

PALAEONTOLOGY, VOLUME 28

The section is similar to one recorded at the quarries by Stephens et al. (1953, pp. 84-85). These
authors identified the sandstone (beds 1-6 of the section detailed above) as the Gaisby Rock, which
was formerly quarried at Gaisby, some 500 m to the north-west of Bolton Wood Quarries. They
regarded the Gaisby Rock (ibid., pp. 75-76) as a locally thickened representative of a thin sandstone
known in the Bradford area as the 80-Yard Rock, but a more likely correlation has been put forward
by C. G. Godwin (pers. comm.). He points out that the Gaisby Rock is about the same thickness as
the ‘main lower leaf of the Elland Flags (terms defined by Godwin 1984, fig. 1), and in the absence of
indications to the contrary, he prefers to correlate the two, as no local variations in thickness are
thereby implied. This correlation is accepted here, with the implication that the Gaisby Rock lies
about 30 m higher in the sequence than is indicated by Stephens et al. (1953, p. 70). On either
correlation, the section lies within the lower half of Westphalian A.

The Coal Measures of the Pennine region are of fluvial or fluvio-deltaic origin (Elliott 1969;
Haszeldine and Anderton 1980) and in this context bed 6, the thick, cross-bedded sandstone with tree
trunks and mudstone pebbles, is likely to represent a river or distr+butary channel deposit. The origin
of beds 1-5 is less certain, though their position beneath the channel deposit suggests that they may
represent a set of crevasse-splay, mouth-bar and barrier-beach sediments. Whether all of these
environments are present here, or only some of them, cannot be determined from the limited
exposures available. The presence of Arenicolites carbonarius suggests that the environment was non-
marine (Hardy 19706, p. 9).

DESCRIPTION OF TRACE FOSSILS

Most of the specimens were collected from the top few centimetres of a loose slab measuring about
6 square metres in surface area. Others were obtained from smaller slabs. The traces were also found
in place but only a few of these could be collected, due to the hardness of the rock. The way-up of the
loose material was determined by comparison with that obtained in situ. All the traces are full reliefs
(endichnia), preserved entirely within the sandstone; there are no epireliefs or hyporeliefs. Cut
sections suggest that all were produced within the sediment and that no traces formed at the
sediment/water interface have been preserved. The structures are of two kinds, about equal in
abundance, with a few connected examples that prove their common origin; they are distinguished as
‘oblique penetration structures’ and ‘horizontal burrows’. An idealized reconstruction of the trace
fossils is shown in Text-fig. 1 .

Oblique penetration structures. These are bedding disturbances, irregularly shaped or roughly
quadrangular as seen from above (PI. 75, figs. 3, 5, 10; Pi. 76, figs. 1, 3) and measuring 10-25 mm in
width by 10-35 mm in length. The average width (22 examples) is 15 mm. A section cut across the
width of one example (PI. 75, fig. 6) shows irregular distortion of dark bedding-laminae within an ill-
defined biconvex area of disturbance. Side views (PI. 75, figs. 1, 2, 4, 9) show that most of the
disturbances extend obliquely through the sediment at angles of about 35° from the horizontal; a few
are steeper. The greatest thickness of sediment penetrated in this way is 30 mm, but the specimen
concerned is incomplete both at top and base. When a specimen is split along the bedding, roughly
similar quadrangular shapes appear at all levels in the structure. In two of the specimens (PI. 75,
figs. 3, 5) the outline is hoof-shaped or lunate with a pointed projection at the centre of the concave
side: strong evidence for a xiphosurid origin. In one of these specimens the lamination bulges
upwards at the convex (head) end and downwards at the pointed (tail) end, but in the other is
depressed throughout. A few other specimens show signs of the ‘xiphosurid’ shape, the asymmetry of
which enables the forward end to be distinguished from the rear. Cut sections of these oriented
examples show that more of them were formed by upward movement through the sand than by
downward movement.

Horizontal burrows. The burrows are ribbon-like structures, flattened parallel to the bedding. They
are straight or variably curved, and incomplete lengths up to 250 mm are present. In width they range

CHISHOLM: WESTPHALIAN XIPHOSURID BURROWS

621

text-fig. 1. Idealized diagram to show structure of Aulichnitesl bradfordensis in side view (a) and on bedding
planes viewed from above (b). Movement of animal from left to right. Horizontal burrows are shown on bedding
planes (1), in lengthwise section (2), and in cross-section (3). Oblique upward penetration structure shown on

bedding (4) and in lengthwise section (5).

from 10 to 19 mm, the majority being about 15 mm across, and the thickness of most examples lies
between 2 and 4 mm; a single specimen is 7 mm thick. The more complete burrows pass at one or both
ends into oblique penetration structures and the outline of these can be used (as described above) to
determine the direction of movement. The anterior end in such cases tends to have a sharp, convex
outline whereas the posterior end is more vague (PI. 75, figs. 7, 8; PI 76, figs. 1, 5).

The outer surfaces of the burrows do not separate cleanly from the matrix and so cannot be
observed directly; all the specimens have split through the middle, revealing the internal structure.
Cross-sections suggest that both the upper and lower surfaces are generally ill defined, but that the
upper are in some cases vaguely bilobate, with a faint median furrow, whereas the lower tend to be flat
or unilobate. The sediment inside the burrows shows a disturbed texture in which dark flaky grains of
mica and carbon have been reoriented from their original position parallel to bedding, and have also
been concentrated at the top of the burrow (PI. 76, fig. 2). In many specimens the disturbed material,
as seen on bedding surfaces, shows a bilobate pattern, with a well-marked median groove and less
regular lateral grooves (PI. 76, figs. 1, 3, 4). Transverse markings, often arranged in a chevron pattern,
are present in parts of some specimens (PI. 76, figs. 1, 3); in the best preserved examples the chevrons
open towards the anterior (PI. 76, fig. 1 ). In one specimen (GSL 1839) the transverse lines are seen to
mark the ends of faint laminae in the burrow fill, but details of this texture are not clear.

INTERPRETATION OF TRACE FOSSILS

The hoof-shaped outlines of some of the penetration structures (PI. 75, figs. 3, 5) compare well
with the shapes produced in soft sediment by juveniles of the Recent horseshoe crab, Limulus
polyphemus (Caster 1938, pi. 12; Miller 1982, fig. 3), and are good evidence for the xiphosurid origin
of the whole burrow assemblage. Burrowing is a normal activity of L. polyphemus (Caster 1938,

622

PALAEONTOLOGY, VOLUME 28

pp. 17-19; Vosatka 1970) and is thought to have been a characteristic feature of xiphosurid behaviour
since Ordovician times (Bergstrom 1975, p. 301). The initial burrowing movements of juvenile
L. polyphemus have been described by Vosatka ( 1 970, p. 283) and by Eldredge ( 1 970). The sediment is
excavated and displaced by the limbs while the prosoma is pushed forwards and downwards into the
sediment by flexing movements of the body. The processes involved in continuous burrowing have not
been described in detail, however, but are supposed by Caster (1938, p. 19) to involve a combination
of leg and tail movements with the ploughing action of the headshield. The deformation produced in
the sediment by burrowing was not described by any of these authors, so that the interpretation of
the Bradford burrows is necessarily conjectural.

The median groove of the burrow-fill was probably made by the telson, the straightness of the
groove indicating that it did not lash from side to side during burrowing but probably was pressed
downwards against the sediment or trailed passively. The internal structure of the burrows is
interpreted as a back-fill texture. It roughly reflects the outline of the rear of the animal (PI. 75,
fig. 7; PI. 76, fig. 1), so was probably produced by repeated backward pressure against the sediment
that had been displaced and passed back under the animal by the action of the limbs. The marginal
furrows were probably produced by the outer margins of the prosoma and the average width of the
burrows, 15 mm, is therefore a measure of the width of this feature.

The penetration structures are not regarded as escape-traces from rapid sedimentation or erosion,
because rising and descending structures are present in the same few centimetres of sandstone; it is
more likely that oblique movements were part of normal burrowing behaviour. This conclusion is
consistent with the lack of any preferred orientation of burrows or penetration structures relative to
the current direction shown by parting lineation (PI . 76, fig. 1 ) and ripple lamination. The assemblage
is thought more likely to have resulted from scavenging for food than from concealment or resting,
these types of behaviour being more probably represented by the closely related trace fossil
Limulicubichnus rossendalensis (Hardy, 1970).

Of the two genera of xiphosurids known to occur in Upper Carboniferous sediments of the region,
Bellinurus and Euproops (Calver 1 968, pp. 159-1 60), the former is thought to have lived in open-water
habitats, whereas the latter probably lived in the coal swamps (van der Heide 1951, pp. 63-65);
E. danae may even have been able to live on subaerial vegetation (Fisher 1 979). Bellinurus is thus more
likely to have produced the burrows described here, which occur in fluvial crevasse-splay or delta-
front sandstones.

DISCUSSION

Burrows that can be regarded as incomplete examples of Aulichnitesl bradfordensis have been found
at other localities in the Upper Carboniferous of the Pennine region. One specimen (British
Geological Survey, Palaeontology Collection, Keyworth, JIC 1259-60) was found by the author in
loose blocks of fine-grained, micaceous, parallel-laminated sandstone at Catlow Quarry, Lancashire
(grid reference SD 8873 3676). The quarry is in the Dyneley Knoll Flags (Earp et a/. 1961, p. 137),
which lie at about the same stratigraphical level as the Elland Flags (Wray 1929, p. 272). The burrow

EXPLANATION OF PLATE 75

Figs. 1-10. Aulichnitesl bradfordensis , ichnosp. nov. 1, 2,4, 9. Oblique penetration structures, vertical sections cut
parallel to direction to travel; note variable nature of lamination disturbance but consistency of angle at which
sediment was penetrated. British Geological Survey, Palaeontology Collection, Keyworth: GSL 1 844, 1845 A ,
1845, 1836, all x 2. 3 and 5. Oblique penetration structures showing ‘xiphosurid’ outline; view from above,
light from top left. 3, paratype, GSL 1832, x 2; 5, GSL 1846, x 1. 6. Oblique penetration structure, section
across width. GSL 1843 A , x 2. 7 and 8. Horizontal burrows, view from above, light from top centre. Bilobate
pattern; ‘head’ at right. 7, GSL 1834, x 1; 8. GSL 1830, x 1 (in situ specimen). 10. Oblique penetration
structure, view from above, light from top left. GSL 1842, x 1 .

PLATE 75

CHISHOLM, Aulichnitesl bradfordensis

10

624

PALAEONTOLOGY, VOLUME 28

is preserved on a bedding plane as an internal section 12 mm broad by 60 mm long, with a well-
marked median furrow but only faint transverse markings. The anterior end has a sharp semi-
circular outline, whereas the posterior end fades off more gradually. L. rossendalensis is common
as hypichnial mounds on blocks in the same quarry. Another example (University of Man-
chester, Department of Geology Collection, MGSF 95), collected by Dr J. E. Pollard from
the Milnrow Sandstone (equivalent of the 80- Yard Rock) at Sycamore Quarry, Kerridge, Cheshire
(SJ 940 766), is about 22 mm wide, with curved transverse chevron markings and a median furrow,
and follows a gently curved course, with one sharper bend, for a distance of about 30 cm. The
traces described from sandy siltstones above the Butterly Marine Band at Standedge, West
Yorkshire, by Eagar et al. (1983, p. 298) under the provisional name ‘ Scolicia ’, are also similar
in many ways to A .? bradfordensis , being endichnial burrows flattened parallel to bedding, with
bilobate structure and a pattern of transverse markings in places. The maximum width of the burrows
is about the same as that of the Bradford traces, but no ‘xiphosurid’ outlines are associated. The
Standedge traces are somewhat older (Kinderscoutian) than those from Bradford but they like-
wise occur in sediments laid down in a river mouth or interdistributary bay environment (Eagar
et al. 1983, pp. 287-288). They are thought to have been produced by small arthropods which, in
view of the similarities with A.? bradfordensis , may well have been xiphosurids. Some bilobed traces
from the Grassington Grit (Pendleian) of West Yorkshire were referred by Eagar et al. (1985,
pi. 1e, f) to cf. Aulichnites. They are more sinuous than any so far mentioned and more closely
resemble A. parkerens is; they appear to be burrows but their greater sinuosity suggests that they were
made by a different arthropod.

Three different aspects of xiphosurid behaviour have now been identified as trace fossils in Upper
Carboniferous sediments of the Pennine region. Resting, alternating with short forward movements,
is represented by L. rossendalensis (Hardy 1970a, b), walking and half-swimming movements by
Kouphichnium aff. variabilis (Chisholm 1983), and burrowing movements by A.? bradfordensis. The
outlines of L. rossendalensis and A.? bradfordensis are very similar in shape and size, suggesting that
they were made by the same animal. All the traces were probably produced by members of the genus
Bellinurus (Hardy 1970a; Chisholm 1983; and above). Bilobate burrows showing a more sinuous
pattern may have been produced by other arthropods.

SYSTEMATIC PALAEONTOLOGY

Ichnogenus aulichnites Fenton and Fenton, 1937

Remarks. The type species, A. parkerensis Fenton and Fenton, 1937, is a bilobate burrow consisting
of an epichnial ridge 5 to 11-4 mm wide with a strong median furrow. The burrow filling is of
sand in oval layers which are concave anteriorly and convex upwards (Fenton and Fenton
1937, pp. 1079-1080). An examination of the holotype by Hakes (1977, p. 218) has shown that
the undersurface is unilobate and convex downwards, the depth below the median furrow being
about 3 mm.

EXPLANATION OF PLATE 76

Figs. 1 -5. Aulichnites ? bradfordensis , ichnosp. nov. 1. Horizontal burrow; ‘head’ at right, chevron ornament at
left. Penetration structures and a second burrow also visible. View from above, light from top centre.
Holotype, GSL 1831, x \. 2. Horizontal burrow, side view of lengthwise section showing disturbed texture
and local concentration of dark material towards top. Undisturbed lamination in lower half of specimen. GSL
1837, x 2. 3. Horizontal burrow showing bilobate structure and ornament; penetration structures also
present. View from above, light from top centre. Paratype, GSL 1833, x \. 4. Horizontal burrow showing
irregular bilobate pattern. View from below, light from left. GSL 1835, x \. 5. Horizontal burrow with
sinuous shape; ‘head’ end at right. View from below, light from top centre. Paratype, GSL 1832, x 1.

PLATE 76

CHISHOLM, Aulichnitesl bradfordensis

626

PALAEONTOLOGY, VOLUME 28
Aulichnitesl bradfordensis ichnosp. nov.
Plates 75 and 76; text-fig. 1

1976 ^Gyrochorte, Hakes, p. 26; pi. 5, fig. 4.

1977 ‘1 Aulichnites sp.. Hakes, p. 218; pi. 1 b.

1983 ‘ Scolicia ’, Eagar et al. , p. 298; pi. 24 a-c.

Derivation of name. From the city of Bradford, where the trace fossils were found.

Holotype. British Geological Survey, Palaeontology Collection, Key worth, Nottingham, GSL 1831.

Paratypes. GSL 1832-1834.

Type locality. Bolton Wood Quarries, Bradford, Yorkshire.

Horizon. Gaisby Rock (but see above); age Westphalian A.

Occurrence. Sparsely in fine-grained sandstone, about 22 m below top of Gaisby Rock.

Diagnosis. Horizontal bilobate burrows linked to obliquely rising or descending traces, some of
which have a hoof-shaped outline in plan view. The bilobate burrows are straight or variously curved
with a median furrow and variable transverse markings, chevron-shaped in places. The width is
generally about 15 mm.

Description. A detailed description of the two main elements, oblique penetration structures and horizontal
burrows, is given on pp. 620-621.

Remarks. The ‘xiphosurid’ outline of the best-preserved penetration structures (PI. 75, figs. 3, 5)
closely resembles, both in size and shape, that of the resting trace L. rossendalensis described by
Hardy (1970u) from an Upper Carboniferous sandstone in Lancashire, but it would be inappropriate
to apply this name to the Bradford structures since these are extensive burrow systems, not short
resting traces.

All burrows collected are full-reliefs made up of material disturbed by the passage of the animal.
The majority are preserved on parting surfaces parallel to bedding as upwardly convex bilobate
ridges and show internal structure rather than the ornament of the top surface. Most bilobate
ichnogenera are based on material preserved as epireliefs or hyporeliefs rather than as full-reliefs, so
the choice of a suitable name is not straightforward; however, the evidence of cross-sections shows
that the structures are closer to epireliefs than to hyporeliefs, in that the upper surfaces are in places
convex and vaguely bilobate, whereas the lower surfaces are poorly defined and appear to be
unilobate or flat. Of the bilobate epirelief ichnogenera Halopoa , Scolicia, Gyrochorte , and
Aulichnites, the first is considered unsuitable because it has an irregular surface ornament, not
notably transverse, and is only bilobate in places (Martinsson 1965, figs. 29-32). The name ‘ Scolicia ’
has been used for trails or burrows of a broadly similar nature from Upper Carboniferous sandstones
elsewhere in the Pennine region (Eagar et al. 1983, p. 298), but is not used here because most examples
of Scolicia (and its synonyms) illustrated by Hantzschel (1975, p. W107) show surface patterns or
internal structures more complex than those of the Bradford burrows. Gyrochorte has a simple,
plaited pattern on the bilobate upper surface but the depth of the structure can be greater than its
width and Heinberg (1973) has suggested that the trace was made by a worm-like animal burrowing
obliquely through the sediment. Because of these differences of structure and interpretation,
Gyrochorte is considered less appropriate than Aulichnites, an ichnogenus with a similar surface
pattern but in which the width of the burrow is greater than the depth (Fenton and Fenton 1 937, pi. 1 ,
fig. 1; Hakes 1977, p. 218).

The new trace fossil is only tentatively assigned to Aulichnites in view of the uncertain nature of the
outer surfaces, which in the type ichnospecies, A. parkerensis, are bilobate above and unilobate below
(Hakes 1977, p. 218). These uncertainties apart, it is clearly different from A. parkerensis in several
notable respects. First, it is linked to hoof-shaped structures; secondly, the pattern of transverse

CHISHOLM: WESTPHALIAN XIPHOSURID BURROWS

627

markings is much less regular, and less consistently present, than in A. parkerensis; thirdly, where the
transverse markings are chevron-shaped these open towards the anterior rather than to the posterior;
and fourthly, the chevrons meet the centre line at a more acute angle.

Acknowledgements. 1 would like to thank Mr C. G. Godwin for clarification of his views on the sequence in the
Bradford area, and Dr J. E. Pollard for many useful comments on a draft manuscript. The paper is published
with the approval of the Director, British Geological Survey (N.E.R.C.).

REFERENCES

Bergstrom, J. 1975. Functional morphology and evolution of xiphosurids. Fossils and Strata , 4, 291-305.
calver, m. a. 1968. Coal Measures invertebrate faunas. In murchison, d. g. and westoll, t. s. (eds.). Coal and
coal-bearing strata, 147-177. Oliver and Boyd, Edinburgh and London.
caster, k. e. 1938. A restudy of the tracks of Paramphibius. J. Paleont. 12, 3-60.

Chisholm, j. i. 1983. Xiphosurid traces, Kouphichnium aflF. variabilis (Linck), from the Namurian Upper
Haslingden Flags of Whitworth, Lancashire. Rep. Inst. Geol. Sci ., No. 83/10, 37-44.
eagar, R. m. c., okolo, s. A. and Walters, G. f. 1983. Trace fossils as evidence in the evolution of Carbonicola.
Proc. Yorks, geol. Soc. 44, 283-303.

— 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, H. A. (ed.). Biogenic structures; their use in interpreting depositional environments. Special
Publication, SEPM.

earp, j. r., magraw, d., poole, E. G., LAND, D. H. and Whiteman, A. J. 1961. Geology of the country around
Clitheroe and Nelson, ix + 346 pp. Mem. Geol. Surv. U.K.
eldredge, n. 1970. Observations on burrowing behaviour in Limulus polyphemus (Chelicerata, Merostomata),
with implications on the functional anatomy of trilobites. Am. Mus. Novitates, 2436, 1-17.
elliott, r. e. 1969. Deltaic processes and episodes: the interpretation of productive coal measures occurring in
the East Midlands, Great Britain. Mercian Geol. 3, 111-135.
fenton, c. L. and fenton, m. a. 1937. Burrows and trails from Pennsylvanian rocks of Texas. Am. Midi. Nat.
18, 1079-84.

fisher, d. c. 1979. Evidence for subaerial activity of Euproops danae (Merostomata, Xiphosurida). In nitecki,
m. H. (ed.). Mazon Creek Fossils, 379-447. Academic Press, New York, San Francisco and London,
xv + 581 pp.

Godwin, c. G. 1984. Mining in the Elland Flags: a forgotten Yorkshire industry. Rep. Br. Geol. Surv.
16 (4), 17 pp.

hakes, w. G. 1976. Trace fossils and depositional environment of four clastic units. Upper Pennsylvanian
megacyclothems, northeast Kansas. Paleont. Contr. Univ. Kans., Article 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. Special Issue No. 9, 209-226.

hantzschel, w. 1975. In teichert, c. (ed.). Treatise on Invertebrate Paleontology. Part W: Miscellanea.

Supplement 1, W 1-269. Geol. Soc. Amer. and Univ. Kansas Press.
hardy, p. g. 1970a. Xiphosurid trails from the Upper Carboniferous of northern England. Palaeontology ,
13, 188-190.

— 1970A Aspects of palaeoecology in the arenaceous sediments of Upper Carboniferous age in the area
around Manchester. Ph.D. thesis (unpubl.), University of Manchester.

haszeldine, r. s. and anderton, r. 1980. A braidplain facies model for the Westphalian B Coal Measures of
north-east England. Nature, Lond. 284, 51-53.

heinberg, c. 1973. The internal structure of the trace fossils Gyrochorte and Curvolithus. Lethaia, 6,
227-238.

martinsson, a. 1965. Aspects of a Middle Cambrian thanatotope on Oland. Geol. For. Stockh. Forh. 87,
181-230.

miller, m. f. 1982. Limulicubichnus: a new ichnogenus of limulid resting traces. J. Paleont. 56, 429-433.
Stephens, j. v., mitchhll, G. h. and Edwards, w. 1953. Geology of the country between Bradford and Skipton,
vii + 1 80 pp. Mem. Geol. Surv. U.K.

628

PALAEONTOLOGY, VOLUME 28

van der heide, s. 1951. Les arthropodes du terrain houiller du Limbourg meridional. Meded. geol. Sticht. (C) IV,
No. 5, 1-84.

vosatka, e. d. 1970. Observations on the swimming, righting and burrowing movements of young horseshoe
crabs, Limulus polyphemus. Ohio J. Sci. 70, 276-283.

wray, D. a. 1929. The Carboniferous succession in the central Pennine area. Proc. Yorks, geol. Soc. 21, 228-287.

Typescript received 27 July 1984

Revised typescript received 18 February 1985

J. I. CHISHOLM

British Geological Survey
Murchison House
West Mains Road
Edinburgh eh9 3la

EARLY CRETACEOUS ISOCRINUS FROM
NORTHEAST JAPAN

by TATSUO OJI

Abstract. Two species of Isocrinus, one of which is new ( Isocrinus ( Chladocrinus ) hanaii sp. nov.), are described
from the upper Aptian of the Miyako Group, Northeast Japan. The skeletons of I. (C.) hanaii are preserved
intact at several localities, so it is presumed that they were rapidly buried either alive or shortly after death, the
animals having lived on a sandy bottom in agitated water. In contrast to the relatively deep-water occurrence of
modern isocrinids (mostly 200-1000 m), those from Miyako lived in shallow water, indicating that certain
stalked crinoids persisted at shallow depths until mid-Cretaceous times. They were among the last isocrinids to
have lived in an almost predator-free environment. Soon afterwards, a diversification of predatory teleostean
fish was reflected in the contemporaneous appearance of new types of isocrinids.

Various marine invertebrates are known from the Lower Cretaceous Miyako Group in Northeast
Japan; crinoids are among the most common, yet they remain undescribed and have even been
ignored in some faunal lists. This study presents systematic descriptions of two species of elegantly
preserved isocrinids and discusses the implications of their occurrence with respect to changes in the
bathymetric distribution of isocrinids through the Mesozoic and the Cainozoic. They are the first
crinoids to be described from the Cretaceous of Japan.

The pioneer work on the stratigraphy of the Miyako Group by Yabe and Yehara (1913) recorded
the occurrence of excellently preserved crinoids which were identified as Pentacrinus. In the late 1 940s,
T. Hanai carried out detailed field surveys and excavated fine specimens of crinoids, including nearly
complete crowns. R. C. Moore examined these specimens during a visit to Japan and suggested their
assignment to Isocrinus. Since then, crinoids from the Miyako Group have been cited or listed as a
species of Isocrinus (e.g. Kobayashi et at. 1954; Tamura 1982). Excavations by Hanai and his
coworkers in the 1960s led to the discovery of additional material. The present study is based mostly
on this material, together with material newly collected by the author.

The Miyako Group ranges from the upper Aptian to the lower Albian. Except for the lowest
unit (Raga Formation) the Miyako Group is characterized by rich marine faunas which include
a variety of bivalves, gastropods, ammonites, and encrusting organisms such as hermatypic
corals and calcareous algae. At present, seventy species of bivalves (Hayami 1975), eighty-seven
species of gastropods (Kase 1984), and approximately sixty species of cephalopods (Obata and
Matsumoto 1977) have been recorded. These fossils indicate a shallow, open marine environment
in a warm climate (Hanai et al. 1968). Recently discovered fossil beachrock indicates an ancient
strand line and further substantiates the dominance of a tropical or subtropical climate (Hanai
andOji 1981).

Isocrinids occur abundantly in almost all the shallow marine strata of the Miyako Group. They
normally occur as disarticulated segments of stem and cirri, as is usually the case with fossil crinoids,
either associated with other marine fauna or scattered throughout the sediments. At a few localities,
however, unusually well-preserved specimens of Isocrinus occur commonly with their arms spread on
bedding planes. In addition to their palaeontological significance, such exceptional modes of
occurrence also give an indication of the nature of the sedimentary environment.

The crinoid fauna of the Miyako Group is of low diversity despite the abundance of specimens, and
contains only a few species of isocrinids and one species of comatulid; only two isocrinid species are
represented by sufficient material for description.

| Palaeontology, Vol. 28, Part 4, 1985, pp. 629-642, pis. 77-79.|

630

PALAEONTOLOGY, VOLUME 28

| , 'H Acidic tuff
Siltstone
tm Silty sandstone
11 Fine sandstone
[ly-yj Coarse sandstone
Conglomerate

•fa Horizons of well-preserved
I. (Chi.) hanaii , n sp

■fe Horizon of I. (Chi.) sp cf
I neocomiensis (Desor)

text-fig. 1 (left). Location of the collecting localities and distribution of the Miyako Group (black). Northeast

Japan.

text-fig. 2 (right). Columnar section of the Miyako Group along the northern coast of Haipe, showing horizons
of collecting sites (left, whole section; right, partly enlarged).

OCCURRENCE AND SEDIMENTARY ENVIRONMENT
Isocrinus (Chladocrinus) hanaii sp. nov.

Specimens are from the lower part of the Hiraiga Formation (uppermost Aptian) in the Tanohata and Moshi
areas (text-fig. I ). In the Tanohata area, almost completely preserved specimens occur in several restricted layers
approximately 10 m above the base of the Eliraiga Formation (Locs. Oj 7543, Hn 0915, Hn 0916, Hn 0920) along
the northern coast of Haipe (text-fig. 2). At these localities a calcareous fine-grained sandstone with large-scale
cross-stratification contains abundant bivalves, gastropods, annelid tubes, and fragmented ossicles of /. (C.)
hanaii sp. nov. The ossicles are mostly concentrated in the lowermost part of a single cross-stratification, or in
continuous thin layers. The highly calcareous, well-washed sandstone, the sedimentary structures, and the mode

OJI: EARLY CRETACEOUS CRINOIDS

631

of occurrence all indicate a high-energy environment in which the disarticulated crinoids were subjected to
transport and sorting.

At Loc. Hn 0920 (the type locality for /. (C.) hanaii sp. nov.) well-preserved articulated specimens with cup,
arms, and column occur sparsely in a single horizon within a light-grey, fine-grained, and well-sorted calcareous
sandstone (text-fig. 3a). The specimens have been brought into relief on the bedding plane by selective natural
weathering and etching of the matrix. Other marine fossils are rare at this horizon. In contrast a bed several
centimetres lower has produced numerous cirrals of I. (C.) hanaii sp. nov. associated with many small-sized
bivalves (text-fig. 3b), identified as Glycymeris ( Hanaia ) densilineata (Nagao), Astarte ( Trautscholdia ) minor
Nagao, A. ( Nicaniella ) semicostata Nagao. These two distinct modes of occurence, nearly complete individuals
and dissociated cirrals, appear in almost the same lithology and are possibly due to different modes of burial of
the skeletons. The former, with well-preserved but fewer specimens, strongly indicates rapid burial of living
crinoids, or shortly after their death. Experiments on several species of Recent comatulid crinoids have shown
that within two or three days of death they collapse into fragments (Cain 1968; Meyer 1971; Liddel 1975), but
there has been no equivalent report on the post-mortem decomposition of Recent isocrinid skeletons. It is
unlikely, however, that the well-preserved Cretaceous isocrinids considered here were subjected to much
transport. On the other hand the occurrence of /. (C.) hanaii sp. nov. almost exclusively as cirrals may possibly
result from the sorting of crinoidal fragments along with sand grains during transport, or during reworking by
wave action.

text-fig. 3. Mode of occurrence of Isocrinus ( Chladocrinus ) hanaii sp. nov. at Loc. Hn 0920, northern coast of
Haipe. a, well-preserved specimens in fine-grained sandstone, b, slightly below the horizon of a, consisting
mostly of cirrals. Both figures are views of bedding plane surfaces. Scale bars 2 cm.

In the Moshi area, /. (C.) hanaii sp. nov. is restricted to an interval 24-27 m above the base of the Hiraiga
Formation (Locs. Hn 4151, Hn 4153, Oj 0257). The specimens are contained in a calcareous, fine-grained
sandstone; while sometimes they retain nearly complete skeletons, they frequently occur as ossicles in calcareous
concretions together with wood fragments and small molluscs such as Mesosaccella insignis (Nagao), A. (N.)
semicostata Nagao, A. ( Freiastarte ) subomalioides Nagao, and the venerid Nagaoella corrugata (Nagao).

Isocrinus (Chladocrinus) sp. cf. Isocrinus? neocomiensis (Desor)

Specimens occur in the middle and upper part of the Tanohata Formation (upper Aptian) of the Tanohata area.
Those from the middle part of the formation are scattered through calcareous sandstone at the following
localities: Loc. Oj 5813, cross-stratified, calcareous, fine-grained sandstone on the northern coast of Hiraiga;
Loc. Oj 7517, pebble conglomeratic, calcareous, coarse-grained sandstone, with fragmented ostreids and other
bivalves, on the northern coast of Haipe. Specimens from the upper part of the formation are in sandy siltstone at
the following localities on the northern coast of Hiraiga: Loc. Oj 5827-2, just below the base of the overlying
Hiraiga Formation; Loc. Oj 5825-2, 4 m below the base of the Hiraiga Formation. The siltstones are extensively
bioturbated; disarticulated columns and brachials occur scattered throughout the sediment. Specimens from the
siltstone are better preserved than those from the sandstone. This species probably lived in relatively calm water.

632

PALAEONTOLOGY, VOLUME 28
DISCUSSION

Direct information on the bathymetry of the sediments within which the specimens occur is not
available. However, the two species described herein are considered to have been shallow-water
isocrinids. Nearly complete specimens of I. (C.) hanaii sp. nov. occur associated with transported
blocks of hermatypic corals and coralline algae at Haipe. Columnals and a series of brachials of I. sp.
cf. /.? neocomiensis are found in siltstone just above fossil beachrock at Hiraiga. Consequently both
species are considered to have been shallow-water dwellers, probably living in depths less than several
tens of metres.

Living stalked crinoids are mainly distributed in relatively deep water, mostly 200-1000 m
(isocrinids), deeper than 200 m (bourgueticrinids), and 500-5000 m (cyrtocrinids) (Breimer 19786,
p. T329). However, a few examples of Mesozoic stalked crinoids (including the species of Isocrinus
described here) are found in apparently shallow-water lithofacies. Because of this, it has been
proposed that the living stalked crinoids are ‘living fossils’ which disappeared from shallow
water to take refuge in deep water. According to Breimer (1978n, p. T10) they are approaching
extinction.

Fossil stalked crinoids for which bathymetric information may be inferred are few. Seirocrinus
subangularis (Miller), an early Jurassic pentacrinitid from Holzmaden, was once regarded as
pseudoplanktonic (Seilacher et al. 1968), but has subsequently been reinterpreted as benthonic by
Rasmussen (1977) who estimated that the crinoid lived at depths of 100-600 m. If Rasmussen is right
this pentacrinitid was present in relatively deep water during the early Jurassic. Nearly complete
specimens of Chariocrinus andreae (Desor) from the Bajocian of Switzerland occur in apparently
shallow-water oolitic sediments (Hess 1972). Other than this occurrence, records of shallow-water
isocrinids are scarce, and the two species of Isocrinus from the Miyako Group described herein
represent the youngest such shallow- water isocrinids so far reported. Apart from the Isocrinidae, a
rather different stalked crinoid, the bathycrinid Dunnicrinus mississippiensis R. C. Moore, is known
from apparently shallow-water facies of Maastrichtian age (Moore 1967). This may suggest that
some bathycrinids persisted in shallow seas after the isocrinids had disappeared from such depths.

Despite differing opinions on their taxonomy, it is possible to split the isocrinids into two groups on
the basis of the kind of articulation in the primibrachials. One group has synarthrial articulation and
the other has cryptosyzygial articulation (including the ‘synostosis’ of several authors). This
subdivision of isocrinids was first proposed by Carpenter as early as 1 882. Synarthry is characterized
by a prominent transverse ridge in the centre of each opposing articular facet, and the ligamentary
bundles are distributed in two fossae on either side of the ridge. Synarthrial articulation allows slight
movement between two connected ossicles. On the other hand, cryptosyzygy is a weak form of
articulation; joint facets are almost flat so that each ossicle is tightly connected. Cryptosyzygy also
occurs in more distal brachitaxes than IBr, and frequently in the proximal part within each
brachitaxis.

Observations of living crinoids show that isocrinid arms are easily disarticulated at the
cryptosyzygy just after specimens have been dredged, and that regeneration of the arms is usually
found at the distal facet of a hypozygal. I have frequently observed this in specimens of the isocrinid
Metacrinus rotundus Carpenter dredged off' the southern coast of Japan. Such regeneration indicates
that disarticulation at the cryptosyzygy occurs while the crinoids are living. Consequently the
cryptosyzygy of isocrinids may be a specialized articulation for autotomy (as is the case for
comatulids) as suggested by Roux (1974). Arms are easily disarticulated at the cryptosyzygy in both
living and dead specimens of Recent isocrinids. There has been no report of regenerated arms
growing from the distal facet of a synarthry.

The group which has synarthrial articulation (named here the ‘old group’) lived from the Triassic
to the Recent, while those which have cryptosyzygy in IBr (named here the ‘new group’) appeared in
the early Cretaceous and diversified after the mid-Cretaceous (Table 1 ). In the present-day Pacific the
‘new group’ is characterized by many species of Metacrinus and Saracrinus living mostly between
100-600 m water depth, while a member of the ‘old group’, Hypalocrinus naresianus (Carpenter),

OJI: EARLY CRETACEOUS CRINOIDS

633

GENERA

IBr

IIBr-

Isocrinus

SYN

SYN + CZ

Hypalocrinus

SYN

SYN *CZ

Chariocrinus

SYN

SY N ♦ CZ

Balanocrinus

SYN

SYN ♦ CZ

Nielsenicrinus

CZ

SY N ♦ CZ

Austinocrinus

CZ

unknown

Isselicrinus

CZ

CZ

Doreckicrinus

CZ

CZ

Cainocrinus

CZ

SYN ♦ CZ

Teliocrinus

CZ

CZGSYN)

E ndoxocri nus

CZ

CZ

Diplocrinus

CZ

CZ

Cenocrinus

CZ

CZ

Metacrinus

CZ

CZ

Saracrinus

CZ

CZ

STRATIGRAPHIC DISTRIBUTION

TRIASSIC JURASSIC CRETACEOUS PALEOGENE NEOGENE OUAT

table 1 . Stratigraphic distribution of ligamentary articulation type within 'old' and ‘new’ groups of Isocrinidae
(see text); genera and stratigraphic information modified after Rasmussen (1978). IBr, type of ligamentary
articulation in primibrachials; IIBr, type of ligamentary articulation in more distal brachitaxes than
secundibrachials; SYN, synarthry; CZ, cryptosyzygy (including synostosis and symmorphy).

is restricted to depths greater than 600 m. Data on distribution are still insufficient, especially
for the ‘old group’, but I conclude that the ‘new group’ lives at shallower depths than the ‘old
group’.

Teleostean fish are considered to be the main predators of crinoids (Meyer and Macurda 1977) and
this may well apply to isocrinids. There is some evidence from a deep-sea photograph of Conan et al.
(1981, fig. 4) which shows predation by a fish on Diplocrinus wyvillethomsoni (Carpenter). An
adaptive radiation of teleostean fish took place in shallow water after the mid-Cretaceous (Thomson
1977); if teleosts were the predators of isocrinids, this radiation would have led to some selection
pressure on these shallow-water stalked crinoids.

I consider that isocrinids of the ‘new group’ probably had an advantage over the ‘old group’ with
regard to predation by bony fish because of their greater ability to shed arms in times of emergency. It
is not apparent whether they shed their arms deliberately or passively, but it is presumed that they
underwent autotomy just as comatulids do. Autotomy or disarticulation of the arms probably
guarded the central soft body of isocrinids from predation. Crinoids of the ‘new group’ have a greater
number of cryptosyzygies near the central soft body mass and therefore are presumably better
protected from predation and better adapted for regeneration. Consequently, some of the ‘new
group’ have overcome the selection pressure and revived in relatively shallow depth (sometimes less
than 100 m) in the present-day Pacific.

Apparent regeneration is absent in Isocrinus from Miyako, strongly suggesting that it seldom if
ever suffered from predation; the late Aptian shallow sea of Miyako was still then a favourable
environment for the ‘old group’.

It has not yet been documented whether the isocrinids shed arms by autotomy, and the above
discussion is based only on the fact that the arms of living or dead specimens of isocrinids are easily
disarticulated at the cryptosyzygy. Further investigation, and particularly experiments on living
specimens, are needed to resolve this matter.

634

PALAEONTOLOGY, VOLUME 28
SYSTEMATIC PALAEONTOLOGY

Family isocrinidae Gislen, 1924
Genus isocrinus von Meyer, 1936

Subgenus chladocrinus Agassiz, 1836 emend. Sieverts-Doreck, 1971

Remarks. Chladocrinus was first introduced by Agassiz, and loosely defined for fossil species of
' Pentacrinus' with relatively numerous internodal plates. Owing to his brief diagnosis and failure to
designate a type species, this genus has long been ignored. Recently, Sieverts-Doreck ( 1971) discussed
the validity of the genus, designated as type species P. basaltiformis Miller, and emended the
diagnosis. She included five early Jurassic species in the genus and treated the two closely related
forms Isocrinus von Meyer, 1936 (type species I. pendulus von Meyer) and Neocrinus Thomson, 1864
(type species P. ( Neocrinus ) decorus Thomson) as distinct genera. According to Sieverts-Doreck,
Chladocrinus is easily distinguished from Isocrinus by its lack of symmorphial articulation, and from
Neocrinus by its more ramified arms. On the other hand, Hess (1972) regarded Chladocrinus as a
subgenus of Isocrinus because the two are differentiated only by the types of ligamentary articulation
between IIBr 3 _ 4 (cryptosyzygy versus symmorphy). Hess (1972) also treated Neocrinus as a distinct
genus. Rasmussen (1978) accepted the emended diagnosis of Sieverts-Doreck (1971) and considered
Chladocrinus a distinct genus. He did not restrict Chladocrinus to the early Jurassic species but
included within it those early Cretaceous to Recent species of Neocrinus whose ligamentary
articulation fitted the diagnosis given by Sieverts-Doreck, i.e. IBr^ (synarthry), IIBr 1 _ 2 (synarthry),
and IlBr 3 ^ 4 (cryptosyzygy).

If Chladocrinus in the sense of Rasmussen (1978) is adopted, what is the essential difference
between Chladocrinus and Isocrinus! Isocrinus possesses symmorphial articulation, generally at
IIBr 3 _ 4 , which is replaced by cryptosyzygial articulation in Chladocrinus. Consequently, the two
genera are separated only by differences in the types of ligamentary articulation.

Both cryptosyzygy and symmorphy are considered to be modified forms of syzygy (Breimer
1978«), differentiated as follows: in cryptosyzygy the joint facets are almost flat; in symmorphy they
are not flat, and instead a prominent transverse culmination of the epizygal brachial fits into a
corresponding depression of the hypozygal to form an interlocked structure. Isocrinus is character-
ized by symmorphy, but the degree of prominence and depression varies between species. For example,
the articular facet of /. cingulatus (Munster) drawn by Hess (1972, fig. 3) has a hypozygal which is
almost flat, the depression being conspicuous only near the margin. I. (C.) hanaii sp. nov. shows an
almost flat suture between IIBr 3 _ 4 (PI. 77, fig. 2), but the articular facets are somewhat undulating;
the distal facet of the hypozygal is more or less concave and the adoral part is slightly elevated (PI. 78,
figs. 7, 8). Similar articulation is also present in an epizygal of I. (C.) sp. cf. /.? neocomiensis (Desor)
(PI. 78, fig. 6). Considering the various degrees of prominence and depression observed in
symmorphy and cryptosyzygy, it seems difficult to draw a line between the two forms of articulation.
I follow Hess (1972) in treating Chladocrinus as a subgenus of Isocrinus , and regard the Miyako
examples as a species of the subgenus Chladocrinus because of their slightly depressed, almost flat
cryptosyzygy.

Another classification of the Isocrinidae is that of Roux (1977). Based on columnal morphology,
Roux emphasized a strong similarity between the fossil Balanocrinus and Recent N. blakei and
N. decorus. Later Roux (1981) placed the two living species in his Balanocrininae; he included fossil
Isocrinus in his Isocrininae. His subdivisions were based mostly on columnal microstructure, and

EXPLANATION OF PLATE 77

Figs. 1-4. Isocrinus ( Chladocrinus ) hanaii sp. nov., lower part of Hiraiga Formation, northern coast of Haipe.
I , ME6945, Loc. Hn 0916, paratype, x 1. 2, ME6936, Loc. Hn 0920, holotype, x2. 3, ME6943, Loc Hn 0920,
paratype, x2. 4, ME6950, Loc. Hn 0916 or Hn 0920 (precise locality unknown), paratype, x 1.

Figs. 1 , 2, and 4 whitened; Fig. 3 submerged in water.

PLATE 77

OJI, Isocrinus (Chladocrinus)

636

PALAEONTOLOGY, VOLUME 28

text-fig. 4. Stereom in columnal; note rounded meshes of a-stereom. a, Isocrinus ( Chladocrinus ) hanaii sp. nov.
from Loc. Hn 0920, light micrograph of thin section of columnal, cut transversely near surface, x 57. b, /. (C.)
sp. cf. /.? neocomiensis (Desor), ME6962 from Loc. Oj 5825-2, SEM micrograph, x 95 (treated in HF following

the method of Sevastopulo and Keegan 1980).

especially on the shape of the meshes of the a-stereom. The two species of Isocrinus ( Chladocrinus )
from the Miyako Group are characterized by rounded-polygonal to rounded meshes, and fall into his
Isocrininae (text-fig. 4). The controversy surrounding the classification of the Isocrinidae, especially
regarding the recognition of reliable taxonomic criteria, remains unsolved.

Isocrinus ( Chladocrinus ) hanaii sp. nov.

Plate 77; Plate 78, figs. 1-5, 7, 8; Plate 79, figs. 1-7; text-fig. 4a

1913 Pentacrinus , Yabe and Yehara, pp. 12, 13.

1913 Pentacrinus(l), Yabe and Yehara, p. 17.

1954 Isocrinus, Kobayashi el al., p. 309.

1966 Isocrinus sp., Masutomi and Hamada, p. 143, pi. 72, fig. 3.

1982 Isocrinus sp., Tamura, p. 20, table 1 .

Diagnosis. Medium- to small-sized species of Isocrinus ( Chladocrinus ) characterized by very low and
tumid columnals, and by regular distribution of slightly embayed cryptosyzygy between the third and
fourth brachials in each brachitaxis.

EXPLANATION OF PLATE 78

Figs. 1-5, 7, 8. Isocrinus ( Chladocrinus ) hanaii sp. nov., lower part of Hiraiga Formation, Matsushima islet in
Moshi (Locs. Hn 4151, Hn 4153), and on northern coast of Haipe (Oj 7543). 1, ME6962, Loc. Hn 4153,
paratype in lateral (a, b) and dorsal (c) views, x 2. 2, ME6964, Loc. Hn 41 51, paratype in lateral (a) and dorsal
(b) views, x 2. 3, ME6949, Loc. Oj 7543, in ventral (a), dorsal ( b ), and lateral (c) views, x 2; arrow indicates
one of distal facets of IIBr 3 (also shown in Fig. 7). 4a, b , ME6955, Loc. Oj 7543, paratype, stereo pair of distal
facet of IBr l5 synarthry, x 8-5. 5, ME6955, Loc. Oj 7543, paratype in ventral view, x 4. la, b, ME6949, Loc.
Oj 7543, paratype, stereo pair of distal facet of IIBr 3 , x 8-2. 8a, b , ME6954, Loc. Oj 7543, paratype, stereo pair
of distal facet of IIBr 3 , x 10.

Fig. 6. Isocrinus ( Chladocrinus ) sp. cf. Isocrinus ? neocomiensis (Desor), ME6974, Loc. Oj 5825-2, proximal facet
of brachial, x 8-3.

PLATE 78

OJI, Isocrinus ( Chladocrinus )

638

PALAEONTOLOGY, VOLUME 28

Types. ME6936, holotype, well-preserved crown and column from Loc. Hn 0920 (PI. 77, fig. 2). ME6937-6959,
paratypes from Locs. Hn 0915, Hn 0916, Hn 0920, and Oj 7543, northern coast of Haipe. ME6960-6964,
paratypes from Locs. Hn 4151, Hn 4153, and Oj 0257, Moshi. All the types are housed in the University
Museum, University of Tokyo.

Description. Cup bowl-shaped, low, and wide. Basals small, triangular, and separated on surface. Radials low
and broad, with smooth and evenly curved dorsal surface. Articular facet for IBr inclined 50-60°. Dorsal
ligament fossa relatively large, approximately one half as large as the articular facet.

Arms more slender towards distal ends, but almost uniform in diameter within each brachitaxis. Nearly
isotomous branching until fourth axillary (IVAx). In paratype ME6950, after IVAx, one of two VBrr forks to
form VIBrr, and the other remains undivided (PI. 77, fig. 4). If there is no further division of arms distally and
every brachitaxis follows such a pattern of ramification, the finials can number 1 20. Brachial plates generally thin
and smooth on surface. IBr! very thin. Articulation IB^ 2 embayed synarthrial (PI. 78, fig. 4). On dorsal surface,
distal facet of IBr! slightly concave and in contact with strongly convex proximal facet of IBr 2 . IBr 2 axillary.
Dorsal surface of IBr, nearly rhombic in outline, a little wider than high. Secundibrachials twelve or thirteen,
rarely eleven or fourteen in number. IIBrj strongly wedge-shaped. Articulation IIBr^ embayed synarthrial as
between IBr,^. IIBr 2 a little higher than the other secundibrachials. First pinnules on IIBr 2 . IIBr 2 _ 3 oblique
muscular. 1 1 B r 3 and II Br 4 wedge-shaped, a little thinner than the other secundibrachials. Articulation 1 1 Br 3 _ 4
slightly embayed cryptosyzygial, showing nearly straight suture on dorsal surface. Distal facet of IIBr 3 smooth
and marginal crenulation invisible (PI. 78, figs. 7, 8). Ligamentary articulation of this kind developed between
third and fourth brachials of every brachitaxis, with only one exception (one secundibrach of ME6962).
Tertibrachials seventeen, nineteen, or twenty-one in number. Quartibrachials twenty-one or twenty-five, rarely
thirty-seven. Proximal pinnules composed of approximately ten pinnulars, which are rectangular in lateral view.

Column short and slender; proximal part highly variable in diameter but gradually modified distally to
uniform diameter. Distal part slightly larger than the proximal part. Internodal plates low and tumid, with
pentagonal to rounded subpentagonal outline in transverse section. Suture almost straight on surface. Articular
facet of columnals petaloid; floors subguttiform, outwardly a little broad, surrounded by culmina and crenellae.
Crenulae vary in size from the adradial to the marginal position, with the greatest length between radial and
marginal areas. Culmina small in number for the diameter; two or three along margin, three or four along radius
in columnals of diameter 4- 1 mm. Canal narrow and rounded. Number of internodal plates seven or eight, rarely
nine in the distal part. Radial pore developed in the proximal part of paratype ME6947, but otherwise absent.
Nodal plates higher (approximately twice as high as internodal), wider, and more stellate in outline than
internodal. Interradial edge sharp and arched. Articulation between nodal and infranodal synostosial. Periphery
of the distal facet of nodal slightly raised. Cirrus sockets always five in number, large and wide, occupying almost
entire height and width of lateral side of nodal, but not interfering with adjacent columnals. Articular ridge of
cirrus socket weakly protruding with slightly tuberculated triangular ends. Proximal cirral plates thin and disc-
shaped, with circular to slightly ovate outline in transverse section. Height of cirral plates suddenly increases
distally in several plates, becoming almost as high as wide. Articular ridge slightly raised, canal passing a little
above the centre.

EXPLANATION OF PLATE 79

Figs. 1-7. Isocrinus ( Cliladocrinus ) hanaii sp. nov., lower part of Hiraiga Formation, northern coast of Haipe.
1-5, Loc. Hn 0920, articular facets of columnals, all x 5; 1 and 5, ME6956, paratype; 2, ME6957, paratype;
3, ME6958, paratype; 4, ME6959, paratype. 6, ME6942, Loc. Hn 0920, paratype, lateral view of column, cirri,
and crown, x 2. 7, ME6947, Loc. Hn 0916, paratype, proximal column and cirri, x 2.

Figs. 8-16. Isocrinus ( Cliladocrinus ) sp. cf. Isocrinus ? neocomiensis (Desor), Tanohata Formation.

8, ME6965, Loc. Oj 5813, northern coast of Hiraiga, articular facet (a) and lateral view (b) of columnals, x 4.

9, ME6966, Loc. Oj 5813, northern coast of Hiraiga, articular facet (a) and lateral view (b) of columnals, x 4.

10, ME6967, Loc. Oj 7517, northern coast of Haipe, articular facet ( a ) and lateral view ( b ) of columnals,
x4 11, ME6968, Loc. Oj 5827-2, northern coast of Hiraiga, articular facet of internodal plate, x 5.
12, ME6969, Loc. Oj 5827-2, northern coast of Hiraiga, articular facet of nodal plate, x 5. 13, ME6970,
Loc. Oj 5827-2, northern coast of Hiraiga, underside view of radial plate, x 4. 14, ME6971, Loc. Oj 5825-2,
northern coast of Hiraiga, articular facet (a) and lateral view (b) of columnals, x 4. 15, ME6972, Loc. Oj
5825-2, lateral view of column, x5-9. 16, ME6973, Loc. Oj 5825-2, lateral views of column, (a) x4, (b) x 9-9.

Figs. 1-4, 9b, 10a, 13, and 14a whitened; Fig. 5 submerged in water.

PLATE 79

OJI, Isocrinus ( Chladocrinus )

640

PALAEONTOLOGY, VOLUME 28

Discussion. The description above is based on material from Haipe and Moshi. Specimens from the
two areas are considered conspecific because of the agreement (or strong similarity) in form of the
dorsal cup and brachials, the arrangement of ligamentary articulation, and the pattern of ramification.
Relatively complete specimens from Haipe are almost uniform in size, and may represent an adult
stage. On the other hand, those from Moshi include relatively small columns which may represent a
younger stage; these columns show a more pronounced alternating variation in diameter and
columnals are higher with respect to the diameter. Columns of I. (C.) hanaii sp. nov. almost lack
radial pores and crenulated sutures between columnals, suggesting that their rate of the accretionary
growth was high.

I. (C.) hanaii sp. nov. is similar to N. australis australis (C. Moore) and N. a. albascopularis
(Etheridge) from the lower Cretaceous of Australia in the form of the crown (Moore 1870, pi. 3,
figs. 1-3; Etheridge 1901, pi. 1, fig. 4, pi. 3, figs. 1-3; Etheridge 1902, pi. 4, figs. 7-10) but the surface
of the radials and brachials of I. (C.) hanaii sp. nov. is smooth in contrast to the rugose surface
of theca and brachials of N. australis (which has, in addition, higher IBr 2 and more strongly wedge-
shaped IIBij).

I. (C.) hanaii sp. nov. is distinguished from Percevalicrinus aldingeri Klikushin (Rasmussen 1961,
pi. 10, figs. 1-11 as N. tenellus (Eichwald); Klikushin 1981, pi. 9, figs. 1, 2) by the more tumid
columnals of the former which show an alternating variation in diameter in the proximal part of the
column. The present species is also distinguished from P. tenellus (Eichwald) from the lower
Cretaceous of the USSR (Klikushin 1979, pi. 1 1, fig. 5; text-fig. \d) by the latter’s regular arrangement
of meshes in the a-stereom.

There are many species of isocrinids which are tentatively assigned to Isocrinus only on the basis of
stem elements (e.g. Rasmussen 1961). I. (C.) hanaii sp. nov. differs from all of these Cretaceous species
in having strongly tumid columnals with a smooth lateral surface.

Although the Miyako examples from calcareous sandstone have been subjected to secondary
precipitation of calcite around the ossicles, and the details of the articular facets are more or
less obscured, they often retain an almost complete crown and most of the original pattern of
ramification. In paratype ME6962 the number of secundibrachials is exceptionally small, eleven in
one brachitaxis; this is probably related to the absence of ligamentary articulation at IIBr 3 _ 4 in the
brachitaxis. No apparent regeneration has been observed in any of the material.

Isocrinus ( Chladocrinus ) sp. cf. Isocrinusl neocomiensis (Desor)

Plate 78, fig. 6; Plate 79, figs. 8 16

1961 Isocrinus ? neocomiensis (Desor); Rasmussen, p. 143, pi. 18, figs. 4, 5.

Material. ME6965 and 6966, columns from Loc. Oj 5813, northern coast of Hiraiga. ME6967, column from
Loc. Oj 7517, northern coast of Haipe. ME6968-6970, columnals and a radial plate from Loc. Oj 5827-2,
northern coast of Hiraiga. ME697 1-6974, columns and brachials from Loc. Oj 5825-2, northern coast of Hiraiga.

Description. Brachials smooth on dorsal surface. A brachial (epizygal) with slightly convex and almost smooth
proximal facet of cryptosyzygial articulation, as in IIBr 3 _ 4 of I. (C.) hanaii sp. nov. (ME6974; PI. 78, fig. 6).
Marginal crenulation inconspicuous.

Column fairly constant in diameter in the small- and medium-sized specimens ME6972 and 6973, slightly
variable in diameter in the larger specimens ME6965-6967 and 6971). In the small specimen ME6972, column
pentagonal in section, with columnals almost uniform in height and diameter. In the larger column ME6973, a
row of a few tubercles is aligned horizontally on the side of a large columnal, and also along the suture. In still
larger columns, the kind most commonly represented, internodals are a little tumid and slightly variable in
diameter; larger columnals are almost pentagonal and higher compared with smaller ones, which are slightly
stellate to pentalobate and less high (ME6965-6967 and 6971). Internodals eight in number. Articular facet of
columnals with subguttiform floors surrounded by culmina and crenellae. Crenulae with gradual transition in
size from adradial to marginal position, attaining greatest length between radial and marginal areas. Culmina
seven or eight in number in a column of diameter 5-7 mm. Canal narrow and rounded. Nodal plate higher, wider,
and more stellate in outline than internodal plates, especially in large specimens. Cirrus sockets five in number.

OJI: EARLY CRETACEOUS CRINOIDS

641

large and wide, occupying almost the entire height and width of lateral side of nodal. Articular ridge weakly
protruding with slightly tuberculated triangular ends. Articular ridge of cirral plate weakly raised, with canal
passing a little above centre.

Remarks. This species is assigned to the subgenus Chladocrinus because of the slightly convex
cryptosyzygial articular facet of the epizygal. Small- and medium-sized columns with longitudinal
ridge and tubercles are strongly reminiscent of /.? neocomiensis (Desor) from the Hauterivian of
France (Rasmussen 1961, pi. 18, figs. 4, 5), which is represented only by columns. The present species,
however, occurs in younger strata (upper Aptian). Because the brachial plates of IP. neocomiensis
from Europe are unknown, strict specific identification is difficult.

The present species resembles /. (C.) hanaii sp. nov. in the form of its brachial plates (especially in
the facet of ligamentary articulation) but differs in that its succeeding columnals show little or no
variation in diameter and tubercles are present on the lateral side of small- to medium-sized columns.

Acknowledgements. I sincerely thank Emeritus Professor T. Hanai (University of Tokyo) for introducing me to
the study of crinoids, for providing excellent specimens, and for encouragement and various suggestions
throughout this work. Mr J. C. Grimmer (Scripps Institute of Oceanography) also made helpful suggestions and
sent several reference papers. Thanks are due to Dr R. V. Kesling for reading an earlier draft of the paper. Drs. I.
Hayami and K. Chinzei (University of Tokyo) offered valuable advice and encouraged the author. Dr T. Kase
(National Science Museum) provided specimens from the Moshi area.

REFERENCES

breimer, a. 1978a. General morphology. Recent crinoids, pp. T9-T58. In moore, r. c. and teichert, c. (eds.).
Treatise on invertebrate paleontology. Part T. Echinodermata 2, Crinoidea (1). Geological Society of America
and University of Kansas Press, Boulder, Colorado and Lawrence, Kansas.

19786. Ecology of Recent crinoids. Ibid., T3 16-T330.

cain, j. d. b. 1968. Aspects of the depositional environment and palaeoecology of crinoidal limestones. Scott. J.
Geol. 4 , 191-208.

carpenter, p. h. 1882. The stalked crinoids of the Caribbean Sea. Bull. Mus. comp. Zool. Harv. 10 , 165-181.
conan, g., roux, m. and sibuet, m. 1981. A photographic survey of a population of the stalked crinoid
Diplocrinus ( Annacrinus ) wyvillethomsoni (Echinodermata) from the bathyal slope of the Bay of Biscay.
Deep-Sea Res. 28 , 441-453.

etheridge, r. jr. 1901. Additional notes on the palaeontology of Queensland, part 2. Bull. geol. Surv. Qd,
13 , 1-37.

1902. A monograph on the Cretaceous invertebrate fauna of New South Wales. Mem. geol. Surv. N.S. W.,
Palaeont. Ser. 11, I -50.

hanai, t., obata, i. and hayami, i. 1968. Notes on the Cretaceous Miyako Group. Mem. natn. Sci. Mus., Tokyo,
1, 20-28. [In Japanese.]

and oji, t. 1981. Early Cretaceous beachrock from the Miyako Group, Northeast Japan. Proc. Japan

Acad., B, 57, 362-367.

hayami, i. 1975. Systematic survey of the Mesozoic Bivalvia from Japan. Bull. Univ. Mus. Tokyo,
10 , 1-249.

hess, H. 1972. Chariocrinus n. gen. fur Isocrinus andreae ( Desor) aus dem unteren Hauptrogenstein (Bajocien)des
Basler Jura. Eclog. geol. Helv. 65, 197-210.

kase, T. 1984. Early Cretaceous marine and brackish-water Gastropoda from Japan, 262 pp. National Science
Museum, Tokyo.

klikushin, v. G. 1979. Microstructural features of isocrinid stem. Paleont. J. 13 , 83-90.

1981. Sea lilies of the genus Percevalicrinus. Ibid. 15 , 84-95.

KOBAYASHI, T., HANZAWA, S., ASANO, K., KIMURA, T., HAYASAKA, I., EGUCHI, M., HATAI, K., MAKIYAMA, J.,

matsumoto, t. and nisiyama, s. 1954. ‘ Koseibutsugaku 1 , 372 pp. Asakura-shoten, Tokyo. [In Japanese.]
liddel, w. d. 1975. Recent crinoid biostratinomy. Abstr. Progm geol. Soc. Am. 7, 1 169.
masutomi, k. and hamada, t. 1966. Fossils in color, 258 pp. Hoikusha, Osaka. [In Japanese.]
meyer, D. L. 1971. Post mortem disarticulation of Recent crinoids and ophiuroids under natural conditions.
Abstr. Progm geol. Soc. Am. 3 , 645-646.

and macurda, d. b. jr. 1977. Adaptive radiation of the comatulid crinoids. Paleobiol. 3 , 74-82.

642

PALAEONTOLOGY, VOLUME 28

moore, c. 1870. Australian Mesozoic geology and palaeontology. Q. Jl geol. Soc. Loud. 26, 226-261.
moore, R. c. 1967. Unique stalked crinoids from Upper Cretaceous of Mississippi. Paleont. Contr. Pap. Univ.
Kans. 17, 1-35.

obata, I. and matsumoto, T. 1977. Correlation of the Lower Cretaceous formations in Japan. Scient. Rep. Dept.
Geol. Kyushu Univ. 12, 165-179. [In Japanese.]

rasmussen, h. w. 1961 . A monograph on the Cretaceous Crinoidea. Biol. Skr. 12(1). 428 pp.

— 1977. Function and attachment of the stem in Isocrinidae and Pentacrinitidae: review and interpretation.
Lethaia , 10, 51-57.

— 1978. Systematic description, Articulata, pp. T813-T928. In moore, r. c. and teichert, c. (eds.). Treatise
on invertebrate paleontology. Part T. Echinodermata 2 , Crinoidea (3). Geological Society of America and
University of Kansas Press, Boulder, Colorado and Lawrence, Kansas.

roux, m. 1974. Les principaux modes d’articulation des ossicles du squelette des Crinoi'des pedoncules actuels.
Observation microstructurales et consequences pour l’interpretation des fossiles. C. r. hebd. Seanc. Acad. Sci.,
Paris, D, 278, 2015-2018.

— 1977. The stalk joints of Recent Isocrinidae (Crinoidea). Bull. Br. Mus. nat. Hist. (Zook), 32, 45-64.

— 1981. Echinodermes: Crinoi'des Isocrinidae. Resultats des Campagnes M.U.S.O.R.S.T.O.M. 1. Philip-
pines. Mem. O.R.S.T.O.M. 91,477-543.

seilacher, a., drozdzewski, G. and haude, r. 1968. Form and function of the stem in a pseudoplanktonic
crinoid ( Seirocrinus ). Palaeontology , 11, 275-282.

sevastopulo, G. d. and keegan, j. b. 1980. A technique for revealing the stereom structure of fossil crinoids.
Ibid. 23, 749-756.

sieverts-doreck, h. 1971. Uber Chladocrinus Agassiz (Isocrinidae) und die nomenklatorische Verankerung
dieser Gattung. Neues Jb. Geol. Paldont. Mh. 5, 314-320.
tamura, m. 1982. Isocrinus columnal bearing limestones in Sambosan Belt and Japanese Triassic Isocrinus
columnals. Mem. Fac. Educ. Kumamoto Univ. Nat. Sci. 31, 19-24. [In Japanese.]

THOMSON, K. S. 1977. The pattern of diversification among fishes, 374-404. In hallam, a. (ed.). Patterns of
evolution , as illustrated by the fossil record , 591 pp. Elsevier, Amsterdam, Oxford, and New York.
yabe, h. and yehara, s. 1913. The Cretaceous deposits of Miyako. Scient. Rep. Tohoku Imp. Univ. 2, 1 (2), 9-15.

Typescript received 31 July 1984
Revised typescript received 29 January 1985

TATSUO OJI

Geological Institute
University of Tokyo
Tokyo 1 13, Japan

STRUCTURE AND FUNCTION OF THE PECTORAL
GIRDLE AND FORELIMB OF ST RU T H I O M I M U S
ALTUS (THEROPODA: O RN IT H O M I M I D AE)

by ELIZABETH L. NICHOLLS Gild ANTHONY P. RUSSELL

Abstract. The forelimb and pectoral girdle of Struthiomimus alius are described for the first time. The
ornithomimid pectoral girdle dilfers from that of other theropods in having a higher scapular prominence
(acromion process), an anterior flange on the supraglenoid buttress, and a narrow, attenuated coracoid.

Osteological and myological comparisons with recent reptiles and birds, combined with muscle scar evidence,
suggests that the primary girdle of Struthiomimus was oriented somewhat laterally, as in recent crocodiles and
lizards, and that it was mobile with respect to the body wall. The potential for extensive protraction and
retraction of the humerus is evident, endowing Struthiomimus with extensive forereach abilities, combined with
limited rotational potential.

In the manus the offset first digit differs from the usual theropod condition in being rotated outwards, away
from the midline of the hand. Digits II and III are incipiently coalesced and functioned as a unit. The osteological
evidence suggests that the manus of Struthiomimus operated as a hooking and clamping structure, rather than as
a grasping or raking one.

While the forelimb of ornithomimids is ‘coelurosaurian’ in length, it lacks the raptorial
characteristics of that group, and there has been considerable speculation regarding its function.
Osborn (1916) suggested that the first digit was opposable and described the manus as a grasping
hand. Later workers, however, have questioned both the opposability of digit I and the grasping
ability of the manus (Ostrom 1969; Galton 1971; Osmolska et a/. 1972).

The forelimb of Struthiomimus was figured by Osborn (1916), but only a brief description was given
and the scapulocoracoid was not described. The complete forelimb and shoulder girdle are preserved
in only a few specimens of North American ornithomimids and in most cases these are either
mounted (AMNH 5339, ROM 851) or unprepared (NMC 8632). Consequently most subsequent
comparative work has been based on Osborn’s incomplete description.

A specimen of S. cthus with a well-preserved, articulated pectoral limb and shoulder girdle is now
available (Nicholls and Russell 1981). We here redescribe the forelimb and describe and figure the
scapulocoracoid for the first time. The presence of clear muscle scars on both the humerus and the
scapulocoracoid have prompted us to reconsider forelimb function in this species. Inferences drawn
with respect to function are based upon Recent comparative material described herein.

Abbreviations used in this work are as follows: AMNH, American Museum of Natural History;
BM(NH), British Museum (Natural History); NMC, National Museum of Canada; ROM, Royal
Ontario Museum; TMP, Tyrrell Museum of Palaeontology; UA, University of Alberta; UCMZ,
University of Calgary, Museum of Zoology; USNM, United States National Museum.

SYSTEMATIC PALAEONTOLOGY

Family ornithomimidae Marsh 1890
Genus Struthiomimus Osborn 1916 (Emended Russell 1972)

Struthiomimus altus (Lambe 1902)

UCMZ(VP)1980. 1 . Incomplete skeleton consisting of limbs, girdles, gastralia, and fragments of
vertebral column and ribs. The specimen was collected from the Judith River Formation (Judithian,

[Palaeontology, Vol. 28, Part 4, 1985, pp. 643-677.]

644

PALAEONTOLOGY, VOLUME 28

Upper Cretaceous), of southern Alberta, Canada. The left forelimb and pectoral girdle are complete
(text-fig. 1 ) and were found articulated. A description of the specimen and its taxonomic relationships
appear elsewhere (Nicholls and Russell 1981).

Little comparative material is available. Scattered podial fragments of ornithomimids are
abundant in the bone scrap of the Judith River Formation, but articulated specimens, or even
complete elements, are rare. The two best specimens are ROM 851 ( Ornithomimus edmontonicus) and
AMNH 5339 ( S . altus). ROM 85 1 is so crushed that few surface details are discernible. Both of these
specimens are mounted, making detailed anatomical comparison difficult. They have been described
and figured by Parks (1933) and Osborn (1916) respectively. Comparisons of UCMZ(VP)1980. 1
with other North American ornithomimids is based primarily on examination of the following
specimens: NMC 12441 , 8632, 12228, 8902; UA 16182; ROM 851, 840; and also on literature reports.

Russell (1972) defined three genera of North American ornithomimids: Ornithomimus , Struthiomi-
mus , and Dromiceiomimus. We here recognize only the first two, which may be distinguished on the
basis of the manus (Nicholls and Russell 1981)— the manus of Dromiceiomimus is incompletely
known.

OSTEOLOGY OF THE PECTORAL GIRDLE AND FORELIMB OF
STRUT HIOMIMUS ALTUS

(a) Orientation

Throughout the subsequent descriptions we have attempted to standardize directional terminology.
Due to the difficulty of orienting adult structures in a standard fashion we have chosen to employ
developmental terminology and orientation as they relate to the main body and limb axes. In this
context all structures have developmental dorsal-ventral, anterior-posterior, and either lateral-
medial (limb girdles) or proximal-distal (limbs) axes. All descriptive terminology relates to these axes
(refer to orientation arrows on figures for clarification).

(. b ) Scapulocoracoid

In UCMZ(VP)1980. 1 the left scapulocoracoid is complete, except for the dorsal tip of the scapular
blade (text-fig. 1). Both coracoids are complete and overlap along the ventral midline. As the left
coracoid is partially covered by the right element, the description of the coracoid is a composite,
based on the coracoids of both sides. Measurements of the girdle and forelimb are given in Table 1.

The exact length of the scapula is not known, but assuming the scapulo-femoral ratio of
UCMZ(VP)1980.1 to be the same as that of AMNH 5339 (Osborn 1916), we estimate the length of
the scapula to be 380 mm. It is long and slender and ventrally the shaft is oval in cross-section, but
dorsally becomes compressed and blade-like (text-fig. 2) and the shaft has a slight posterior
curvature. Although the dorsal tip of the blade is missing, there is no evidence that it was significantly
expanded, in accordance with the situation in other ornithomimids.

Situated anteriorly on the scapula, just dorsal to the scapulocoracoid suture, is a compressed, keel-
like prominence (text-fig. 2) that has been referred to as the ‘acromion process' by several authors
(Ostrom 1969, 1978; Osmolska and Roniewicz 1970; Cooper 1981). Its homology with that
tuberosity has not been demonstrated, however, and in the absence of clavicles the existence of an
acromion can only be surmised. For this reason we have chosen to refer to this structure as the
scapular prominence. It is very well developed in Struthiomimus (text-fig. 2) as it is also in
Ornithomimus and Gallimimus. It differs, however, from the situation found in most theropods,
where the scapular prominence is more pronounced anteriorly but does not extend as far dorsally
(text-fig. 3). The anterior edge of the scapular prominence is quite rugose and porous in texture,
suggesting the attachment of either muscle or ligament.

The glenoid fossa is deep and sellar. It is equally developed on both scapula and coracoid and has
prominent supra- and infraglenoid buttresses. The scapular portion of the glenoid bears an anteriorly
directed flange (text-fig. 2), representing an extension of the supraglenoid buttress that resists dorsal
deflection of the humerus during extreme humeral protraction.

NICHOLLS AND RUSSELL: ST R UTH I O M I M U S FORELIMB

645

text-fig. 1. a , the left pectoral girdle and forelimb of Strut hiomimus alius, UCMZ(VP) 1980.1 . The scale bar
represents 150 mm. b, detail of the left manus and wrist. The scale bar represents 50 mm.

646

PALAEONTOLOGY, VOLUME 28

table 1. Measurements of the pectoral girdle and forelimb of
Struthiomimus altus (UCMZ(VP)1980. 1).

Scapula— dorsoventral length 380*

—anteroposterior width at midshaft 46

—dorsoventral length of scapular prominence 62

— anteroposterior width of scapulocoracoid suture 88

Coracoid —dorsoventral length, posterior to glenoid 66

—anteroposterior width 179

— length of biceps tubercle 21

—height of biceps tubercle 9

Humerus— proximodistal length 362

—anteroposterior width at midshaft 40

Ulna— proximodistal length 256

—anteroposterior width at midshaft 17

Radius — proximodistal length 239

—anteroposterior width at midshaft 15

Metacarpals, proximodistal length

— 1 " 102
—II 109

-III 109

Phalanges, proximodistal length

-1-1 127

-1-2 (ungual) 95

— III 40

-11-2 113

-II-3 (ungual) 127*

— III-I " 24

— III-2 29

— III-3 89

— III-4 (ungual) 98*

Dorsoventral articular height

Ungual articular surface dimensions:

Anteroposterior articular width

1-2 21/16

II- 3 18/16

III- 4 17/15

All measurements are in mm, and refer to the left limb and girdle,
except for the coracoid which is represented by the right element.
Measurements marked with an asterisk (*) are estimates.

On the dorsal surface of the scapular lip of the glenoid is a narrow, oval depression. A similar
depression has been noted in other theropods, notably Deinocheirus (Osmolska and Roniewicz 1970)
and Gallimimus (Osmolska et al. 1 972) and probably represents the site of origin of the scapular head
of the triceps.

The coracoid is about three times as long as deep, the majority of the length being due to the
extension of the posterior coracoid process beyond the glenoid fossa (text-fig. 2). This process does
not terminate in a curved apex as it does in most theropods, but is truncated posteriorly. The coracoid
is thickest along the dorsal edge of the posterior coracoid process, where it forms a conspicuous
infraglenoid buttress.

On the lateral surface of the coracoid plate, ventral to the glenoid, is a pronounced elongate
tuberosity, the biceps tubercle (text-fig. 2; Table 1). This corresponds to the ‘coracoid tuber’ described
by Osmolska et al. (1972) for Gallimimus and to the ‘biceps tubercle’ of Deinonychus (Ostrom 1974)

NICHOLLS AND RUSSELL: ST RUT H IO M I M U S LORELIMB

647

text-fig. 2. Left scapulocoracoid of Struthiomimus altus, UCMZ(VP) 1980.1. Ventro-
medial curvature removed.

648

PALAEONTOLOGY, VOLUME 28

Struthiomimus Gallimimus

Albertosaurus

Allosaurus

Deinonychus

text-fig. 3. Representative theropod scapulocoracoids compared with that of Struthiomimus. All are
drawn to approximately the same length for ease of comparison. The scale bar represents 50 mm. In
Struthiomimus note the narrow, attenuated posterior coracoid process, the anterior flange on the
supraglenoid buttress, and the height of the scapular prominence. The latter two characteristics are shared
by Gallimimus. Diagrams are based on ROM 762 ( Albertosaurus libratus), ROM 5091 ( Allosaurus
fragilis ), and a cast of Gallimimus bullatus at the ROM. Data on Deinonychus from Ostrom (1974).

and the prosauropod Massospondylus (Cooper 1981). It is also well developed in Dromeosaurus
(TMP 79.20.1) and is present in most long-armed theropods. Ostrom (1974) suggested that the
relative size of the biceps tubercle may be related to forelimb length, but this seems unlikely in view of
Madsen’s (1976) comment that, at least in Allosaurus , the development of the biceps tubercle is
extremely variable.

There are three clearly defined areas of muscle attachment on the coracoid plate. The most
prominent of these is a triangular depression on the dorsal edge of the posterior coracoid process,
ventral to the infraglenoid buttress (text-figs. 2 and 11). The depression is very broad and deep,
narrows posteriorly, and its surface is quite smooth. This region is interpreted as being the site of
origin of the M. coracobrachialis brevis (see below).

The other two areas indicative of muscle attachment both lie on the lateral surface of the coracoid
plate. The first of these is a broad depression anterior to the glenoid (text-figs. 2 and 1 1 ). It overlies the
region of the scapulocoracoid suture and the coracoid foramen, extending from the ventral edge of
the scapular prominence to the biceps tubercle. This is interpreted as being the site of origin of the
M. supracoracoideus (see below).

Posterior to the biceps tubercle, the lateral surface of the long posterior coracoid process bears
a heavily striated scar (text-figs. 2 and 11), here interpreted as the site of origin of the M.
coracobrachialis longus (see below). Identical muscle scars were reported by Osmolska et al. (1972)
for Gallimimus.

In overall form the scapulocoracoid of S. altus (UCMZ(VP)1980.1) is very like that of the other
North American ornithomimids. It does, however, differ considerably from that of other theropods,
as noted by Sternberg (1933) in his description of O. edmontonicus. In most theropods the depth of the
coracoid plate greatly exceeds its length, and the posterior coracoid process is short, terminating in
a curved apex ventral to the glenoid (text-fig. 3). A long, shallow coracoid plate with an attenuated
coracoid process, an anterior flange extending from the supraglenoid buttress, and a high scapular
prominence are all characteristic of the ornithomimid scapulocoracoid. These features are also
present in the Mongolian ornithomimids G. bullatus (Osmolska et at. 1972) and Archaeornithomimus
asiaticus (USNM 6567, as figured by Gilmore 1933).

NICHOLLS AND RUSSELL: STRUTH JO M IM US FORELIMB

649

Deinocheirus , however, sometimes considered to be an ornithomimid (Ostrom 1 976c/, 1978), has
a more typical theropod scapulocoracoid. The scapular prominence is broken in Deinocheirus , but
appears to extend considerably far dorsally, as in ornithomimids. There is, however, no accessory
flange on the supraglenoid buttress and the coracoid is very deep dorsoventrally, and exhibits little
extension of the posterior coracoid process.

(c) Humerus

The left humerus (text-fig. 4) of UCMZ(VP)1980. 1 is complete, although the middle of the shaft has
been crushed dorsoventrally. It closely resembles the humerus of Gallimimus and Deinocheirus ,
although the deltopectoral crest is not as strongly developed as that of the latter genus.

The distal end of the humerus is set at an angle of approximately 40° to the proximal end, a higher
degree of torsion than is usual in ornithomimids. In Gallimimus this angle is 25-30° (Osmolska et al.
1972) and in NMC 8632, 12441, and ROM 840 it is closer to 20°. Osborn (1916) does not mention
the degree of torsion in AMNH 5339 (S. altus), although his figures 7 and 8 indicate that some
torsion is present. The high degree of torsion in UCMZ(VP)1980. 1 may be due to post-mortem
deformation, as humeral torsion is also high in ROM 851 (35-40°), which has been crushed in a
similar manner.

The anterior tuberosity is fully as high as the head and proximally bears an elongate articular
surface that juxtaposes the articular surface of the head (text-fig. 4). This accessory articular surface
also encroaches on to the dorsal surface of the humerus and is as well developed as the head. It fits
beneath the anterior flange of the glenoid during extreme humeral protraction. Distally, the anterior
tuberosity merges gradually with the deltopectoral crest. The latter is poorly developed compared
with that of other theropods, its apex being located less than one-fifth of the way along the humeral
shaft. It is set at an angle of about 40° to the proximal end of the humerus. On the dorsal surface of the
edge of the crest is a thickened lip which possibly marks the separation of the insertion of the
M. pectoralis and the M. supracoracoideus.

Posterior to the deltopectoral crest, on the dorsal surface, a shallow depression extends along the
humeral shaft, possibly marking the insertions of the deltoideus musculature.

The posterior tuberosity is only moderately developed. It does not project far out from the shaft, as
it does in Deinonychus or AUosaurus , but extends further along the shaft.

The distal end of the humerus is expanded into a pair of condyles. The ulnar, or posterior, condyle
is the larger of the two, extends the furthest distally and is bulbous and symmetrical in plan. The
anterior, (radial) condyle is narrow, elongated, and continuous with the ectepicondylar ridge. The
two condyles are separated ventrally by a broad fossa. Dorsally the olecranon fossa is present only as
a faint depression. The entire distal end of the bone has a rugose, porous surface texture, suggesting
the presence of extensive articular cartilage.

( d ) Ulna

The ulna of UCMZ(VP)1980.1 resembles that of 5. altus, as figured by Osborn (1916, fig. 8). It is
triangular in cross-section and gently curved, being convex toward the radius (text-fig. 5). The
olecranon process is long, extending 20 mm proximal to the articular surface of the radius. It is
significantly deflected from the ulnar shaft, and manipulation of the osteological preparation at the
elbow joint indicates that full extension of the forearm was possible. The concave articular facet for
receipt of the radius is deep and well developed. The form of both the proximal and distal radioulnar
articulations is suggestive of the presence of syndesmotic unions in life. Such joints, binding the
elements by way of collagenous fibres, would permit slight play between the elements but limit
rotatory ability.

The distal end of the ulna is crescentic and its anterior edge is flattened along its syndesmotic
contact with the radius. On the ventral surface are two condyles separated by a broad, shallow groove
(text-fig. 5). The anterior of these is only weakly developed, while the posterior one is larger. A small
convexo-concave pisiform is situated adjacent to the posterior condyle. The concave surface of the

650

PALAEONTOLOGY, VOLUME 28

50mm

text-fig. 4. Left humerus of Struthiomimus altus, UCMZ(VP)1980.1. Dorsal view.
M. supracoracoideus and Mm. deltoideus complex refer to the implied insertional areas for
these muscles (see text for details).

NICHOLLS AND RUSSELL: STRU T H lO M I M U S FORELIMB

olecranon process

text-fig. 5. Radius and ulna of Struthiomimus altus, UCMZ(VP)1980. 1 . a, dorsal

view; B, ventral view.

652

PALAEONTOLOGY, VOLUME 28

pisiform fits snugly against the posterior ulnar condyle, forming a rotational surface between the ulna
and the third metacarpal (text-fig. 6). The entire distal end of the ulna bears numerous striations,
suggesting the presence of articular cartilage.

(< e ) Radius

The radius is almost straight, except at the proximal end where it curves toward the ulna. The
proximal articular surface is oval and flat and thus fits perfectly the syndesmotic articular facet of the
ulna, allowing little rotation (text-fig. 5).

Distally the radius is oval in cross-section, except at its contact with the ulna where it is extensively
flattened. The distal end of the radius terminates almost 1 0 mm short of that of the ulna when the two
elements are articulated, a point not evident in Osborn’s (1916) illustration. This discrepancy in
length reflects the form and disposition of the carpals, most of which are concentrated distal to the
radius (text-fig. 6).

The distal articular surface of the radius is convex for articulation with the radiale, and the entire
distal end of the radius is heavily striated.

(/) Carpus

All of the carpals are excellently preserved, three of them adhering to the distal end of the radius and
one to the proximal surface of the metacarpals (text-fig. 6).

The radiale is ovoid in outline, convex distally and concave proximally where it fits the convex
surface of the radius. The distal surface of the intermedium is convex and its dorsal outline triangular.
The apex of the triangle forms a low ridge which extends along its proximal surface between the
radius and ulna. Distally, between the radiale and intermedium, is a small, disc-like bone, probably
a centrale. It thins rapidly toward its ventral surface and fits in a slight depression on the proximal end
of metacarpal I. The fourth carpal is extremely flattened and closely adherent to the proximal surface
of metacarpals I and II. It is so broad and irregularly shaped that it may represent two or more distal
carpals in fusion and is here interpreted as distal carpals 2 and 3. The fifth carpal bone is the pisiform,
already described under the consideration of the ulna.

Compared with the carpus of other theropods, that of Struthiomimus is most like that of
Albertosaurus , as described by Lambe (1917), although the distal carpals appear to be more
specialized. The carpus lacks the well-defined articular facets present in the carpals of Deinonychus
(Ostrom 1969), and to a lesser extent Allosaurus (Madsen 1976). Apparently the carpus of
Struthiomimus operated as a hinge-joint, permitting little or no rotation, but was not as 'stiff as
indicated by Gregory (in Osborn 1916).

Galton (1971) illustrated six carpals in Syntarsus and briefly pared them with the carpals of
Struthiomimus, although no attempt was made to describe them or to identify the individual
elements. The broad, flat anterior distal carpal of Syntarsus (Galton 1971, figs. 1 and 3) resembles the
fused distal carpals of Struthiomimus. The proximal carpals of Syntarsus, however, are much flatter
than the corresponding elements of Struthiomimus, and there is no concentration of the carpals distal
to the radius in Syntarsus.

Only one carpal is preserved in Gallimimus. This was considered by Osmolska et al. (1972) to be the
radiale, but it bears no resemblance to any of the carpals of UCMZ(VP)1980.1. In ROM 840 the
pisiform and radiale are preserved in situ and are like the corresponding bones of Struthiomimus altus.

(g) Manus

The manus of UCMZ(VP)1980.1 is very like that of AMNH 5339, as illustrated by Osborn (1916,
fig. 3). The three metacarpals are subequal in length, metacarpal I being only slighly shorter than the
others. All the metacarpals are tightly adpressed proximally and slightly arched. Distally metacarpal
I is strongly divergent and its articular surface is rotated anteriorly.

Two types of joint structure are present in the manus (text-fig. 7). The articulations between the
metacarpals and the proximal phalanges are of the ball and socket type, allowing considerable
flexion, extension, and rotational movement. The interphalangeal joints, in contrast, are ginglymoid

NICHOLLS AND RUSSELL: S T R UT H 1 0 M I M U S FORELIMB

653

I

50 mm

-i

text-fig. 6. Carpus of Strut hiomimus altus , UCMZ( VP) 1980.1 . a, dorsal view; b, distal view
of radius and ulna with attached carpals; c, proximal view of metacarpals with attached
carpals. Note, c is represented as a mirror image to keep the alignment for the three parts ol

the figure constant.

654

PALAEONTOLOGY, VOLUME 28

and permit extension and flexion, but little or no rotation, as is usual in the manus of theropods. In
the more typically raptorial theropods (e.g. Ornitholestes, Allosaurus, Chirostenotes ), groove and keel
articulations are present on the metacarpals of both digits I and II. The smoothly rounded distal
metacarpal articulation is found on all three digits only in ornithomimids. This type of joint surface
permits rotational movement and considerable hyperextension of the digits, but little or no flexion
below the horizontal. Most of the flexion in the manus occurred at the interphalangeal joints.

Digit I

Metacarpal I is closely applied to metacarpal II for a little more than two-thirds of its length, and
its posterior edge is flattened along this contact (text-fig. 7). Distally it diverges from the rest of the
metacarpus and the distal articular surface is rotated anteriorly and dorsally. Most of the articular
surface is smoothly convex. On the dorsal surface, however, there is a deep groove which guides the
phalanx anterodorsally (away from the other digits) on extension (text-fig. 7). On flexion, digit I
converges on the other two digits. Pits for the collateral ligaments are not as well developed as in
Osborn’s figure (1916, fig. 3). The pit is moderately developed on the posterior side (which is rotated
to face dorsally), but very poorly developed on the anterior side.

The distal articular surface of metacarpal I in Struthiomimus differs from that of Ornithomimus
(ROM 851, NMC 8632), in which metacarpal I is longer and not offset. In Ornithomimus the three
metacarpals are parallel throughout and their distal articular surfaces are rotated posteriorly. The
first metacarpal of Struthiomimus more closely resembles that of Gallimimus, which is also shortened
and rotated anteriorly.

The first phalanx of digit I is the longest phalanx in the manus. Its proximal articular surface is
concave and rotated anteriorly. Dorsally there is an enlarged tubercle which fits into the dorsal
groove on the distal end of its metacarpal. Distally the grooved articular surface encroaches
considerably on to the ventral surface of the phalanx, allowing the ungual to be flexed up to 65-70°
below the horizontal. Pits for the collateral ligaments are very deep.

Digit II

The second metacarpal is flattened anteriorly along its contact with the first. On the proximal
articular surface there is a low, broad ventral tubercle. During flexion of the wrist this tubercle fits
into the shallow groove between the condyles on the distal end of the ulna (text-fig. 6). Distally the
tubercle extends as a stout ridge on the ventral side of the metacarpal.

The distal articular surface of metacarpal II lacks any suggestion of the dorsal groove present on
metacarpal I. The articular surface is slightly asymmetrical, extending further on the posterior than
the anterior side. Pits for the collateral ligaments are only moderately developed.

In the first phalanx the proximal articular surface is smoothly concave, with ventral tubercles for
the attachment of the collateral ligaments. The anterior tubercle is much more strongly developed
than the posterior one. When articulated with the metacarpal, extension moves the phalanx dorsally
and posteriorly away from digit I, and flexion moves it anteriorly and ventrally toward it. The
grooved distal articular surface is well developed ventrally, allowing considerable flexion. Pits for
collateral ligaments are only moderately developed.

The penultimate phalanx in digit II is very long, being longer than the metacarpal. Pits for the
collateral ligaments are well developed and the distal articular surface is of the normal groove and
keel type, extending far on to the ventral surface, permitting considerable flexion of the ungual.

Digit III

The third metacarpal is very slender and is closely adherent to metacarpal II. The distal articular
facet is broadly rounded and symmetrically developed. Pits for the collateral ligaments are well
developed and open distally, forming a broad groove. On the first phalanx the anterior tubercle for
the collateral ligament is more strongly developed than the posterior one, making the proximal
articular surface asymmetric in the same manner as the corresponding phalanx in digit II. Extension
moves the phalanx posterodorsally.

NICHOLLS AND RUSSELL: STRUTHIOMIMUS FORELIMB

655

text-fig. 7. Manus of Struthiomimus altus, UCMZ
(VP) 1980.1. a, dorsal view of entire manus; b, distal
articular surfaces of the three metacarpals in natural
position; c, proximal articular surfaces of phalanx 1
for the three digits. Note, for ease of comparison c is
represented as a mirror image so that the alignment of
the digits in the three parts of the figure may be kept
constant.

post

post

post

656

PALAEONTOLOGY, VOLUME 28

The remaining phalanges in digit III all have the normal symmetrical phalangeal ridge and groove
articulations. The grooves are very deep and form tightly interlocking joints which permit no lateral
displacement. The pits for the collateral ligaments are correspondingly reduced. They are practically
non-existent in phalanges 1 and 2, although they are strongly developed in the penultimate phalanx.

Unguals

In UCMZ( VP)1980. 1 the ungual of digit I is complete but the extreme distal tips of unguals II and
III have been broken. The unguals are very long and, when covered with a horny sheath during life,
must have constituted more than one-third of the length of the manus. The unguals differ from the
narrow, highly curved talons of typically raptorial theropods. They are longer, straighter, and
broader, being slightly expanded proximoventrally. The articular surface covers the entire proximal
end of the ungual and is strongly keeled. The flexor tubercle is not situated directly ventral to the
articular surface as in most theropods, but is instead displaced distally about one-quarter of the
distance along the phalanx. This greatly enhanced the mechanical advantage of the ungual flexor
muscles.

All of the unguals were capable of being highly flexed, forming an angle of almost 70° with the long
axis of the penultimate phalanx. The ungual of digit I is the most trenchant of the three, being
narrower and more sharply curved than the other two.

The unguals of Strut hiomimus differ slightly from those of other ornithomimids. Compared to
Omithomimus , as defined by Russell (1972), the unguals of Struthiomimus are much more robust and
more strongly curved. (This, however, may be an allometric feature, as all the specimens identifiable
as Omithomimus are smaller animals.) In Gallimimus the unguals are shorter and curved, resembling
the first ungual of Struthiomimus. Osmolska et al. (1972) indicated that only a minimal amount of
flexion of the unguals was possible in Gallimimus , in contrast to the situation in Struthiomimus (see
above).

Isolated ornithomimid unguals were illustrated and discussed by Ostrom (1969, 1978) in his
consideration of the manus of Deinonychus and Compsognathus. The parameters he used to compare
the unguals of theropods were the somewhat equivocal ratio of the length (extension) of the ungual
relative to its height, and the angle formed between the cutting edge of the ungual and its arc of
rotation. The ornithomimid ungual he used for comparison was that of O. sedens , and his figure
clearly shows the long, straight, non-raptorial nature of the unguals of Omithomimus. In text-fig. 8
the unguals of UCMZ(VP)1980.1 are compared with those of other theropods. Employing Ostrom’s
(1969) criteria for claw form and function, the ungual of digit I appears ‘subraptoriaf and is
comparable to those of Ornitholestes and Compsognathus. The unguals of digits II and III are
decidedly non-raptorial.

Another parameter useful in comparing unguals is the width of the articular surface, relative to the
height, and this is a more reliable criterion for evaluating claw function than the degree of curvature.
The true curvature of a claw is formed by its horny sheath, which is seldom preserved in the fossil
record. In highly raptorial theropods the unguals are very narrow, with their articular height being
almost twice their articular width. As can be seen from Table 1 the unguals of Struthiomimus are very
broad, their width being almost equal to their height. Again this indicates that they are non-raptorial.

COMPARATIVE RESULTS-THE BASIS FOR A FUNCTIONAL ASSESSMENT
(a) Levels of comparison

In order to attempt to gain a mechanically feasible assessment of the form and possible functional
attributes of the breast-shoulder apparatus and forelimb of Struthiomimus, comparisons of various
types were made. Two ‘obvious’ comparative models come to mind. The first is a comparison with the
equivalent structures of crocodilians in an attempt to investigate, as far as possible, similarities due to
the existence of homologies (see, for example, Coombs 1978u; Gardiner 1982; Lauder 1981).

In overall body form, however, crocodilians and ornithomimid dinosaurs are relatively dissimilar
and, by inference, it seems reasonable to suspect that their way of life was relatively dissimilar also.

NICHOLLS AND RUSSELL: ST RU T H 1 0 M I M U S LORELIMB

657

Ornithomimus

v

Compsognathus

Ornitholestes

text-fig. 8. The manual unguals of Struthiomimus altus compared with those of other theropods. The
ungual of digit I is sub-raptorial and comparable with that of Compsognathus. The unguals of digits II and
III are decidedly non-raptorial. The horizontal scale represents 20 mm. a, S. altus , UCMZ(VP)1980.1,
digits I, II, and III; b, Gallimimus bull at us , cast at ROM; c, Ornithomimus edmontonicus , ROM 851;
D, C. longipes, adapted from Ostrom (1978); e. Deinonychus antirrhopus , adapted from Ostrom (1969);

F, Allosaurus fragilis, ROM 5091; G, Ornitholestes hermanni, adapted from Ostrom (1969).

This leads to the second comparison— that with ratite birds. Similarity of form here is not founded
upon congruence of homologies but on homoplastic resemblance (similarity due to convergence).
This comparison between ornithomimids and ratites can be considered to be somewhat ‘classical’ in
approach, but one in which the basic assumptions have never been tested. The similarity of the ratite
pectoral girdle to that of theropod dinosaurs has been most recently discussed by McGowan ( 1982).

To arrest the comparison at this point, however, and include only Struthiomimus , Alligator , and
Struthio cannot fail to produce the expected result- that the breast-shoulder apparatus of the former
is structurally and functionally similar to that of Struthio due to the great resemblance of form. Such
an outcome may or may not be reasonable, but the addition of a third comparison allows a more
objective assessment of the resemblances. This third comparison is with chameleons. Chameleons are
unusual among normal-limbed lizards in the morphology of their breast-shoulder apparatus and in
their mode of progression (Gasc 1963; Peterson 1984). Ostrom ( 1976a) stated that the narrow form of
the scapula found in Archaeopteryx occurs only in obligate bipeds (birds. Archaeopteryx , and
theropod dinosaurs). This is not accurate and, indeed, the primary shoulder girdle of chameleons
bears a striking resemblance to that of coelurosaurs, a resemblance that has not gone unnoticed in the

658

PALAEONTOLOGY, VOLUME 28

past (Peterson 1973; Bakker 1975). Such similarity, without the influence of potential ancestral-
descendant relationships affecting interpretation, forms the basis of the third level of comparison.

(. b ) Comparative material

The following specimens were dissected (numbers in parentheses refer to number of individuals):
Alligator mississippiensis* (1), Caiman sclerops* (1), Crocodylus niloticus (BMNH 62.1.24.52) (2),
Struthio camelus * (1), Dromiceius novaehollandiae* (2), Chamaeleo sp. (3), Chamaeleo jacksoni* (1).
Those specimens marked with an asterisk (*) form part of the University of Calgary Museum of
Zoology anatomical collection. Representative skeletal material for all taxa represented was also
examined.

(c) The breast-shoulder apparatus: general considerations

In order to be able to more fully appreciate and assess the structural and functional attributes of the
breast-shoulder apparatus of Strut hiomimus, it is necessary to first outline some general points of
shoulder structure.

The primary subject of this paper, Struthiomimus alt us , was apparently an obligate biped. Here the
forelimbs and breast-shoulder apparatus have been released from their traditional role in
quadrupedal locomotion and exhibit certain features associated with the relative freedom of the
limbs. Such differences reflect not only different functions but also different mechanical potentials of
the system.

In its most complete (primitive) form the breast-shoulder apparatus consists of paired primary
girdles (the scapulocoracoid complexes) of endoskeletal origin, the secondary girdle complex,
consisting of paired clavicles and a median unpaired interclavicle, of dermal origin, and an axial
endoskeletal component, the sternum (together with its associated ribs). The costosternal complex
forms an integral part of the breast-shoulder apparatus. As the primary girdle does not contact the
vertebral column, the secondary girdle and costosternal complexes act as a system of braces
preventing excessive displacement of the primary girdles but, at the same time, permitting a limited
amount of movement with respect to the body wall. The relative structure of the various components,
and the nature of the joints between them are, to a large extent, indicative of the functional potentials
of the breast-shoulder apparatus (Dvir and Berme 1978).

The release of the forelimb from weight-bearing and its retention as a well-developed structure
in an obligately bipedal, non-brachiating form such as Struthiomimus has influenced the structure
of the breast-shoulder apparatus considerably. The multiple comparisons discussed below attempt
to place the form of the apparatus seen in Struthiomimus into a biomechanically consistent
framework.

(d) Osteological comparative material

(i) Pectoral girdle of Alligator (text-fig. 9)

The breast -shoulder apparatus of crocodilians consists of the scapula, coracoid, interclavicle, and
a costosternal system (Kalin 1929). The clavicle is absent and the scapula and coracoid are not fused
into a scapulocoracoid plate. The scapula is flattened and blade-like, considerably broader than the
scapula of the other forms being discussed, but is only slightly expanded distally. The blade faces
laterally and is gently curved to fit the contours of the body wall. The anterior scapular prominence
extends well above the level of the glenoid, as in Struthiomimus. Posteroventrally the scapula
broadens into a stout supraglenoid buttress, on the dorsal lip of which is an oval, roughened area
representing the point of origin of the scapular head of the triceps.

The glenoid fossa, to which both the scapula and coracoid contribute equally, is located at the
posterior edge of the girdle and faces posterolaterally. The infraglenoid buttress of the coracoid is
situated ventral, anterior, and slightly medial to the supraglenoid buttress.

The coracoid is short and deep, its greatest dimension being in the dorsoventral plane. There is no
biceps tubercle. The infraglenoid buttress is strongly developed and is situated ventral, anterior, and
slightly medial to the supraglenoid buttress. This part of the coracoid plate faces laterally.

NICHOLLS AND RUSSELL: ST RUT H IO M I M U S FORELIMB

659

Chamaeleo Alligator Struthio

text-fig. 9. Basic form of the primary girdle of the three comparative models discussed in
the text. All are viewed from the developmentally lateral aspect regardless of their actual
orientation in life. The torsion of the primary girdle of Struthio has been artificially removed
for ease of presentation. Upper row, primary girdles only; lower row, girdles with major
ligament systems superimposed. Ligaments outlined in dashed lines and with their expanse
stippled are located on the medial face of the girdle. Each girdle has been drawn to the same
dorsoventral linear dimension for ease of comparison. Abbreviations: ant., anterior; lig.
coll, sterncor., ligamentum collateralia sternocoracoidea; lig. sterncor. lat . , ligamentum
sternocoracoideum laterale; M. coractric., M. coracotriceps; M. scaptric., M. scapulotriceps;

med., medial.

Ventral to the glenoid the coracoid curves ventrally and medially to meet the sternum. Posteriorly
it forms a small but distinct posterior coracoid process. The ventral edge of the coracoid abuts the
lateral edge of the sternum in a frontal plane. The coracosternal angle (the angle the coracosternal
articulation makes with the sagittal axis of the sternum) is low.

(ii) Pectoral girdle of Struthio (text-fig. 9)

In most birds the forelimb is modified for flight and the morphology of the sternum, coracoid, and
limb elements are considerably altered from the ‘typical’ tetrapod form. In ratites flight has been

660 PALAEONTOLOGY, VOLUME 28

secondarily lost (Cracraft 1974), the scapula and coracoid are fused into a single plate and the
humerus is less severely rotated.

In Struthio the secondary girdle is absent. The pectoral girdle consists only of the scapula and
coracoid, fused into a single plate. In this respect the shoulder girdle of Struthio resembles that of
Struthiomimus and other theropods, although no suture is present between the two bones. The distal
parts of the forelimb of Struthio are much reduced and the shoulder structure is reflective of this. The
glenoid fossa is very small relative to the size of the girdle and faces laterally. There is a marked flexure
between the scapula and the coracoid, with about 60° of torsion between the plane of the two bones.

Deserving of particular mention is the orientation of the scapulocoracoid plate in life. Whereas in
Alligator the coracoid is elongated and oriented ventromedially to abut the coracoid sulcus of the
sternum, the coracoid of Struthio , as is the case in birds in general, has been rotated anteromedially, in
association with the reorientation of the coracoid sulcus of the sternum. Here the coracoid abuts the
functionally anterodorsal aspect of the sternum and the coracosternal angle is very high. The broad
coracoid plate is thus oriented anterolaterally and, as a consequence, the glenoid has come to lie in
a much more lateral position. Apart from its torsion the scapulocoracoid plate is essentially planar,
the coracoid showing little angulation. The ventral border of the coracoid is straight, not curved, and
the posterior coracoid process is reduced.

The coracoid plate is very broad and triangular. There is a large centrally located foramen, but it is
not homologous with the coracoid foramen of reptiles and is a derived feature of ratites (Cracraft
1974). Broom (1906) indicated that, developmentally, it is formed by an anterior extension of the
scapula (‘prescapular process’) which extends ventrally to join the coracoid at the sternum.

Two distinct tuberosities are present on the scapulocoracoid plate of Struthio. The larger of the two
was referred to as the ‘coracoid tuber’ by Cracraft (1974) and was stated to be unique to some ratites.
Broom ( 1 906) referred to this tuberosity as the ‘acromion process’ and McGowan ( 1 982), in his work
on the shoulder girdle of kiwis, referred to it as ‘the acromial tuberosity’. It is the site of origin of much
of the deltoid musculature and it is here referred to as the ‘scapular prominence’, although its
homology with the scapular prominence in other aclaviculate forms has not been demonstrated.

The second tuberosity is situated on the coracoid, ventral and slightly anterior to the glenoid. It is
oval in shape with its long axis oriented dorsoventrally . McGowan ( 1 982) calls a similar tuberosity in
the kiwi the ‘acrocoracoid process’. This structure, by correlation with muscle origins, is the biceps
tubercle. It is situated much closer to the glenoid than is the biceps tubercle in Struthiomimus.

(iii) Pectoral girdle of Chamaeleo (text-fig. 9)

In outline the pectoral girdle of Chamaeleo resembles that of both Struthio and theropod
dinosaurs. The secondary girdle is absent and the pectoral girdle consists only of a scapula and
coracoid, which are fused into a single scapulocoracoid plate. Skinner (1959) reported the transient
appearance of clavicles and an interclavicle during chameleontid development, but stated that they
disappeared rapidly.

The scapula is a long, thin rod which lacks fenestrae. Ventrally it expands into a scapular
prominence anteriorly and the glenoid fossa posteriorly. The scapular prominence is well developed
(Siebenrock 1893; Skinner 1959), but does not extend far dorsally, reaching only a little above the
level of the supraglenoid buttress. The scapula contributes to a little over one-third of the glenoid
fossa and there is a well-developed supraglenoid buttress.

The remaining two-thirds of the glenoid is formed by the unfenestrated coracoid. There is a well-
developed infraglenoid buttress which is situated ventral and slightly medial to the supraglenoid
buttress. The coracoid is relatively flat, lies in the same plane as the scapula, and faces laterally. The
ventral edge is extended posteriorly to form a posterior coracoid process. While this posterior
coracoid process is not as well developed as in Struthiomimus , it is more prominent than in either
Struthio or Alligator and is comparable to that of many theropods (e.g. Dromeosaurus r TMW19 .29 A ,
Alhertosaurus NMC 2120).

Anterior to, and considerably ventral to, the glenoid is the biceps tubercle. It is in a similar position
to the biceps tubercle in Struthiomimus , although it is not as prominent as in that genus. It is not at all

NICHOLLS AND RUSSELL: STRUTH 10 M IM US FORELIMB

661

like the biceps tubercle of Struthio , which is an elongate prominence situated much closer to the
glenoid.

The coracoid abuts the anterolateral edge of the sternum via a dorsally facing coracoid sulcus. The
coracosternal angle is low, being about 30°.

Lecuru (1968u, b) distinguished particular features of the lacertilian breast-shoulder apparatus
associated with arboreal locomotion, including reduction in the number of scapulocoracoid
fenestrae, a tall, narrow scapular blade, modifications of the anteroventral border of the coracoid,
and a relatively ventral acromion process (or scapular prominence). Such features reflect adaptation
for mobility of the primary girdle on the body wall (Peterson 1973).

( e ) Ligament systems of the breast-shoulder apparatus

The ligaments of the breast-shoulder apparatus are seldom considered in studies of the shoulder
region, but form an extremely important part of this apparatus when considered as a functional
complex. Indeed, consideration of the breast-shoulder apparatus without consideration of the
ligament systems means that the functional potential of this apparatus cannot be fully appreciated.
Obviously, such systems cannot be reconstructed for fossil forms in any detail, but an appreciation of
their architecture in living forms permits some predictive statements to be made.

(i) Alligator

In crocodilians a stout anterior scapulosternal ligament is present (text-fig. 9). It arises from the
ventral aspect of the scapular prominence and from here fans out as it passes ventrally. Anteriorly it is
thickened and forms a stout band which attaches to the anteriormost extremity of the interclavicle.
From here it passes posteriorly as a thin sheet which attaches to the interclavicle and the ventral
border of the coracoid sulcus of the sternum. It restricts the degree of excursion that can occur at the
coracosternal articulation, especially when the humerus is depressed and the limbs become semi-
erect. The stout anterior band has the orientation of a clavicle but represents a tensile rather than
a compressive structure. Muscular origin from the anterior scapulosternal ligament is meagre.

The medial scapulosternal ligament is continuous with the anterior one on the medial face of the
primary girdle. The medial ligament spans the dorsal aspect of the coracosternal articulation and
tapers as it passes dorsally across the medial face of the coracoid. At the point where the coracoid
curvature is most pronounced it separates from the anterior ligament and attaches to the medial face
of the coracoid. From here a slender strand continues dorsally across the coracoscapular joint,
passing anterior to the glenoid, and attaches to the medial face of the scapula on its posterior aspect,
relatively high up on the shaft. Just dorsal to the glenoid a band of tissue associated with the origin of
the M. scapulotriceps diverges at right angles from the main course of the ligament, and slightly
ventral to this are bands associated with the M. coracotriceps.

(ii) Struthio

The ligaments of the breast-shoulder apparatus of birds are complex and their nomenclature
profuse (see Baumel 1979, pp. 148-151). Essentially, however, the arrangement of ligaments about the
coracoid is quite similar to that found in Alligator. The furcula, when present, is involved, but the
absence of this structure in Struthio relieves some of the complication.

Two primary sheets can be recognized. The membrana sternocoracoclavicularis extends from the
rostral border of the sternum and crosses the coracosternal articulation to attach to the anterior part
of the coracoid and the region of the relatively reduced scapular prominence (text-fig. 9). The
membrana sternocoracoclavicularis can be topographically equated with the anterior scapulosternal
ligament described for Alligator (above). In the case of Struthio the interclavicle is absent and the
orientation of the coracoid on the sternum is different, but essentially the same topographical points
are interconnected. The ligamentum sternocoracoideum laterale of Struthio governs the ventrolateral
aspect of the coracosternal articulation (text-fig. 9).

The developmentally dorsal lip of the coracoid sulcus of the sternum is spanned by the ligamentum
collateralia sternocoracoidea in Struthio (text-fig. 9). It passes on to the medial face of the coracoid

662

PALAEONTOLOGY. VOLUME 28

for much of its length. There is no association with the M. scapulotriceps and no extension up on to
the scapular shaft. The M. coracotriceps, generally a vestigal muscle in birds (Berger 1966), is absent
in Struthio and thus also has no association with the ligamentum collateralia sternocoracoidea.

Topographically the ligamentum collateralia sternocoracoidea of Struthio occupies the same basic
position as the medial scapulosternal ligament of Alligator (above), but morphologically it is more
equivalent to the same ligament of Chamaeleo (see below). The three ligaments together (membrana
sternocoracoclavicularis, ligamentum sternocoracoideum laterale, and ligamentum collateralia
sternocoracoidea) govern the mobility at the coracosternal articulation. Mobility of this joint in
birds, however, differs from that typically seen in reptiles and the implications of this with respect to
coracoid shape and orientation will be more fully considered below.

(iii) Chamaeleo

In Chamaeleo the coracoidal arm of the medial scapulosternal ligament of other lizards is absent,
permitting greater mobility at the coracosternal articulation (Peterson 1973) (text-fig. 9). There is no
connection between the forearm extensor musculature and the medial scapulosternal ligament and
the absence of a secondary girdle has, as a correlate, the absence of the mesocleidosternal ligament of
other lizards (Peterson 1973).

Among lizards, only in chameleons has an anterior scapulosternal ligament been reported. It is
very similar in form to that described for Alligator (see above), but no interclavicle is present. The
anterior scapulosternal ligament arises from the scapular prominence and passes ventrally along the
anterior margin of the scapulocoracoid plate, becomes free of the anterior margin of the girdle, and
passes slightly anteroventrally to meet its fellow of the opposite side in the ventral midline. The
bilateral ligaments fuse to give rise to a short sagittal ligament which passes posteriorly to attach to
the ventral lips of the sternal grooves where they approach each other. Broad bands of fascia connect
the transverse and longitudinal arms of the ligament (Peterson 1973).

(/) Comparative myological material

No attempt has been made to reconstruct all the muscles of the pectoral region of Struthiomimus;
rather we have restricted our work to those muscles for which there is good evidence in the form of
muscle scars. The muscles considered are the following: (i) M. deltoides scapularis; (ii) M. deltoides
clavicularis; (iii) M. supracoracoideus; (iv) M. coracobrachialis; (v) M. biceps brachii; (vi) M.
scapulotriceps. Terminology used is that of Romer (1944), unless otherwise noted.

(i) M. deltoides scapularis (text-fig. 10a) (M. teres major, Haughton 1867a; M. dorsalis scapulae,
Fiirbringer 1876)

In Alligator this muscle arises from the anterolateral surface of the scapular blade. It has
a tendinous insertion on the anterodorsal surface of the humerus, just distal of the head.

In Struthio the scapulodeltoid (M. deltoides major, Berger 1960) arises from the scapular
prominence and the anteromedial edge of the scapulocoracoid plate. There is a small secondary head
arising from the biceps tubercle. It inserts along the dorsal surface of the humerus, extending more
than half way along its shaft. Similar origin and insertion patterns, except for the small accessory

text-fig. 10. The form of the muscles discussed for the three comparative models. All are viewed from the
(primitively) lateral aspect of the primary girdle. The humerus has been represented in simplified form as
a cylindrical rod. a, the Mm. deltoideus complex; b, M. coracobrachialis with the humerus represented
in a protracted and depressed attitude. Dashed portions represent parts of the muscle lying ventral to the
humerus; c, M. supracoracoideus; d, M. biceps brachii. Dashed portions represent parts of the muscle lying
ventral to the humerus; e, M. scapulotriceps. Abbreviations: brev., brevis; long., longus; M. corachum. ant.,
M. coracohumeralis anterior; M. coracobrach., M. coracobrachialis; M. delt. clav., M. deltoides clavicularis;
M. delt. scap., M. deltoides scapularis; M. humerorad., M. humeroradialis; M. sternohum. ant., M.
sternohumeralis anterior; M. supracorac., M. supracoracoideus; M. suprascap., M. suprascapularis; pars

corac., pars coracoideus; pars scap., pars scapularis.

NICHOLLS AND RUSSELL: STRUTHIO MIM US LORELIMB

663

664

PALAEONTOLOGY, VOLUME 28

head, which appears not to have been reported previously, are described for the emu and rhea by
Haughton (18676, c). The accessory head may be the homologue of the M, deltoideus minor, caput
ventrale of carinate birds (Van den Berge 1979, p. 200).

The absence of an origin of the scapulodeltoid from the scapular blade and the increased
insertional length of this muscle are typical avian characteristics (Berger 1960). The origin of this
muscle has been brought to lie in a plane which essentially runs along the scapulocoracoid suture and
the glenohumeral joint. In so doing its leverage has been altered. Its involvement in humeral elevation
is decreased but its part in protraction is enhanced. Its transformation in birds may be associated with
the reorientation of the scapulocoracoid and the relative immobility of this element.

In Chamaeleo the scapulodeltoid arises from the anterolateral surface of the scapular blade and
inserts at the proximal end of the deltopectoral crest of the humerus.

(ii) M. deltoides clavicularis (text-fig. 10a) (M. deltoides scapularis inferior, Fiirbringer 1876;

M. scapulohumeralis anterior, Romer 1922; Coombs 1978a).

None of the forms under consideration have a clavicle. In Alligator this muscle arises from the
external surface of the scapular prominence, crosses the glenohumeral joint, and inserts on to
the dorsal surface of the humerus, medial to the deltopectoral crest. Its insertion interdigitates with the
origin of the M . humeroradialis ( Alix 1 874), a developmental derivative of this muscle ( Romer 1 944).

In birds this muscle is known as the M. tensor propatagialis brevis (Van den Berge 1979), or the
Mm tensores patagii longus et brevis (Berger 1966). It is absent in Struthio , possibly in association
with the loss of flight and the reduction of the forelimb and patagium.

In Chamaeleo the clavicular deltoid is considerably altered from the usual lacertilian situation and
has been subdivided into the M. coracohumeralis anterior and the M. sternohumeralis anterior
(Skinner 1959). Their continuous origin is from the anterior scapulosternal ligament, and their
insertion is on to the dorsal surface of the deltopectoral crest.

(iii) M. coracobrachialis (text-fig. 10 b)

In crocodilians the M. coracobrachialis longus is usually considered to be absent (Fiirbringer 1876;
Romer 1944; Holmes 1977). The most prominent component of the M. coracobrachialis (M.
coracobrachialis brevis, Romer 1944) takes origin from much of the lateral surface of the coracoid
plate. It inserts on the proximal ventral surface of the humerus, between the deltopectoral crest and
the posterior tuberosity. Dissection of both Crocodylus and Alligator , however, indicates that this
muscle arises by way of two heads — the M . coracobrachialis brevis (described above) arising from the
broad, external surface of the coracoid plate, and the M. coracobrachialis longus having its origin
from the posterior coracoid process. The muscle can also be separated into two heads at its insertion
on the humerus.

In Struthio the single coracobrachialis (M. coracobrachialis externus, Romer 1944; M. coraco-
brachialis posterior, Berger 1960; M. coracobrachialis cranialis, McGowan 1982) is considered to be
homologous, at least in part, to the M. coracobrachialis brevis of reptiles (Romer 1944). It arises on
the posterolateral edge of the coracoid plate, ventral to the glenoid, and inserts on the ventral surface
of the proximal end of the humerus, between the deltopectoral crest and the posterior tuberosity. In
Struthio it is strongly developed.

The reorientation of the coracoid on the sternum, the reduction of the posterior coracoid process,
and the lateral orientation of the glenoid have had a profound influence on the functioning of the
M. coracobrachialis in Struthio. Leverage in humeral retraction is markedly reduced while its role in
depression has been enhanced. As in Chamaeleo (see below) it has essentially become a part of the
glenoid cuff musculature, playing a role in control at the glenohumeral articulation.

In Chamaeleo the M. coracobrachialis longus arises from the posteolateral surface of the coracoid
plate and inserts on the entepicondyle of the humerus. The regression of the posterior coracoid
process has reduced its leverage in humeral retraction. The M. coracobrachialis brevis arises on the
posterior edge of the coracoid, ventral of the glenoid. It inserts on the ventral surface of the humerus,
about half-way along the shaft.

NICHOLLS AND RUSSELL: STRUTHIOMIMUS FORELIMB

665

(iv) M. supracoracoideus (text-fig. 10c)

In Alligator the supracoracoideus is divided into two heads. The pars scapularis (Furbringer 1876)
arises on the lateral surface of the scapula, ventral to the scapular prominence. The pars coracoideus
(Furbringer 1 876) arises from the medial surface of the coracoid, ventral to the scapular prominence.
It curves around the anterior edge of the coracoid to join with the pars scapularis. Together they
insert on the deltopectoral crest of the humerus, medial to the insertion of the M. pectoralis.

In Struthio the M. supracoracoideus arises from the broad, anterior surface of the coracoid plate,
covering the large coracoid foramen. It soon becomes tendinous, its tendon passing between the
scapular prominence and the biceps tubercle. The tendon inserts on the dorsal surface of the
humerus, just distal to the head. The importance of the biceps tubercle in supporting the tendon of
the M. supracoracoideus in birds was noted by Walker (1972) and Ostrom (19766). The particular
morphology of the M. supracoracoideus in Struthio may again be related to the secondary absence of
flight capabilities. Its role is in humeral elevation and protraction, but its effectiveness in the former is
limited by the relatively ventral position of the biceps tubercle. Similar descriptions of this muscle
have been furnished by Haughton (18676, c) for the emu and rhea.

In Chamaeleo the supracoracoideus is subdivided into a dorsal M. suprascapularis and
a ventral M. supracoracoideus (Skinner 1959; Peterson 1973). The suprascapularis originates
on the anterolateral surface of the scapular ramus, adjacent to the anterior border of the M.
scapulodeltoideus. The ventral M. supracoracoideus takes origin from the anterolateral surface of
the coracoid. The two branches have a common insertion at the humeral head, in the plane of the
glenohumeral joint.

(v) M. biceps brachii (text-fig. 10 d)

There is little variation in the biceps in all the forms considered. It arises on the external surface
of the coracoid, anteroventral to the glenoid, and inserts on the proximal end of the radius and
ulna. In Struthio it is considerably reduced in size, probably in association with the reduction in
size of the antebrachium. Macalister (1867) reported a separate slip of this muscle arising from the
M. coracobrachialis in Struthio , but we did not locate this, and neither did Haughton (1867c) in his
examination of the rhea. In Struthio and Chamaeleo a distinct biceps tubercle is present.

(vi) M. scapulotriceps (text-fig. 10 e)

In Alligator this muscle (M. anconeus scapulae lateralis externus, Furbringer 1876; M. triceps caput
scapularis, Romer 1922) has a complex origin. It arises by way of three tendons: from the medial
surface of the scapula, dorsal to the glenoid; from a branch of the medial scapulosternal ligament;
and from the lateral surface of the scapula, at the supraglenoid buttress. It inserts on the olecranon of
the ulna.

In both Struthio and Chamaeleo the M. scapulotriceps arises by a single head from the lateral
surface of the scapula, dorsal to the glenoid fossa. It inserts on the olecranon process. In Struthio it is
greatly reduced in size and takes origin considerably further dorsal on the scapula with respect to the
glenoid. In Chamaeleo origin is close to the glenohumeral joint, but has no connection with the medial
scapulosternal ligament. This lack of connection with the ligaments of the breast-shoulder apparatus
is a major contributory factor to the enhancement of forereach.

DISCUSSION

(a) The form and orientation of the primary girdle of Struthiomimus in the context of the
comparative models

On reviewing the three comparative models, it is immediately apparent that, as expected, there are
few overt similarities between the shoulder girdle of Struthiomimus and that of Alligator. The
shoulder girdle of Alligator must fulfil the role of both locomotion and support of the animal, and
neither of these demands apply to the bipedal Struthiomimus.

666

PALAEONTOLOGY, VOLUME 28

In comparing the shoulder girdle of Struthiomimus with that of Struthio and Chamaeleo , a number
of similarities are apparent. In all three forms the secondary girdle is absent, there is a single
scapulocoracoid plate, a long thin scapular blade, and a biceps tubercle.

On closer comparison, however, these similarities in Struthio appear to be rather superficial. The
scapulocoracoid of Struthio consists of a very large coracoid plate but a greatly reduced scapula. This
is reflective of the reduced forelimb in the ostrich. The coracoid faces anteriorly and there is about 60°
of torsion between the planes of the scapula and the coracoid. The coracosternal articulation is
basically a hinge-type structure (Baumel 1979) with a markedly transverse orientation in association
with a similar orientation of the coracoid sulci. Manipulation of articulated elements and reports of
rhea (Porteilje 1925; Raikow 1969, fig. 3) and ostrich (Sauer and Sauer 1966, figs. 16, 18, and 20)
behaviour indicate that the mobility of the forelimbs, important in courtship and aggression, is
restricted to the glenohumeral joint and joints distal to this. The ability of the coracoid to slide in the
coracoid sulcus is severely limited, and is tightly bound by the ligaments of the breast-shoulder
apparatus.

In contrast the scapulocoracoids of both Struthiomimus and Chamaeleo have well-developed
scapulae. There is no torsion and the entire scapulocoracoid faces laterally. Indeed the pectoral girdle
of Chamaeleo so closely approaches the typical theropod condition that the only obvious difference is
that of size. The pectoral girdle of Chamaeleo , however, differs markedly from that of terrestrial,
sprawling lizards. In terrestrial forms, such as Iguana , the secondary girdle is present and the
scapulocoracoid is shorter, broader, fenestrate and the coracosternal angle is high. These differences
were discussed by Peterson (1971, 1973), who pointed out that the features typical of Chamaeleo are
associated with the mobility of the scapulocoracoid during arboreal locomotion.

Mobility of the scapulocoracoid plate relative to the sternum plays an important role in the
locomotion of many lizards (see Jenkins and Goslow 1983 for an account of locomotion in Varanus )
but it is particularly well developed in chameleons. The forelimbs of chameleons have been brought
closer under the body, in what Bakker (1971) called the ‘semi-erect’ stance. The absence of a clavicle,
the low coracosternal angle, and the modifications of the ligaments of the breast-shoulder apparatus
permit the coracoid to slide anteriorly and posteriorly in the coracoid sulcus, concomitantly rotating
the girdle in the parasagittal plane. The long scapular blade of chameleons increases the leverage of
the muscles that attach dorsally and rotate the girdle. The outcome of girdle rotation in Chamaeleo is
to increase forereach of the pectoral limb as the animal moves through a discontinous network of
branches (Peterson 1973, 1984).

The relevance to dinosaurs of Peterson's work on chameleonid lizards was recognized by Bakker
(1975), but he applied it to quadrupedal dinosaurs. These would incur problems of weight support
irrelevant to chameleons because of their small size, and unencountered in bipedal dinosaurs. The
biomechanical implications of scapular rotation in quadrupedal dinosaurs were evaluated by
Coombs (19786) and found to be incompatible with the basic morphology of these forms. In
Struthiomimus , however, the forelimbs are freed from the role of locomotion and weight support.

Considerable mobility also exists in the primary girdle of crocodilians. As in both Chamaeleo and
Struthiomimus , the primary girdle faces laterally and the coracosternal angle is low. At moderate
speeds crocodilians exhibit the high walk, essentially a trotting gait (Sukhanov 1 968; Whetstone and
Whybrow 1983). With increasing speed the high walk gives way to the gallop (Zug 1974; Webb and
Gans 1982). Lateral bending in the trunk is not pronounced, this being reflected in the nature of the
intervertebral articulations (Hofstetter and Gasc 1969). The intergirdle distance is relatively short
(only 15 vertebrae between girdles) and the animals are short-coupled (Peabody 1959). In the relative
absence of whole body movements which shift the primary girdle (Daan and Belterman 1968), the
rotation of the girdle in the parasagittal plane, with respect to the sternum, performs a similar
function.

Thus, the scapulocoracoid plate in crocodilians is mobile in association with speed, while in
chameleons it is associated with movement through a discontinuous substrate (Peterson 1984). In
both cases, however, the end result is increased forereach.

Another example taken from within the lizards serves to corroborate the utility of the comparisons

NICHOLLS AND RUSSELL: ST RUTH IO M IM US FORELIMB

667

based upon Chamaeleo. Most geckos are dorsoventrally depressed lizards that exhibit sprawling
locomotion with the proximal segments of the limbs held out to the sides of the body (Russell 1975).
In one genus, however, there has been a remarkable departure from this typical facies —Uroplatus is
quite chameleon-like in its habits and overall appearance (Angel 1 942). In association with this many
of the attributes of the breast-shoulder apparatus of Chamaeleo are duplicated or closely approached
(Wellborn 1933, pp. 159-160, 193-196, fig. 38). Our interpretation of the breast-shoulder apparatus
of Struthiomimus indicates similar structural attributes for promoting forereach and forelimb
mobility, particularly in the anterior quadrants of the glenohumeral joint.

The problem of orientation of the scapulocoracoid plate on the body wall in theropods has been
a persistent one. Ostrom (1974) discussed the scapulocoracoid of Deinonychus and proposed an
orientation for it similar to that in modern birds, with the scapulocoracoid oriented anteriorly, as in
Struthio. No consideration, however, was given to the problem of why the coracosternal articulation
should be immobile in a form with highly raptorial forelimbs, and it is clear that this would impose
severe limitations on forereach. Subsequently, he has represented the scapulocoracoid orientation of
Deinonychus in both a Chamaeleo- like manner (Ostrom 1976c, fig. 2) and an anterior orientation, as
in birds (Ostrom 1976#, b). The latter orientation has been disputed by Tarsitano and Hecht (1980).

The long forelimb and well-developed manus of Struthiomimus are inconsistent with an immobile,
anteriorly oriented scapulocoracoid. Taken over all, and in comparison with recent models, the
osteological evidence suggests that the primary girdle of Struthiomimus was oriented somewhat
laterally, as in Chamaeleo , and that considerable excursion was possible with respect to the body wall.

In UCMZ(VP)1980. 1 parts of the vertebral column and ribs are preserved. As articulated, the
scapular blade curves dorsally and posteriorly, making an angle of approximately 35° with the axis of
the preserved vertebrae. The ventral edge of the coracoid plate lies in the same plane as the gastralia
and ossified xiphisternal processes (Nicholls and Russell 1981) and this was probably its orientation
during life. It is consistent with orientation of the scapulocoracoid in other articulated ornithomimids
(ROM 851, NMC 8632, AMNH 5339). In this orientation the medial surface of the scapular blade
lies flat against the dorsal ribs and the blade faces laterally. The ventral edge of the scapular
prominence curves medially, resulting in an anterior inclination of the scapulocoracoid plate anterior
to the glenoid. The glenoid fossa itself faces posterolaterally.

In the region of the scapulocoracoid suture the lateral face of the coracoid plate faces
anterolaterally (see above). In the region of the biceps tubercle, however, the coracoid plate curves
medially and the external face of the posterior coracoid process faces ventrally while its dorsal edge
faces laterally. Consequently the coracoid plate of Struthiomimus lies in two distinct planes:
anterolaterally, ventral to the scapular prominence and ventrally, in the region of the posterior
coracoid process.

The same condition is present in NMC 8632, 8902 and ROM 851, 840. Tarsitano and Hecht (1980),
in their discussion of Archaeopteryx , state that the coracoid of this genus is more complex than that of
other reptiles, in that it lies in two different planes. In fact, the flexure of the coracoid plate that they
describe for Archaeopteryx is very similar to that present in Struthiomimus. Tarsitano and Hecht’s
statement ( 1 980, p. 163) that the coracoid of theropods ‘is a fairly simple plate lying essentially in the
same plane as the scapula’ is incorrect. Indeed the coracoid of many theropods is curved medially,
although not to the extent of that of Struthiomimus.

The structure of the sternum in Chamaeleo is suggestive of a possible solution for the orientation of
the primary girdle in Struthiomimus , being compressed, with dorsally turned coracoid sulci. A sternum
of this basic type was figured for a specimen of Albertosaurus (NMC 2120) by Lambe (1917, figs. 29
and 30). Along the anterior edges of this are notches that appear to represent the coracoid sulci. The
curved nature of this sternum (Lambe 1917, fig. 29a) places the sulci in an anterolateral orientation,
directed dorsally, quite similar to the situation in Chamaeleo. Unfortunately, the restoration of the
sternum (Lambe 1917, fig. 30) represents it as a flat plate, artificially placing the coracoid sulci in an
anterior orientation, reminiscent of that of birds. That the sternum was a curved or angulated
structure is much more likely, however, as only in the former case could the scapular blade take up its
orientation against the ribs, as indicated by Lambe (1917, figs. 5, 7, and 49). A similar form for the

668

PALAEONTOLOGY, VOLUME 28

sternum of Massospondylus is hinted at by Cooper (1981). The coracoid of Albertosaurus is discussed
by Lambe (1917, p. 47) as being a curved structure. The medial inflection of the coracoid plate would
have permitted articulation with the coracoid sulcus, provided a sliding articulation, and placed the
scapulocoracoid plate in an appropriate orientation to follow the contour of the body wall . In such an
orientation the glenoid would have faced posterolaterally but would have been relatively laterally
situated, as in Chamaeleo. Orientation of the glenoid into a relatively more lateral situation provides
the potential for greater anterior excursion of the forelimb, allowing the humeral head to move more
freely into the anterior quadrants of the glenoid.

It is likely, given the structure of the primary girdle and the nature of its ventral coracoid surface,
that the sternum of Struthiomimus was similar in form to that described for Albertosaurus (above).
Suggestions that the coracoids of theropods may have overlapped along the midline in life (referred
to as arcifery and reminiscent of the pectoral structure in arciferal frogs) (Osmolska and Roniewicz
1970), seem anatomically untenable.

Given this mode of orientation of the scapulocoracoid plate, comparative analysis suggests that
the ligament systems of the breast-shoulder apparatus of Struthiomimus probably bore most
resemblance to those of Chamaeleo and Alligator. The medial scapulosternal ligament would
probably have exhibited the basic form and relationships seen in Chamaeleo, permitting freer
coracosternal movement and also allowing more degrees of freedom to the mobility of the humeral
head. It is also probable that an anterior scapulosternal ligament was present, taking on the form of
that seen in Chamaeleo and helping to maintain the integrity of the coracosternal articulation during
extreme forereach and retraction (Peterson 1973).

(b) Myological reconstruction of the primary girdle of Struthiomimus, with reference to humeral
mobility

A rather pessimistic view of the utility of muscle scars in muscle reconstruction has been presented by
McGowan (1979, 1982). He carried out a detailed study of the musculoskeletal system of the fore and
hind limbs of the brown kiwi and stated that it would be impossible to reconstruct musculature from
the available osteological data. This view is in general accordance with our own dissections. Only
a few muscles were found to leave a discernible scar in A. mississippiensis and Struthio camelus. Brown
(1981), in his work on the Upper Jurassic Plesiosauroidea, expressed the opinion that there was no
osteological evidence for the detailed muscular reconstruction of the plesiosaur humerus as presented
by Watson (1924) and Robinson (1975). The question of the propriety of reconstructing the muscles
of extinct vertebrates on osteological features is thus raised.

We feel that a compromise probably exists between the views of McGowan and Brown on the one
hand and those of other workers who have apparently employed muscle scar evidence to great
advantage. Reconstruction of anatomical minutiae seems beyond the scope of muscle scar evidence,
but in strongly sculptured regions such as the shoulder, bone architecture is probably a reasonable
indicator of at least the major muscles once present in the area. We have restricted ourselves to the
attempted reconstruction of major muscles for which good evidence, in the form of scars or major
topographical features, seem to exist.

In considering the breast-shoulder apparatus, the location of each muscle in relation to the glenoid
and the nature of the muscle belly are important in functional interpretations. The location of
muscles with respect to the glenoid will determine the leverage and control over the glenohumeral
joint and will influence adduction, abduction, protraction, and retraction. The muscles considered
are discussed in the context of the comparative models and their potential actions assessed.

In Struthiomimus no distinct muscle scar is present to indicate the point of origin or the M. deltoides
scapularis. The unreduced nature of the laterally facing scapular blade, however, is reminiscent of the
situation in Chamaeleo and Alligator. It is thus probable that the scapulodeltoid of Struthiomimus
arose from the anterolateral surface of the scapular blade, as it does in the former two genera
(text-figs. 10a, I 1a), and promoted extensive humeral protraction and elevation.

Given the carriage of the forelimbs of Struthiomimus, their probable involvement in prehension
(see below) and their unreduced state, it is unlikely that the M. deltoides clavicularis was reduced, as it

NICHOLLS AND RUSSELL: STRUT H IO M I M U S FORELIMB

669

is represented as a simple cylinder, retracted in a, b, and d, protracted in c. a, the M. deltoideus complex;
b, the M. supracoracoideus, the dashed portion is an anatomical equivalent to the dorsal portion of this
muscle in Chamaeleo (see text-fig. 10c), but its presence is conjectural; c, the M. coracobrachialis,
dashed portions represent parts of the muscle lying ventral to the humerus; d, the M. biceps brachii
and M. scapulotriceps. For further details see text.

is in Struthio. The form of the scapular prominence in Struthiomimus most closely resembles that of
Alligator and it is probable that the clavodeltoid arose from the lateral surface of the scapular
prominence (text-figs. 10a, 1 1a). With this configuration the M. deltoides clavicularis would operate
primarily as a protractor and elevator of the humerus, and the height of the scapular prominence
suggests that these actions were well developed. In comparison with Chamaeleo (text-fig. 10a) the
M. deltoides clavicularis of Struthiomimus appears to be functionally equivalent to the M.
coracohumeralis anterior, but not to the M. sternohumeralis which is concerned with humeral
protraction in the parasagittal plane.

One of the most prominent features of the coracoid of Struthiomimus is the elongate posterior
coracoid process. Its lateral surface is heavily striated and there is a deep trough along its dorsal edge.
This suggests that the M. coracobrachialis was bipartite. The M. coracobrachialis brevis probably
arose from the trough-like depression ventral to the glenoid (text-fig. 1 1 c). This branch would have
been most effective in humeral adduction. The M. coracobrachialis longus probably took origin from
the lateral face of the extensive posterior coracoid process. Such a posterior origin, relative to the
glenoid, suggests powerful humeral retraction and humeral adduction.

Previous considerations of the pectoral myology of dinosaurs (e.g. Ostrom 1974, Borsuk-
Bialynicka 1977, Coombs 1978u, Cooper 1981) have generally considered the M. coracobrachialis
longus to be absent. This follows Romer’s (1944) contention that it is absent in crocodilians. The
two heads of the coracobrachialis complex in Alligator (text-fig. 10b), and the morphology of the
coracoid plate in Struthiomimus suggests that the latter bore both branches. As it is closer in its
overall orientation to the M. coracobrachialis complex of Alligator than Chamaeleo or Struthio (text-
fig. 10b) it is indicative that the humerus could not be elevated greatly above the horizontal plane.

In Struthiomimus it is unlikely that the M. supracoracoideus would have been largely tendinous in
the region of the glenoid, as it is in Struthio. The latter is an avian characteristic and has been much
discussed by Ostrom (1976(7) and Olson and Feduccia (1979). Ostrom (19766) associated a prominent
biceps turbercle with translating the direction of pull of a ventrally situated M. supracoracoideus, but
this need not always be the case, as is evident from Chamaeleo (text-figs. 9, 10c).

670

PALAEONTOLOGY, VOLUME 28

It is more likely that the supracoracoideus in Struthiomimus was a bipartite muscle, as in both
Alligator and Chamaeleo (text-figs. 10c, 1 1 b). The broad depression anterior to the glenoid (text-
fig. 2) appears to indicate the origin of the M. supracoracoideus, but this would correspond only to
the ventral part of the muscle in Chamaeleo. The structure of the glenoid in Struthiomimus suggests
that the humeral protractors were well developed. Under these circumstances it seems likely that the
supracoracoideus would resemble the condition in Chamaeleo , extending dorsally along the scapular
blade (text-fig. 10c). This would have provided greater versatility of humeral protraction as well as
significant elevation. It would have promoted movement of the humeral head into the anterior
quadrants of the glenoid, thus enhancing forereach.

The M. biceps brachii of Struthiomimus probably arose from the biceps tubercle (text-fig. 11d),
inserting on the radius and ulna. The presence of a well-developed antebrachium and manus implies
that it would have been a strongly developed muscle, and not a reduced one as in Struthio. There is no
evidence for a humeral origin of the biceps, as suggested for Massospondylus by Cooper (1981).

The anatomical evidence available for Struthiomimus suggests that the condition of the M. biceps
brachii was intermediate between that of Chamaeleo and Struthio (text-fig. 10d). Action over the
glenohumeral joint would be small and leverage at the elbow strong. It would have been most
effective in elbow flexion with the brachium in the forereach position and the humeral head occupying
the anterior quadrants of the glenohumeral joint. The closeness of origin to the glenoid would mean
that the role of this muscle in protraction would be reduced. The position of the biceps tubercle in
Struthiomimus may indicate a slightly more medial orientation of the scapulocoracoid plate than is
found in Chamaeleo.

In Struthiomimus a distinct scar is present on the dorsal lip of the supraglenoid buttress close
to the glenohumeral joint, that probably represents the point of origin of the M. scapulotriceps
(text-fig. 11d). The morphology of the primary girdle suggests that the M. scapulotriceps of
Struthiomimus would have resembled most closely that of Chamaeleo (text-fig. 1 0e). The M.
coracotriceps was probably absent and the humeral head was, therefore, probably endowed with
greater degrees of freedom of movement. In summary, the muscles that have been reconstructed by
way of comparison with the alligator, ostrich, and chameleon, indicate some particular features of the
mobility of the humerus. In Struthiomimus the head of the humerus had considerable mobility within
the glenoid cavity. The prominent supra- and infraglenoid buttresses of the glenoid and the well-
developed anterior and posterior tuberosities of the humeral head suggest that significant degrees of
anteroposterior displacement of the humerus, via rotation and translation, were possible, but
elevation and depression were much more restricted. Elevation of the humerus could not extend
much above the horizontal, but the deltoid complex and the supracoracoideus were developed to
bring about powerful elevation to this level, coupled with strong protraction, pulling the humeral
head into the anterior quadrants of the glenoid. The biceps was well placed to produce strong flexion
of the elbow when the humerus was elevated and protracted. Combined with the mobility of the
primary girdle on the sternum (see above), these activities would bring about extensive forereach,
resulting in the antebrachium and manus being pushed forward at, or just above, shoulder height,
with the potential for the biceps to draw the antebrachium and manus towards the head when the
neck was extended.

Acting in an antagonistic fashion the coracobrachialis complex was positioned to bring about
powerful humeral retraction and depression, enhanced by the origin arising from the elongate
posterior coracoid process. It is unlikely that the M . pectoralis was well developed as the deltopectoral
crest is relatively small, compared with that of other theropods, and the angle the crest makes with the
proximal articular surface is low. These factors were taken by Ostrom (1969, p. 109) to be indicative
of only poorly developed adduction and retraction of the humerus in ornithomimids. It is likely,
however, that the coracobrachialis complex assumed a good deal of the responsibility for humeral
retraction, but with the humerus in a more horizontal orientation.

In conjunction with this, the scapulotriceps was positioned to extend the elbow, while the absence
of the coracotriceps would greatly increase the mobility of the primary girdle on the sternum by
eliminating one of the chief stabilizing ligaments.

NICHOLLS AND RUSSELL: ST RUT H 10 M I M U S LORELIMB

671

(c) Significance of antebrachial and carpal mobility

The elbow joint of Struthiomimus was essentially a hinge joint and the radius and ulna were not
freely mobile on each other, but functioned as a unit (see above). Thus, with the humerus elevated and
protracted the antebrachium could be flexed and extended to move the manus away from or toward
the head.

The concentration of the carpal elements distal to the radius is worthy of comment. That this is
their natural position is indicated by the close fit of the proximal ridge of the intermedium between the
radius and ulna, and a distinct depression on the proximal end of metacarpal I for reception of the
centrale. The same orientation of carpals is seen in AMNH 5339. This concentration of bones
proximal to metacarpal I would act like a wedge to displace the metacarpus posteriorly, when the
wrist was extended, thus effectively increasing the reach of the first digit.

(, d ) Functional significance of the manus

Ostrom (1969) presented a functional analysis of the forelimb of Deinonychus , and compared it with
that of ornithomimids. He concluded that the forelimb of Struthiomimus could not be considered
a raptorial grasping structure, as digit I was not truly opposable, the carpus was relatively inflexible
and both the internal (our posterior) tuberosity and the deltopectoral crest were poorly developed.
Subsequent workers have described the manus of ornithomimids as being relatively weak, and have
suggested that the long straight claws were used for lightly raking the surface of the ground to gain
access to food (Osmolska et al. 1972; Russell 1972). Osmolska et al. (1972) have indicated that only
very limited flexion was possible in the manus of Gallimimus.

We agree with Ostrom that the manus of Struthiomimus cannot be considered a grasping or
raptorial structure, in the manner of that of Deinonychus , but neither can it be considered to be the
ineffective, weak structure typical of Gallimimus (Osmolska et al. 1972). Considerable flexion of all
the digits in the manus was possible (text-fig. 12). In a typically raptorial manus, such as that of
Deinonychus or Allosaurus , digits II and III are of unequal length and there is considerable divergence
between them. Significant pronation and supination of the manus was also possible, as is evidenced
by the articular surfaces of the carpus (at least in Deinonychus) and freedom of movement between the
radius and ulna. In contrast the forelimb of Struthiomimus is characterized by an absence of any
significant rotation. The carpus operated as a simple hinge joint and the radius and ulna were firmly
syndesmotically united and incapable of independent rotation.

Unique features of the manus of Struthiomimus are the nearly equal length of all the digits, the
incipient coalescence of digits II and III, and the extreme divergence between digit I and the two
lateral digits. The incipient coalescence of digits II and III and their nearly equal length results in
these two digits acting in unison. In this regard the manus of Struthiomimus resembles that of
chameleonid lizards and tree sloths. In chameleons the digits are arranged in opposing sets, while in
tree sloths they are in a single set. In both cases, however, the digits are of equal length, are closely
applied, and are enclosed in a common sheath of skin that extends to the base of the ungual, holding
the digits parallel and causing them to operate as a single unit. Such a sheath may have been present
around digits II and III in the manus of Struthiomimus.

The divergence of the first digit in Struthiomimus has been considered to be a typical theropod
characteristic (Ostrom 1969), and Gilmore (1920, p. 61) compared it to that of Ornitholestes and
Allosaurus. While the distal end of the first metacarpal is offset anteriorly in these two genera, as it is in
Struthiomimus , the nature of the metacarpophalangeal articulation is quite different. In both
Allosaurus and Ornitholestes this articular surface is rotated posteriorly (text-fig. 13). Consequently
hyperextension carries digit I dorsally and posteriorly, towards the other digits. When flexed the first
digit moves away from the midline of the hand, spreading the digits and broadening the grasp of the
raptorial manus.

In Struthiomimus the opposite situation exists. The distal articulation of metacarpal I has been
rotated anteriorly (text-fig. 13). Hyperextension carries digit I dorsally and anteriorly, resulting in
a very wide divergence between digit I and digits II and III. When flexed, digit I moves posteriorly.

672

PALAEONTOLOGY, VOLUME 28

towards the midline of the hand, effectively narrowing the manus. Flexion also brings digits II and III
anteriorly, towards the midline of the hand. With the manus fully flexed the three digits lie close
together with digits II and III parallel to one another and digit I converging on these.

The anterior rotation of the distal articular surface of metacarpal I was responsible for Gabon’s
(1971) misinterpretation of Osborn’s (1916, fig. 3) figure of the manus of Struthiomimus. Galton
( 1971, pp. 5, 6) suggested that Osborn had incorrectly represented the articular orientation of the first
phalanx of digit I, and proposed that instead this phalanx should be rotated clockwise (for the left
manus, viewed distally) through 45°. Such a reorientation would result in a more typical theropod
configuration of the first digit. Such a reorientation, however, does not accord with the available
anatomical evidence. Indeed, Osborn’s (1916) interpretation of the orientation of digit I was correct,
and the same configuration is evident in UCMZ(VP) 1 980. 1 (text-fig. 1 ). Rotation of the first phalanx
to the position suggested by Galton (1971) results in an anatomically untenable disposition, with the
articular surfaces of the first metacarpal and first phalanx not being aligned and, upon extension, the
dorsal edge of the first phalanx cutting across the collateral ligament pit on the metacarpal.

Convergence of the digits in flexion is not what would be expected if the manus operated as a raking
structure, as has been suggested by Osmolska et al. (1972) and Russell (1972). In a rake the prongs are
spread to cover a wide area, and a certain amount of digital splay would be expected. Similarly, the
manus should be capable of enough pronation to bring the palmar surface parallel to the ground. As
has been shown, however, the manus of Struthiomimus was capable of little pronation and when fully
flexed, the digits of Struthiomimus form a very effective hook. The hook-like effect of the manus is
enhanced by the elongation of the penultimate phalanx. While this is a characteristic of many
theropods, it is particularly well developed in Struthiomimus. (The penultimate phalanx makes up

NICHOLLS AND RUSSELL: STRUTHIOMIMUS FORELIMB

673

text-fig. 13. Distal articular surfaces of left metacarpals of Struthiomimus compared with
those of Allosaurus. The arrows indicate the directional path of the first phalanx on
hyperextension. In Struthiomimus the distal articular surface of metacarpal I has been rotated
in a counter-clockwise direction, away from the midline of the hand. Hyperextension results
in a very wide divergence between digit I and the remaining digits. In Allosaurus , as in most
theropods, the distal articular surface of digit I has been rotated in a clockwise direction.
When fully extended, the three digits converge. A, S. altus, UCMZ(VP)1980. 1; b, A.fragilis,

ROM 5091. (Not drawn to scale.)

674

PALAEONTOLOGY, VOLUME 28

29% of the length of digit II on UCMZ(VP) 1980.1, compared with 25-6% in Deinonychus and 24-6%
in Deinocheirus. Data from Ostrom 1969, and Osmolska and Roniewicz 1970, respectively.) In an
effective grasping structure the phalanges would be of more uniform length, allowing the digits to curl
around their object.

In the elongation of the penultimate phalanx, and in the long, straight unguals, the ornithomimid
manus resembles the hook-like manus of tree sloths and anteaters (pers. obs.). The joint structure of
the two is different, however. In the edentates there is very little flexion between the short, proximal
phalanges, most of the proximal flexion being restricted to the metacarpophalangeal joints
(Humphry 1869). In Struthiomimus the reverse is true. In both, however, there is considerable
potential for flexion of the unguals. The end result is similar— a hook is formed by the elongate
unguals. In addition, however, the convergence of the first digit on the second and third in
Struthiomimus , would also add a clamping function to the manus.

Exactly what Struthiomimus was hooking and clamping with its manus can only be surmised. It
lacks both the fossorial and the suspensory specialization of living edentates. With their flat,
edentulous beaks it is most likely that ornithomimids were herbivorous. Jarzen (1982) has discussed
the palynology of the Judith River Formation and indicated that ferns (Polypodiaceae) and tree ferns
(Cyatheaceae and Dicksoniaceae) were a significant part of the flora. Cycads were also still abundant.
If these were utilized as a food source the hooking action of the ornithomimid manus could be
employed to pull fronds, sporangia, or even small branches within reach of its mouth.

CONCLUSIONS

The comparative evidence presented has enabled the breast-shoulder apparatus and forelimb of
Struthiomimus to be reconsidered structurally and functionally. Previous considerations of
coelurosaurs and related forms, based chiefly upon avian models, appear to be inadequate. The avian
breast-shoulder apparatus exhibits a number of unique specializations associated with the flight
mechanism, and these impose several limitations upon the mobility of certain parts of the
breast-shoulder apparatus. Extending the comparison to include Alligator and Chamaeleo has led to
the postulation that the primary girdle of Struthiomimus was mobile with respect to the body wall and
that the forelimb had considerable degrees of freedom of movement.

It is unlikely that the humerus of Struthiomimus could have been elevated much above the
horizontal, due to the presence of a strong dorsal glenoid buttress, but it could be depressed
considerably. The comparative and reconstructed mylogy, in association with skeletal and
ligamentous data, suggest that the forelimb was capable of undergoing considerable protraction with
the humerus in the horizontal or semi-vertical position, and that humeral retraction and adduction
were relatively powerfully developed. The absence of stabilization of the primary girdle by
a secondary girdle and the structural attributes of the forelimb skeleton are consistent with the
concept of a highly mobile forelimb. Restriction of primary girdle mobility, as seen in recent birds, is
not supported by available morphological evidence.

S. altus had elongate forelimbs. In UCMZ(VP) 1980.1 the forelimbs are 58% as long as the
hindlimbs. Unlike other long-armed coelurosaurs, however, the long forelimbs are decidedly non-
raptorial. Rotational movements in the forelimb are limited, there is little digital splay in the manus,
and the claws are long and comparatively broad. The major movements of the forelimb appears to
have been protraction and retraction at the glenohumeral joint and extension and flexion of the
antebrachium and manus.

The manus of Struthiomimus was adapted for neither grasping nor raking the ground, as has been
previously suggested, but instead appears to have been a specialized clamping and hooking structure.
The extreme divergence between digit I and the two lateral digits, when the manus was fully extended,
is unequalled in other theropods. The incipient coalescence of digits II and III, and their ability to be
strongly flexed suggests that they were enclosed in a common sheath of skin. This hook-like structure
of the manus, combined with the extensive forereach of the limb as a whole, suggests that

NICHOLLS AND RUSSELL: ST RUT H 1 0 M I M U S FORELIMB

675

Struthiomimus may have used its long arms for hooking small branches, or the fronds of ferns and
cycads, and pulling them within reach of its long neck and edentulous head.

Acknowledgements. We would like to thank Dr D. A. Russell, Dr R. C. Fox, and Dr P. Currie for allowing
E. L. N. to examine specimens in their charge. We are grateful to the staff at the Royal Ontario M useum for their
time and assistance in making their collections of ornithomimids available for study and to Dr Christopher
McGowan of that institution for helpful discussions on ratite shoulder musculature. Dr H. Osmolska kindly
supplied us with excellent photographs of the Mongolian ornithomimid material. Drs E. N. Arnold, B. G.
Gardiner, P. L. Forey, and J. Thomason provided helpful comments on earlier drafts of the manuscript. Typing
of the various drafts was capably handled by Mrs Ronica Pacholok and Miss Marg Hunik. Financial support for
this work was provided by University of Calgary Special Projects grants 69-7994 and 69-7940 to A. P. R. and by
NSERC grant A-9745 to A. P. R.

REFERENCES

alix, E. 1874. Essai sur I'appareil locomoteur des oiseaux. Paris: G. Masson.
angel, F. 1942. Les Lezards de Madagascar. Mem. Acad. Malagache. 36, 1-193.

barker, R. T. 1971. Dinosaur physiology and the origin of mammals. Evolution , Lancaster , Pa. 25 , 636-658.

- 1975. Dinosaur renaissance. Scient. Am. 232, 58-78.

baumel, j. j. 1979. Arthrologia, pp. 123-173. In baumel, j. j., king, a. s., lucas, a. m., breazile, j. j., and
evans, h. e. (eds.). Nomina Anatomica Avium. Academic Press, New York.
berger, a. j. 1960. The musculature, pp. 301 344. In marshall, a. j. (ed. ). Biology and Comparative Physiology
of Birds. Vol. 1 . Academic Press, New York.

-- 1966. The musculature, pp. 224-500. In george, j. c. and berger, a. j. (eds.). Avian Myology. Academic
Press, New York.

borsuk-bialynicka, m. 1977. A new camarasaurid sauropod Opisthocoelicaudia skarzynskii gen. n., sp. n. from
the Upper Cretaceous of Mongolia. Results of the Polish -Mongolian palaeontological expeditions, pt. VI I .
Palaeont. Pol. 37, 5-64.

broom, r. 1906. On the early development of the appendicular skeleton of the ostrich with remarks on the origin
of birds. Trans. S. Afr. Phil. Soc. 16 , 355-368.

brown, d. s. 1981. The English Upper Jurassic Plesiosauroidea (Reptilia) and a review of the phylogeny and
classification of the Plesiosauria. Bull. Br. Mus. Nat. Hist. A, Geol. 35, 253-347.
coombs, w. p. jr. 1978a. Forelimb muscles of the Ankylosauria (Reptilia: Ornithischia). J. Paleont. 52 , 642-657.

19786. Theoretical aspects of cursorial adaptation in dinosaurs. Q. Rev. Biol. 53 , 393-418.
cooper, m. r. 1981. The prosauropod dinosaur Massospondylus carinatus Owen from Zimbabwe: its biology,
mode of life and phylogenetic significance. Occ. Pap. Natn. Mus. Monuments , Rhodesia. Series B, Natural
Sciences. 6, 689-840.

cracraft, j. 1974. Phylogeny and evolution of the ratite birds. Ibis, 1 16, 494-521.

daan, s. and belterman, T. 1968. Lateral bending in locomotion of some lower tetrapods I and IF Proc. K. ned.
Akad. Wet. Series C 71 , 245-266.

dvir, z. and berme, n. 1978. The shoulder complex in elevation of the arm: a mechanism approach. J. Biomech.
11,219-225.

eurbringer, m. 1876. Zur vergleichenden Anatomie der Schultermuskeln. Morph. Jb. 1, 636-816.
galton, p. m. 1971. Manus movements of the coelurosaurian dinosaur Syntarsus and opposability of the
theropod hallux (sic). Arnoldia (Rhodesia), 15 , 1 -8.

Gardiner, b. g. 1982. Tetrapod classification. Zook J. Linn. Soc. 74 , 207-232.

gasc, j.-p. 1963. Adaptation a la marche arboricole chez le cameleon. Archs. Anal. Hist. Embryo!. 46, 81 115.
gilmore, c. w. 1920. Osteology of the carnivorous Dinosauria in the United States National Museum. Bull. U.S.
Natn. Mus. 110 , 1 1 59.

— 1933. On the dinosaurian fauna of the Iren Dabsu formation. Bull. Amer. Mus. Nat. Hist. 67, 23-78.
haughton, s. 1867a. Notes on animal mechanics No. VI— On the muscular anatomy of the crocodile. Proc. R.
Ir. Acad. 9, 268-277.

— 1867 b. Notes on animal mechanics No. X Muscular anatomy of the emu (Dromaeus Novae Hollandiae).
Ibid. 487-497.

— 1 867 c. Notes on animal mechanics No. XI -Muscular anatomy of the rhea ( Struthio rhea). Ibid. 497-504.
hoftstetter, r. and gasc, j.-p. 1969. Vertebrae and ribs of modern reptiles, pp. 210-310. In gans, c., bellairs,

a. d’a., and parsons, t. s. (eds.). Biology of the Reptilia , Volume 1. Academic Press, London.

676

PALAEONTOLOGY, VOLUME 28

holmes, r. 1977. The osteology and musculature of the pectoral limb of small captorhinids. J. Morph. 152 ,
101-140.

Humphry, g. m. 1869. The myology of the limbs of the Unau, the Ai, the 2-toed anteater and the pangolin. J.
Anat. Physiol., Loud. 4 , 17-78.

jarzen, d. m. 1982. Palynology of Dinosaur Provincial Park (Campanian) Alberta. Syllogeus, 38, 1-69.
jenkins, f. a. jr. and goslow, g. e. jr. 1983. The functional anatomy of the shoulder of the savannah monitor
lizard (Varanus exanthematicus). J. Morph. 175, 195-216.
kalin, j. a. 1929. Uber den Brustschulterapparat der Krokodile. Actci Zool., Stockh. 10 , 343-399.
lambe, j. m. 1902. On the Vertebrata of the mid-Cretaceous of the Northwestern Territory. Part 2. New genera
and species from the Belly River series (mid-Cretaceous). Geological Survey of Canada Separate Report 774.
Contr. Can. Palaeont. 2, 25-81.

— 1917. The Cretaceous theropodous dinosaur Gorgosaurus. Canada Department of Mines, Memoir, 100 ,
1-84. Government Printing Bureau, Ottawa.

lauder, g. v. 1981. Form and function: structural analysis in evolutionary morphology. Paleobiol. 7 , 430-442.
lecuru, s. 1968«. Remarques sur le scapulo-coracoide des lacertiliens. Annals. Sci. Nat., Zoologie et Biologie
Animale, Paris 12 e serie 10,475-510.

1968b. Etude des variations morphologiques du sternum, des clavicules et de finterclavicule des
lacertiliens. Ibid. 511-544.

macalister, a. 1867. On the anatomy of the ostrich ( Struthio camelus). Proc. R. Ir. Acad. 9, 1-25.
mcgowan, c. 1979. The hind limb musculature of the brown kiwi. Apteryx australis mantelli. J. Morph. 160 ,
33-74.

1982. The wing musculature of the brown kiwi Apteryx australis mantelli and its bearing on ratite affinities.
J. Zool., Lond. 197 , 173-219.

madsen, j. H. jr. 1976. Allosaurus fragi/is: A revised osteology. Bull. Utah Geol. Miner. Surv. 109, 1-163.
marsh, o. c. 1890. Description of new dinosaurian reptiles. Amer. J. Sci. 39, 81 -86.

nicholls, e. l. and russell, a. p. 1981 . A new specimen of Struthiomimus altus from Alberta, with comments on
the classificatory characters of ornithomimids. Can. J. Earth Sci. 18 , 518-526.
olson, s. L. and feduccia, a. 1979. Flight capability and the pectoral girdle of Archaeopteryx. Nature, Lond. 278 ,
247-248.

OSBORN, h. F. 1916. Skeletal adaptations of Ornitholestes, Struthiomimus and Tyrannosaurus. Bull. Amer. Mas.
Nat. Hist. 35, 733-771.

osmolska, h. and roniewicz, e. 1970. Deinocheiridae, a new family of theropod dinosaurs. Results of the
Polish-Mongolian palaeontological expeditions, pt. II. Palaeont. Pol. 21, 5-19.

— and barsbold, r. 1972. A new dinosaur Gallimimus bullatus, N.gen., N.sp. (Ornithomimidae) from the
Upper Cretaceous of Mongolia. Results of the Polish-Mongolian palaeontological expeditions, pt. IV.
Palaeont. Pol. 27 , 103-143

ostrom, j. h. 1969. Osteology of Deinonychus antirrhopus, an unusual theropod from the Lower Cretaceous of
Montana. Bull. Peabody Mus. Nat. Hist. 30, 1 165.

-1974. The pectoral girdle and forelimb function of Deinonychus (Reptilia: Saurischia): A correction.
Post ilia, 165, 1—11.

— 1976a. Archaeopteryx and the origin of birds. Biol. J. Linn. Soc.8,9\ 182.

1976 b. Some hypothetical anatomical stages in the evolution of avian flight. Smithson. Contr. Paleobiol.
27 , 1-21.

1 976 c. On a new specimen of the Lower Cretaceous Theropod dinosaur Deinonychus antirrhopus. Breviora,
439, 1-21.

1978. From dinosaur to bird: The missing link. Natn. Geogr. Mag. 154 , 152-185.

parks, w. a. 1933. New species of dinosaurs and turtles from the Upper Cretaceous formations of Alberta.
Univ. Toronto Stud. Geol. Ser. 34, 1-33.

peabody, f. e. 1959. Trackways of living and fossil salamanders. Univ. Calif. Pubis. Zool. 63, 1-72.
peterson, J. A. 1971. Functional morphology of the shoulder in Chamaeleo and Anolis. Amer. Zool. 11 ,
704-705.

1973. Adaptation for arboreal locomotion in the shoulder region of lizards. Ph.D. Thesis, University of
Chicago.

1984. The locomotion of Chamaeleo (Reptilia: Sauria) with particular reference to the forelimb. J. Zool.,
Lond. 202 , 1 -42.

porteilje, A. F. j. 1925. Zur Ethologie bzw. Psychologie der Rhea americana L. Ardea, 14 , 1 14.
raikow, r. j. 1969. Sexual and agonistic behaviour of the common rhea. Wilson Bull. 81 , 196-206.

NICHOLLS AND RUSSELL: STRUT H 10 M I M U S FORELIMB

677

robinson, j. a. 1975. The locomotion of plesiosaurs. Neues Jb. Geol. Palaont. Abh. 149, 286-332.
romer, a. s. 1922. The locomotor apparatus of certain primitive and mammal-like reptiles. Bull. Amer. Mus. Nat.
Hist. 46 , 517-603.

- 1944. The development of tetrapod limb musculature— the shoulder girdle of Lacerta. J. Morph. 74 , 1 - 41 .
russell, a. p. 1975. A contribution to the functional analysis of the foot of the tokay, Gekko gecko (Reptilia:
Gekkonidae). J. Zoo/., Loud. 176,437-476.

russell, d. a. 1972. Ostrich dinosaurs from the late Cretaceous of Western Canada. Can. J. Earth Sci. 9,
375-402.

sauer, e. g. f. and sauer, e. m. 1966. The behavior and ecology of the South African ostrich. Living Bird, 5,
45-75.

siebenrock, F. 1893. Das skelet von Brookesia super ciliaris Kuhl. Sber. Akad. Wiss. Wien. 102 , 71-1 18.
skinner, j. h. 1959. Ontogeny of the breast-shoulder apparatus of the South African lacertilian, Microsaura
pumila pumila (Daudin). Annale Univ. Stellenbosch. 35, 5-66.

Sternberg, c. m. 1933. A new Ornithomimus with complete abdominal cuirass. Can Fid. Nat. 47 , 79-83.
Sukhanov, v. b. 1968. General system of symmetrical locomotion of terrestrial vertebrates and some features of
movement of lower tetrapods. Amerind Publishing Co., New Delhi.
tarsitano, s. and hecht, m. k. 1980. A reconsideration of the reptilian relationships of Archaeopteryx. Zool.
J. Linn. Soc. 69 , 149-182.

van den berge, j. c. 1979. Myologia, pp. 175-219. In baumel, j. j., king, a. s., lucas, a. m., breazile, j. e. and
evans, h. e. (eds.). Nomina Anatomica Avium. Academic Press, New York.
walker, a. d. 1972. New light on the origin of birds and crocodiles. Nature, Lond. 237 , 257-263.
watson, d. m. s. 1924. The elasmosaurid shoulder-girdle and forelimb. Proc. Zool. Soc. Lond. 1924 , 885-917.
webb, g. j. w. and gans, c. 1982. Galloping in Crocodylus johnstoni — a reflection of terrestrial activity? Rec.
Aust. Mus. 34 , 607-618.

wellborn, v. 1933. Vergleichende osteologische Untersuchungen an Geckoniden, Eublephariden und
LToplatiden. Sber. Ges. Naturf. Freunde Berl. 1933 , 126-199.
whetstone, k. N. and whybrow, p. j. 1983. Cursorial crocodilian from the Triassic of Lesotho (Basutoland),
Southern Africa. Occ. Pap. Mus. Nat. Hist. Univ. Kansas, 106 , 1-37.

ZUG, G. R. 1974. Crocodilian galloping: A unique gait for reptiles. Copeia, 1974 , 550-552.

ELIZABETH L. NICHOLLS
Department of Biology
The University of Calgary
2500 University Drive N.W.
Calgary, Alberta, Canada T2N 1N4

ANTHONY P. RUSSELL

Department of Ichthyology and Herpetology
Royal Ontario Museum
Toronto, Ontario, Canada M5S 2C6
and

Typescript received 13 August 1984
Revised typescript received 20 February 1985

Department of Biology
The University of Calgary
2500 University Drive N.W.
Calgary, Alberta, Canada T2N 1N4

CARADOC-ASHGILL CONODONT FAUNAS
FROM WALES AND THE WELSH BORDERLAND

by N. M. SAVAGE CUld M. G. BASSETT

Abstract. Twenty-two multi-element conodont species are described from Caradoc-Ashgill (Ordovician)
strata sampled throughout Wales and the Welsh Borderland. Plectodina bullliillensis , P. bergstroemi , and
Prioniodus deani are described as new species. Strata from the Costonian to Soudleyan stages appear to lie within
the tvaerensis conodont Biozone, with Longvillian to ?Onnian strata in the superbus Biozone, and Pusgillian to
Hirnantian strata in the ordovicicus Biozone, although in no sections are the zonal boundaries sharply defined.
Contrasts in Colour Alteration Index suggest that the Caradoc-Ashgill rocks of the Welsh trough have been
subjected to much higher geothermal temperatures than those on the Welsh Borderland platform.

T he starting point for modern work on conodonts of all ages in the British Isles was the description
by Rhodes ( 1 953) of four Lower Palaeozoic faunas from Wales and the Welsh Borderland, three from
Ordovician rocks and one from Silurian. Given this initial impetus in Ordovician studies, and the fact
that numerous faunas of this age have been described subsequently from virtually all parts of the
world, it is somewhat surprising at first sight that very few further investigations have been made in
the historical type areas of the Ordovician System within the Welsh region. In fact only three
additional publications (Lindstrom 1959; Bergstrom 1964; Orchard 1980) contain systematic
accounts of Ordovician conodont faunas from this area, covering seven discrete stratigraphical units
over and above those described by Rhodes. To these should also be added Bergstrom’s (197 la) wider
biostratigraphical review, which does contain comments on undescribed faunas from other horizons
and localities, plus brief mention of the correlative significance of a late Llandeilo ( Nemagraptus
gracilis Biozone) fauna from limestone blocks in the Garn Formation of Anglesey (Bergstrom 19816)
and other Llandeilo elements from the type area (Bergstrom et al. 1984). Some elements from South
Wales and Shropshire were also noted and illustrated by Bergstrom (1983, pp. 50, 51, fig. 6a-h) in a
discussion of Llandeilo-Caradoc correlations.

This somewhat limited systematic coverage has inevitably restricted the interpretation in the Welsh
area of various aspects of Ordovician history based on conodonts, such as biostratigraphy,
palaeobiogeography, and palaeotemperatures (e.g. see Bergstrom 1973, 1977, 198 In' for summaries).
Flowever, the relative paucity of data is certainly not a reflection of lack of interest in the area but is a
result of one dominant factor— the absence of stratigraphically and geographically widespread
carbonate units suitable for the techniques of acid digestion that are now standard practice in
conodont studies. Throughout both the platform and trough areas of the Wales-Welsh Borderland
depositional basin, Ordovician sediments are developed mainly in clastic facies ranging from boulder
down to clay grade, with a predominance of fine sands and silts; the relatively few, locally developed
carbonate units in the sequence were naturally those first subjected to study for conodonts in the
investigations referred to above. This has resulted in a somewhat piecemeal and restricted analysis of
the faunas.

Notwithstanding the problems of extraction and stratigraphical/geographical coverage, our
studies suggest that the potential for collecting further data on Ordovician conodont faunas in Wales
and the Welsh Borderland is far from exhausted. Apart from the carbonate units already
investigated, there are a number of other thin, lensoid limestone levels that have not been analysed,
and there are also units of calcareous-cemented clastic strata in thin macro-shelly partings that can be
broken down by standard techniques. This paper represents an attempt to expand our knowledge of

(Palaeontology, Vol. 28, Part 4, 1985, pp. 679-713, pis. 80-86.|

680

PALAEONTOLOGY, VOLUME 28

Ordovican conodont faunas from the region, first by concentrating on a limited part of the sequence
(Caradoc and Ashgill) throughout the whole area, and secondly by systematic collecting through
vertical sequences rather than from spot samples.

STRATIGRAPHY, LOCALITIES, COLLECTIONS

The geographical and stratigraphical coverage of our collections from Caradoc and Ashgill strata
throughout Wales and the Welsh Borderland is shown in text-figs. 1-3. Within this coverage we
collected 1 10 samples, ranging in weight from 1 -5 kg up to about 4 0 kg, and of these there were forty-
eight that yielded no conodont faunas. Text-fig. 4 shows the stratigraphical ranges of species from the
forty more productive horizons represented in the collections, and Table 1 gives the yields of all
elements in the same samples. Yields from all sixty-two productive samples are plotted in Table 2. As
a means of reviewing and interpreting data from previous publications, these samples include new
collections of those Caradoc-Ashgill faunas already known from the area. It must be admitted that
the results of intensive collecting are still disappointing, and that there is still far from adequate
knowledge of conodonts in the region, but our new faunas do represent a substantial increase in data
and provide a further base for additional investigations. We believe it likely that many of the
seemingly intractable sediments will produce biostratigraphically significant faunas if very large
samples are collected in the future.

Williams et al. ( 1 972) summarized the distribution of Caradoc-Ashgill sequences in Wales and the
Welsh Borderland, and our sampling programme was based initially on this review. In the light of our
conodont studies, any departures from this correlation scheme are shown in text-fig. 2 and discussed
below. Full details of the stratigraphy and locality of all our 1 10 samples have been deposited with the
British Library, Boston Spa, Yorks UK, as Supplementary Publication No. SUP 14024 (21 pages).
They may be purchased from the British Library, Lending Division, Boston Spa, Wetherby, Yorks
L23 7BQ. In addition, they are available from the Department of Geology, National Museum of
Wales (NMW Geology Open File No. 2). For immediate reference, a summary of the data is given in
the Appendix (p. 710). All collections referred to in this paper are housed in the National Museum of
Wales (NMW) under Accession Numbers 81.4G, 81.5G, and 81.6G. In the text below, numbers in
brackets refer to the sample numbers indicated in text-figs. 1-4 and Tables 1 and 2.

Stratigraphy and correlation (text-fig. 2)

South Wales. Our oldest productive samples (5, 6) from the south-western areas of the Caradoc-
Ashgill outcrop (text-fig. 1) were from the Robeston Wathen Limestone in its type development to
the west of St Clears (see Price 1973, pp. 231-234, figs. 2 and 3). On the basis of its apparent
stratigraphical conformity below calcareous siltstones equivalent to the Sholeshook Limestone that
bear a rich shelly fauna, the Robeston Wathen Limestone is dated firmly as early mid Ashgill in age,
within Cautleyan zones 1-3 (Price 1973, fig. 6). From these beds we have recovered Amorphognathus
ordovicicus , which accords well with the general acceptance of this species as an indicator of the
Ashgill (e.g. Bergstrom 1983, fig. 1). Samples (1-4) in the type development of the Sholeshook
Limestone further to the west near Haverfordwest were unproductive. Older beds in this region are
mostly in graptolitic shales, but a report by Addison (in Williams et at. 1972, p. 36) suggests that
some limestones previously thought to be of Llandeilo age include basal Caradoc (Costonian) shelly
faunas; the same report mentions a conodont fauna from this horizon indicative of a gracilis
graptolitc Biozone age, but details are as yet unpublished and we have not collected further samples.

In the Llandeilo region. Price (1973, p. 244; 1984, p. 103) has summarized the shelly faunal evidence
for assigning a late Pusgillian to early Cautleyan age to the Birdshill and Crug limestones (see also
Owens 1973, p. 48). Correlation of these levels based on conodonts has been the subject of consider-
able discussion (cf. Bergstrom 1971a; Orchard 1980) but they have been placed consistently by all
authors within the superbus Biozone, although Orchard ( 1980, p. 13) suggested that the Crug may be
the slightly younger of the two and close to the superbus /ordovicicus boundary. Our samples from
both these horizons (II 17, Crug Limestone; 19-25, Birdshill Limestone) contain similar faunas

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

681

text-fig. 1 . Outcrop of Caradoc-Ashgill strata in Wales and the Welsh Borderland showing the geographical

distribution of conodont samples 1-1 10.

p

hJ

C/3

p

OQ*

P

r ^

C /5 3 ^

*-*■ 'C

II p

ssS.

— • O
a> O

L.

CD ^

O ^

Sr

O r- '-

— .
o
D-
3
o

a> L

3 r

C/5 C/5

JO ~

c

p

e

s

XS

•-t

o
a

C

o «
-• O
< 3

re (t

"3

c

A F

A

c

) 0

c

A S H

G S L

COSTONIAN

HARNAGIAN

SOUDLEYAN

LONGVILLIAN

WOOLSTONIAN

MARSHBROOKIAN

ACTONIAN

ONNIAN

PUSGILLIAN

CAUTLEYAN

RAWTHEYAN

HIRNANTIAN

> D- O-

n

p

»-t

p

O-

o

0

1

>

c/5

sr

CTQ

o

3

C/5

5'

p^

rD

C/5

P

3

D-

"S. O ^

O co

o

p

3

a.

Crt

3“

O

5

S'

OQ

3

re

3

CO 3 .

fJQ

r 3
£>a

" fi-
co P.

D- <T
<« <
3 re

O

r 3.

3- o

3 o

a> 3
oo O

o g-

3 °
a t3

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

683

(Table 1) suggesting a closely similar age, and include what we identify as abundant specimens of
A. ordovicicus (see discussion below); there are no specimens identified as A. superbus. On this basis
we see no evidence for a superbus age in either the Crug or Birdshill limestones; their assignment to the
ordovicicus Biozone is consistent with the early Ashgill age suggested by the shelly faunas and
removes this otherwise anomalous extension of the superbus Biozone well above the top of the
Caradoc. Fragmentary specimens illustrated by Orchard (1980, pi. 4, figs. 23, 26) as A. cf. A. superbus
from the type Ashgill area of northern England (Cautleyan) are inadequate to show the characteristic
features of the species.

An isolated limestone lenticle at Pen-y-banc (sample 18) to the north-west of Llandeilo, generally
included previously in the Birdshill outcrop (e.g. see Pringle and George 1948, fig. 1 1 ), contains an
anomalous fauna; here we identify A. superbus but no A. ordovicicus. We consider that these beds are
probably of mid Caradoc age (text-fig. 2) and are unrelated to the Birdshill-Crug limestones.

Welsh Borderland and eastern Wales. Most of the stages and lithostratigraphical divisions of the
Caradoc Series in the south Shropshire type area (samples 26-76) have yielded conodonts (text-figs. 2
and 3), but they are rarely abundant in the dominantly clastic sediments. No diagnostic zonal forms
have been recovered, and correlation with the conodont zonal scheme continues to be based on
broader shelly faunal and graptolitic evidence in various areas of Europe (e.g. Bergstrom 1971a, b ,
1977, 1978, 1983). Our collections in the upper part of the succession are based on the revised
stratigraphy of Hurst (1979).

In the Gwern-y-Brain section near Welshpool (Cave 1 965), two samples (77, 78) from the Nod Glas
Formation have yielded fairly common faunas whose age interpetations are problematical. The
uppermost Nod Glas beds here contain the trilobite Onnia gracilis indicative of the Onnian, while the
unconformably underlying Gaerfawr Formation contains a Woolstonian shelly fauna (Cave 1965),
so that the limits of the Nod Glas Formation are fairly well circumscribed by comparison with the
Shropshire succession (text-fig. 2). However, conodonts from the basal 50 cm (sample 77) of the
phosphoritic limestones making up the Nod Glas here contain fairly common Plectodina bu/lhillensis
sp. nov., which in Shropshire occurs only in samples from the Costonian to the Woolstonian. In
addition the same sample contains Amorphognathus aff. A. tvaerensis (see p. 694), which again might
support an earlier Caradoc age (?pre-mid Soudleyan — text-fig. 2). The full stratigraphical range of
P. bulhillensis is not yet known from continuous, productive sections, but it cannot be ruled out that
the base of the Nod Glas at Gwern-y-Brain is older than has been supposed previously, although the
limit may be dictated by the Gaerfawr shelly fauna; it is also possible that the lowest Nod Glas
conodonts are reworked although the specimens themselves show no sign of this, and further studies
will be necessary to resolve the problem. The top 30 cm of the Nod Glas phosphoritic limestones
(sample 78) below the nodular beds (Cave 1965, fig. 1) yielded abundant specimens of A. ordovicicus.
If the trilobite evidence from here is definitive of the Onnian, then it could be that the superbus-
ordovicicus Biozone boundary is as low as this below the Ashgill (see discussion above) as we propose
tentatively in text-fig. 2, although further investigations are again required to test these relationships.

North Wales. In the Berwyn Hills (79-82) and Bala district (97-100), well-dated Longvillian-
Woolstonian formations contain superbus Biozone faunas, including P. bullhillensis that gives further
weight to correlations with the Shropshire succession. Various discontinuous limestone units of
Cautleyan-Rawtheyan ages (83-95, 97-105) (text-fig. 2) generally yielded sparse faunas, but all are
indicative of the ordovicicus Biozone.

None of our samples of Hirnantian age has yielded conodonts. Large samples from the type
Hirnant Limestone of Bala (106) and from calcareous-cemented arenites of the Conway Castle Grits
at Deganwy (107-110) were processed in the hope of isolating post -ordovicicus Biozone faunas to
compare with the Gamachignathus faunas described from North America and northern England (e.g.
McCracken and Barnes 1981a, b ; McCracken et al. 1980; Orchard 1980; Uyeno and Barnes 1983);
the potential of the latter for wider correlation is still unknown as the diagnostic forms have not yet
been identified in other regions. Gamachignathus is a subjective synonym, apparently junior, of

[ Text continues on page 689]

table 1. Number and distribution of individual conodont elements recovered in the forty more productive samples.

684

PALAEONTOLOGY, VOLUME 28

? Ozarkodina pseudofissilis

Plectodina bergstroemi

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

685

0 © fD O On '■O

\D OO lO V-) CTv

01 ^ O VO N -

- - (N

■| G < cc

•5 <2 d dj

"3

2 o c S a 1

kj to
a o

O TO

Bj X) O 3 CTj .O ctf

D-Cu^C/iC/^OO-^D-D-^t/D

i Xi w

■a.a. 2 i

: aj jd =r ?3 .o — C:

;PPc/)' 2 pfl, 2 i/) c

; § ^ cu <
, d d

5T TO TO

-2 -s .2

"3 d d

; a d d

Soo

table 1. Number and distribution of individual conodont elements recovered in the forty more productive samples.

non-gemculalc

10:arkodtnu pirudofisjilii

graciliform

Plrcindina berpsiroeml

IPItciodina tp.
Sc

scandodifonn

SAVAGE AND BASSETT ORDOVICIAN CONODONTS

table 2. Number and distribution of conodont species in all 1 10 samples from Wales and the Welsh Borderland, with details of Colour Alteration

Index for each sample.

686

PALAEONTOLOGY, VOLUME 28

Conodont Colour Index - 4-5 - - - - 4- -- -- 5- 5- 45 5 5 - - - 5-5 5-5 5-5 - Total

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

687

ON'tOsT'O—

— — - - J — I O' o — • >

'O —

fNOv — — — — (

^ Q 5 i

Si ^ O *3 <3

*• ffl 3

i, ?

; < dj?

1 g- S "S

| § § ~ S*

a ^ f 1 S; © ©

• 3 J3 2 12 c s* -V
2 ^ a 3 3 3

' S.-S -2

0"T3 T3

- 3 3

d g

< CQ

•o -a

■2 -2 5

g <S 2 ri d ,

'^ 5 ;^

table 2. Number and distribution of conodont species in all 1 10 samples from Wales and the Welsh Borderland, with details of Colour Alteration

Index for each sample.

lOzarkodina pstudofusilis

P hullhillcnsis
IPIeciodma sp.

Prioniodiu sp.

ISiauffcrrlla sp.

Conodont Colour Index

A iff A. uaerensii
Drepanolslndus suhtrtclus

Pandcradus cf. P gracilis

P bullhdlcnns
IPleaodtna sp.

Rhodeiagnalhus elegans
ISlaufferclla sp.

text-fig. 3. Geological map and stratigraphical section of the Caradoc Series in the Onny valley, Shropshire,
showing detail of the distribution of conodont samples; stratigraphical thicknesses are based mainly on
measurements made along the section on the north side of the A489 road. Samples plotted on the left of the
vertical column are from the A489 road section, those on the right are from the Onny river and old railway
section. Details of strike/dip are omitted from the map for clarity, but the dip is fairly constant throughout the
section from 8° to 15° (maximum observed 22°) along a bearing of 150°.

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

689

Birksfeldia Orchard, 1980, which Orchard recovered from the probably Rawtheyan Cystoid
Limestone and Upper Keisley Limestone in northern England, and the middle to late Ashgill
Abercwmeiddaw Mudstone, Rhiwlas Limestone, and Robeston Wathen Limestone in North and
South Wales. The date of publication of Orchard’s paper is August 1 980, whereas that of McCracken
et al. is November 1980. The only other described British Hirnantian conodonts are zonally
undiagnostic forms reported by Orchard (1980, pp. 12, 14) from the Swindale Limestone of Cross
Fell, northern England.

Conodont biostratigraphy

Text-fig. 4 shows the approximate boundaries between the ordovicicus, superbus , and tvaerensis
biozones in relation to our more productive samples. In detail these boundaries are unreliable
because of the pronounced facies difference between the Shropshire and Welsh localities. The
contrast between the composition of the collections is striking. The Shropshire material has a
predominance of Plectodina and virtually no specimens of Amorphognathus. The Welsh material has
an abundance of Amorphognathus. Because elements of the Amorphognathus lineage are used as zonal
indices through this interval elsewhere (e.g. Bergstrom 1983, fig. 1), their absence in the
predominantly shelly facies characteristic of the Caradoc type section is a significant drawback when
trying to correlate from Shropshire even to Wales as well as to most other regions. Significant
conodont occurrences in our samples are discussed below.

Prioniodus deani sp. nov. is known from only seventeen elements from the Coston Formation (26)
and the Hoar Edge Grit (71-73). It is characterized by a strongly arched, short. Pa element and may be
a useful indicator of pre-Harnagian age. It appears to be restricted to the Costonian. Plectodina tenuis
(Branson and Mehl) is fairly easily recognized in our collections by its strongly arched Pb element and
a Sb element possessing a denticle beside the main cusp which is larger than that cusp. It occurs as
low as the Coston Formation (26) and other Costonian localities and as high as the limestone at

text-fig. 4. Stratigraphical range and biostratigraphical zonation of conodonts from the forty more productive
samples in Wales and the Welsh Borderland. Within any one Stage there may be some stratigraphical overlap
between samples from different regions and sections, but the samples are plotted individually as a means of
depicting general ranges; there is no overlap in samples from different Stages; relative durations of Stages are not
necessarily true but simply reflect an artificial spacing necessary to accommodate the number of samples plotted.

690

PALAEONTOLOGY, VOLUME 28

Pen-y-banc (18) in the probable Marshbrookian. Both ends of its range in Britain are much lower
than presently recorded in North America where Sweet (19816) shows it from the late Kirkfieldian to
late Richmondian.

P. bullhillensis sp. nov. is the most abundant species in our collections with almost 700 elements,
mostly from the Hoar Edge Grit (71-74) in the Costonian to early Harnagian but with a range
extending up to the Cymerig Limestone (97-99) in the probable Longvillian-Woolstonian. It is
characterized by a very small Pb element which has a high anterior blade and by a Pa element which is
usually large but rather delicate just behind the main cusp and therefore usually broken at that point.
In some samples the Pa element is smaller but is still recognizable by its three anterior denticles,
backwardly inclined main cusp, and the frequent breakage just behind that cusp. The species is not
known from elsewhere but may be a useful indicator of early-mid Caradoc age deposits in Britain.

P. bergstroemi sp. nov. is known mainly from the Smeathen Wood Formation (31, 32) in the
Harnagian but also from a few specimens from the lower Cheney Longville Formation (47) in the
probable Woolstonian. The complete apparatus is still not known, partly because of the relatively
few elements available and partly because most of the known specimens are broken. Nevertheless, it
is quite distinctive. Icriodella superba Rhodes is a variable species and therefore difficult to subdivide
in a meaningful way. It appears to range throughout the upper Caradoc and most of the Ashgill.
A. aff. A. tvaerensis Bergstrom is known only from nineteen elements from the lower part of the Nod
Glas Formation (77). It has some of the characteristics of A. inaequalis (the Pa element) and some of
A. tvaerensis (the M and Pb elements). As discussed above, the occurence of this form along with
P. bullhillensis in this fauna suggests that the lower part of the Nod Glas Formation might be much
older than the upper beds which include A. ordovicicus.

The type locality of Rhodesognathus elegans (Rhodes) is at Rhodes’s Locality 3, Pen-y-garnedd,
Wales (Rhodes 1953) (our samples 79, 80). According to Lindstrom (1977) and Sweet et al. ( 1971) the
species ranges from the low Caradoc to the end of the Ashgill. In our collections it ranges from the
Woolstonian Cheney Longville Formation (47) to the Pusgillian as represented in the Crug and
Birdshill samples (11-17, 19, 20). Thus we seem to be in approximate agreement with Lindstrom and
others.

The type locality of A. superbus is Rhodes’s Locality 2 (Rhodes 1953, p. 264) in the Cymerig
Limestone (our samples 93-100). The holotype is an M element and these are rare and may vary more
than earlier suspected. Difficulties in distinguishing A. superbus from A. ordovicicus may be reduced
by noting the differences between the respective Pb elements, as described in the systematic section
below. It appears that A. ordovicicus is present in the Glyn Formation (92-95), the Crug Limestone
(11-17), the Birdshill Limestone (8-10), the limestones at Dryslwyn (23) and Llanegwad (25), the
Robeston Wathen Limestone (5, 6), and the upper part of the Nod Glas Formation (78), whilst
A. superbus is present in the limestone at Pen-y-banc (18) and the Cymerig Limestone (97, 98).

lOzarkodina pseudofissilis (Lindstrom) probably does not belong to Ozarkodina but is referred to
that genus until its affinities can be clarified. The apparatus is currently known only from Pa and Pb
elements. The species is characteristic of the Crug Limestone at Crug (11-17), which is the type
locality, and also of the Birdshill Limestone (19, 20), and the limestones at Dryslwyn (22, 23) and
Llanegwad (25). It appears to be an indicator of early Ashgill age. Aphelognathus rhodesi (Lindstrom)
occurs in our samples only from the Crug Limestone (11-17) and may be a useful Pusgillian indicator.

GEOTHERMAL HISTORY OF THE CONODONTS

The colour of conodonts indicates the thermal history of the deposits and largely reflects the
thickness and duration of the overburden, although igneous and hydrothermal activity also may be
factors. Dynamic (regional) metamorphism should not influence the colour of conodonts. In general
the colour, measured as the Colour Alteration Index (CAI) of Epstein et al. (1977), will be the same
for conodonts of the same age in the same area. Where distances of several tens of kilometres are
involved the overburden thickness may have varied sufficiently for conodont colour differences to
result (Savage 1983).

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

691

The Caradoc age conodonts from Shropshire all exhibit a Colour Alteration Index of 1 to 2
(Table 2), suggesting low geothermal temperatures. The Caradoc conodonts of Central and North
Wales, and the Ashgill conodonts from South Wales, all have a CAI of 4-5 or more (Table 2),
suggesting much higher geothermal temperatures. Bergstrom (1981 a, p. 385) noted the CAI contrast
in the region and suggested that volcanism or heat flow associated with tectonism is a more likely
explanation for the high Welsh indices than overburden. If one accepts the calculations of Epstein
et al. (1977), the overburden on the deposits of the Welsh trough necessary to produce the thermal
changes would need to have been several kilometres greater than that on the Borderland platform
deposits.

SYSTEMATIC DESCRIPTIONS

In this study we use the element notation of Sweet (1981 a) adopted in the recent conodont volume of
the Treatise on Invertebrate Paleontology. We have attempted to identify dextral and sinistral
elements in the descriptions and plate explanations, assuming laterally organized pairs with the
elements pointed away from the observer. Illustrations on the plates are arranged in stratigraphical
and geographical order, with the older faunas from the Caradoc area on Plates 80-82, and the
younger late Caradoc and Ashgill faunas on Plates 83-86. This arrangement is made to emphasize
associations in the different areas and to assist in facies comparisons.

Order conodontophorida Eichenberg, 1930
Family balognathidae Hass, 1959
Genus amorphognathus Branson and Mehl, 1933c

Type species. Amorphognathus ordovicica Branson and Mehl, 1933c, p. 127.

Remarks. The orientation of the dextral and sinistral Pa elements used here is that used by
Bergstrom (1978, pi. 80, figs. 1-3) but the reverse of that used by Lindstrom (1977) and by Orchard
(1980). Many workers state that dextral elements are distinguished by high anterior blades
(Schopf 1966; Lindstrom 1977; Orchard 1980). In our material both sinistral (PI. 85, figs. 25
and 26; PI. 86, figs 25 and 26) and dextral (PI. 83, figs. 1 and 2; PI. 84, figs. 1 and 3) elements may
have high or low blades.

Amorphognathus ordovicicus Branson and Mehl, 1933c

Plate 84, figs. 1 -21; Plate 85, figs. 1-26; Plate 86, figs. 1-13

1933c Amorphognathus ordovicica n. sp. Branson and Mehl, p. 127, pi. 10, fig. 38 [holotype] [Pa element],
1933c Ambalodus triangularis n. sp. Branson and Mehl, p. 128, pi. 10, figs. 35-37 [Pb element].

1959 Goniodontus superbus n. sp. Ethington, p. 278, pi. 40, figs. 1 and 2 [M element].

1959 Trichonodella inclinata Rhodes; Ethington, p. 290, pi. 41, fig. 6 [Sa element].

1959 Tetraprioniodus parvus n. sp. Ethington, p. 288, pi. 40, fig. 8 [Sb element].

1959 Eoligonodina elongata ( Rhodes); Ethington, p. 277 , pi. 40, fig. 5 [Sc element].

1959 Keislognathus simplex n. sp. Ethington, p. 280, pi. 40, figs. 9 and 10 [Sd element],

1978 Amorphognathus ordovicicus Branson and Mehl; Bergstrom, pi. 80, figs. 1-11 [multielement
apparatus].

Remarks. Amorphognathus ordovicicus is most easily distinguished from earlier species of the genus
by its smaller and more robust Pb elements (PI. 84, figs. 4 10) as well illustrated in Bergstrom and
Sweet 1966, pi. 28, figs. 7 and 8. The M element is also characteristic (PI. 84, figs. 15 and 16), and is
well illustrated in Bergstrom 1971 a, p. 93, fig. 4(5), but this element is infrequently recovered and
is not generally convenient for diagnosis.

692

PALAEONTOLOGY, VOLUME 28

Amorphognathus superbus (Rhodes, 1953)

Plate 83, figs. 119

1953 Holodontus superbus n. sp. Rhodes, p. 304, pi. 21, figs. 125 127 [holotype] [M element].

1953 Amorphognathus ordovicicus Branson and Mehl; Rhodes, p. 283, pi. 20, figs. 47-49 [Pa element].

1953 Ambolodus triangularis var. indentatus n. var. Rhodes, p. 280, pi. 20, figs. 35-37 [Pb element].

1953 Ligonodina elongata n. sp. Rhodes, p. 305, pi. 21, figs. 130 and 131 [Sc element].

1953 Ligonodina extensa n. sp. Rhodes, p. 306, pi. 21, figs. 128 and 129 [Sc element],

1953 Trichonodella gracilis n. sp. Rhodes, p. 314, pi. 21, figs. 144, 147-150 [Sa element].

Remarks. In this species of Amorphognathus the sinistral Pb element is large, deeply excavated, thin-
walled, and has a deeply indented and strongly sinuous aboral inner margin (PI. 83, fig. 8). The
dextral Pb element has an aboral inner margin which is deeply indented but non-sinuous (PI. 83,
fig. 6). The M element has an arched aboral margin and bears a low cusp and from one to three
subequal denticles (PI. 83, figs. 1 1-14).

Confusion about the distinctions between this species and A. ordovicicus may have added to the
difficulty in correlating the Caradoc and Ashgill in Britain and elsewhere. The type species of
A. superbus is from the Cymerig Limestone at Locality 2 of Rhodes (1953) in ‘the shallow quarry
1750 ft due north of Plas Rhiwaldog and 600 ft east of Y-Garnedd, Merioneth’ (our samples
98-100), and the holotype is a M element. M elements of Amorphognathus are usually very rare and
seem to vary enough within a single fauna to make several specimens desirable for species

EXPLANATION OF PLATE 80

All figs, x 40.

Figs. 1-5. Plectodina tenuis (Branson and Mehl, 1933). 1, inner lateral view of sinistral Pa element,

NM W 81.6G.1; 2 and 3, inner lateral and outer lateral views of sinistral M element, NMW 81 .6G.2; 4 and 5,
outer lateral and inner lateral views of dextral Sc element, NMW 81.6G.3; all from sample 58.

Figs. 6 and 7. Pseudooneotodus sp. A. Lateral and upper views, NMW 81.6G.4, from sample 58.

Figs. 8-14. Icriodella superba Rhodes, 1953. 8 10, outer lateral, upper, and lower views of dextral Pa element,
NMW 81.6G.5; 11 and 12, outer lateral and inner lateral views of dextral Pb element, NMW 81.6G.6;
13 and 14, anterolateral and posterolateral views of inner side of dextral M element, NMW 81.6G.7; all from
sample 58.

Figs. 15-22. Plectodina bullhillensis sp. nov. 15, inner lateral view of sinistral Pa element, NMW 81.6G.8;
16, outer lateral view of dextral Pb element, NMW 81.6G.9; 17, inner lateral view of dextral Sb element,
NMW 81.6G.10; 18, outer lateral view of sinistral Sc element, NMW 81.6G.il; 19, inner lateral view of
dextral M element, NMW 81 .6G. 12; 20, posterior view of Sa element, NMW 81.6G. 13; 21, outer lateral view
of dextral Pa element, NMW 81.6G.14; 22, outer lateral view of sinistral Sc element, NMW 81.5G.1.
Figs. 15-17 from sample 74; figs. 18-21 from sample 73; fig. 22 from sample 74.

Figs. 23-39. Prioniodus deani sp. nov. 23 and 24, outer lateral and inner lateral views of sinistral Pa element,
NMW 81.5G.2 (holotype); 25 and 26, inner lateral and outer lateral views of dextral Pa element,
NMW 81.5G.3; 27-29, outer lateral, lower, and inner lateral views of dextral Pb element, NMW 81.5G.4;
30-32. lower, outer lateral, and inner lateral views of sinistral Pb element, NMW 81.6G. 15; 33 and 34, inner
lateral and outer lateral views of sinistral Pb element, NMW 8 1 ,6G. 1 6; 35-37, inner lateral, outer lateral, and
upper views of sinistral Pb element, NMW 81.6G.17; 38 and 39, lower and upper views of S element,
NMW 81.5G.48. Figs. 23-29, 38 and 39 from sample 26; figs. 30-32 from sample 73; figs. 33-37 from
sample 71.

Figs. 40, 41. '? Plectodina sp. Inner lateral and outer lateral views of dextral M(?) element, NMW 81.6G.18,
from sample 3 1 .

Figs. 42-47. Panderodus cf. P. gracilis (Branson and Mehl, 1933). 42, outer view of sinistral compressiform
element, NMW 81.6G.19; 43, inner view of dextral compressiform element, NMW 81.5G.5; 44 and 45, outer
and inner views of dextral graciliform element, NMW 81.6G.20; 46 and 47, outer and inner views of dextral
graciliform element, NMW 81.6G.21. Fig. 42 from sample 74; fig. 43 from sample 26; figs. 44 47 from
sample 71.

PLATE 80

SAVAGE and BASSETT, Caradoc conodonts

694

PALAEONTOLOGY, VOLUME 28

identification. The Pb elements recovered from the type locality were named by Rhodes Ambolodus
triangularis var. indentatus n. var. In our collections from Y Garnedd the M and Pb elements look the
same as the Rhodes specimens (Rhodes 1953, figs. 35-37, 125-127). The Pb element is thin-walled,
deeply excavated, and deeply indented, with a strongly sinuous inner margin. In contrast, the Pb
element of Amorphognathus ordovicicus is small and stout. The illustration by Rhodes of Pb elements
of A. superbus (Rhodes 1953, pi. 20, figs. 30-37), from his Locality 2, with Pb elements from his
Locality 3 (Rhodes 1953, pi. 20, figs. 28-31), which is ‘a disused quarry about 900 ft east of the Powis
Arms at Pen-y-garnedd, Montgomeryshire’ (our samples 79, 80) appears to have led to some
confusion. These Pb elements from Pen-y-garnedd are the small, stout kind, very similar to those of
A. ordovicicus but in this case most likely belonging to A. complicatus Rhodes, the holotype of which
also comes from Rhodes’s Locality 3. A. complicatus is thought by Bergstrom (1971a, 1983) to have a
range corresponding with the upper half of that of A. superbus. This distinction between the Pb
elements of A. superbus and A. complicatus does not seem to have been recognized sufficiently
hitherto and it has become common to show the forms, named by Rhodes as Ambolodus triangularis
var. indentatus n. var. and A. triangularis Branson and Mehl, as synonyms (Bergstrom 1964, p. 56;
Lindstrom 1977, p. 44; Orchard 1980, p. 16). Any confusion is probably compounded by the
inclusion by Lindstrom of a small, stout Pb element as part of the illustration of the Amorphognathus
superbus apparatus in the Catalogue of Conodonts (Lindstrom 1977, pp. 40, 41). The material used in
this illustration is not from the type area in Wales but is from Banklick Creek in Kentucky (Bergstrom
and Sweet 1966, pp. 273, 296, 426, pi. 28, figs. 7 and 8) and probably does not belong to A. superbus.
Orchard assigned material from the Crug and Birdshill limestones to A. superbus and A. aff.
A. superbus but illustrated only single views of four broken elements (Orchard 1980, pi. 4, figs. 19-21,
and 24). These specimens inadequately support his determination and probably belong variously to
A. ordovicicus and Prioniodus sp.

Amorphognathus aff. A. tvaerensis Bergstrom, 1962
Plate 86, figs. 25-33

aff. 1962 Amorphognathus tvaerensis n. sp. Bergstrom, p. 36, pi. 4, figs. 7 and 8 [holotype], 9 and 10
[Pa element].

aff. 1962 Tvaerenognathus ordovicica n. gen. et sp. Bergstrom, p. 57, pi. 1, figs. 1-5 [M element].

aft'. 1 962 Ambolodus triangularis erraticus n. ssp. Bergstrom, p. 26, pi. 3, figs. 15-17 [Pb element].

EXPLANATION OF PLATE 81

All figs, x 40.

Figs. 1 17. Plectodina bullhillensis sp. nov. 1-3, inner lateral, outer lateral, and lower views of sinistral Pa
element, NMW 81.6G.22 (holotype); 4-6, inner lateral, lower, and outer lateral views of sinistral Pb element,
NMW 81.6G.23; 7 and 8, outer lateral and inner lateral views of dextral Sb element, NMW 81.6G.24; 9 and
10, lower and inner lateral views of sinistral M element, NMW 81.6G.25; 1 1 and 12, inner lateral and outer
lateral views of sinistral Sc element, NMW 81.6G.26; 13-15, lower, inner lateral, and outer lateral views of
asymmetric sinistral Sa element, NMW 8 1 .6G.27; 1 6 and 1 7, posterior and postero-lower views of Sa element,
NMW81.6G.28; all from sample 72.

Fig. 18. Drepanoistodus suberectus (Branson and Mehl, 1933). Inner lateral view of NMW 81.6G.29, from
sample 72.

Figs. 19-35. Plectodina tenuis (Branson and Mehl, 1933). 19-21, inner lateral, outer lateral, and lower views of
Pa element, NMW 81 .5G.6; 22 and 23, lateral and lower views of sinistral Pb element, NMW 81.5G.7; 24 and
25, inner lateral and lower views of sinistral Sb element, NMW 8 1 .5G.8; 26-28, outer lateral, lower, and inner
lateral views of dextral M element, NMW 8 1 5G.9; 29 and 30, outer lateral and inner lateral views of dextral Sc
element, NMW 81.5G.10; 31, inner lateral view of dextral Sb element, NMW 81.5G.il; 32 and 33, inner
lateral and lower views of asymmetric sinistral Sa element, NMW 81.5G.12; 34 and 35, posterior and lower
views of Sa element, NMW 81.5G.13; all from sample 26.

PLATE 81

SAVAGE and BASSETT, Plectodina , Drepanoistodus

696

PALAEONTOLOGY, VOLUME 28

aff. 1962 Ambalodus triangularis suecicus n. ssp. Bergstrom, p. 28, pi. 3, figs. 11-14 [Pb element],
aff. 1983 Amorphognathus tvaerensis Bergstrom; Bergstrom, p. 47, fig. 4h-p [multielement apparatus].

Remarks. A few specimens of Amorphognathus from sample 77 comprise what is probably a new
species of the genus. The Pa, M, and Pb elements resemble those of the early to middle Caradoc
species A. tvaerensis. As discussed above (p. 715), the age of sample 77 in the lower Nod Glas
Formation was earlier thought to lie within the range of A. superbus but the absence of that species
from our fauna and the presence of A. aff. A. tvaerensis suggests that it may be older.

Genus rhodesognathus Bergstrom and Sweet, 1966
Type species. Ambolodus elegans Rhodes, 1953, p. 278.

Rhodesognathus elegans (Rhodes, 1953)

Plate 82, figs. 34-37; Plate 83, figs. 26 and 27; Plate 84, figs. 28 and 29; Plate 85, figs. 36-39; Plate 86, figs. 23 and 24

1953 Ambolodus elegans n. sp. Rhodes, p. 278, pi. 20, figs. 22 and 24 [holotype], 21, 23, and 25
[Pa element],

1953 Ambolodus pulcher n. sp. Rhodes, p. 279, pi. 20, figs. 38-41 [Pb element].

1953 Ambolodus robustus n. sp. Rhodes, p. 279, pi. 20, figs. 26, 27, 32, and 33 [Pa element].

1977 Rhodesognathus elegans (Rhodes); Lindstrom, p. 535 [multielement apparatus].

Remarks. Sweet (1979u, p. G21) and Orchard (1980, p. 25) have commented on the apparatus of
Rhodesognathus , and particularly on the possibility that R. elegans may have possessed ramiform
elements like those of Amorphognathus. Orchard (1980) found no evidence of these ramiform
elements but considers (pers. comm. 1985) such an apparatus quite likely. We have no
ramiform elements with our Pa and Pb elements and we are particularly conscious that the twelve Pa
and nine Pb elements recovered from our Cheney Longville Formation sample 50 have no ramiform
elements associated with them. In this fauna the absence of Amorphognathus and its ramiform
elements is helpful in reducing the number of possible associations.

EXPLANATION OF PLATE 82

All figs, x 30.

Figs. 116. Plectodina bergstroemi sp. nov. 1 and 2, outer lateral and inner lateral views of dextral Sb element,
NMW 81.6G.30; 3 and 4, outer lateral and inner lateral views of dextral Pb element, NMW 81.6G.31; 5-7,
inner, lower, and outer views of sinistral Sb element, NMW 81.6G.55; 8 and 9, inner lateral and outer lateral
views of dextral Sc element, NMW 81.6G.56; 10-12, inner lateral, outer lateral, and lower views of dextral Pb
element, NMW 81.6G.32 (holotype); 13-15, inner lateral, outer lateral, and lower views of dextral Pa element,
NMW 81.6G.33; 16, inner lateral view of broken M element, NMW 81.6G.34; all from sample 32.

Figs. 17-27. Plectodina bullhillensis sp. nov. 17-19, inner lateral, outer lateral, and lower views of dextral
Pa element, NMW 81.5G.14; 20-22, outer lateral, lower, and inner lateral views of dextral Pb element,
NMW 81.5G.15; 23 and 24, inner lateral and outer lateral views of dextral Sc element, NMW 81.5G.16;
25, inner lateral view of dextral M element, NMW 81.5G.17; 26 and 27, posterior and anterior views of
Sa element, NMW 8 1 . 5G . 1 8; all from sample 45.

Figs. 28-33. Icriodella superba Rhodes, 1953. 28 and 29, inner lateral and outer lateral views of dextral
M element, NMW 81.5G.19; 30-32, upper, outer lateral, and lower views of sinistral Pa element,
NMW 81.5G.20; 33, outer view of sinistral Pb element, NMW 81.5G.21. Figs. 28-32 from sample 46; fig. 33
from sample 50.

Figs. 34-37. Rhodesognathus elegans (Rhodes, 1953). 34, outer lateral view of dextral Pa element,

NMW 81.5G.22; 35 and 36, outer lateral and inner lateral views of dextral Pb element, NMW 81.5G.23;
37, outer lateral view of sinistral Pb element, NMW 81.5G.24; all from sample 50.

Fig. 38. Drepanoistodus suberectus (Branson and Mehl, 1933). Inner lateral view of NMW 81.5G.25, from
sample 50.

Figs. 39-41. IStaufferella sp. Lower, inner lateral, and outer lateral views of NMW 81.5G.26, from sample 45.

PLATE 82

SAVAGE and BASSETT, Caradoc conodonts

698

PALAEONTOLOGY, VOLUME 28

Family cyrtoniodontidae Hass, 1959
Genus aphelognathus Branson, Mehl and Branson, 1951

Type species. Aphelognathus grandis Branson, Mehl and Branson, 1951, p. 9.

Aphelognathus rhodesi (Lindstrom, 1959)

Plate 84, figs. 34-46

1959 Ozarkodina rhodesi n. sp. Lindstrom, p. 441, pi. 1, figs. 1 and 2 [holotype], 3-9 [Pb element],

1959 Prioniodus pulcherrima n. sp. Lindstrom, p. 442, pi. 3, figs. 28-30 [Pa element].

1959 Cordylodus cf. spurius Branson and Mehl; Lindstrom, pi. 4; figs. 19 and 20 [Sc element],
21 [M element].

1959 Zygognathus crugensis n. sp. Lindstrom, p. 451, figs. 1 1-27 [Sb element],

1959 Trichonodella parabolica n. sp. Lindstrom, p. 450, figs. 18-22 [Sa element].

1980 Aphelognathus nudus sp. nov. Orchard, p. 18, pi. 2, figs. 1 and 2 [Sc element], 3 [Sb element],
4 [Pb element], 8 and 1 1 [M element],

19816 Aphelognathus rhodesi (Lindstrom); Sweet, p. 49, figs. 1-6, p. 51, 52 [multielement apparatus].

Remarks. Lindstrom (1959) described and figured as form species all the elements of this species from
the Crug Limestone of South Wales (our samples 11-17). The first of these form species is Ozarkodina
rhodesi on p. 441 and this should thus determine the multielement name. Orchard (1980) believed that
the Pa and Pb elements are indistinguishable from those described by Hinde (1879) from Ontario,
Canada, and tentatively named the apparatus Aphelognathus furcatus (Hinde) based on the Pa
element. Sweet (19796, p. 66) has presented evidence to show that Prioniodus furcatus Hinde is part of
an Oulodus apparatus, and it now seems best to assign the Crug material to A. rhodesi. A. rhodesi is
easily distinguished from other species of Aphelognathus by its short Sb element, lack of denticles
anterior of the cusp on the M and Sc elements, and more reclined denticles on the Pa element.

Genus plectodina Stauffer, 1935
Type species. Plectodina dilata Stauffer, 1935, p. 152.

explanation of plate 83

All figs, x 40; all from sample 97.

Figs. 1 19. Amorphognathus superbus (Rhodes, 1953). 1-3, upper, outer lateral, and lower views of dextral
Pa element, NMW 81.5G.27; 4-7, outer lateral, upper, inner lateral, and lower views of dextral Pb element,
NMW 81.5G.28; 8-10, inner lateral, lower outer lateral, and upper views of sinistral Pb element,
NMW 81.5G.29; 11 and 12, inner lateral and outer lateral views of sinistral M element, NMW 81.5G.30;
13 and 14, outer lateral and inner lateral views of dextral M element, NMW 81.5G.31; 15, inner lateral views
of dextral Sc element, NMW 81.5G.32; 16, inner lateral view of Sa element, NMW 81.5G.33; 1 7 and 18, inner
lateral and outer lateral views of dextral Sc element, NMW 81.5G.34; 19, lateral view of Sd element,
NMW 81.5G.35.

Figs. 20-25. Iciodella superba Rhodes, 1953. 20 and 21, outer lateral and inner lateral views of sinistral
M element, NMW 81.5G.36; 22-24, upper, lower, and inner lateral views of sinistral Pa element,
NMW 81.5G.37; 25, outer lateral view of sinistral Pb element, NMW 81.5G.38.

Figs. 26 and 27. Rhodesognathus elegans (Rhodes, 1953). 26, outer lateral view of dextral Pb element,
NMW 81.5G.39; 27, outer lateral view of dextral Pa element, NMW 81.5G.40.

Figs. 28-35. Plectodina bullhillensis sp. nov. 28, outer lateral view of sinistral Pa element, NMW 81.5G.41;
29 and 30, inner lateral and outer lateral views of sinistral Sc element, NMW 81 .5G.42; 31 and 32, inner lateral
and outer lateral views of dextral M element, NMW 81.5G.43; 33 and 34, inner lateral and outer lateral views
of dextral Sb element, NMW 81.5G.44; 35, anterior view of Sa element, NMW 81.5G.45.

Figs. 36 and 37. Pseudooneotodus sp. B. Lateral and upper views of NMW 81.5G.46.

Fig. 38. Gen. et sp. indet. B. Lateral view of element, NMW 81.5G.47.

PLATE 83

SAVAGE and BASSETT, Caradoc and Ashgill conodonts

700

PALAEONTOLOGY, VOLUME 28

Plectodina bergstroemi sp. nov.

Plate 82, figs. 1-16

Diagnosis. A species of Plectodina in which all the elements have broad bases containing conspicuous
white matter. The denticles are typically discrete and widely spaced.

Holotype. Pb element NMW 81 .6G.32, sample 31 (Smeathen Wood Formation, Ragleth, Shropshire); Plate 82,
figs. 10-12.

Description. The Pa element bears stout, nodular denticles and a main cusp which is inwardly curved with a
distinct ridge up its inner face (PI. 82, figs. 13-15). The Pb element has large, strongly inclined denticles and an
inwardly curved main cusp below which the basal cavity is deeply cupped (PI. 82, figs. 3, 4, 10-12). A narrow
ridge runs up the inner face of the main cusp from the basal margin ( PI . 82, figs. 4,10). The M element is strongly
curved and has sharp anterior and posterior edges; its basal cavity is widely expanded inward (PI. 82, fig. 1 6). The
Sc element has a gently recurved main cusp and long, discrete denticles along its posterior bar (PI. 82, figs. 8, 9).
The Sb element consists of a large median cusp curving posteriorly and extending below its junction with the
anterior and posterior bars. These bars slope sharply downward to enclose an angle of about 90° and bear widely
spaced denticles (PI. 82, figs. 1, 2, 5-7). The Sa element has not been recovered. All the elements have thick,
conspicuous white matter in their broadly excavated aboral surfaces. This material does not grade into the
darker material comprising the remainder of the element (PI. 82, figs. 3, 6, 11).

Remarks. This species is named after Stig Bergstrom in recognition of his contributions to knowledge
of Ordovician conodonts.

EXPLANATION OF PLATE 84

All figs, x 40.

Figs. 1-21. Amorphognathus ordovicicus Branson and Mehl, 1933. 1-3, upper, lower, and upper outer lateral
views of dextral Pa element, NMW 81.4G. 1; 4-6, lower, outer lateral, and upper views of sinistral Pb element,
NMW 8 1 .4G.2; 7 and 8, outer lateral and inner lateral views of sinistral Pb element, NMW 8 1 .4G.3; 9 and 10,
inner lateral and upper views of dextral Pb element, NMW 81.4G.4; 11, lateral view of Sd element,
NMW 81.4G.5; 12, lateral view of Sd element, NMW 81.4G.6; 13 and 14 outer lateral and upper views of
sinistral Pb element, NMW 81.4G.7; 15 and 16, outer lateral and inner lateral views of dextral M element,
NMW 81.4G.8; 17, lateral view of Sd element, NMW 81.4G.9; 18, lateral view of sinistral Sb element,
NMW 81.4G.10; 19 and 20, lateral view of Sd element, NMW 81.4G.il; 21, outer lateral view of dextral Sc
element, NMW 81.4G.12. Figs. 1-12, 17, and 18 from sample 14; figs. 13, 14, and 19-21 from sample 11;
figs. 15 and 16 from sample 17.

Figs. 22-25. Icriodella superba Rhodes, 1953. 22, upper view of anterior fragment of sinistral Pa element,
NMW 81.4G.13; 23, outer lateral view of sinistral Pa element, NMW 81.4G.14; 24 and 25, inner lateral and
outer lateral views of sinistral M element, NMW 81.4G.15; all from sample 14.

Figs. 26 and 27. Prioniodus sp. Inner lateral and outer lateral views of dextral Sc element, NMW 81.4G.16, from
sample 14.

Figs. 28 and 29. Rhodesognathus elegans (Rhodes, 1953). 28, outer lateral view of sinistral Pb element,
NMW 81.4G.17, from sample 11; 29, outer lateral view of dextral Pa element, NMW 81.4G.57, from
sample 14.

Figs. 30-33. lOzarkodina pseudofissilis (Lindstrom, 1959). 30, inner lateral view of dextral Pa element,
NMW 81.4G.18; 31 and 32, inner lateral and lower views of sinistral Pa element, NMW 81.4G.19; 33, outer
lateral view of dextral Pb element, NMW 81.4G.20. Figs. 30 and 33, from sample 1 1; figs. 31 and 32, from
sample 14.

Figs. 34-46. Aphelognathus rhodesi (Lindstrom, 1959). 34 and 35, lower and inner lateral views of sinistral
M element, NMW 81 .4G.21; 36 and 37, inner lateral and lower views of sinistral Pb element, NMW 81.4G.22;
38 and 39, inner lateral and lower views of sinistral Pb element, NMW 81.4G.23; 40 and 41, lower and
posterior views of Sa element, NMW 81.4G.24; 42, inner lateral view of dextral Pa element, NMW 81.4G.25;
43 and 44, inner lateral and lower views of dextral Sc element, NMW 81.4G.26; 45 and 46, inner lateral and
outer lateral views of sinistral Sb element, NMW 81.4G.27. Figs. 34, 35, 38-46 from sample 14; figs. 36 and 37
from sample 1 1 .

PLATE 84

SAVAGE and BASSETT, Caradoc and Ashgill conodonts

702

PALAEONTOLOGY, VOLUME 28

Plectodina bullhillensis sp. nov.

Plate 80, figs. 15-22; Plate 81, figs. 1-17; Plate 82, figs. 17-27; Plate 83, figs. 28-35

Diagnosis. A species of Plectodina in which the Pa element bears discrete, stout denticles, a main cusp
more posteriorly inclined than the anterior denticles, and a basal cavity which is widely expanded
along the posterior two-thirds of the element. The Pb element is relatively small with an anterior
blade twice the height of the posterior blade.

Holotype. Pa element NMW 81 .6G.22, sample 72 (Hoar Edge Grit, Bullhill Gutter, Shropshire); PI. 81, figs. 1-3.

Description. The Pa element shows some ontogenetic range. It may be large and broad, in which case it bears
about three short, stout anterior denticles and four to six fairly stout posterior denticles. In these large specimens
the entire basal cavity is widely expanded, and particularly so along the posterior two-thirds of the element
(PI. 81, figs. 1-3). In smaller specimens the cusp and denticles are less thickened and the posterior part of the
element usually is broken off (PI. 80, fig. 1 5; PI. 82, figs. 1 7-19). The Pb element is consistently small and delicate.
It has an anterior blade twice the height of the posterior blade and a basal cavity expanded beneath the fairly
upright main cusp (PI. 81, figs. 4-6). The M element has a long blade-like main cusp which is sharply twisted
inward and backward near its base. There are no anterior denticles. The basal margin of the cusp is quite sharply
inclined at an angle of about 1 10° relative to the basal margin of the posterior bar (PI. 81, fig. 10). The posterior
bar is thin, particularly where it attaches to the cusp, so that in many specimens it is broken from the cusp. It is
straight or slightly recurved, tapers posteriorly, and bears seven or more posteriorly inclined denticles which are
in contact at their bases but are otherwise discrete. A posteriorly expanded basal cavity is present below the cusp.
It is sharply constricted where it extends to the posterior bar but then continues along the bar as a pronounced
groove (PI. 81, figs. 9 and 10). The Sc element has a gently reclined main cusp and a long, straight posterior bar
which bears eight or more well-spaced denticles (PI. 81 , figs. 1 1 and 12). The cusp is laterally compressed apart

EXPLANATION OF PLATE 85

All figs, x 40.

Figs. 1-26. Amorphognathus ordovicicus Branson and Mehl, 1933. 1, upper view of sinistral Pa element,

NMW 81.4G.28; 2, upper view of sinistral Pa element, NMW 81.4G.29; 3-5, outer lateral and inner lateral
views of Sd element, NMW 81.4G.30; 6 and 7, posterior and anterolateral views of Sa element,
NMW 81.4G.31; 8 and 9, inner lateral and outer lateral views of dextral Sb element, NMW 81.4G.32; 10 and
1 1, outer lateral and inner lateral views of sinistral Sb element, NMW 81 .4G.33; 12 and 13, outer lateral and
inner lateral views of dextral Pb element, NMW 81.4G.58; 14, inner lateral view of dextral Sc element,
NMW 8 1 .4G.34; 1 5, inner lateral view of dextral Sc element, NMW 8 1 .4G.35; 1 6, outer lateral view of dextral
Sb element, NMW 81.4G.36; 17, inner lateral view of dextral Sc element, NMW 81.4G.37; 18-20, anterior
and lateral views of Sa element, NMW 81.4G.38; 21, upper view of sinistral Pb element, NMW 81.4G.39; 22,
inner lateral view of sinistral Pb element, NMW 81.4G.40; 23, upper view of sinistral Pa element,
NMW 81.4G.41; 24, upper view of dextral Pa element, NMW 81.4G.42; 25 and 26, upper and upper lateral
views of sinistral Pa element, NMW 8 1 ,4G. 43. Figs. 1 1 1 and 22 from sample 19; figs. 12-20 from sample 21;
figs. 21 and 23-26 from sample 20.

Fig. 27. lOzarkodina pseudofissilis (Lindstrom, 1959). Inner lateral view of sinistral Pa element, NMW 81 .4G.44,
from sample 2 1 .

Figs. 28-35, 44, and 45. Prioniodus sp. 28-30, upper, outer lateral, and inner lateral views of sinistral Pa
element, NMW 81.4G.45; 31 and 32, outer lateral and inner lateral views of sinistral Pb element,
NMW 81.4G.46; 33, outer lateral view of dextral Pb element, NMW 81.4G.47; 34 and 35 outer lateral and
inner lateral views of sinistral Sc element, NMW 81.4G.48; 44 and 45, lateral views of M element,
NMW 81.4G.49. Figs. 28-32, 34, and 35 from sample 21; figs. 33, 34, and 45 from sample 19.

Figs. 36-39. Rhodesognathus elegans (Rhodes, 1953). 36, inner lateral view of dextral Pb element,
NMW 81.4G.50; 37, outer lateral view of sinistral Pb element, NMW 81.4G.51; 38, outer lateral view of
sinistral Pa element, NMW 81.4G.52; 39, outer lateral view of dextral Pa element, NMW 81.4G.53.
Figs. 36 and 37 from sample 21; fig. 38 from sample 20; fig. 39 from sample 19.

Figs. 40-43. Icriodella superba Rhodes, 1953. 40, upper view of fragmentary Pa element, NMW 81.4G.54;
41 and 42, lateral and upper views of fragmentary Pa element, NMW 81 .4G.55; 43, lateral view of Pb element,
NMW 81.4G.56. Figs. 40 and 43 from sample 20; figs. 41 and 42 from sample 19.

PLATE 85

SAVAGE and BASSETT, Caradoc and Ashgill conodonts

704

PALAEONTOLOGY, VOLUME 28

from a gentle expansion extending down its inner side (Pi. 8 1 , tig. 11). There are no anterior denticles. The basal
cavity beneath the cusp is narrow and continues back along the posterior bar with only minor decrease in width.
The Sb element has a weakly sinuous and posteriorly recurved main cusp with basal lips projecting downward on
both the anterior and posterior sides (PI. 8 1 , figs. 7 and 8). The lateral bars are of unequal length and diverge at an
angle of 75° to 80°. The shorter lateral bar, which is flexed slightly posteriorly and arched, bears three or four
compressed denticles. The longer lateral bar is straighter and bears three or four more widely spaced denticles.
The Sa element is usually symmetrical with lateral processes enclosing an angle of about 65°. The posterior face
of the cusp bears a median ridge which is cleft where it runs into the basal cavity (PI. 81, fig. 16). Occasionally the
Sa element is slightly asymmetrical and then has the lateral processes more open, to enclose an angle of about
85°, and the cusp ridge twisted sidewards (PI. 81, figs. 13-15).

Remarks. This species is characterized by its broad Pa element, bearing only three or four anterior
denticles, and its contrastingly small, delicate Pb element. At present it is known from the lower part
of the Costonian to the upper part of the Woolstonian.

Plectodina tenuis (Branson and Mehl, 1933c)

Plate 80, figs. 1-5; Plate 81, figs. 19-35

1933c Ozarkodina tenuis n. sp. Branson and Mehl, p. 128, pi. 10, figs. 19 [syntype], 20, 21, and 23
[Pa element].

1933c Prioniodus ( ?) flex uosus n. sp. Branson and Mehl, p. 130, pi. 10, fig. 16 [M element].

1933c Cordylodusl delicatus n. sp. Branson and Mehl, p. 129, pi. 10, figs. 14 and 15 [Sc element],

1933c Phrcigmodus minis n. sp. Branson and Mehl, p. 123, pi. 10, fig. 12 [Sb element],

1933c Trichognathus tenuis n. sp. Branson and Mehl, p. 131, pi. 10, fig. 18 [Sa element],

1966 Plectodina furcata (Hinde); Bergstrom and Sweet, p. 377, pi. 32, figs. 17-19; pi. 33, figs 1-4; pi. 34,
figs. 9-12 [part of multielement apparatus],

71981/? Plectodina tenuis (Branson and Mehl); Sweet, p. 274, figs. 10-18, pp. 287-290 [multielement
apparatus: but note reversal of Pa and Pb notation compared with that used herein and in Sweet
1979Z>, 1981a].

Remarks. This species has been referred frequently to Plectodina furcata (Hinde) but it now appears
that the holotype of furcata is part of an Oulodus apparatus as discussed above in the remarks on

EXPLANATION OF PLATE 86

All figs, x 40.

Figs. 1-13. Amorphognathus ordovicicus Branson and Mehl, 1933. 1, upper view of sinistral Pa element,

NMW 81 .6G.35; 2, upper view of sinistral Pa element, NMW 81.6G.36; 3, upper view of sinistral Pa element,
NMW 81.6G.37; 4 and 5, upper and lower views of sinistral Pa element, NMW 81.6G.38; 6, upper view of
sinistral Pb element, NMW 81.6G.39; 7, upper view of sinistral Pb element, NMW 81.6G.40; 8 and 9, upper
and outer lateral views of sinistral Pb element, NMW 81.6G.41; 10, upper view of dextral Pb element,
NMW 81.6G.42; 11-13, anterior and posterior views of Sa element, NMW 81.6G.43; all from sample 95.

Fig. 14. Gen. et sp. indet. A. Lateral view of element, NMW 81.6G.44, from sample 95.

Fig. 15. Protopanderodus liripipus Kennedy, Barnes and Uyeno, 1979. Lateral view of NMW 81.6G.45, from
sample 95.

Figs. 16 and 17. Phragmodus cf. P. undatus Branson and Mehl, 1933. 16, inner lateral view of dextral S element,
NMW 81.6G.46; 17, inner lateral view of sinistral M element, MNW 81.6G.47; both from sample 95.

Figs. 18-22. Icriodella superba Rhodes, 1953. 18-21, upper, lower, inner lateral, and outer lateral views of
dextral Pa element, NMW 81.6G.48; 22, inner lateral view of sinistral M element, NMW 81.6G.49; all from
sample 77.

Figs. 23 and 24. Rhodesognatlws elegans (Rhodes, 1953). Inner lateral and outer lateral views of dextral
Pa element, NMW 81.6G.50, from sample 77.

Figs. 25-33. Amorphognathus all. A. tvaerensis Bergstrom, 1962. 25-27, inner lateral, upper, and lower views
of sinistral Pa element, NMW 8I.6G.51; 28 and 29, posterior and anterior views of sinistral M element,
NMW 81 .6G.52; 30, lateral view of Sa element, NMW 81 .6G.53; 3 1 -33, outer lateral, inner lateral, and lower
views of sinistral Pb element, NMW 81.6G.54; all from sample 77.

PLATE 86

SAVAGE and BASSETT, Caradoc and Ashgill conodonts

706

PALAEONTOLOGY, VOLUME 28

Aphelognathus rhodesi. The elements of Plectodina in the Branson and Mehl fauna now take the name
P. tenuis (Branson and Mehl, 1933c). These Shropshire occurrences of this species are from the
Costonian to Woolstonian or possibly lowest Marshbrookian. The Costonian occurrences appear to
extend the lower range of the species, previously thought to commence in the late Kirkfieldian
(Sweet 19816).

The type material of this species is from the Maquoketa Formation, Clarksville, Missouri. Our
material is very similar to the specimens illustrated from that formation by Branson and Mehl
(1933c). The multielement apparatus assigned to P. tenuis and figured by Sweet (19816, p. 274) in the
Catalogue of Conodonts is from several localities in Kentucky and one in Ohio. The M element is
different from the Branson and Mehl specimen (1933c, pi. 10, fig. 16) in having smaller and more
upright denticles. The Sa element illustrated by Sweet (19816, p. 275, figs. 1 1 and 12) also appears to
be more slender and sinuous than the Branson and Mehl specimen (1933c, pi. 10, fig. 12), although
the latter is incomplete and difficult to compare.

? Plectodina sp.

Plate 80, figs. 40 and 41

Remarks. This M(?) element is relatively short, with closely set inclined denticles along the bar. It
cannot be assigned to any known species but is probably part of a Plectodina apparatus.

Family icriodontidae Muller and Muller, 1957
Genus icriodella Rhodes, 1953

Type species. Icriodella superba Rhodes, 1953, p. 288.

Icriodella superba Rhodes, 1953

Plate 80, figs. 8 14; Plate 82, figs. 28-33; Plate 83, figs. 20-25; Plate 84, figs. 22-25; Plate 85, figs. 40-43

1953 Icriodella superba n. sp. Rhodes, p. 288, pi. 20, figs. 62, 63, 78 [holotype], 54, 58, and 65
[Pa element].

1953 Sagittodontus robustus n. sp. Rhodes, p. 31 1, pi. 21, figs. 141 and 142 [M element],

1953 Sagittodontus robustus var. erectus n. var. Rhodes, p. 311, pi. 21, figs. 143, 151, and 152
[M element],

1953 Sagittodontus robustus var. distaflexus n. var. Rhodes, p. 312, pi. 21, figs. 137 and 138
[M element],

1953 Trichonodella divaricate n. sp. Rhodes, p. 313, pi. 21, figs. 140, 145, and 146 [Pb element].

1953 Icriodella superba var. acuta n. var. Rhodes, p. 288, pi. 20, figs. 59, 60, 64, 65, 71-73, and 77
[Pa element],

1953 Icriodella plana n. sp. Rhodes, p. 287, pi. 20, figs. 67, 74, and 76 [Pa element],

1953 Icriodella n. sp. Rhodes, p. 288, pi. 20, fig. 61 [Pa element],

1953 Icriodella deforma n. sp. Rhodes, p. 286, pi. 20, figs. 67-70 [Pa element],

1953 Icriodella elongata n. sp. Rhodes, p. 287, pi. 20, figs. 79-81 [Pa element],

1981 Icriodella superba Rhodes; Klapper and Bergstrom, pp. W125, W126, fig. 74(1 a-f) [multielement
apparatus: but note that this figured material is not from the type locality and is quadrimembrate
compared with our trimembrate apparatus].

Remarks. Although specimens of Icriodella occur at several horizons in Shropshire and Wales, no
consistent differences are evident by which the genus can be subdivided. All the specimens are
referred to I. superba Rhodes, originally described from the Cymerig Limestone of North Wales.
Orchard (1980) recognized I. superba superba Rhodes and I. superba deforma Rhodes from the
Cymerig Limestone, the latter being diagnosed by uneven development of the anterior denticles of the
Pa elements. Although some specimens in the collections described herein have an uneven
development, we are not able to recognize consistent differences. The Pa elements from any particular
locality show sufficient range of variation of anterior denticle development, asymmetry of anterior

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

707

platform, inclination of denticles and main cusp, and total length of the unit, that the taxonomic
subdivisions of Rhodes ( 1 953) commonly have been synonymized by later workers (Bergstrom 1 964;
Schopf 1966; Orchard 1980).

Family phragmodontidae Bergstrom, 1981c
Genus phragmodus Branson and Mehl, 19336

Type species. Phragmodus primus Branson and Mehl, 19336, p. 98

Phragmodus cf. P. undatus Branson and Mehl, 19336
Plate 86, figs. 16 and 17

cf. 19336 Phragmodus undatus n. sp. Branson and Mehl, p. 115, pi. 8, figs. 22-24 [S elementl.
cf. 1966 Phragmodus undatus Branson and Mehl; Bergstrom and Sweet, p. 369, pi. 28, figs. 1 3 20
[Pa, Pb, M, and S elements = multielement apparatus].

Remarks. Only a few specimens of each of the S and M elements have been recovered and these are
broken. Nevertheless, the S element is sufficiently diagnostic to make tentative assignation to
Phragmodus undatus possible.

Family prioniodontidae Bassler, 1925
Genus prioniodus Pander, 1856

Type species. Prioniodus elegans Pander, 1856, p. 29.

Prioniodus deani sp. nov.

Plate 80, figs. 23-39

Diagnosis. A species of Prioniodus in which the Pa elements are strongly arched and relatively short.
The dextral Pb element has straight lower margins and a prominent lateral process. The sinistral Pb
element has arched lower margins and a weak lateral process.

Holotype. Pa element NMW 81.5G.2, sample 26 (Coston Formation, Shropshire); Plate 80, figs. 23 and 24.

Description. The dextral Pa element is short and strongly arched with a small, posteriorly inclined main cusp
(PI. 80, figs. 25 and 26). At least seven small denticles are present on the anterior process and four on the posterior
process. The outer lateral process arises from the base of the main cusp. The sinistral Pa element is more strongly
arched but otherwise very like the dextral element (PI. 80, figs. 23 and 24). The dextral Pb element has straight
lower margins, a large, slightly inclined main cusp, and a prominent lateral process (PI. 80, figs. 27-29). The
sinistral Pb element has arched lower margins, a relatively small main cusp which is inclined posteriorly and
inwards, and a very weak lateral process which arises just anterior of the cusp (PI. 80, figs. 29-37). The S element
has a very weak cusp, one edenticulate process, and two denticulate process (PI. 80, figs. 38 and 39). The
M element has not been recovered.

Remarks. This species is known at present only from the Costonian and may be a useful indicator of a
pre-Flarnagian age. The specific name recognizes W. T. Dean’s contributions to the Ordovician
geology of Shropshire.

Prioniodus sp.

Plate 84, figs. 26 and 27; Plate 85, figs. 28-35, 44, and 45

1980 Prioniodus sp. nov. A, Orchard, p. 24, pi. 6, figs. 5, 9, 11, and 12

Remarks. This species is represented by specimens from the Birdshill Limestone and Crug Limestone.
Unfortunately, all the Pa and Pb elements in the collection are broken. Orchard (1980, p. 24)
discussed specimens from the Birdshill Limestone which are probably conspecific with this material.

708

PALAEONTOLOGY, VOLUME 28

Family panderodontidae Lindstrom, 1970
Genus panderodus Ethington, 1959

Type species. Paltodus unicostatus Branson and Mehl, 1933ft, p. 42.

Panderodus cf. P. gracilis (Branson and Mehl, 1933ft)

Plate 80, figs. 42-47

cf. 1933ft Paltodus gracilis n. sp. Branson and Mehl, p. 108, pi. 8, figs. 20 and 21 [graciliform element],
cf. 1933ft Paltodus compressus n. sp. Branson and Mehl, p. 109, pi. 8, fig. 19 [compressiform element],
cf. 1976 Panderodus gracilis (Branson and Mehl); Dzik, p. 428, fig. 15 a, ft, e , / (multielement
apparatus].

Remarks. Workers who have attempted to distinguish multielement apparatuses of Panderodus
include Barrick (1977), Cooper (1975), Dzik (1976), and Barnes et al. (1979). We have chosen to
adopt a simple distinction in dealing with our specimens and to separate only the more flattened
‘compressiform' elements from the narrower ‘graciliform’ elements.

Family scolopodontidae Bergstrom, 1981c
Genus staufferella Sweet, Thompson and Satterfield, 1975

Type species. Distacodus falcatus Stauffer, 1935, p. 142.

? Staufferella sp.

Plate 82, figs. 39-41

Remarks. This species is represented in our collections by a single specimen. It is rounded in cross-
section and has a weak carina on one side. It may be a unicarinate element of Staufferella.

Family drepanoistodontidae Bergstrom, 1981c
Genus drepandoistodus Lindstrom, 1971

Type species. Oistodus forceps Lindstrom, 1955, p. 574.

Drepanoistodus suberectus (Branson and Mehl, 1933ft)

Plate 81, fig- 18; Plate 82, fig. 38

1933ft Oistodus suberectus n. sp. Branson and Mehl, p. 1 1 1, pi. 35, figs. 22-27 .

1979ft Drepanoistodus suberectus (Branson and Mehl); Sweet, p. 79, figs. 7-21, 23, and 30
[multielement apparatus].

Remarks. Only non-geniculate elements of this species have been recovered. They range throughout
the full stratigraphical coverage of our samples.

Family protopanderodontidae Lindstrom, 1970
Genus protopanderodus Lindstrom, 1971

Type species. Acontiodus rectus Lindstrom, 1955, p. 549.

Protopanderodus liripipus Kennedy, Barnes and Uyeno, 1979

Plate 86, fig. 15

1979 Protopanderodus liripipus n. sp. Kennedy, Barnes and Uyeno, p. 546, pi. 1, figs. 9-19 [multi-
element apparatus].

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

709

Remarks. The sixteen specimens in the sample 78 collection include symmetrical and asymmetrical
protopanderodiform elements and a single scandodiform element. The elements are very similar to
the type material from the early Caradoc age Tetaugouche Group, New Brunswick.

Family polygnathidae Bassler, 1925
Genus ozarkodina Branson and Mehl, 1933a

Type species. Ozarkodina typica Branson and Mehl, 1933a, p. 51.

? Ozarkodina pseudofissilis (Lindstrom, 1959)

Plate 84, figs. 30-33; Plate 85, fig. 27

1959 Ctenognathns pseudofissilis n. sp. Lindstrom, p. 439, pi. 4, figs. 1-9 [Pa element],

1959 Ozarkodina pseudotypica n. sp. Lindstrom, p. 441, pi. 4, figs. 17 and 18 [Pb element].

Remarks. Numerous well-preserved Pa elements of this species have been recovered by us from the
Crug quarry (11-17), which is the type locality, and also farther west along the Towy Anticline. The
Pb elements are rare. No other elements of a possible apparatus have been recovered and the species is
referred only tentatively to Ozarkodina. In these South Wales localities this species is commonly
associated with Amorphognathus ordovicicus and appears to indicate an Ashgill age.

Orchard (1980) reconstructed this species to include the Pa and Pb elements and, as a possible Sc
element, the single Hindeodellal sp. specimen recovered from the Crug Limestone by Lindstrom (1959,
pi. 1, fig. 10). It is surprising that other parts of the apparatus have failed to show up in any of our
thirteen samples that yield the Pa or Pb elements, and that Lindstrom and Orchard had much the
same result. There remains the possiblity that the species is bimembrate.

Family unknown

Genus pseudooneotodus Drygant, 1974
Type species. Oneotodusl beckmanni Bischoff and Sannemann, 1958, p. 98.

Pseudooneotodus sp. A
Plate 80, figs. 6 and 7

1967 Genus et species ind. B Serpagli, p. 107, pi. 29, figs, la, b.

Remarks. This material appears to be identical to that figured by Serpagli (1967) from the Ashgill of
the Carnic Alps. The form described as Pseudooneotodus aff. P. beckmanni by Orchard (1980), from
the Ashgill of the English Lake District, appears to have a more acutely angled cone.

Pseudooneotodus sp. B
Plate 83, figs. 36 and 37

Remarks. This single element differs from Pseudooneotodus sp. A in having a subrectangular outline,
a difference which additional specimens might show to be of no taxonomic significance if a range of
variation can be demonstrated. The specimen illustrated by Branson and Mehl (19336, pi. 9, fig. 3)
from the Ordovician Plattin Formation of Missouri is even more angular.

Gen. et sp. indet. A
Plate 86, fig. 14

Remarks. This element has a sharply angled triangular cross-section near the base and in this respect
is similar to some of the specimens assigned to Walliserodus debolti (Rexroad) by Serpagli

710 PALAEONTOLOGY, VOLUME 28

(1967, pi. 31). However, the element figured herein is more flared posteriorly and bears some small
denticles on this posterior edge.

Gen. et sp. indet. B

Plate 83, fig. 38

Remarks. This element is flattened on its inner side and is convex, with a weak carina on its outer side.
The outer base is flared.

Acknowledgements. We thank Professor C. R. Barnes, Dr R. J. Aldridge, and Dr M. J. Orchard for critically
reading drafts of the manuscript. Dr D. Price and Dr L. R. M. Cocks gave advice on some collecting localities
and stratigraphy. Dr L. Cherns helped with field collecting, Mrs B. J. Savage and Mrs K. Bryant helped with
sample preparation; facilities for processing were provided by the University of Oregon and the University of
Kent at Canterbury (N.M.S.) and the Department of Geology, University College, Cardiff (M.G.B.). Fieldwork
expenses for N.M.S. were defrayed by National Science Foundation grant EAR 77-12908. Contributions from
the NSF and the National Museum of Wales towards the costs of publication are gratefully acknowledged.

APPENDIX

Summary of stratigraphical horizon (Formation/Member) and locality of samples 1110 for immediate
reference to data plotted in text-figs. 1 4 and Tables 1 and 2. Locality details are given here as a generalized field
name plus an eight figure National Grid Reference. Full topographical descriptions and stratigraphical plots of
individual samples within sections are given in the Supplementary Publication deposited in the British Library
(see p. 680).

SOUTH WALES
Sample No.(s)

I- 3 Sholeshook Limestone, Sholeshook railway cutting, SM 9681 1705.

4 Sholeshook Limestone, Prendergast quarry. SM 9564 1662.

5, 6 Robeston Wathen Limestone, Robeston Wathen quarry, SN 0841 1615.

7 Sholeshook Limestone, Llanddowror quarry, SN 2538 1429.

8 Sholeshook Limestone, Fron quarry, SN 171 1 1711.

9 Sholeshook Limestone, Trefenty quarry (Foxhole), SN 2962 1355.

10 Sholeshook Limestone, Parke trackside, SN 2392 1335.

II- 17 Crug Limestone, Crug quarry and farm, SN 6270 2306.

18 Un-named limestone, Pen-y-banc quarry, SN 6081 2385.

19-21 Birdshill Limestone, Birdshill quarry, SN 6014 2312.

22, 23 Birdshill Limestone, Dryslwyn Castle, SN 5545 2034.

24 Birdshill Limestone, Ty-picca quarry, SN 5395 2064.

25 Birdshill Limestone, Llanegwad quarry, SN 5151 2130.

WELSH BORDERLAND AND EASTERN WALES
Sample No.{s)

26, 27 Coston Formation, Onny section quarry, SO 41 18 8624.

28, 29 Coston Formation, A489 quarry, SO 4119 8623.

30-32 Smeathen Wood Formation, Ragicth quarry, SO 4456 9135.

33, 34 Horderley Sandstone Formation, A489 road section, SO 4151 8599.

35 Horderley Sandstone Formation, A489 road section, SO 4153 8595.

36, 37 Horderley Sandstone Formation, A489 road section, SO 4175 8583.

38 Horderley Sandstone Formation, Longlane quarry, SO 4129 8422.

39, 40 Horderley Sandstone Formation, High Wood quarry, SO 41 19 8522.

41 Alternata Limestone, Onny river, S0 4179 8575.

42-44 Alternata Limestone, Soudley quarry, SO 4772 9182.

45 Alternata Limestone, Marshbrook railway cutting, SO 4403 9049.

46 Alternata Limestone, Onny railway, SO 4175 8570.

47 Cheney Longville Formation, Glynboro Member, Soudley quarry, SO 4772 9182.

48, 49 Cheney Longville Formation, Glynboro Member, A489 road section, SO 4193 8575.

50 Cheney Longville Formation, Glynboro Member, Cheney Longville roadside, SO 4185 8501.

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

71 1

51 Cheney Longville Formation, Glynboro Member, A489 road section, SO 4201 8571.

52 Cheney Longville Formation, Crosspipes Member, Cheney Longville roadside, SO 4191 8495.
53, 54 Cheney Longville Formation, Crosspipes Member, A489 road section SO 4209 8566.

55 Cheney Longville Formation, Crosspipes Member, A489 road section, SO 4219 8559.

56, 57 Acton Scott Formation, Ragdon Member, A489 road section, SO 4232 8552.

58 Acton Scott Formation, Ragdon Member, Hatton stream SO 4645 9013

59 Acton Scott Formation, Ragdon Member, Hatton stream SO 4642 9018.

60 Acton Scott Formation, Ragdon Member, A489 road section, SO 4239 8551.

61 Acton Scott Formation, Wistanstow Member, A489 road section, SO 4241 8550.

62, 63 Acton Scott Formation. Wistanstow Member, Ragdon stream, SO 4507 9061.

64, 65 Acton Scott Formation, Wistanstow Member, Onny river, SO 4236 8539.

66, 67 Acton Scott Formation, Wistanstow Member, Ragdon stream, SO 4505 9054.

68 Acton Scott Formation, Henley Member, Acton Scott church quarry, SO 4496 8955.

69, 70 Onny Shale Formation, Onny river, SO 4259 8535.

71, 72 Hoar Edge Grits, Bullhill Gutter stream, SO 5122 9825.

73, 74 Hoar Edge Grits, Black Dicks Coppice quarry, SO 5092 9787.

75, 76 Alternata Limestone, Chatwall farmyard, SO 5135 9741 .

77,78 Nod Glas Formation, Gwern-y-Brain stream, SJ 2181 1268.

NORTH WALES
Sample No.(s)

79, 80 Pen-y-garnedd Formation, Pen-y-garnedd quarry, SJ 1090 2375.

81, 82 Pen-y-garnedd Formation, Greenhall Park quarry, SJ 1572 1888.

83 85 Dolhir Formation, Dolhir Limestone Member, Ddol-hir quarry, SJ 2029 3675.

86, 87 Dolhir Formation, Dolhir Limestone Member, Gelli Farm quarry, SJ 2360 1939.

88 Dolhir Formation, Dolhir Limestone Member, Gelli Farm yard, SJ 2368 1924.

89-91 Dolhir Formation, Dolhir Limestone Member, Cefngoed quarry, SJ 2127 3643.

92-95 Glyn Formation, Glyn Limestone Member, Ty-draw hill quarry, SJ 2062 3704.

96 Nant Hir Mudstones, Derfel Limestones Member, Aberderfel stream, SH 8508 3920.

97 Gelli Grin Ashes, Cymerig Limestone Member, Gelli Grin quarry, SH 9449 3401 .

98-100 Gelli Grin Ashes, Cymerig Limestone Member, Y Garnedd quarries, SH 9451 3537.

101 Moelfryn Mudstones, Rhiwlas Limestone Member, Rhiwlas river section, SH 9231 3690.

102-105 Moelfryn Mudstones, Rhiwlas Limestone Member, Creigiau Bychan crags, SH 9200 3150.

106 Foel y Ddinas Mudstones, Hirnanl Limestone Member, Cwm Hirnant quarry, SH 9510 2963.
107-1 10 Conway Castle Grits, Deganwy quarry, SH 7853 7908.

REFERENCES

barnes, c. r., Kennedy, d. j., mccracken, a. d., nowlan, G. s. and tarrant, g. a. 1979. The structure and
evolution of Ordvician conodont apparatuses, Leihaia , 12 , 125-151.
barrick, j. e. 1977. Multielement simple-cone conodonts from the Clarita Formation (Silurian), Arbuckle
Mountains, Oklahoma. Geologica Palaeont. 11 , 47-68.

bassler, r. s. 1925. Classification and stratigraphic use of conodonts (abstr.). Bull. geol. Soc. Am. 36, 218-220.
bergstrom, s. m. 1962. Conodonts from the Ludibundus Limestone (Middle Ordovician) of the Tvaren area
(S.E. Sweden). Ark. Miner. Geol. 3, 1 -61.

— 1964. Remarks on some Ordovician conodont faunas from Wales. Acta Univ. Lund. 2 (3), 1-66.

1971a. Conodont biostratigraphy of the Middle and Upper Ordovician of Europe and eastern North

America. Mem. geol. Soc. Am. 127 , 83-157.

— 19716. Correlation of the North Atlantic Middle and Upper Ordovician conodont zonation with the
graptolite succession. Mem. Bur. Reck. geol. minier. 73, 177-187.

— 1973. Ordovician conodonts, 47-58. In hallam, a. (ed.). Atlas of palaeobiogeograpliy. Elsevier Scientific
Publishing Company, Amsterdam, London, New York.

1977. Early Paleozoic conodont biostratigraphy in the Atlantic borderlands, 85-1 10. In swain, f. m. (ed.).
Stratigraphic micropaleontology of Atlantic Basin and Borderlands. Elsevier Publishing Company, Amsterdam.

1978. Middle and Upper Ordovician conodont and graptolite biostratigraphy of the Marathon, Texas,
graptolite zone reference standard. Palaeontology , 21 , 723-758.

1981a. Conodonts as paleotemperature tools in Ordovician rocks of the Caledonides and adjacent areas in
Scandinavia and the British Isles. Geol. For. Stockh. Fork. 102 [for 1980], 377-392.

19816. Biostratigraphical and biogeographical significance of conodonts in two British Middle Ordovician
olistostromes. Abstr. Progm. geol. Soc. Am., North-Central Section, 13 , 271.

712

PALAEONTOLOGY, VOLUME 28

Bergstrom, s. m. 1981c. Families Phragmodontidae and Scolopodontidae, W129, W141. In clark, d. l. et al.
Treatise on Invertebrate Paleontology, W, Supplement 2, Conodonta. Geological Society of America and
University of Kansas Press, Lawrence.

— 1983. Biogeography, evolutionary relationships, and biostratigraphic significance of Ordovician platform
conodonts. Fossils Strata , 15, 35-58.

— Rhodes, f. h. t. and lindstrom, m. 1984. Conodont biostratigraphy of the type Llandeilo and associated
strata in the Ordovician of Wales. Abstr. Progm geol. Soc. Am., Southeastern and North-Central Sections,
16(3), 125.

-and sweet, w. c. 1966. Conodonts from the Lexington Limestone (Middle Ordovician) of Kentucky, and
its lateral equivalents in Ohio and Indiana. Bull. Am. Paleont. 50 (229), 271-441.
bischoff, g. and sannemann, d. 1958. Unterdevonische Conodonten aus dem Frankenwald. Notizbl. hess.
Landesamt. Bodenforsch. Wiesbaden , 86, 87-1 10.

branson, e. b. and mehl, m. G. 1933c/. Conodonts from the Bainbridge (Silurian) of Missouri. Univ. Mo. Stud.
8, 39-52.

— 19336. Conodonts from the Plattin (Middle Ordovician) of Missouri. Ibid. 101-119.

— 1933c. Conodonts from the Maquoketa-Thebes (Upper Ordovician) of Missouri. Ibid. 121-132.
-and branson, c. c. 1951. Richmond conodonts of Kentucky and Indiana . J. Paleont. 25, 1-17.
cave, r. 1965. The Nod Glas sediments of Caradoc age in North Wales. Geol. J. 4, 279-298.
cooper, b. j. 1975. Multielement conodonts from the Brassfield Limestone (Silurian) of southern Ohio.
J. Paleont. 49, 984-1008.

drygant, d. m. 1974. Simple conodonts from the Silurian and lowermost Devonian of the Volyno-Podolia area.
Paleont. Sb. 10, 64-70. [In Russian.]

dzik, j. 1976. Remarks on the evolution of Ordovician conodonts. Acta palaeont. pol. 21 (4), 395-455.
eichenberg, w. 1930. Conodonten aus dem Culm des Harzes. Palaont. Zeitschr. 12, 177-182.
ethington, r. l. 1959. Conodonts of the Ordovician Galena Formation. J. Paleont. 33, 257-292.
epstein, a. G., epstein, j. b. and Harris, l. d. 1977. Conodont color alteration— an index to organic
metamorphism. Prof. Pap. U.S. geol. Surv. 995, 1-20.
hass, w. h. 1959. Conodonts from the Chappel Limestone of Texas. Ibid. 294-J, 365-400.
hinde, g. J. 1879. On the conodonts from the Chazy and Cincinnati Group oftheCambro-Silurian, and from the
Hamilton and Genesee-Shale divisions of the Devonian, in Canada and the United States. Q. Jl geol. Soc.
Lond. 35, 351-369.

hurst, j. m. 1979. The stratigraphy and brachiopods of the upper part of the type Caradoc of south Salop. Bull.
Br. Mas. nat. Hist. 32, 183-304.

Kennedy, d. j., barnes, c. R. and uyeno, t. t. 1979. A Middle Ordovician conodont faunule from the
Tetagouche Group, Camel Back Mountains, New Brunswick. Can. J. Earth Sci. 16, 540-551.
klapper, g. and bergstrom, s. m. 1981. Family Icriodontidae, W125. In clark, d. l. et al. Treatise on
Invertebrate Paleontology, W, Supplement 2, Conodonta. Geological Society of America and University of
Kansas Press, Lawrence.

lindstrom, m. 1955. Conodonts from the lowermost Ordovician strata of south-central Sweden. Geol. For.
Stockh. Forh. 76, 517-603.

— 1959. Conodonts from the Crug Limestone (Ordovician, Wales). Micropaleontology, 5, 427-452.

1970. A suprageneric taxonomy of the conodonts. Lethaia, 3, 427-445.

— 1971. Lower Ordovician conodonts of Europe. In sweet, w. c. and bergstrom, s. m. (eds.). Symposium on
conodont biostratigraphy. Mem. geol. Soc. Am. 127, 21 61.

— 1977. Amorphognathus\ Rhodesognathus, 21-52; 531-537. In ziegler, w. (ed.). Catalogue of Conodonts,
Vol. 3. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart.

mccracken, a. d. and barnes, c. R. 1981 a. Conodont biostratigraphy and palaeoecology of the Ellis Bay
Formation, Anticosti Island, Quebec, with special reference to late Ordovician-early Silurian chronostrati-
graphy and the systemic boundary. Bull. geol. Surv. Can. 329, 51-134.

19816. Conodont biostratigraphy across the Ordovician-Silurian boundary, Ellis Bay Formation,
Anticosti Island, Quebec, 61-69. In lesperance, p. j. (ed.). Field Meeting, Anticosti-Gaspe, Quebec, 1981.
Vol. II: Stratigraphy and Paleontology. IUGS, Subcommission on Silurian Stratigraphy, Ordovician-Silurian
Boundary Working Group.

— nowlan, G. s. and barnes, C. R. 1980. Gamachignathus, a new multielement conodont genus from the latest
Ordovician, Anticosti Island, Quebec. In Current Research, Part C. Pap. geol. Surv. Can. 80-1C, 103-1 12.

muller, k. j. and muller, e. m. 1957. Early Upper Devonian (Independence) conodonts from Iowa, Part 1.
J. Paleont. 31, 1069 1108.

SAVAGE AND BASSETT: ORDOVICIAN CONODONTS

713

orchard, m. j. 1980. Upper Ordovician conodonts from England and Wales. Geologica Palaeont. 14, 9-44.
OWENS, R. M. 1973. British Ordovician and Silurian Proetidae (Trilobita). Palaeontogr. Soc. [ Monogr ]. 1-98.
pander, c. H. 1 856. Monographic der fossilen Fische des silurischen Systems des russisch-baltischen Gouverne-
ments. Akademie der Wissenschaften, St. Petersburg, 91 pp.
price, d. 1973. The age and stratigraphy of the Sholeshook Limestone of southwest Wales. Geol. J. 8, 225-246.
1984. The Pusgillian Stage in Wales. Geol. Mag. 121, 99-105.

pringle, J. and GEORGE, T. N. 1948. British Regional Geology. South Wales. (2nd edn.), vi + 100 pp. Geological
Survey and Museum, HMSO, London.

Rhodes, F. h. t. 1953. Some British Lower Palaeozoic conodont faunas. Phil. Trans. R. Soc. B237, 261-334.
savage, n. m. 1983. The use of conodont occurrence and thermal alteration data as an aid in distinguishing
tectonostratigraphic terranes. Progm Abstr. a. Mtg Geol. Ass. Can. 8, A60.
schopf, t. j. m. 1966. Conodonts of the Trenton Group (Ordovician) in New York, southern Ontario, and
Quebec. Bull. N.Y. St. Mus. 405, 1-105.

SERPAGLI, E. 1967. 1 conodonti deH'Ordoviciano Superiore (Ashgilliano) delle Alpi Carniche. Boll. Soc. Paleont.
it ah 6, 30-111.

Stauffer, c. R. 1935. Conodonts from the Glenwood beds. Bull. geol. Soc. Am. 46. 125-168.
sweet, w. c. 1979 a. Conodonts and conodont biostratigraphy of Post-Tyrone Ordovician rocks of the
Cincinnati Region. Prof. Pap. U.S.geol. Surv. 1066-G, 1-26.

— 19796. Late Ordovician conodonts and biostratigraphy of the western Midcontinent Province. In
Sandberg, c. a. and Clark, d. l. (eds.). Conodont biostratigraphy of the Great Basin and Rocky Mountains.
Geology Stud. Brigham Young Univ. 26, 45-86.

— 1981a. Macromorphology of elements and apparatuses. In clark, d. l. et al. W5-W20. Treatise on
Invertebrate Paleontology , W , Supplement 2 , Conodonta. Geological Society of America and University of
Kansas Press, Lawrence.

— 19816. Aphelognathus; Plectodina. In ziegler, w. (ed.). Catalogue of conodonts. Vol. 4 , 27-56; 271-290.
Schweizerbart’sche Verlagsbuchhandlung, Stuttgart.

— ethington, r. l. and barnes, c. R. 1971. North American Middle and Upper Ordovician conodont faunas.
In sweet, w. c. and bergstrom, s. m. (eds.). Symposium on conodont biostratigraphy. Mem. geol. Soc. Am.
127, 163-193.

— Thompson, T. L. and Satterfield, i. r. 1975. The Cape Limestone of Missouri. Rep. Invest. Mo. geol. Surv.
57, 1-60.

uyeno, t. t. and barnes, c. r. 1983. Conodonts of the Jupiter and Chicotte Formations (lower Silurian),
Anticosti Island, Quebec. Bull. geol. Surv. Can. 355, i-viii, 1-49.

WILLIAMS, A., STRACHAN, I., BASSETT, D. A., DEAN, W. T., INGHAM, J. K., WRIGHT, A. D. and WHITTINGTON, H. B. 1972.
A correlation of Ordovician rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 3, 1 -74.

N. M. SAVAGE

Department of Geology
University of Oregon
Eugene, Oregon 97403
USA

M. G. BASSETT

Typescript received 29 September 1984
Revised typescript received 8 March 1985

Department of Geology
National Museum of Wales
Cardiff CF1 3NP
Wales, UK

CAMBRIAN ELEUTHEROZOAN ECHINODERMS
AND THE EARLY DIVERSIFICATION OF
EDRIOASTEROIDS

by ANDREW B. SMITH

Abstract. The five genera and thirteen named species of edrioasteroid from the Cambrian are reviewed and,
where necessary, redescribed. All are interpreted as sessile suspension feeders, with an external system of radial
water vessels and tube feet that functioned in opening the cover plate sheets. The relationships of these five
genera to one another and to other edrioasteroid groups is analysed and a revised classification of the
Edrioasteroidea is proposed. Stromatocystites has fewest autapomorphic characters and is placed as primitive
sister group to the other four genera. These fall into two distinct groups: Totiglobus and Walcottidiscus are rather
globular with a reduced dorsal surface, while Cambraster and Edriodiscus are discoidal with a prominent
marginal frame. Three orders are recognized within the Edrioasteroidea: the Edrioasterida, which includes
Totiglobus , Walcottidiscus , and the family Edrioasteridae; the Cyathocystida, for the families Cyathocystidae
and Pyrgocystidae; and the Isorophida, which is expanded to include Cambraster, Edriodiscus , and the families
Cyclocystoididae and Agelacrinitidae.

In the Cambrian there are a small number of echinoderms that have conventionally been classified
as edrioasteroids, of which the best known is Stromatocystites. Although Pompeckj first described

S. pentangularis from the Middle Cambrian of Czechoslovakia in 1 896, these animals have remained
relatively poorly understood. Yet Stromatocystites has held an important place in theories of
echinoderm phylogeny. Bather (1900, 1915) considered Stromatocystites to be ancestral to all
eleutherozoan groups and an intermediate between asteroids, echinoids, and holothuroids on the one
hand, and cystoids on the other. This view fell largely out of favour when Fell (1962, 1963) proposed
that eleutherozoan groups were polyphyletic in origin, and in the Treatise on Invertebrate
Paleontology (Regnell 1966) Stromatocystites is treated as no more than a primitive edrioasteroid.
More recently, however, the phylogenetic status of Stromatocystites has been reassessed by both
Termier and Termier (1969, 1980), who provided a highly novel interpretation, and myself (Smith
1984#, b\ Paul and Smith 1984) where a return to the more traditional view is argued for.

Four other genera of edrioasteroid in addition to Stromatocystites have been described from the
Cambrian: Cambraster , Walcottidiscus , Totiglobus , and Edriodiscus. Because of the importance of
these genera to theories of the early history of eleutherozoan echinoderms, a detailed appraisal of
their morphology and relationships seemed long overdue. Some species have been described recently
in considerable detail, such as C. elegans by Ubaghs (1971), C. tastudorum by Jell et at. (1985),

T. nimius by Bell and Sprinkle (1978), and S. walcotti by Paul and Smith (1984). Others are, however,
still poorly known. This paper therefore sets out to review what is known about the morphology of
these animals and to examine their relationships, not only with one another but also with other echino-
derm groups. However, a detailed comparison between Cambrian edrioasteroids and primitive
starfish such as Archegonaster will be dealt with in a subsequent paper and is not considered here.

MORPHOLOGY AND ANATOMY

General organization

Although Cambrian edrioasteroids are variable in their overall shape they all share basically the same
body plan. All have a skeletal system that is clearly differentiated into dorsal and ventral surfaces. On

| Palaeontology, Vol. 28, Part 4, 1985, pp. 715-756, pis. 87-89.|

716

PALAEONTOLOGY, VOLUME 28

Totiglobus

text-fig. 1. Diagrammatic cross-sections through the five Cambrian genera considered
here and Camptostroma to show the relative development of dorsal (vertical hatching) and
ventral (dashed line) plate surfaces.

the ventral surface there is a central mouth and a marginal anus. Five ambulacral zones radiate from
the mouth and are separated by wedge-shaped interambulacral zones. The dorsal surface is
composed of a marginal ring of somewhat larger plates surrounding a pavement of flat polygonal
plates. In Stromatocystites and Cambraster, dorsal and ventral surfaces are more or less equally
developed so that the boundary between them coincides more or less with the ambitus (text-fig. 1).
The two genera differ in that C. tastudorum appears to have a double layer of plates outside the
marginal ring (Jell et al. 1985) which Stromatocystites lacks. Edriodiscus has a single layer. In
Walcottidiscus the ventral surface has become enlarged relative to the dorsal surface and the
boundary between them now lies sub-ambitally. Totiglobus has become even more extreme and its
dorsal plating is reduced to a relatively minute disc (text-fig. 1); the ventral surface is expanded and
extends well below the ambitus. Thus the relative development of dorsal and ventral surfaces is
largely responsible for the differences that exist in overall shape. All are sub-circular to sub-
pentagonal in outline but whereas Stromatocystites , Edriodiscus , and Cambraster have a fairly flat
profile, Walcottidiscus has a depressed ovoid profile and Totiglobus is almost globular.

Digestive system

The mouth lies centrally on the ventral surface at the point of convergence of the five ambulacra. It
does not open directly to the exterior, but is roofed over by a series of cover plates which may or may
not have been able to open in life. The peristomial opening is relatively small and in S. walcotti,
Cambraster , and Totiglobus is surrounded by a fixed mouth frame of ambulacral plates. In
S. pentangular is, however, the mouth may have been more flexible as there is some evidence that
the two columns of plates in each ambulacrum were able to separate along the perradial suture

SMITH: CAMBRIAN EDRIOASTEROIDS

717

proximally, as in many primitive starfish. Unfortunately, the detailed arrangement of plates in the
oral area of this species is still unknown.

The anus also opens through the ventral surface and is situated marginally in the C/D
interambulacrum. In Stromatocystites and Cambraster the periproct is a moderately large circular
zone of lath-like plates arranged radially (text-figs. 2 and 7). A similar periproctal system, but with
fewer and more organized plates, is present in Totiglobus. The periproctal plates he either flush with
the surrounding interambulacral plates or are slightly raised to form a low conical mound.

Since both mouth and anus lie on the same surface, the digestive system must, at the very least, have
been looped. The ubiquity of torsion in the digestive tract of all echinoderms, both larval and adult,
suggests that the digestive tract in edrioasteroids was also coiled. Bather (1915) argued that the gut
probably coiled anticlockwise in edrioasteroids because of a certain asymmetry in the arrangement
of periproctal plates. No such asymmetry can be detected in the Cambrian species and the direc-
tion of coiling must remain uncertain. However, an anticlockwise coiling seems the most likely since
the hydropore/gonopore opens close to the peristome on the right-hand side of the C/D inter-
ambulacrum, and such a direction of coiling of the digestive tract would leave more room for the
internal gonad in this position. By inference from living groups the digestive system presumably
consisted of a short oesophagus descending from the mouth, leading to a coiled intestine and then to a
short rectum.

Dorsal surface

Plates of the dorsal surface can be divided into two groups: those that form the marginal ring and
those that form the remainder of the dorsal surface. In Stromatocystites plates of the marginal ring
are poorly differentiated from the other dorsal plates and lie supra-ambitally. They are moderately
large, weakly geniculate plates with a rounded adoral edge and flatter, broader adapical edge which
faces the substratum. There are no sutural epispires along any of the plate margins; although adjacent
plates abut, their zone of contact is narrow and they would not have been able to form a rigid
marginal frame. Internally these marginal plates are gently concave and lack internal processes or
other evidence of muscle attachment areas. Marginal plates in Cambraster are greatly enlarged and
form a prominent frame to the ventral surface (text-fig. 2). There are eighteen to twenty marginal
ossicles in the ring. These ossicles are triangular in cross-section and adjacent ossicles firmly abut.
Their lateral faces are smooth, without crenulation, so presumably the ossicles were bound together
by collagenous ligament to form a fairly inflexible marginal frame. Ossicles at each radius have a
V-shaped notch where the ambulacral flooring plates extend on to the marginal frame (text-fig. 2).
Totiglobus also possesses a stout marginal ring of abutting ossicles; this ring would have been rather
poorly flexible. Marginal ossicles are hardly distinguishable from the exterior but have large internal
processes (see Bell and Sprinkle 1978, text-fig. 4) which presumably were associated with muscle
attachment. As the processes face towards the centre of the dorsal surface, these could well have been
attachment sites for radially arranged dorsal muscle fibres necessary if the dorsal surface acted as
a suction pad. Details of the marginal plates in Walcottidiscus are poorly known. Externally the
marginal plates are not clearly differentiated and, as in Totiglobus , there is no recognizable tesselate
ring of plates. Whether the marginal ring plates were more clearly differentiated internally is
unknown.

There is an irregular pavement of tesselate, polygonal plates within the marginal ring. The entire
dorsal surface is plated in all but Walcottidiscus where there is a central uncalcified zone. All of the
dorsal plates in Stromatocystites are similar in thickness. There is a large equant plate at the centre
of the dorsal surface which is surrounded by a variable number of large, radially elongate plates
(text-fig. 4). Dorsal plates decrease in size away from the centre of the disc and, immediately adjacent
to the marginal ring, there is a zone of tiny lath-like platelets that may not fully abut. This is the region
of plate addition where new dorsal plates were added immediately inside the marginal ring. It is also a
region of high flexibility. Dorsal plating in Totiglobus is very like that in Stromatocystites except that
there is no clearly distinguishable central plate. Only a small number of polygonal plates lie within the
marginal ring and these are smooth externally but weakly ridged and grooved internally (see Bell and

718

PALAEONTOLOGY, VOLUME 28

text-fig. 2. Cambraster cannati (Miquel, 1894), camera lucida drawing, ventral surface of
holotype of C. elegans Termier and Termier, 1969, from a latex, BM(NH) E63 1 35 (see text-fig. 19a).
Original is in the collections of the Museum of le Havre, France.

Sprinkle 1978, text-fig. 4b). All plates are tessellate and plate size decreases quite noticeably towards
the outer edge.

Cambraster has a central zone of small polygonal plates surrounded by one or two circles of large
polygonal plates. Over much of the central region of the aboral surface the plates are relatively thin
and sub-tesselate, fitting together to form a continuous pavement. All known specimens of C. cannati
show rather distorted and jumbled plating but there are two clear circles of large, radially elongate,
polygonal plates which either abut or have small plates intercalated in between (text-fig. 2). There is
only one circle of plates in C. tastudorum , all of which are contiguous (Jell et al. 1985). Plates become
thicker and more obviously polygonal towards the margin, but at the very distal edge there are small
lath-like plates which probably extended slightly beyond the marginal ossicles (text-fig. 2; see also

SMITH: CAMBRIAN EDRIOASTEROIDS

719

text - fig . 3. Cambraster cannati (Miquel, 1894), camera lucida drawing of
dorsal surface of holotype of Eikosacystis miqueli Termier and Termier,
1969, from a latex, BM(NH) E63136 (see text-fig. 19c, d). Original is in the
collections of the Museum of le Havre, France.

720

PALAEONTOLOGY, VOLUME 28

text-fig. 4. Stromatocystites walcotti Schuchert, 1919, USNM 66443, camera lucida drawing of
the holotype, showing the dorsal surface and part of the interior of the ventral surface where the

dorsal surface has been damaged.

Ubaghs 1971). The dorsal surface is slightly larger than the ventral surface, so that the marginal
ossicles lie just above the ambitus and are hidden from sight when viewed from beneath.

Edriodiscus, like Stromatocystites and Cambraster, is fully plated, but here all the plates within the
marginal ring are of very much the same size. These are flat, polygonal, and tesselate. There is a ring
of flat-based, cylindrical marginal ossicles which form a frame, as in Cambraster. Here, however, the
marginal ossicles are inserted into the dorsal surface plating so as to separate a peripheral skirt of
smaller plates from the main central pavement (text-fig. 20). All plates are covered with radially
arranged ridges and grooves, very reminiscent of the ornamentation on the lower surface of
peripheral skirt plates in isorophid edrioasteroids.

The structure of the outer zone of calcite plates in Walcottidiscus is not clearly seen. There appears
to be a large number of small lath-like plates which, in general appearance, resemble the dorsal
plating found in Edrioaster.

It is quite clear that neither Stromatocystites nor Cambraster had any means of attaching to the
substratum by their dorsal surfaces, at least as adults, and must simply have lain unattached on the

SMITH: CAMBRIAN EDRIOASTEROIDS

721

sea floor. Walcottidiscus , on the other hand, had a large uncalcified dorsal pad which presumably
acted as some form of adhesion disc. Edriodiscus has prominent radially arranged ridges and grooves,
like those found on the lower surface of peripheral plates in agelacrinitids. This structure is
undoubtedly associated with attachment but the precise function of the ridges remains uncertain. The
question remains whether Totiglobus had any means of attachment. Bell and Sprinkle (1978) thought
that the entire dorsal surface within the marginal ring may have acted as a suction pad. Certainly it is
hard to imagine how a globular animal such as Totiglobus could have been stable with such a small
base unless it was able to attach itself. This might have been achieved through various means, but the
large internal processes on the marginal ring ossicles suggest that adhesion was achieved through
suction. If the dorsal surface was moderately flexible immediately within the marginal ring, then
radially arranged muscles running from the centre of the dorsal surface to these internal processes
might have been able to pull the dorsal pad inwards to create suction.

Ventral surface

Interambulacra/ areas. The interambulacral zones in Stromatocystites and Cambraster are composed
of a large number of polygonal, tesselate plates (text-fig. 13). Epispires are present along the plate
sutures except immediately adjacent to ambulacral plates, marginal ossicles, and plates of the
periproct. Epispires presumably provided egress for external finger-like extensions of the body
coelom that functioned in gaseous exchange and presumably resembled papillae of Recent sea-stars.
Totiglobus , Walcottidiscus , and Edriodiscus all lack interambulacral epispires and their plating is
tesselate to sub-tesselate. New interambulacral plates were added adjacent to the marginal ring.
Ambulacral zones. The five ambulacra are arranged in a 2:1:2 pattern. Close to the mouth the
A and B and the D and E rays combine to make a T-shaped junction with ambulacrum C above the
mouth (text-fig. 15). The ambulacra in Cambraster and Stromatocystites are straight and continue
right to the edge of the ventral surface. The tips of the ambulacra are actually inserted between the
ossicles of the marginal ring in both genera and could therefore grow only at the same rate as the body
as a whole. Edriodiscus has weakly curved ambulacra which still reach the marginal ring. Totiglobus
has straight ambulacra but, although these extend below the ambitus, they do not reach the marginal
ring. The ambulacra in Walcottidiscus extend to the ambitus and then curve sinistrally. Ambulacra
are initially straight in juveniles but curve to grow around the body in adults. Ambulacra form the
animals’ food-gathering surfaces, so their length relative to body size is obviously crucial. Totiglobus
and Walcottidiscus arrived at different solutions to the problem of how to increase their food-
gathering surface relative to general body size. Totiglobus increased its ambulacra by expanding the
entire ventral surface relative to the dorsal surface, whereas Walcottidiscus evolved curved ambulacra
so that their growth was not constrained by the marginal ring but could continue around the
periphery.

The detailed structure of the ambulacra is very similar in all genera. Each ambulacrum consists of
a biserial column of flooring plates which are roofed over by two multiplated series of cover plates.
The morphology of the flooring plates is known in detail only for Stromatocystites , Cambraster , and
Totiglobus. Similar flooring plates are probably present in Walcottidiscus but here only the narrow
external face of the flooring plates can be seen. Individual flooring plates are rather squat,
particularly in Stromatocystites. There is a small rectangular abradial face which is exposed
externally between the interambulacral plates and the cover plates (text-figs. 6 and 8c). This is
generally smooth and flat, and adjacent faces abut. The rest of the flooring plate lies hidden beneath
the cover plate series. Flooring plates are arranged alternately and the two columns meet perradially
along a weakly zigzagged suture. This suture becomes almost straight distally in Cambraster , but in
Stromatocystites and Totiglobus the ambulacral plates firmly interlock along the mid-line. There is
a moderately large sutural pore between successive plates in each column, passing from the interior of
the test to the ambulacral channel. The pore is circular to ovoid in outline and is more or less equally
shared between the two plates. It occupies the outer half of the adradial face (text-fig. 6) and therefore
opens immediately beneath the cover plates. On the upper face, this pore is sometimes surrounded by
a slight rim. In the central region between sutural pores the flooring plate is expanded into a broad

722

PALAEONTOLOGY. VOLUME 28

flooring plates

peripheral

platelets

5 mm

text-fig. 5. Stromatocystites pentangularis Pompeckj, 1896, camera
lucida drawings, a, BM(NH) E16004, ventral surface, b, BM(NH)
E63138 (latex kindly supplied by Professor G. Ubaghs), dorsal
surface of a specimen in the collection of Dr Krantz ol Bonn.

SMITH: CAMBRIAN EDRIOASTEROIDS

723

V-shaped ridge (text-fig. 6). This ridge has a small but distinct pit situated immediately beneath the
cover plates and opening upwards; it is presumably a ligament pit for the collagenous fibres necessary
to bind the cover plates to their flooring plates. The perradial portion of the flooring plate is smooth,
weakly concave, and forms a shallow channel. The inner face of the flooring plates is unnoteworthy,
except in S', walcotti where there are paired lateral prongs (text-fig. 7). What the three dimensional
shape of these flooring plates is and what purpose the paired prongs served is not at all clear. Where
the perradial suture is obviously zigzag, the ambulacra] flooring plates must have been more or less
rigidly fixed, but in Cambraster, where this suture becomes almost linear towards the tips of the arms,
it is quite possible that the ambulacra had a certain amount of flexibility and could widen or narrow
the ambulacral groove as necessary.

sutural gap

position of radi
water vessel

perradius

al

abradial face
(exposed externally)

cover plates
rest here

ligament pit

a

text-fig. 6. Ambulacral flooring plates, ventral view, a , Cambraster cannati (Miquel, 1894), detail of
one plate from the specimen illustrated in text-fig. 2. b, c, Stromatocystites pentangularis Pompeckj,
1896; b, BM(NH) E63099; c, BM(NH) E63103. V, Totiglobus nimius Bell and Sprinkle, 1978, drawn

from Bell and Sprinkle (1978, pi. 2, fig. 2).

Cover plates are arranged as a multiple series and attach along the inner border of the flooring
plates (text-fig. 14). Distinctly larger primary cover plates are always developed, and towards the
distal tip of the ambulacra may be the only cover plates present. Primary cover plates are pentagonal
in outline with a broad flat base that rests on the flooring plate, and a distal point. One primary cover
plate sits on each flooring plate, except in Walcottidiscus where there appear to be two per flooring
plate (PI. 89, fig. 4). Primary cover plates lie either directly above their corresponding flooring plate or
slightly offset, as in some specimens of Totiglobus (Bell and Sprinkle 1978). There may be smaller
plates occasionally inserted between the primary cover plates in Stromatocystites , but in other genera
the primary cover plates abut each other.

The inner face of each primary cover plate has a distinct ridge that runs slightly obliquely from the
proximal edge to the distal point. This becomes less prominent away from the flooring plate and
fades. A second, less pronounced ridge can also be made out in some, convergent with the first. This
defines a central triangular area on the inner face of the cover plate that possibly marks the position of
ligament attachment.

Distal to the primary cover plates comes an irregular array of smaller, secondary cover plates.
These are most numerous in Walcottidiscus , where there may be four or five irregular rows, whereas
in Stromatocystites and Cambraster they form only a narrow band. In Totiglobus there is only a single
secondary cover plate inserted between the distal edges of adjacent cover plates. In well-preserved
specimens the paired cover plate series meet along the perradius to enclose the ambulacral groove; the
perradius often appears slightly sinuous.

724

PALAEONTOLOGY, VOLUME 28

flooring plates

marginal plate

periproct

5 mm

text-fig. 7. Stromatocystites walcotti Schuchert, 1919, USNM 376690, camera lucida drawing, showing the
interior of the ventral surface and oral frame structure; no locality data, but the matrix is similar to other
specimens of S. walcotti from Bonne Bay, Newfoundland.

Oral area. The precise arrangement of plates in the oral area is largely unknown in the majority of
species. In all except possibly S. pentangularis the most proximal ambulacral flooring plates are firmly
bound together to form a fixed oral frame. In S. walcotti only the most proximal flooring plate in each
column, together with the large elongate hydropore plate, are involved in the oral frame (text-fig. 7).
The first flooring plates from adjacent ambulacra meet interradially so as to exclude any inter-
ambulacral plates from the peristome margin. The hydropore plate is large and asymmetrical, with
a broad inward sloping face on the right hand side. This plate lies in interambulacrum C/D. As yet no
external hydropore opening has been observed and the stone canal may have opened directly into the
peristomial cavity. Both Cambr aster and Totiglobus have an oral frame that consists of five large
interradial elements. These appear to have formed through fusion of the first two ambulacral plates in
adjacent ambulacral columns (i.e. they are composed of four plates in total). These plates are very

SMITH: CAMBRIAN E D R I O ASTEROI DS

725

distinctive, having a broad triangular external face and an adradial face with two ridges and pits on
each side (text-fig. 2). The pits may represent passageways that have been lost in fusion. The
hydropore in Totiglobus opens as a slit-like pore between two plates, one of which is the oral frame
element in the C/D interray, the other being the next proximal flooring plate in ambulacrum C or an
adjacent interambulacral plate (Bell and Sprinkle 1978, text-fig. 2). In C. cannati the oral frame is
disrupted in the best preserved specimen and the exact location of the hydropore is uncertain.
However, there is a large interambulacral plate lying close to the peristome in the C/D interray which
is almost certainly the hydropore plate (text-fig. 2); this was probably in contact with the mouth
frame plate of that interray in life. The hydropore is clearly seen in C. tastudorum where the open-
ing lies between the C/D ambulacral mouth frame and the hydropore plate (Jell el at. 1985). As in
many edrioasteroids the arrangement of cover plates in the oral area is undifferentiated from that of
the ambulacra and no large plates stand out.

The water vascular system

The presence of radially arranged ambulacra and a hydropore shows that these Cambrian genera
possessed a water vascular system, like all other echinoderms. From the single, off-centred hydropore
a short stone canal would have descended to the circum-oesophageal ring. There is little direct
evidence as to where this ring lay, but the shallow ledge on the adoral face of the mouth frame plates
in Totiglobus and Cambraster may mark its position.

The ambulacra were unquestionably associated with radial water canals that ran from the circum-
oesophageal ring to the tip of each ambulacrum. The position of this water vessel has been
interpreted in two contrasting ways. Bather (191 5), when faced with an almost identical arrangement
in Edrioaster , thought that the radial water vessel lay above the flooring plates and that the sutural
passageways between adjacent flooring plates connected external tube feet to their internal ampullae
(text-fig. 8a). This arrangement was also favoured by Paul and Smith (1984) for Stromatocystites.
Bell (1977), however, argued that edrioasteroids possessed an internal radial water vessel that gave
rise to a rather complex arrangement of tubes and bulbs (text-fig. 8b): Bell and Sprinkle (1978)
applied this model to Totiglobus. Bell produced three arguments in support of his interpretation.
First, he noted that in some isorophid edrioasteroids there are pores between adjacent cover plates
that lead not into the ambulacral grooves but directly into the thecal interior. These he interpreted as
being passageways for tube feet. However, isorophids are a highly derived group with flooring plates
that are completely different from those of Edrioaster and Stromatocystites. Even if Bell is correct in
interpreting these cover plate pores as passageways for tube feet, nothing equivalent is found in the
Cambrian genera and the arrangement of the water vascular system in the highly derived isorophids
is not necessarily the same as in more primitive groups.

Bell's second observation was that in edrioasteroids the hydropore lies on the outer side of the
mouth frame, making it rather implausible that the stone canal looped under the mouth frame to
reach the externally situated circum-oesophageal ring. He thought it much more likely that the stone
canal would simply have descended to an internal ring vessel which then gave rise to internal radial
water vessels. Yet this arrangement with external radial water vessels and hydropore situated distal to
the mouth elements is precisely what occurs in pelmatozoans, asteroids, and primitive ‘ophiuroids’.

Thirdly, Bell suggested that the ambulacral pore lay too close to the cover plates to have allowed
the tube feet room to function efficiently, and therefore that the ambulacral pore was for egress of the
radial water vessel. This argument collapses if the function of the tube feet in these animals was to
open the cover plate skirt, as Paul and Smith ( 1 984) suggested. The tube foot would then need to have
been positioned close to the cover plate attachment zone, and possibly even connected to the cover
plate sheet. The flooring plate passageways would then have led from the tube feet to their internal
ampullae which served as fluid reservoirs and allowed the tube feet to inflate and deflate
independently (text-fig. 8c). As embryological studies have shown that an internal radial water vessel
is a derived character, and as all pelmatozoans and the more primitive cleutherozoans have external
radial water vessels, it seems much more probable that in these Cambrian genera the radial water
vessel lay above the flooring plates in the ambulacral tunnel. One of the prime reasons for having

726

PALAEONTOLOGY, VOLUME 28

text-fig. 8. Diagrammatic cross-sections through ambulacra showing flooring
plates, cover plates, and the inferred arrangement of the water vascular system
according to: a. Bather (1915); b. Bell (1977); c, this paper.

a series of cover plates above the ambulacral tunnel must surely have been to provide protection for
the radial water vessel and associated nerves. The smooth perradial channel on the floor of the
flooring plates presumably marks the position of this vessel.

MODE OF LIFE

From the preceding description it would appear that Cambrian edrioasteroids were sessile, low-level
suspension feeders. They lived either unattached on the sea floor or fixed to the substratum by their
dorsal surface, which in some was modified into a suction pad. All lived with their oral surface facing
upwards away from the sea floor. This is suggested by their shape, since they have a flat dorsal surface
and a weakly to strongly convex ventral surface, and by the presence of ventral epispires in some
genera. Epispires are directly connected with gaseous exchange and no echinoderm has respiratory
structures on the lower rather than the upper surface. Food must have been captured by the
ambulacra and transported to the mouth along the ambulacral grooves, since these converge upon
the mouth. When the animal was feeding the cover plate sheets must have been open to expose the
ambulacral grooves. The simplest mechanism for raising the cover plate sheets would have been
through inflation of the closely adpressed tube feet. If the prominent pit on the outer face of each
cover plate is correctly interpreted as a ligament pit, then connective tissue running from this pit to
a corresponding depression on the inner face of primary cover plates could have acted like a tension

SMITH: CAMBRIAN EDRIOASTEROIDS

727

spring, pulling the cover plate skirt back down over the ambulacral groove when the tube foot
deflated. As both mouth and food gathering surfaces were situated on the upper surface, these
animals could not have been detritus feeders but must have eaten fine particulate matter in
suspension, like pelmatozoan echinoderms. Presumably they were ciliary mucus feeders, using ciliary
currents to draw particulate matter into the ambulacral grooves and mucus to trap this material and
pass it to the mouth. Both functions could have been performed by the epithelial lining of the flooring
plates and the inner surface of the cover plates, although it is possible that there may have been
additional crinoid-like tube feet arising directly from the radial water vessel to assist in food capture.

Termier and Termier ( 1 969) suggested that Stromatocystites and Cambraster were infaunal, living
beneath a thin covering of sediment and drawing water with suspended nutrients into permanently
sealed ambulacral grooves through orifices at the tip of each ambulacrum. However, the distal
orifices that they described are post-mortem artifacts caused by the collapse of the distalmost cover
plates in two ambulacra in the holotype of S. walcotti. These animals are also unlikely to have lived
infaunally since, unlike the asteroids with which the Termiers drew comparison, they had no paxillae
to maintain a water-filled space around their buried body, which would have enabled the papillae to
continue functioning.

PHYLOGENETIC ANALYSIS

The evolutionary relationships of Cambrian edrioasteroids are important for the analysis of the early
history of eleutherozoan echinoderms. The major eleutherozoan groups, such as the asteroids,
echinoids, and holothuroids, do not appear in the fossil record until after the Cambrian; the same is
true fora number of important edrioasteroid groups. So, can any of these lineages be traced back into
the Cambrian? The relationship of Stromatocystites to other Lower Cambrian echinoderms has
already been analysed by Paul and Smith (1984) who also briefly sketched the later diversification of
eleutherozoans. Here I attempt a more detailed analysis of how the Cambrian edrioasteroids relate to
later groups.

For this phylogenetic analysis outgroup comparison has been made with the Lower Cambrian
genus Camptostroma. The reasons for placing this genus as one of two primitive sister groups of
Stromatocystites were fully dealt with by Paul and Smith (1984). They considered Camptostroma to
be intermediate between Stromatocystites and the Cambrian pelmatozoans. The characters that seem
to be important in determining relationships between the five Cambrian genera are listed in Table 1 .
Despite our lack of knowledge about certain morphological attributes in some of these genera,
a reasonable cladogram can be drawn up (text-fig. 9). All five genera differ from Camptostroma in
having a flattened dorsal surface, rather than a squatly conical dorsal surface capable of some degree
of spiral elongation and contraction. Their body-wall skeleton is also only one layer thick and
composed of tessellate plates dorsally, whereas in Camptostroma the skeleton is multilayered and
dorsal plating is clearly imbricate.

Of the five Cambrian genera, Stromatocystites has the fewest novel characters in comparison with
Camptostroma and other Lower Cambrian echinoderms, and is placed as the primitive sister group to
the rest. The remaining four genera fall into two groups of two. Walcottidiscus and Totig/obus share
the synapomorphies of having a rather globular body form (as a result of relative expansion of the
ventral surface and reduction of the dorsal surface), the loss of interambulacral epispires, the
presence of a basal suction pad, and ambulacra whose distal tips are not inserted into the marginal
ring of ossicles. Cambraster and Edriodiscus are quite different in shape, with an extensive flat dorsal
surface and a slightly smaller ventral surface. Their marginal ossicles, which are relatively poorly
developed in the other genera, are large and stout with a roughly triangular cross-section; this is taken
as a synapomorphy for the group. In both Cambraster and Edriodiscus the dorsal pavement of plates
extends slightly beyond the marginal ring to form a peripheral sheet, although Edriodiscus is more
advanced in having the marginal ossicles actually inserted into the dorsal pavement. Totig/obus and
Cambraster share two important characters which are not present in Stromatocystites but which, by
implication, should also be present in Edriodiscus and Walcottidiscus (although these are too poorly

728

PALAEONTOLOGY, VOLUME 28

text-fig. 9. Cladogram of character distribution for the
five genera of Cambrian eleutherozoans; characters 1-13
are given in Table 1 .

known at present): 1, an identical mouth frame composed of five interradial elements, each of which is
formed by fusion of four proximal flooring plates, two from each adjacent ambulacrum; 2, a discrete
hydropore plate that lies adjacent to the mouth frame but is not incorporated into it.

Relationships with other groups

Edrioasterids. The Ordovician family Edrioasteridae currently contains just two genera: Edrioaster
and Edriophus. They have the same overall shape and plating arrangement as Totiglobus and
Walcottidiscus , with ambulacra that extend sub-ambitally, a ring of enlarged ossicles around the
dorsal surface, and an uncalcified zone at the centre of the dorsal surface. Their mouth frame consists

table I . Distribution of characters in Cambrian genera of edrioasteroids. Camptostroma is taken as being

primitive for outgroup comparison.

Derived character state

Stromatocystiti

Cambraster

Edriodiscus

Totiglobus

Walcottidiscus

1 . Flat, tesselate dorsal surface

X

X

X

X

X

2. Dorsal plating one layer thick

X

X

X

X

X

3. Marginal ring of plates present

X

X

X

X

X

4. Mouth frame composed of five fused ambulacral
flooring plates, interradially positioned

X

7

X

?

5. Hydropore plate excluded from oral frame

—

X

7

X

?

6. Interambulacral plates tesselate; no epispires

—

—

7

X

X

7. Dorsal surface acting as suction pad

—

—

—

X

X

8. Dorsal growth reduced relative to ventral growth

—

-

-

X

X

9. Ambulacral tips free of marginal ring

—

-

7

X

X

10. Central region of dorsal surface uncalcified

—

-

—

—

X

1 1 . Marginal ossicles stout, forming a prominent ring

—

X

X

—

—

12. Peripheral skirt extending beyond marginal ossicles

—

X

X

—

—

13. Marginal ossicles inserted into dorsal pavement

—

—

X

—

—

SMITH: CAMBRIAN EDRIOASTEROIDS

729

of five interradial elements, each formed through the fusion of proximal flooring plates, but they are
unique in also having five radially positioned elements which underlie the flooring plates (see
Bell 1976), a feature which can be taken as a synapomorphy for the two genera. The interambulacral
zones are identical to those of Totig/obus and Walcottidiscus', ambulacral zones are similar but
possess enlarged primary cover plates and greatly reduced secondary cover plates. Bell and Sprinkle
(1978) suggested that Totiglobus might be ancestral to edrioasterids. However, Walcottidiscus is
taken as the primitive sister group here rather than Totiglobus since Walcottidiscus and edrioasterids
share two advanced characters: 1, ambulacra that curve around the ambitus; 2, a central uncalcified
zone to the dorsal surface.

Isorophids. The majority of isorophid edrioasteroids have been reviewed recently by Bell (1976) in
admirable detail and the group is therefore relatively well understood. Isorophids share the following
four advanced characters that distinguish them from other edrioasteroids: 1 , uniserial flooring plates
lacking sutural passageways; 2, an oral frame composed of the first flooring plate in each
ambulacrum; 3, ambulacral cover plates that extend intrathecally and completely conceal the
flooring plates; 4, an uncalcified dorsal surface. These are all autapomorphies of the group.
Bell (1976) divided isorophids into two suborders and four families, carefully listing the diagnostic
features of each. However, he did not distinguish between symplesiomorphic and synapomorphic
character states and, in the light of what is now known about the Cambrian edrioasteroids, it is worth
examining these groupings cladistically. Table 2 gives an analysis of the characters used by Bell (1976)
and identifies those that are apomorphic and can be used in constructing a cladogram (text-fig. 10).
Using Stromatocystites for outgroup comparison, as primitive, it is apparent that Bell’s group
Isorophina is based largely on symplesiomorphic characters and that the only possible synapo-
morphy that they share is the presence of four enlarged primordial cover plates, at least primitively.
Of the two families included within the Isorophina, the Agelacrinitidae form a monophyletic group
with the following autapomorphies: 1 , hydropore structure formed of a large number of small plates,
without a clearly differentiated hydropore plate; 2, ambulacral cover plates arranged in distinct
cycles. The genus Lispidecodus , which Kesling (1967) considered to be sufficiently distinct to merit
its own family, also belongs here in my opinion. The other family. Bell’s Isorophidae, is paraphyletic
as it lacks any unique synapomorphy.

Members of the second suborder, Lebetodiscina, share two good synapomorphies: 1 , the loss of all
secondary cover plates in all but the oral area; 2, having sutural passageways between adjacent cover
plates. The great majority also possess enlarged primordial cover plates, especially in the C/D
interray. In the family Carneyellidae the oral area is dominated by three large primordial cover plates;
both the secondary cover plates and the shared oral cover plates have been lost from the oral area.
The dominance of the primordial cover plates in the oral region and the extremely simple
arrangement of ambulacral cover plates are features typical of juvenile lebetodiscinids and suggest
that this family may have evolved through heterochrony. The Lebetodiscinae have a less well-
differentiated peripheral rim than any other isorophid family, and this may be a shared derived
character for the group.

Isorophids are usually considered to have evolved from edrioasterids (Bather 1915; Bockelie and
Paul 1983), although Bell and Sprinkle (1978) thought that they were independently derived from
some unknown Cambrian group. Some primitive isorophids such as Savagella retain a stout ring of
marginal ossicles that are triangular in cross-section and very reminiscent of those in Cambraster and
Edriodiscus. Isorophids also have a peripheral skirt of plates lying outside the marginal ring of
ossicles, as found in Edriodiscus. These apomorphic features suggest that Cambraster , Edriodiscus ,
and isorophid edrioasteroids belong to the same clade. Jell et al. (1985) have recently reported an
isorophid edrioasteroid from the Upper Cambrian of Australia, but this is too poorly preserved to
add much to our understanding of their early evolution.

Cyclocystoids. This group has been revised and reinterpreted recently by Smith and Paul (1982).
They chose to maintain cyclocystoids as a separate class but suggested that they were closely
related to both Cambraster and isorophid edrioasteroids. Cyclocystoids are a rather peculiar and

730

PALAEONTOLOGY, VOLUME 28

table 2. Analysis of character distribution within Bell’s (1976) Order Isorophida. Synapomorphic characters
1-12 (see text-fig. 10) are in bold. Outgroup comparison has been made with Cambrian edrioasteroids.

Suborder lebetodiscina Bell, 1976

1. Anal structure a periproct

2. Three oral primary cover plates enlarged or none

4. Primary cover plates only

5. Cover plate sutural passageways

Suborder isorophina Bell, 1976

Anal structure a valvular cone

3. Four oral primary cover plates enlarged or none
Primary and secondary cover plates
No cover plate sutural passageways

Suborder isorophina
Family isorophidae Bell, 1976
Theca domal

Four primary oral cover plates

Hydropore incorporated into central oral rise 6.

Cover plates alternately large and small 7.

Interambulacra squamose and imbricate

Bell, 1976

Family agelacrinitidae Chapman, 1860

Theca domal or clavate

Four primary oral cover plates or none

Hydropore structure isolated or semi-integrated;

composed of many plates
Cover plates arranged in multiple cycles

Interambulacra squamose, imbricate or tesselate

Suborder lebetodiscina Bell, 1976

Family lebetodiscidae Bell, 1976

Theca domal, discoidal or clavate
Lateral shared cover plates present
Secondary oral cover plates present

10. Hydropore plate adjacent to ambulacrum

Ambulacra high pronounced ridges
Cover plate passageways near vertical

11. Peripheral plates squamose, marginals poorly

differentiated

Cover plates hardly differentiated in oral area

Family carneyellidae Bell, 1976
Theca domal

8. Lateral shared cover plates absent

9. Secondary shared cover plates absent

Hydropore plate forming part of oral area
Ambulacra low and rounded
Cover plate passageways oblique
Peripheral plates geniculate with differentiated
marginals

12. Three very large primary cover plates in oral area

Isorophina L«b®todlsc!na

text-fig. 10. Cladogram of character
distribution for the major groups within
Bell’s (1976) group Isorophida. Characters
1-12 are given in Table 2.

SMITH: CAMBRIAN EDRIOASTEROIDS

731

poorly understood group of echinoderms that have all been placed within the single family
Cyclocystoididae. They share a number of unique characters which distinguish them from other
echinoderms, including the possession of perforate marginal ossicles that have distal cupules roofed
over by a movable peripheral skirt of plates, a dorsal plated surface composed of annular plates with
a polygonal outline, and the absence of a ventral periproct. However, they share with Cambraster ,
Edriodiscus , and isorophids such as Savagella the ring of stout marginal ossicles and the peripheral
skirt of plates; they share uniserial ambulacral flooring plates with isorophids alone. This suggests
that cyclocystoids and isorophids are sister groups.

Cyathocystids. Cyathocystis and Cyathotheca are two very distinctive genera that were placed in
their own family Cyathocystidae by Bather (1899). Although Bell (1980) suggested that Timeischytes
and Hadrochthus should also be included in this family, Bockelie and Paul (1983) convincingly
demonstrated that they did not belong here and were in fact isorophids. Bockelie and Paul (1983)
identified a number of autapomorphies for this group: 1, the dorsal surface is a single calcite
element which is cup-shaped and moulded to the substratum; 2, the five primordial cover plates
remain in contact throughout growth; 3, a single large deltoid plate occupies each interambulacral
zone; 4, there are no ambulacral flooring plates. They therefore suggested that cyathocystids re-
presented a line of descent from Stromatocystites independent of the main edrioasterid-isorophid
lineage.

Not all of their supposed synapomorphies are unique to this group, however, since the five
primordial cover plates also remain in contact throughout growth in pyrgocystids (text-fig. 1 1 ) and
pyrgocystids also possess only a single interradial plate in each interambulacrum. Furthermore, the
fact that these deltoid plates surround the peristome, abut along the distal edge at the perradial
suture, and carry articulating cover plates strongly indicates that these are homologous not with
interambulacral plates but with the fused ambulacral mouth frame plates present in Cambraster ,
Totiglobus , and edrioasterids. Bockelie and Paul (1983) were therefore mistaken in suggesting that
cyathocystids lacked flooring plates; rather they have lost all but the most proximal fused ambulacral
flooring plates that form the mouth frame elements. Cyathocystids are so highly modified that their
precise phylogenetic position is difficult to ascertain. However, the presence of oral frame elements
composed of fused ambulacral flooring plates is a synapomorphy shared with both edrioasterids and
the Cambraster- isorophid clade. In the cladogram they are therefore placed in a trichotomy with
these two groups (text-fig. 12).

Pyrgocystids. Pyrgocystids are fully plated and turret-shaped edrioasteroids with an elongate stalk
composed of imbricate dorsal plates and a small ventral surface surmounting the stalk (text-fig. 1 1 ).
The marginal ossicles lie at the top of the turret and surround the ventral surface. On each marginal
plate there is a pair of internal processes similar to those in Totiglobus (text-fig. 1 1). These were
presumably attachment sites for the dorsal muscles responsible for contraction of the turret. In some
there is a basal sac of minute platelets embedded in a coriaceous membrane. Unfortunately very little
is known about the plating of the ventral surface. A single large interradial plate lies in each
interambulacrum except posteriorly, where a small number of smaller plates are found (see Holloway
and Jell 1984). Pyrgocystids have tall, narrow cover plates very much like those of cyathocystids;
the primordial cover plates meet above the oral area (text-fig. 1 1 b), also like cyathocystids. If, as in
cyathocystids, the large interradial ‘deltoid plates’ turn out to be mouth frame plates formed through
fusion of proximal flooring plates, as is strongly suspected, then cyathocystids and pyrgocystids are
best considered to be sister taxa. The perradial margin of the deltoid plates is scalloped exactly as in
cyathocystids and no flooring plates have yet been seen in partially disarticulated specimens. Both
groups have the same turret-shaped theca and almost identical cover plate arrangements. The
principal difference is that in cyathocystids dorsal plating is formed of a single calcite element,
whereas in pyrgocystids it is composed of many plates. Possibly dorsal plating in cyathocystids has
become fused.

Text-fig. 12 provides a character matrix for the various groups of edrioasteroid discussed above
and a summary of their relationships derived from cladistic analysis of this matrix.

732

PALAEONTOLOGY, VOLUME 28

cover plates

deltoid plate

marginal ossicle

muscle attachment process
(internal)

primordial cover
plates

deltoid

plate

basal coriaceous sac

text-fig. II. Rhenopyrgus grayae (Bather), BM(NH) E23470, Ashgill, Upper Ordovician,
Girvan, Scotland; camera lncida drawings of the holotype to show the basic morphological
features of pyrgocystids: a, lateral; b, oral surface.

SMITH: CAMBRIAN EDRIOASTEROIDS

733

CLASSIFICATION

Edrioasteroids have always been considered a natural grouping and, since 1899, have been assigned
the status of class (Bather 1 899). However, current classification schemes of this group, such as those
of Bell (1980), do not reflect the hierarchical groupings identified in the preceding section and
summarized in text-fig. 1 2. It is therefore necessary to rationalize the classification as a whole in order
to place the Cambrian genera into their appropriate monophyletic groups. If the characters used to
define edrioasteroids are examined critically, it is difficult to identify any advanced characters that are
not also present in other primitive eleutherozoan groups. Instead, edrioasteroids are characterized by
having retained primitive features of crown-group echinoderms, such as the ventral periproct,
unmodified ambulacral cover plates, an oral mouth frame that is fixed, and an attached mode of life
as adults; but they lack the pelmatozoan synapomorphies of a dorsal stem and an exothecal
ambulacral subvective system. Having interradial mouth frame elements formed by fusion of
proximal ambulacral flooring plates is an important shared derived character that unites Totiglobus,
Cambraster, edrioasterids, cyathocystids, and probably pyrgocystids but which is not found in
Stromatocystites or primitive asteroids and ophiuroids. This implies that all edrioasteroids, with
the exception of Stromatocystites , form a natural clade which can be considered as a plesion within
the stem group of the Eleutherozoa (see Smith 1 984/?). As this plesion corresponds more or less to the
class Edrioasteroidea Bather as currently accepted (though with the removal of Stromatocystites and
the inclusion of cyclocystoids), this seems the obvious name and taxonomic rank to maintain,
respecting historical tradition. Stromatocystites is then the primitive sister group to all other known
eleutherozoan echinoderms and is best classified as a separate and distinct plesion.

Within the plesion Edrioasteroidea there are three subgroups forming a trichotomy in the
cladogram (text-fig. 1 2). At present, there are insufficient morphological data to solve this trichotomy
convincingly and the three groups identified are best assigned equal rank, even though two of them
contain just four genera each whereas the third contains some thiry-five genera and is considerably
more diverse. This last group includes Cambraster , Bell’s Isorophida, and cyclocystoids. Cyclo-
cystoids have in recent years been separated off at class level, even though the few genera known all
belong to a single family, the Cyclocystoididae. Similarly, isorophids, as defined by Bell (1976),
display an extremely limited range of morphological variation and are a very conservative group. Bell
(1976) elevated this group to the rank of order, recognizing two suborders each with two families.
However, in comparison with taxonomic assignment in other echinoderm groups. Bell’s order
Isorophida is more comparable to a family, and the various subdivisions to subfamilies and tribes. It
might be more sensible therefore to reduce Bell’s Isorophida to a lower rank such as family, where
the name Agelacrinitidae Chapman 1860 would have priority. The two suborders erected by Bell
could then be transformed to the rank of subfamily; Lebetodiscinae (for the Lebetodiscina) and
Isorophinae (for the Isorophina). This would make cyclocystoids and the Agelacrinitidae sister
groups. Although Cambraster , cyclocystoids, and agelacrinitids differ in a number of striking
features, they appear to form a monophyletic group and should therefore be classified together.
Taxonomic debasement can be avoided by uniting them within one order and the most obvious
available name for this group is the order Isorophida. This then requires the expansion of Bell’s
original diagnosis for the Isorophida to include the cyclocystoids and Cambraster.

If this clade is assigned the rank of order, then the two other monophyletic sister groups should also
be ranked as orders. There is already an order Edrioasterida for the genera Edrioaster , Edriophus , and
Totiglobus ; it is only necessary to include Walcottidiscus here and expand the diagnosis for the family
Edrioasteridae in consequence. An order Cyathocystida is also currently available, although at
present it contains only the single family Cyathocystidae. If the pyrgocystids are correctly interpreted
as the sister group to the cyathocystids then the Cyathocystida needs to be redefined to contain two
families, the Pyrgocystidae and the Cyathocystidae.

A phylogenetic classification of the plesion Edrioasteroidea, using the conventions recommended
by Wiley ( 1 979), is presented in Table 3, together with a more traditional scheme for comparison. The
precise taxonomic levels that have been chosen are, and always will be, open to dispute, but they are

734

PALAEONTOLOGY, VOLUME 28

table 3. Classification of the Edrioasteroidea.

Traditional classification ( based on Bel 1 1980 , but with later
additions)

Order stromatocystitoida Termier and Termier, 1969
Family stroma tocystitidae Bassler, 1936
Family cambrasteridae Termier and Termier, 1969
Order edrioasteroida Bell, 1976
Family edrioasteridae Bather, 1898
Family totiglobidae Bell and Sprinkle, 1978
Order isorophida Bell, 1976
Suborder lebetodiscina Bell, 1976
Family lebetodiscidae Bell, 1976
Family carneyellidae Bell, 1976
Suborder isorophina Bell, 1976
Family hemicystitidae Bassler, 1936 ( = isorophidae
’ Bell, 1976)

Family agelacrinitidae Chapman, 1860
Suborder uncertain
Family pyrgocystidae Kesling, 1967
Family lispidecodidae Kesling, 1967
Family rhenocystidae Holloway and Jell, 1984
Order cyathocystida Bockelie and Paul, 1983
Family cyathocystidae Bather, 1899

Revised classification (this paper)

Genus stroma tocystites Pompeckj, 1896
Plesion (Class) edrioasteroidea Billings, 1858
Order edrioasterida Bell, 1976 (sedis mutabilis)

Family totiglobidae Bell and Sprinkle, 1978
Family edrioasteridae Bather, 1898
Order isorophida Bell, 1976 (emend.) (sedis mutabilis)
Genus cambraster Cabibel, Termier and Termier, 1 958
Genus edriodiscus Smith, 1985
Family cyclocystoididae Miller, 1882
Family agelacrinitidae Chapman, 1860
Subfamily isorophinae Bell, 1976
Subfamily lebetodiscinae Bell, 1976
Order cyathocystida Bockelie and Paul, 1983 (emend.)
(sedis mutabilis)

Family pyrgocystidae Kesling, 1967
Family cyathocystidae Bather, 1898

the least important part of the classification scheme. It is the hierarchical and listing orders that
convey information and on which the classification should be judged.

Emended taxonomic definitions

Plesion (Class) edrioasteroidea Billings, 1 858. Stem group eleutherozoans with a fixed mouth frame
of ambulacral flooring plates (primitively five interradial elements). Retained plesiotnorphic
characters include the sessile mode of life as adults, the periproct opening ventrally in the posterior
interambulacrum, and a largely unmodified arrangement of cover plates protecting the ambulacral
groove.

Order edrioasterida Bell, 1976. Sub-globular edrioasteroids with biserial flooring plates, a dorsal
surface that is reduced relative to the ventral surface, and ambulacra that extend sub-ambitally.

text-fig. 12. Character distribution and the derived cladogram for principal edrioasteroid groups.
Characters 1-22 are as follows: 1, oral frame composed of fused ambulacral flooring plates
positioned interradially, transformed to oral frame of radially positioned flooring plates; 2, loss of
epispires from interambulacral zones; 3, dorsal growth retarded relative to ventral growth;
4, ambulacral tips free of the marginal ring; 5, central part of dorsal surface uncalcified; 6, ambu-
lacra curve around ambitus; 7, dorsal surface a tiny, plated suction pad; 8, five radial elements in
oral frame in addition to the fused flooring plates; 9, secondary cover plates greatly reduced or lost;
10, anal structure a periproct; 11, all five primordial cover plates meet centrally; 12, single large
deltoid ( = fused ambulacral flooring plates?) occupies interambulacral zones; 13, dorsal surface
expanded into a stalk; 14, dorsal surface a single (?fused) calcite element; 15, marginal ossicles stout,
forming a distinct frame to the ventral surface; marginal ossicles as imbricate ring (reversed
character state); 16, marginal ossicles inserted into the dorsal pavement to form a clearly
demarcated peripheral skirt; 17, flooring plates uniserial; 18, loss of sutural pores for internal
ampullae; 19, perforate marginal ossicles; 20, dorsal plates annular with central perforation; 21,
peripheral skirt modified into a protective canopy for the cupule zone; 22, cover plates extend
intrathecally at the interradial suture to conceal the flooring plates externally.

SMITH: CAMBRIAN E D R I O A STE RO I DS

735

Plesion (Class) Edrioasteroidea

Order Cyathocystida
Order Edrioasterida

Order Bsorophida

Family Edrioasteridae

Family Agelacrinifidae

©

©

■o

05

V.

©

©

05

3

©

>»

©

©

4m

o

Q.

O

V

05

©

o

s

4*1

o

o

«Q

©

O)

4*0

O

O

£

%

>

o

©

O)

©

©

V.

•O

E

©

O

s

*©

©

>.

>t

©

■©

UJ

Ul

o

Q.

o

UJ

©

m

2

2

©

©

>»

©

©

o

3*

O

©

©

©

©

<0

I e

£ £

© ©

© ©

©

736

PALAEONTOLOGY, VOLUME 28

Order cyathocystida Bockelie and Paul, 1983. Turret-shaped edrioasteroids with a greatly
expanded dorsal surface, a single large deltoid plate in each interambulacrum (fused ambulacral
flooring plate?), and primordial ambulacral cover plates that meet above the peristome. The dorsal
stalk may be either multiplated or composed of a single (?fused) calcite element.

Order isorophida Bell, 1976. Discoidal to clavate edrioasteroids with a distinct peripheral rim of
plates extending beyond the ring of marginal plates.

Family agelacrinitidae Chapman, 1860. Isorophids with uniserial flooring plates and generally
unbranched arms, cover plates with adradial intrathecal extensions that conceal the flooring plates
externally, and an uncalcified dorsal surface within the peripheral skirt.

Subfamily lebetodiscinae (= Suborder lebetodiscina Bell, 1976). Agelacrinitids with sutural
passages between adjacent cover plates.

Subfamily isorophinae ( = Suborder isorophina Bell, 1976). Agelacrinitids with primary ambulacral
cover plates in two or more sizes, arranged cyclically, and primitively with four enlarged primordial
cover plates.

SYSTEMATIC PALAEONTOLOGY

Repositories of specimens referred to below are abbreviated as follows: BM(NH), British Museum (Natural
History), London; CPC, Commonwealth Palaeontological Collections, Bureau of Mineral Resources,
Canberra, Australia; MCZ, Museum of Comparative Zoology, Harvard University, Cambridge, Massachu-
setts; NYSM, New York State Museum, Albany; USNM, United States National Museum, Smithsonian
Institution, Washington D.C.

Genus strom atocystites Pompeckj, 1896

1896 Stromatocystites Pompeckj, p. 506.

1899 Stromatocystites Pompeckj; Jaekel, p. 42.

1900 Stromatocystites Pompeckj; Bather, p. 206.

1911 Chilocystis Perner Mss; Zelizko, p. 6.

1918 Stromatocystites Pompeckj; Jaekel, p. 1 12.

1919 Stromatocystites Pompeckj; Schuchert, p. 1.

1936 Stromatocystites Pompeckj; Bassler, p. 3.

1966 Stromatocystites Pompeckj; Regnell [purs], p. U133.

1969 Stromatocystites Pompeckj; Termier and Termier, p. 133.
non 1905 Stromatocystites Miquel, p. 476 [= Cambraster],

Diagnosis. Theca flattened, discoidal, or weakly pentagonal in outline and fully plated; ambulacra
straight, reaching edge of ventral surface; cover plates arranged as multiserial sheets; interambulacral
plates with sutural epispires; dorsal plating polygonal, tesselate, with large centro-dorsal.

Type species. Stromatocystites pentangularis Pompeckj, 1896, by original monotypy.
Other species. S. walcotti Schuchert, 1919.

EXPLANATION OF PLATE 87

Pigs. 1-6. Stromatocystites pentangularis Pompeckj, 1896. 1, 2, and 5, BM(NH) E16004, Middle Cambrian,
Jince, Czechoslovakia: 1 , general view showing three individuals, two of which are dorsal surface uppermost,
x 1; 2, one individual, ventral surface uppermost, x 2; 5, same individual, dorsal surface, x 2. 3 and 4,
BM(NH) E29830, Lower Cambrian, Bonne Bay, Newfoundland: 3, oral, x 2; 4, aboral, x 2. 6, specimen in
the collection of Dr Krantz of Bonn, latex BM(NH) E631 38, Middle Cambrian, Pod Trnim, Czechoslovakia,
dorsal surface, x 2.

All photographs are of latex casts whitened with ammonium chloride sublimate.

PLATE 87

SMITH, Stromatocystites

738

PALAEONTOLOGY, VOLUME 28

Stratigraphical age and distribution. Uppermost Lower Cambrian ( Olenellus Beds) to Middle Cambrian
( Paradoxides gracilis Zone) from Jince, Czechoslovakia, Bonne Bay, Newfoundland, and Rostock, northern
Germany (in glacial drift, presumably derived from Scandinavia).

Remarks. Stromatocystites was first described by Pompeckj in 1896 and his figures have been copied
by all later workers. S. pentangularis and S. walcotti are each known from a large number of
specimens and are redescribed below. A third species, S. balticus Jaekel, 1899, is represented by only
two specimens whose whereabouts are no longer known and it is treated as a nomen dubium.

Stromatocystites is distinguished from other Cambrian edrioasteroids by having dorsal and ventral
surfaces similar in size, sutural epispires in interambulacral zones, straight narrow ambulacra with
multiserial cover plate sheets, poorly developed marginal ring plates, and simple unfused oral frame
plates. Cambraster likewise has interambulacral epispires but differs in having compound oral frame
plates and a stout marginal ring of ossicles. Walcottidiscus and Totiglobus both lack epispires and
have a reduced dorsal surface.

Stromatocystites is stratigraphically the oldest known genus of eleutherozoan echinoderm.

Stromatocystites pentangularis Pompeckj, 1896

Plate 87; Plate 89, fig. 1; text-figs. 5, 6b, c, 13-15

1896 Stromatocystites pentangularis Pompeckj, p. 506, pi. 13, figs. 1-6.

1899 Stromatocystites pentangularis Pompeckj; Jaekel, p. 42, figs. 5-8.

1900 Stromatocystites pentangularis Pompeckj; Bather, p. 206, fig. 1.

191 1 Chilocystis bohemica Perner Mss, in Zelizko, p. 6.

1919 Stromatocystites pentangularis Pompeckj; Schuchert, p. 2, fig. Ie.

1936 Stromatocystites pentangularis Pompeckj; Bassler, p. 3, pi. 1, figs. 6 and 7.

1958 Stromatocystites pentangularis Pompeckj; Cabibel, Termier and Termier, p. 283, fig. 2.

1966 Stromatocystites pentangularis Pompeckj; Regnell, p. LI 160, text-fig. 126.

1969 Stromatocystites pentangularis Pompeckj; Termier and Termier, p. 133, figs. 1-3.

Diagnosis. A species of Stromatocystites with a domed ventral surface and pentagonal outline;
marginal ossicles undifferentiated; ambulacra narrow, without V-shaped intrathecal extensions;
centro-dorsal plate surrounded by six or seven plates; oral frame large, pentagonal.

Types. Pompeckj (1896) based his description on forty examples from ‘Pod Trnim bei Tejrovic’ of which six were
illustrated. These were stated to belong to the Geologischen Reichsanstalt, Vienna, but are now housed in the
Czechoslovakian Geological Survey, Prague. No holotype was designated by Pompeckj, thus all are syntypes.
The holotype of Chilocystis bohemica is P/77916, Narodni Muzeum, Prague; a cast, BM(NH) E16008, of this
specimen was studied (PL 89, fig. 1).

Material studied. The following description is based largely on eighteen well-preserved individuals, BM(NH)
El 5897, El 5898, El 6004, and E29380, and on the two individuals, USNM 56665. The collections in the Narodni
Muzeum, Prague, have also been studied, as have three latex casts of specimens in the private collection of
Dr Krantz of Bonn, kindly given to me by Professor Ubaghs (BM(NH) E63 1 37-63 139).

Stratigraphical age and distribution. BM(NH) E29380 (four individuals) comes from the Olenellus Beds,
Taconian, high Lower Cambrian of Bonne Bay, Newfoundland. Other material comes from the Paradoxides
gracilis Zone, Upper Jince Beds, Middle Cambrian of Tyrovic and Jince, Czechoslovakia.

Description. All specimens examined are preserved as natural moulds of external surfaces. The theca is
pentagonal in outline with the ambulacra running to the five rounded corners. In a number of specimens the
interambulacral portions are gently concave suggesting that there may have been internal radially arranged
postural muscles present. The entire theca is plated. On the ventral surface the ambulacra form narrow ridges
which extend to the margin. These meet centrally in a 2 : 1 : 2 pattern with ambulacra B + C and D + E paired and
united on either side of the mid-line. Ambulacrum A is unpaired and set almost at right angles to the paired
lateral ambulacra (text-fig. 1 5). The ambulacra form prominent ridges on the theca while interambulacral zones
are generally depressed. The ambulacra are straight and composed of biserial flooring plates with left and right
cover plate series. Flooring plates are relatively short (in the perradial/interradial direction) and broad (in the

SMITH: CAMBRIAN EDRIOASTEROIDS

739

text-fig. 13. Stromatocystites pentangularis
Pompeckj, 1896, BM(NH) E16008, camera
lucida drawing, showing the ventral plating in
one interambulacrum.

flooring plate

primary
cover plate

2 mm

adoral/adambital direction) (text-fig. 6b, c). The perradial edge of each plate is weakly angled and the flooring
plates alternate to produce a zigzag perradial suture. Interradially a small part of the flooring plate is exposed
between the interambulacral plates and the cover plate sheets. On the adoral/adambital suture, immediately
perradial to the site where the cover plates attach, is a small ovoid pore, shared equally between the two flooring
plates. The flooring plates appear to be steeply inclined so as to form a deep and narrow ambulacral tunnel.

Cover plates are arranged as two multiserial sheets that attach towards the adradial edge of the flooring plates,
one on either side of the ambulacrum. A large pentagonal primary cover plate rests directly on each flooring
plate. Between these, more or less directly above the flooring plate sutures, are smaller diamond-shaped cover
plates. A further two or so irregular rows of small cover plates lie above the primary and diamond-shaped cover
plates (text-fig. 14). The two cover plate sheets meet along the mid-line and, in well-preserved material, form
a crenulated crest to the ambulacral ridge. In the oral area the arrangement of cover plates appears identical
with that of the more distal ambulacral areas and there are no distinguishably larger plates. The flooring
plates and cover plate series are more or less identical to those in helicoplacoids (see Paul and Smith 1984).

In almost all specimens compaction has pressed dorsal and ventral surfaces together so that narrow ridges
mark the position of the underlying flooring plates on the dorsal surface. The five ridges do not meet in a 2 : 1 : 2
pattern, as might be expected, but bifurcate near the centre and unite interradially to form a pentastellate ring
(PI. 87, figs. 4-6). This suggests that adorally the individual columns of flooring plates could have separated
along the perradial suture and that the most proximal flooring plates were united interradially to form a large
and flexible peristomial opening, as in primitive asteroids. Unfortunately, no specimens examined showed the
internal face of the ventral surface and the precise arrangemen t of flooring plates in the oral region must remain
speculative.

Interambulacral zones on the ventral surface are composed of small polygonal and tesselate plates of variable
size and shape (text-fig. 13). There are up to fifteen plates abreast in an interambulacral zone. At the sutures
between plates there are small oval or slit-like epispires with neither internal nor external rims. Epispires are
absent from a narrow zone bordering the ambulacra and immediately adjacent to the marginal plates. The
periproct lies more or less centrally in the C/D interray. It is composed of about twelve to fifteen lath-shaped

740

PALAEONTOLOGY, VOLUME 28

1 mm

Y t

■cover plates

text-fig. 14. Stromatocystites pentangularis
Pompeckj, 1896, camera lucida drawings of
cover plate arrangements: a, USNM 56665b;
b, BM(NH) El 6004.

plates arranged radially. These plates and those immediately adjacent lack epispires. They generally lie more or
less flush with the surrounding interambulacral plates.

No third aperture has been identified amongst the plates of the oral area, probably because no specimen is
sufficiently well preserved in this region to reveal such a structure. The dorsal surface is usually concave and
pressed against the ventral surface, although in life it was presumably more or less flat. All specimens show
a sharp marginal rim suggesting that the large marginal plates formed a more rigid frame, as in S. walcotti.
Immediately inside this rim there is a very narrow zone of tiny plates. The remainder of the dorsal surface is
covered in polygonal, tesselate plates with a pitted surface texture. These plates are more or less irregularly
arranged (text-fig. 5b), but a large equant plate at the centre of the disc is surrounded by six or seven other large
plates. There is no evidence for any attachment structure by which adult Stromatocystites might have attached
itself to a hard substratum.

Remarks. S. pentangularis is not uncommon at certain horizons and clusters of individuals are
preserved together on bedding surfaces. As individuals showing both dorsal and ventral surfaces are
found together on the same slab, they probably represent individuals that have been current
transported and buried alive. The original description by Pompeckj (1896) was fairly good and, for
the most part, later workers have simply repeated his findings without adding any new information.

Stromatocystites walcotti Schuchert, 1919
Plate 88; text-figs. 4, 7, 16

1919 Stromatocystites walcotti Schuchert, p. 3, pi. 7, figs. 1-3; text-fig. 1.

1966 Stromatocystites walcotti Schuchert; Regnell, p. U160, text-fig. 126.

1969 Stromatocystites walcotti Schuchert; Termier and Termier, p. 137, pi. 8, figs. 1-4.

1984 Stromatocystites walcotti Schuchert; Paul and Smith, p. 452, text-figs. 6 and 7.

Diagnosis. A species of Stromatocystites discoidal in shape and with a pentagonal outline; marginal
ring plates present and visible externally; ambulacral flooring plates with V-shaped intrathecal
extensions; oral frame not large and pentastellate; centro-dorsal plate surrounded by a circle of ten or
eleven plates.

Types. Holotype, USNM 66443 (PI. 88, fig. 1); paratypes, USNM 66444.

Material examined. In addition to the holotype and paratypes, twenty slabs, USNM 376690 and 384978-384984,
with a total of thirty-seven individuals have been examined.

Stratigrapliical age and distribution. Olenellus Beds, Taconian, upper Lower Cambrian; all specimens come from
the eastern arm of Bonne Bay, western coast of Newfoundland.

SMITH: CAMBRIAN EDRIOASTEROIDS

741

A

text-fig. 15. Stromatocystites pentangularis Pompeckj, 1896, camera lucida drawings of
plating in the oral area: a, USNM 56665b; b, BM(NH) E16004.

742

PALAEONTOLOGY, VOLUME 28

text-fig. 16. Stromatocystites walcotti Schuchert, 1919,
USNM 384978, camera lucida drawing of the dorsal plated
surface of a juvenile showing the prominent marginal ring
plates.

Description. Specimens are 9-22 mm in diameter, weakly pentagonal in outline, and have a flattened profile
without the ventral convexity characteristic of S. pentangularis. Ambulacra are long and straight and form ridges
on the ventral surface. They extend to the ambitus and the tip of each ambulacrum is inserted between the large
plates forming the marginal ring. Flooring plates are short and broad and are arranged alternately in two
columns with a weakly zigzag perradial suture. Each plate appears to have a pair of short processes that extend
mtrathecally, but is otherwise very like those of S. pentangularis. The arrangement of cover plates is not clearly
seen in any of the specimens, but it is quite obviously a multiplated sheet. The ambulacra meet in a 2 : 1 : 2 pattern
over the oral area and there is no apparent differentiation of the cover plates in this region. The internal aspect of
the ventral surface is seen in USNM 376690 and shows the oral frame. The oral frame is transversely elongate
and made up of the ten most adoral unfused flooring plates, together with one large asymmetrical inter-
ambulacral plate in the C/D interray (the hydropore plate) (text-fig. 7). Unlike S. pentangularis, S. walcotti shows
no evidence of having perradial slits and a flexible mouth frame. There is no obvious hydropore groove in the
mouth frame, such as is found in agelacrinitids, but the large hydropore plate is embayed on the D ray side,
possibly to allow passage for some soft tissue structure.

The interambulacral zones of the ventral surface are composed of irregularly arranged polygonal plates with
epispires. Individual plates are comparatively larger and fewer in number than in S. pentangularis , with up to
eight plates abreast in any one interambulacrum. The periproct opens in the C/D interray close to the ventral
margin and slightly offset towards the C ray. It consists of a large number of elongate plates arranged more or
less radially. The periproct is always flush with the surrounding interambulacral plating.

Marginal plating is more strongly developed than in 5. pentangularis and is sometimes the only part to be
preserved. There is a ring of abutting plates which are thickened along the dorsal facing edge. They are held
nearly perpendicular to the sea floor around the margin of the theca.

The dorsal surface is covered by a pavement of polygonal, tesselate plates, except immediately adjacent to the
marginal ossicles where there are two or three irregular rows of small, loosely fitting plates. At the centre of the
disc there is a large, almost circular centro-dorsal plate surrounded by a ring of ten or eleven large, radially
elongate plates. Elsewhere the dorsal plating is somewhat smaller and totally irregular in its arrangement.

EXPLANATION OF PLATE 88

Figs. 1 -5. Stromatocystites walcotti Schuchert, 1919. 1, USNM 66443, holotype, Lower Cambrian, Bonne Bay,
Newfoundland; dorsal surface (see text-fig. 4), x 2-5. 2, USNM 384978, Lower Cambrian, Bonne Bay,
Newfoundland; dorsal surface of a juvenile showing the prominent marginal ring at this stage, x 5. 3-5,
USNM 376690, no locality data: 3, general view of the interior of the ventral surface, x 1-8; 4, detail of the
periproct and ambulacral flooring plates (internal), x 4; 5, detail of the oral frame, x 5.

All photographs are of latex casts whitened with ammonium chloride sublimate.

PLATE 88

SMITH, Stromatocystites

744

PALAEONTOLOGY, VOLUME 28

Remarks. The original description by Schuchert (1919) is accurate in most respects. Schuchert used
the name S. walcotti for six larger specimens (including the holotype) but placed all the smaller
specimens in a different group which he designated as variety minor. As the two forms are
indistinguishable and simply represent individuals at different growth stages, Schuchert’s variety is
not a biologically useful concept.

S. walcotti is similar to S. pentangularis in many respects, but is less inflated ventrally, has a smaller,
more transversely elongate mouth frame, and has a larger number of dorsal plates in contact with the
centrodorsal. Termier and Termier (1969) believed that they could recognize inhalent orifices at the
ambulacral tips, suggesting that the ambulacral passageways were enclosed and only opened to the
exterior via these terminal orifices. In my opinion such openings did not exist in life and the Termiers
were misled by one specimen in which the very terminal cover plates had been displaced during burial
and compaction in a couple of the ambulacra.

S. walcotti was tentatively assigned to the genus Cambr aster by Jell et al. (1985) because C. tastu-
dorum and S. walcotti have rather similar dorsal plating, and S. walcotti is supposed to have a ring
of marginal ossicles like Cambraster. Although marginal ossicles are present in S. walcotti and are
more prominent than in S. pentangularis , particularly in juveniles, they appear to be much less well
developed than they are in Cambraster. Furthermore, the development of a central zone of large
polygonal plates on the dorsal surface is not restricted to Cambraster and S. walcotti but is also a
feature of 5. pentangularis (see PI. 87). Therefore, although S. walcotti may well be directly ancestral
to C. tastudorum , it lacks crucial features (such as a marginal skirt and stout marginal ring) that
characterize Cambraster and cannot be placed within this genus.

Stromatocystites balticus Jaekel, 1899

Remarks. This species was erected on the basis of two specimens, both of which have since been lost.
Jaekel’s original description is rather vague and the distinction between this species and Stromato-
cystites pentangularis is not at all clear. Regnell (1945) has reviewed all that is known about this
species and I can add nothing else. Since no specimens now exist and the original description is too
generalized to distinguish S. balticus from S. pentangularis , I recommend that S. balticus be treated as
a nomen dubium. The original specimens came from the Middle Cambrian Paradoxides para-
doxissimus Zone of the Baltic and were collected from a glacial erratic block in northern Germany.

Genus walcottidiscus Bassler, 1935

1935 Walcottidiscus Bassler, p. 3.

1936 Walcottidiscus Bassler; Bassler, p. 2.

1943 Walcottidiscus Bassler; Bassler and Moody, p. 209.

1966 Walcottidiscus Bassler; Regnell, p. U161 .

Diagnosis. Theca sub-ovoid, circular to roundedly pentagonal in outline; dorsal surface relatively
small compared to ventral surface, consisting of an outer zone of tiny platelets and a central
uncalcified zone; ambulacra curving sinistrally at the ambitus, extending sub-ambitally, composed of

EXPLANATION OF PLATE 89

Fig. 1. Stromatocystites pentangularis Pompeckj, 1896. BM(NH) E16008, latex cast of P/77916, Narodni
Muzeum, Prague, holotype of Chilocystis bohemica , detail of interambulacral zone, x 6.

Figs. 2-6. Walcottidiscus typicalis Bassler, 1935. Burgess Shale, Middle Cambrian. 2 and 3, USNM 90754,
holotype, part and counterpart (see text-fig. 17), x4. 4-6, USNM 90755, holotype of W. magister Bassler,
1 936; 4 and 6, details of cover plate arrangement in ambulacra (f = flooring plates, external portion; c = cover
plates), x 6; 5, general view, x 1.

All specimens whitened with ammonium chloride sublimate.

PLATE 89

SMITH, Stromatocystites , Walcottidiscus

746 PALAEONTOLOGY, VOLUME 28

biserial flooring plates and multiserial cover plate sheets; interambulacral areas subtesselate, lacking
epispires; periproct subambital.

Type species. Walcottidiscus typicalis Bassler, 1935, by original monotypy.

Stratigraphical age and distribution. Middle Cambrian Burgess Shale, Bathyuriscus-Elrathina Zone, British
Columbia, Canada.

Remarks. Bassler (1935) described and figured a small edrioasteroid from the Burgess Shale under the
name W. typicalis. His description was rather sketchy and, in places, somewhat misleading; the
specimen has not since been redescribed. A year later, Bassler (1936) described a second specimen
from the same locality under the name W. magister. This is a much larger specimen and superficially
seems to be quite distinct. However, the differences between the two specimens are almost certainly
a result of size; I interpret W. typicalis as a juvenile and W. magister as an adult of the same species.

Neither specimen is particularly well preserved and details of ambulacral plating around the oral
area cannot be made out. Walcottidiscus has certain characteristics that distinguish it from all other
Cambrian edrioasteroids but which it shares with the Ordovician edrioasterids Edrioaster and
Edriophus. These include the sub-oval shape, the strongly curved ambulacra that run around the
ambitus, and the reduced dorsal surface with its central uncalcified zone. Walcottidiscus differs from
both Edrioaster and Edriophus in the arrangement of ambulacral cover plates. In Edrioaster and
Edriophus there is a series of large primary cover plates, usually with a series of very much smaller
secondary cover plates inserted distally, whereas in Walcottidiscus cover plates are arranged in
a complex multiserial sheet, as in Stromatocystites.

Walcottidiscus typicalis Bassler, 1935

Plate 89, figs. 2-6; text-fig. 17

1935 Walcottidiscus typicalis Bassler, p. 3, pi. 1, fig. 1.

1936 Walcottidiscus magister Bassler; Bassler, p. 2, pi. 2, fig. 2.

1943 Walcottidiscus typicalis Bassler; Bassler and Moody, p. 210.

1943 Walcottidiscus magister Bassler; Bassler and Moody, p. 210.

1966 Walcottidiscus typicalis Bassler; Regnell, p. U161 .

Diagnosis. As for genus.

Types. Holotype, USNM 90754, part and counterpart (PI. 89, figs. 2 and 3). Holotype of Walcottidiscus magister,
USNM 90755 (PI. 89, figs. 4-6).

Stratigraphical age and distribution. Middle Cambrian Burgess Shale, Bathyuriscus-Elrathina Zone, Burgess
Pass, British Columbia, Canada.

Description. The preservation of the holotype, USNM 90754, is rather difficult to interpret as both part and
counterpart show a mixture of mould and original plating. The more complete part shows the exterior of the
underside and part of the interior of the upper surface. The less complete counterpart shows the external mould
of the underside and the external surface of the upper side (text-fig. 17). The theca is somewhat distorted but
appears to have had a rounded pentagonal outline. The maximum diameter of the theca is 15-5 mm and the
minimum 12-0 mm, so the diameter in life must have been approximately 14-0 mm. The other specimen, USNM
90755, shows only the exterior of the upper surface and is damaged in places. It has an anterior-posterior
diameter of 64 0 mm. Some of the calcite plating is preserved in this specimen. In both specimens the ambulacra
extend subambitally and, in the holotype, the dorsal surface can be shown to be smaller than the thecal diameter.
W. typicalis presumably had a sub-oval to flattened sub-globular shape in life.

Ambulacra are arranged in a 2 : 1 : 2 pattern around the mouth. Over much of the upper surface the ambulacra
are more or less straight, but on approaching the ambitus all five ambulacra curve sinistrally. In the smaller
specimen, ambulacra B, C, and D appear on the lower surface, following a slightly oblique course to the margin.
Adjacent ambulacra do not approach one another closely and their tips almost reach the edge of the dorsal
plating. The ambulacra in the larger specimen are strongly flexed at the ambitus and run around the outer edge of
the theca. The tip of one ambulacrum almost reaches the next ambulacrum to the left (PI. 89, fig. 5).

SMITH: CAMBRIAN EDRIOASTEROIDS

747

?marginal ring

cover plates (external)

platelets
at margin of
dorsal surface

cover plates
flooring plates

external lower surface

text-fig. 17. Walcottidiscus typicalis Bassler, 1935, USNM 90754, camera lucida drawing of
the holotype: a, part; b, counterpart.

The detailed structure of the ambulacral flooring plates cannot be made out with any certainty on either
specimen. Flooring plates are biserially arranged and alternate in each ambulacrum. On the outer surface the
adradial portion of each flooring plate is exposed between the cover plate sheets and the interambulacral plate, as
in Stromatocystites. This portion of the flooring plate is flat and rectangular in outline, and adjacent faces abut to
form a regular border to the arch of cover plates (PI. 89, figs. 4 and 6). The arrangement of cover plates is best

748

PALAEONTOLOGY, VOLUME 28

seen in the larger specimen, but can also be made out towards the tips of ambulacra B and C in the holotype.
Cover plates are arranged as a multiserial sheet of abutting plates; they are largest immediately adjacent to the
flooring plates and progressively decrease in size towards the free edge. Towards the distal tip in the smaller
specimen, the cover plate sheet consists of just two or possibly three irregular rows, but in the larger specimen
there are four or possibly five irregular rows of cover plates in each sheet. The first row of cover plates is well
defined and these are the largest. They are broader than tall and more or less rectangular but with an angled
upper edge making them pentagonal. There are approximately two of these primary cover plates to each flooring
plate. The succeeding cover plates are much more randomly arranged and become progressively more equant in
shape towards the free edge (PI. 89, fig. 4).

The oral area is more or less totally obscured in both specimens and nothing can be seen of the oral frame. The
location of the hydropore is possibly indicated by a small raised circular rim with a central depression seen on
both part and counterpart of the holotype. This lies in the C/D interray, slightly removed from the oral area.

Interambulacral areas are relatively broad and are composed of large sub-polygonal and semi-tesselate plates
on the upper surface, with smaller, more imbricate plates sub-ambitally. The plates imbricate adorally so that the
more distal plates overlap slightly on to their more proximal neighbours. At the ambitus there are three or four
plates abreast in each interambulacrum. There are no epispires.

The holotype exhibits an indistinct area of radially arranged plates in the C/D interray that may be the
periproct; this lies sub-ambitally close to the edge of the ventral surface. The dorsal surface is seen only in the
holotype. Much of the dorsal surface appears to have been uncalcified, since both part and counterpart reveal
the inner face of the upper surface centrally. However, around the edge there is a border, 1-2 mm in
breadth, composed of minute plates (3-4 per mm). The diameter of the dorsal surface is approximately 9 10 mm
and so is only about 70 % of the diameter of the theca. There is a suggestion in the A/B and C/D interambulacra
that the dorsal surface may be ringed by a border of broader, rather rectangular plates.

Remarks. Bassler (1935, 1936) placed these two specimens in different species. The principal
differences between them are in size and degree of ambulacral curvature. Unfortunately, neither is
particularly well preserved and many of the structural details are still unknown. In W. typicalis,
although the ambulacra curve sinistrally at the ambitus, they are not particularly long and are well
separated from each other. In W. magister , which is the larger of the two specimens, the ambulacra
also curve at the ambitus to run sinistrally around the thecal margin, but in this specimen they are
much longer and the tip of one ambulacrum almost touches the neighbouring ambulacrum.
W. typicalis is only a quarter of the size of W. magister , so this difference could simply be a factor of
growth. In all edrioasteroids with curved ambulacra, ontogenetic series show that ambulacra are
initially straight in juveniles and become progressively more curved as growth proceeds. The two
specimens are therefore probably no more than juvenile and adult of the same species and
W. magister is treated as a junior synonym of W. typicalis.

Genus totiglobus Bell and Sprinkle, 1978
1978 Totiglobus Bell and Sprinkle, p. 247.

Diagnosis. Theca sub-globular and fully plated; dorsal surface greatly reduced, less than half thecal
diameter, circular in outline with marginal ring of stout plates surrounding a pavement of small
polygonal plates; ambulacra more or less straight, extending sub-ambitally; flooring plates biserial,
alternate, with sutural pores; cover plates in two series; oral frame composed of five compound
ambulacral flooring plates; interambulacral plates imbricate to sub-tessellate, without epispires;
periproct sub-ambital, close to dorsal surface.

Type species. Totiglobus nimius Bell and Sprinkle, 1978, by original designation.

Stratigraphic at age and distribution. Early Middle Cambrian of Nevada, USA.

Remarks. This is probably the best understood of the Cambrian edrioasteroids because of the careful
and detailed work carried out by Bell and Sprinkle ( 1 978). Totiglobus can be distinguished easily from
Stromatocystites and Cambraster by its interambulacral zones, which are devoid of epispires, and by
its relatively small dorsal surface and swollen ventral surface; it lacks a peripheral skirt of plates,
unlike Edriodiscus. Totiglobus most closely resembles Walcottidiscus , but is distinguished by having

SMITH: CAMBRIAN EDRIOASTEROIDS

749

more or less straight ambulacra at all sizes and a fully plated dorsal surface. Ambulacra in
Walcottidiscus curve distally around the ambitus in larger individuals, and the central part of the
dorsal surface is uncalcified.

Totiglobus nimius Bell and Sprinkle, 1978
1978 Totiglobus nimius Bell and Sprinkle, p. 247, pis. 1-6; text-figs. 1-4.

Diagnosis. As for genus.

Types. Holotype, MCZ 983; paratypes, MCZ 984-996 and NYSM 13293-13326.

Stratigraphical age and distribution. Chisholm Shale, Glossopleura Zone, Middle Cambrian of Nevada, USA.

Description. A full and detailed description of this species, together with full locality data, was given by Bell and
Sprinkle (1978).

Genus cambraster Cabibel, Termier and Termier, 1958

1894 Trochocystites Miquel [non Barrande], p. 9.

1905 Stromatocystites Miquel [non Pompeckj], p. 476.

1923 Cambraster Jaekel, p. 344 [nomen nudum],

1935 Cambraster Jaekel; Stubblefield and Spencer in Thoral, p. 35 [nomen nudum],

1958 Cambraster Jaekel; Cabibel, Termier and Termier, p. 284.

1958 Eikosacystis Cabibel, Termier and Termier, p. 286.

1966 Stromatocystites Pompeckj; Regnell [ pars ], p. U160.

1969 Cambraster (Jaekel) emend. Cabibel, Termier and Termier; Termier and Termier, p. 137.

1969 Eikosacystis Cabibel, Termier and Termier; Termier and Termier, p. 141.

1971 Cambraster Cabibel, Termier and Termier; Ubaghs, p. 182.

Diagnosis. Fully plated, disc shaped edrioasteroid with a pentagonal frame of stout marginal ossicles;
ambulacra straight, inserted distally between the marginal ossicles; dorsal surface plated, slightly
larger than ventral surface, and composed of a sub-tesselate pavement of plates arranged in cycles;
interambulacral zones with epispires.

Type species. Cambraster cannati (Miquel, 1894), by original designation.

Other species. C. tastudorum Jell, Burrett and Banks, 1985.

Stratigraphical age and distribution. Paradoxides mediterraneus Zone, Middle Cambrian of the Montagne Noire,
southern France; medial Middle Cambrian of Tasmania.

Remarks. This important genus remained poorly understood until Ubaghs (1971) published
a detailed morphological description of well-preserved material. The name Cambraster was first used
by Jaekel (1923) for a species that he believed to be a primitive asteroid from the Middle Cambrian
sandy schists of the departement of Herault, France. He interpreted Cambraster as intermediate
between edrioasteroids and the Ordovician asteroid Archegonaster , but failed to describe or figure
this animal, or even to designate a specific name. Prior to this, specimens of Cambraster had been
wrongly identified as the carpoid Trochocystites or as Stromatocystites by Miquel (1894, 1905).
Although later authors have been uncertain as to the identity of Jaekel’s Cambraster , Stubblefield
and Spencer (in Thoral 1 935) suggested that it was the same as Miquel’s ( 1 905) species S. cannati , also
from the Middle Cambrian of Herault. Amongst the papers in the possession of W. K. Spencer at his
death, and now in the archives of the British Museum (Natural History), is an original line-drawing
signed by Jaekel and labelled Cambraster (text-fig. 18), proving that Jaekel was indeed referring to the
same animal as Miquel.

The name Cambraster was finally validated by Cabibel, Termier and Termier (1958) who provided
a generic diagnosis and gave a rather sketchy description of the type species, C. cannati (Miquel); they
also described a new genus Eikosacystis, with two species. These were all redescribed in slightly
greater detail by Termier and Termier (1969) who placed Cambraster and Eikosacystis in their own

750

PALAEONTOLOGY, VOLUME 28

o u- , i„-. ( •

- To/.; ; _ , + ^

' * Coywt, T. t w ? frsCClc d >cl H-v.'T? C- * x .‘. &•/. fofu*. lit* /® fc .

text-fig. 18. Original line drawing by Otto Jaekel of a specimen said to be in the Berlin Museum and signed as
Cambraster gal/icus Jkl, Middle Cambrian, Coulounnais, Herault.

families, the Cambrasteridae and the Eikosacystidae respectively. However, as Ubaghs (1971)
pointed out, Cambraster and Eikosacystis are synonymous and differ only in their state of
preservation. Specimens with the plating of the oral surface preserved more or less intact, and
therefore showing ambulacra, were placed by the Termiers into Cambraster while more distorted
specimens that had either a jumbled muddle of dissociated plates within the marginal ring or had lost
the ventral plating altogether (and so revealed the interior of the dorsal surface) were classified as
Eikosacystis , which was thus believed to lack ambulacra. Ubaghs’s (1971) careful and detailed
observations on Cambraster have greatly clarified the status of this genus while further morpho-
logical information has been added by the discovery of additional material from the Cambrian of
Tasmania by Jell et al. (1985).

Cambraster is easily distinguished from other Cambrian edrioasteroids by its prominent marginal
ring of ossicles which, unlike those of Edriodiscus , are not inserted into the dorsal pavement.

Cambraster cannati (Miquel, 1894)

Text-figs. 2, 3, 6a, 18, 19

1894 Trochocystites cannati Miquel, p. 9.

1905 Trochocystites sp. Miquel, p. 475.

1905 Stromatocystites cannati (Miquel); Miquel, pp. 476, 482, pi. 15, fig. 5.

SMITH: CAMBRIAN ED R I O ASTE ROI DS

751

1935 Stromatocystites(l) Cannati Miquel; Thoral, p. 35.

1935 Trochocystites(J) nov. sp. Miquel; Thoral. p. 138, pi. 5, fig. 3.

1958 Cambraster cannati (Miquel); Cabibel, Termier and Termier, p. 284, pi. 1, figs. 2-4; text-
figs. 3 and 4.

1958 Eikosacystis couloumanensis Cabibel, Termier and Termier, p. 286, pi. 1, fig. 5; text-fig. 5.

1958 Eikosacystis? ferralsensis Cabibel, Termier and Termier, p. 287, pi. 1, figs. 6 and 7.

1969 Cambraster cannati (Miquel); Termier and Termier, p. 138, pi. 9, figs. 1-3; text-fig. 5.

1969 Cambraster elegans Termier and Termier, p. 139, pi. 10, figs. 1-4; text-fig. 6.

1969 Eikosacystis couloumanensis Cabibel, Termier and Termier; Termier and Termier, p. 141, pi. 12,
fig. 1; text-fig. 7.

1969 Eikosacystis miqueli Termier and Termier, p. 142, pi. 11, figs. 1 -4; text-fig. 8.

1969 Eikosacystis courtessolei Termier and Termier, p. 143, pi. 12, fig. 2; text-fig. 9.

1969 Eikosacystis ferratsensis Cabibel, Termier and Termier; Termier and Termier, p. 144, pi. 1 2, figs. 3
and 4; text-fig. 10.

1971 Cambraster elegans Termier and Termier; Ubaghs, p. 1 82, text-figs. 8 11.

Diagnosis. As for genus.

Types. The holotype of Cambraster cannati , C. elegans , and Eikosacystis miqueli are currently held at the
Museum of Le Havre, France. Latex casts (BM(NH) E63135, 63136, and 63153) of these specimens were
studied (text-fig. 19). The whereabouts of E. ferralsensis, E. couloumanensis , and E. courtessolei are currently
unknown. They were in the possession of Mme G. Termier at the Departement de Geotectonique, Universite
Pierre et Marie Curie, Paris, but, on enquiry there, could not be located.

Stratigraphical age and distribution. All (he material comes from Beds E and F of Courtessole (1973) in the
Middle Cambrian of the Montagne Noire, in the vicinity of Coulouma and Ferrals-les-Montagnes, France.

Description. A complete and detailed description of this species was given by Ubaghs (1971), under the
name C. elegans , and I have only two additional observations to make. First, although dorsal plating is
generally disrupted to a greater or lesser degree in all of the specimens examined, plates do have an angular
outline and would have formed a continuous pavement of closely fitting plates in life, as in Stromatocystites.
These plates are arranged in circles around the centre of the disc with al least two rings of larger plates
separated by zones of smaller plates. Secondly, the peripheral zone of dorsal plating does not extend much
beyond the ring of marginal ossicles and only a very narrow rim of dorsal plates would have been visible
from above.

Remarks. Termier and Termier (1969) recognized two species of Cambraster and four of Eikosacystis ,
all of which are treated here as synonymous. Two of their species, C. cannati and C. elegans, are
moderately well preserved and show the plating of the oral surface. These two species were
distinguished by the fact that in C. elegans rather more of the dorsal surface plating is visible around
the margin. This is a post-mortem artifact produced by displacement of some of the marginal ring
ossicles and, as the two specimens are otherwise identical, I regard them as synonymous. Two of
the Termier’s species which show only the marginal ring and dorsal plating, E. miqueli and
E. courtessolei, are indistinguishable from one another. Furthermore, as Ubaghs (1971) pointed out,
E. miqueli and C. elegans are specimens of the same species in different states of preservation, one
showing the marginal ring and aboral surface, the other showing the marginal ring and oral surface.
Specimens referred to as E. ferralsensis are large distorted individuals in which most of the plating
within the marginal ring lies jumbled together or has been lost. Finally, E. couloumanensis is known
from only a single small specimen, 12 mm in diameter. This has a stout marginal ring of equant
ossicles but the plating within the marginal ring is poorly preserved. I have not been able to examine
this specimen but, as it comes from the same beds and the same locality as adult C. cannati, I suspect
that it is a juvenile specimen of this species.

Cambraster tastudorum Jell, Burrett and Banks, 1985

Remarks. This species has only recently been described by Jell et al. (1985) from the Cateena Group,
medial Middle Cambrian, of Tasmania. It differs from Cambraster cannati in having epispires

752

PALAEONTOLOGY, VOLUME 28

text-fig. 19. Cambraster cannati (Miquel, 1894). a, BM(NH) E63135, latex cast of the holotype of C. elegans
Termier and Termier, 1969, ventral surface, x 3 (see text-fig. 2). b, BM(NH) E63153, latex cast of the holotype
of Trochocystites cannati Miquel, 1894, x 1-4. C, D, BM(NH) E63136, latex casts of the holotype of Eikosacystis
miqueli Termier and Termier, 1969 (see text-fig. 3): c, aboral surface, exterior, x 3; d, counterpart showing
the ventral surface of the marginal ring and the interior of the dorsal surface (all ventral disc plating having

been lost), x 3.

SMITH: CAMBRIAN EDRIOASTEROIDS

753

developed only towards the centre of the oral surface and in having a prominent contiguous circle of
large polygonal plates near the centre of the aboral surface.

Genus edriodiscus Jell, Burrett and Banks, 1985

1971 Cyclocystoides Henderson and Shergold [non Salter and Billings], p. 706.

1985 Edriodiscus Jell, Burrett and Banks, p. 190.

Diagnosis. Aboral surface fully plated, composed of small polygonal plates with radial ridging;
marginal ring circular, composed of forty to fifty plates, surrounded by a peripheral skirt; ventral
surface unknown.

Type species. Cyclocystoides primotica Henderson and Shergold, 1971, by original designation.

Stratigraphical age and distribution. Early Middle Cambrian of West Queensland, Australia.

Remarks. This genus is very poorly known and contains just one species represented by two
specimens, neither of which reveals the ventral surface. It is, however, distinct from any other known
echinoderm and clearly deserves generic separation. The species was originally placed in the genus
Cyclocystoides by Henderson and Shergold (1971), but Smith and Paul (1982) pointed out that it
lacked important cyclocystoid features such as perforate marginal ossicles, and removed it from that
group.

Edriodiscus , with its stout marginal ring of abutting ossicles, most closely resembles Cambraster.
However, there are only fifteen to twenty marginal ossicles in Cambraster , and these are hidden from
view on the lower surface by the pavement of dorsal plates. Marginal ossicles in Edriodiscus are more
numerous and are inserted into the dorsal pavement. No other Cambrian echinoderm has a
peripheral skirt of plates so well developed. The radial ridges that are such a prominent feature of this
animal are highly reminiscent of the radial ridging present on the lower surface of peripheral plates of
agelacrinitids, and presumably served the same adhesive function. Nothing is known about the oral
surface.

Edriodiscus primotica (Henderson and Shergold, 1971)

Text-fig. 20

1971 Cyclocystoides primotica Henderson and Shergold, p. 706, pi. 138, figs. 1-3.

1985 Edriodiscus primotica (Henderson and Shergold); Jell, Burrett and Banks, p. 190, figs, la-c and 8.

Diagnosis. As for genus.

Types. Holotype, CPC 1 1395 (text-fig. 20); paratype, CPC 1 1396.

Stratigraphical age and distribution. Yelvertoft Beds, late Ordian, early Middle Cambrian of West Queensland,
Australia.

Description. A full description of the lower surface of this species was given by Henderson and Shergold (1971)
and 1 have nothing new to add. New material, showing details of the ventral surface, has recently been found by
Peter Jell and a more complete description has been given by him (Jell et at. 1985).

Acknowledgements. Porter Kier (Smithsonian Institution, Washington), Rudolf Prokop (Narodni Muzeum,
Prague), A. Prieur (Office National de Gestion des Collections Paleontologiques, Lyon), the Abbe Courtessole
(Carcassone), G. Breton (Museum of Le Havre), and R. A. Henderson (James Cook University, Australia) have
allowed me free access to material in their collections. Peter Jell (Victoria National Museum, Australia), Chris
Paul (Liverpool University), and Dick Jefferies (British Museum (Natural History)) have all been kind enough
to read an earlier draft of this paper and made many helpful suggestions. Linally, 1 thank Professor G. Ubaghs
for not only providing me with latexes of specimens in his possession but also for being extremely helpful
throughout the duration of this work. Part of this research was carried out with the support of NERC research
grant GR3/4732.

754

PALAEONTOLOGY, VOLUME 28

text-fig. 20. Edriodiscus primoticus (Henderson and Shergold, 1971), CPC 1 1395, camera lucida drawing of the

holotype, dorsal surface.

SMITH: CAMBRIAN EDRIOASTEROIDS

755

REFERENCES

bassler, r. s. 1935. The classification of the Edrioasteroidea. Smithson, misc. Colins , 93, 1-11.

1936. New species of American Edrioasteroidea. Ibid. 95, 1 33.

— and moody, M. w. 1943. Bibliographic and faunal index of Palaeozoic pelmatozoan echinoderms. Spec.
Pap. geol. Soc. Am. 45, 733 pp.

bather, F. A. 1899. A phylogenetic classification of the Pelmatozoa. Rep. By. Ass. Advmt Sci., D, 916-923.

1900. The Echinoderma. In lankester, e. r. (ed.). A Treatise on Zoology , part 3, 216 pp. Black, London.

— 1915. Studies in Edrioasteroidea I-IX , 403 pp. Published by the author at 'Fabo’, Marryat Road,
Wimbledon, England.

bell, b. m. 1976. A study of North American Edrioasteroidea. Mem. N. Y. St. Mas. Sci. Serv. 21, 447 pp.

1977. Respiratory schemes in the Class Edrioasteroidea. J. Paleont. 51, 619-632.

— 1980. Edrioasteroidea and Edrioblastoidea. In broadhead, t. w. and waters, j. a. (eds.). Echinoderms,
notes for a short course. Univ. Tennessee , Dept. geol. Sci. Stud. Geol. 3, 158-174.

-and sprinkle, j. 1978. Totig/obus, an unusual new edrioasteroid from the Middle Cambrian of Nevada.
J. Paleont. 52, 243-266.

bockelie, f. and paul, c. r. c. 1 983. Cyathotheca suecica and its bearing on the evolution of the Edrioasteroidea.
Lethaia , 16, 257-264.

cabibel, j., termier, h. and termier, g. 1958. Les Echinoderms mesocambriens de la Montagne Noire. Ann/s
Paleont. 44,281-294.

chapman, E. j. 1860. On a new species of Agelacrinites and on the structural relations of that genus. Can. J. Ind.
Sci. & Art, 5, 358-365.

courtessole, r. 1973. Le Cambrien Moyen de la Montagne Noire: biostratigraphie. Laboratoiro de Geologie
CEARN, Toulouse.

fell, h. b. 1962. A classification of the echinoderms. Tuatara , 10, 138-140.

1963. Phylogeny of sea-stars. Phil. Trans. R. Soc. B246, 381 -435.

1965. The early evolution of the Echinozoa. Breviora , 219, 1-17.

Henderson, r. a. and shergold, j. h. 1971. Cyclocystoides from the early Middle Cambrian rocks of north-
western Queensland, Australia. Palaeontology, 14, 704-710.
holloway, d. j. and jell, p. a. 1984. Silurian and Devonian edrioasteroids from Australia. J. Paleont. 57,
1001-1016.

jaekel, o. 1899. Stammesgeschichte der Pelmatozoen, 1: Thecoidea und Cystoidea, 442 pp. Berlin.

1918. Phylogenie und System der Pelmatozoen. Paldont. Z. 3, 1-128.

1923. Zur Morphologie der Asterozoa. Ibid. 5, 344 350.

jell, p. a., burrett, c. f. and banks, m. r. 1985. Some Cambrian and Ordovician echinoderms from eastern
Australia. Alcheringa, 9, 183-208.

kesling, R. v. 1967. Edrioasteroid with unique shape from Mississippian strata of Alberta. J. Paleont. 41,
197-202.

miquel, j. 1894. Note sur la geologie des terrains primaires du departement de l’Herault. Le Cambrien et
l’Arenig. Bull. Soc. Etude Sci. nat. Beziers, 17, 5-36.

— 1905. Essai sur le Cambrien de la Montagne Noire. Coulouma, l’Acadien. Bull. Soc. geol. Fr. (4), 5,
465-483.

paul, c. R. c. and smith, a. b. 1984. The early radiation and phylogeny of echinoderms. Biol. Rev. 59, 443-481 .
pompeckj, j. f. 1896. Die Fauna des Cambrium von Tejrovic und Skrej in Bohmen. Jb K.-K. geol. Reichsanst.,
men, 45, 495-614.

regnell, G. 1945. Non-crinoid Pelmatozoa from the Palaeozoic of Sweden, a taxonomic study. Meddn. Lunds
geol.-min. Instn, 108, 255 pp.

— 1966. Edrioasteroids, pp. U136-U172. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part U.
Echinodermata 3. Geological Society of America and University of Kansas Press, New York and Lawrence,
Kansas.

schuchert, c. 1919. A Lower Cambrian edrioasteroid Stromatocystites walcotti. Smithson, misc. Colins, 70, 1 -7.
smith, a. b. 1984u. Echinoid palaeobiology, 190 pp. George Allen & Unwin, London.

1984A Classification of the Echinodermata. Palaeontology, 27, 431-459.

— and paul, c. R. c. 1982. A revision of the class Cyclocystoidea (Echinodermata). Phil. Trans. R. Soc., B296,
577-684.

termier, h. and termier, G. 1969. Les Stromatocystitoides et leur descendance: essai sur F evolution des premiers
echinoderms. Geobios, 2, 131-156.

756

PALAEONTOLOGY, VOLUME 28

termier, h. and termier, g. 1980. Modalites de devolution des Echinodermes au Cambrien. In jangoux, m.

(ed. ). Echinoderms: present and past, 59-65. A. A. Balkema, Rotterdam.
thoral, M. 1935. Contribution a 1' etude paleontologique de POrdovicien interieur de la Montague Noire et revision
sommaire de la fauna cambrienne de la Montague Noire , 362 pp. Imprimerie de la Manufacture de la Charite,
Montpellier.

ubaghs, G. 1971. Diversite et specialisation des plus anciens echinodermes que Ton connaisse. Biol. Rev. 46 ,

wiley, e. o. 1979. An annotated Linnaean hierarchy, with comments on natural taxa and competing systems.
Syst. Zoo!. 28 , 308-337.

zelizko, J. c. 1911. Nove prispevky ke studiu jineckeho kambria. Rozpr. ceske Akad. 20 (10), 1-7.

157-200.

Typescript received 7 October 1984
Revised typescript received 25 March 1985

ANDREW B. SMITH
Department of Palaeontology
British Museum (Natural Elistory)
Cromwell Road
London SW7 5BD

OPTIMUM PREPARATION, PRESERVATION, AND
PROCESSING TECHNIQUES FOR G R APTOLITE
ELECTRON MICROSCOPY

by l. w. dumican and R. b. rickards

Abstract. Published guidance on the preparation of ultrathin sections of biological material rarely gives
sufficient detail to enable fossil material in general and graptolite material in particular to be prepared to a
consistently high standard. The crucial steps relative to pre-microtoming, ultramicrotoming, and post-
microtome work, including museum storage, are described together with the ‘tricks of the trade’ that, taken in
total, result in successful micrographs.

Transmission electron microscopy (TEM) has been an important technique in biological
and medical science for many years but its application to fossil material poses rather different
problems and such material is generally more difficult to prepare. Urbanek (1978) suggested that
ultrastructural research may help to resolve some of the problems highlighted by earlier
morphological and phylogenetical studies on graptolites. Wide discussion of the mechanism of
periderm secretion in graptolites has suggested that TEM studies may aid a better understanding of
the relations between inferred soft parts and the skeletal material.

Williams (1965, pp. H254-H255) and Nye et al. (1972) described a technique which may be used
simultaneously to section both hard and soft tissues for study by reflected light, a technique which
may be applied, for example, to recent brachiopods. The initial stages of impregnation with epoxy
resins, described by Nye et al. (1972), are similar to those which we have used in the preparation of
graptolites for TEM.

The use of ultrathin sections in palaeontology has been slowly adopted, largely because of the
potential preparation difficulties involved and partly because of costs. Diamond knives commonly
exceed £1000 sterling and are not much cheaper to have sharpened. In contrast, stereoscan electron
microscopy (SEM) is now used routinely in most groups and especially so in graptolites (Rickards et
at. 1 982). There is no logical reason why TEM studies should not become equally routine, for they are
a necessary complement to SEM work, contributing towards our understanding of skeletal
morphology, as has been confirmed by recent studies of graptolites (e.g. Crowther 1981 ). This paper
describes how to isolate graptolites from the original matrix, and their subsequent preparation for
scanning and transmission electron microscopy. Photographic techniques employed to obtain
maximum contrast in the final micrograph are then described briefly and relevant problems of
museum documentation discussed.

Graptolite material is quite variable in its preservation of ultrastructural detail, sometimes
exquisite but often showing varied forms of degradation. Even the best techniques, coupled with
infinite care, occasionally give barely adequate results, whilst some specimens seemingly respond
well to very primitive technology (Berry and Takagi 1970). The current procedures evolved during
tenureship by the authors of a Research Grant from the Natural Environment Research Council
on graptolite ultrastructure, although considerable progress was made in the 1970s and summarized
by Crowther and Rickards (1977). A number of the pre-sectioning techniques have not been
described before but some were developed many years ago by Professor O. M. B. Bulman and by the
junior author.

| Palaeontology, Vol. 28, Part 4, 1985, pp. 757-766.|

758

PALAEONTOLOGY, VOLUME 28

CHEMICAL ISOLATION

Most SEM and TEM work is enhanced by using specimens chemically freed from the rock matrix.
Adhering rock or pyrite can seriously damage diamond knives; but for SEM studies, graptolites in
the matrix have been used successfully (Rickards etcil., 1971; Crowther and Rickards 1977) in part to
test the effect of chemical preparation on the ultrastructural detail, which appears to be negligible.

The procedure for chemical isolation follows two basic routes depending upon whether the matrix
is highly calcareous or not. It should always be carried out in a fume cupboard. Many limestones
containing graptolites can be immersed as small pieces (a few cm across) in dilute hydrochloric acid
(10 % HC1; text-fig. 1), in glass beakers, releasing the graptolites often in a matter of hours. There is
no way of determining in advance whether the periderm is strong enough to withstand the effer-
vescence, or whether the graptolites may be isolated by any method, other than by trial and error.
If the effervescence appears to be breaking up the emerging specimens, then the acid may be diluted
to slow down the action, or a non-effervescing or less violently effervescing acid (such as acetic
acid, CHjCOOH) can be used.

A second procedure may be used when the terriginous content of the limestone is so great that a
relatively firm rottenstone remains, failing to release the graptolites. When this happens the blocks
must be treated with 1 0 % HC1 for up to three weeks, changing the acid daily, until all possible CaC0 3
has been removed. At that stage the rottenstone can be treated with 60% hydrofluoric acid, in
polythene beakers, which usually releases the graptolites within a few hours. In all the above
treatments, two or three pieces of rock can be placed in up to 500 cc of HC1 or up to 250 cc of E1F (text-
fig. 1), or the whole process can be made into an assembly line using large containers and quantities
of fluid.

The next stage is the most laborious and critical if the full suite of isolated specimens is to be
retained. Some specimens will have floated in an oily scum, often adhering to the meniscus. These
need to be carefully pipetted off, using either glass or polythene hand pipettes (for HC1 or HF
respectively), into a container of distilled water in which they may still float. The remaining fluid is
then carefully decanted and washed away using copious supplies of water from the fume cupboard
taps. Specimens ‘floating’ in mid-fluid can be pipetted out at this stage. As a rule, a majority of
specimens lie in the muddy debris at the bottom of the beaker. Distilled water should be added to the
beaker, stirring the sunken specimens and mud, and the whole of the above decanting process
repeated several times until the acid has been washed out. Pipetting can take place at the same time
but eventually the whole remaining mass of mud and graptolites needs to be carefully decanted into a
wide, shallow, preferably white-bottomed picking dish. At no stage should the specimens be allowed
to dry out or to support their own weight in air, for many collapse, especially if the original rock came
from a tectonized region. However, the most delicate rhabdosomes may be supported by fluid and
can easily be held in that medium until they are transferred to the long term resting medium, namely
viscous glycerine.

Initial picking by hand pipette is from the picking tray into distilled water in glass beakers. The
graptolites must not be stored in distilled water for longer than about twenty-four hours, since this
encourages the growth of fungus. The main problem at this stage is the number of fragments and
early growth stages that may be masked by the mud and missed altogether. Specimens should be
washed in several changes of distilled water for a few hours.

The best way to transfer delicate specimens into viscous glycerine is to reduce the amount of
distilled water over them by pipetting it off until only a few mm cover the specimens. Then glycerine
is added gently around the margins of the container, thus pushing the specimens away from the sides.
Eventually, glycerine can be dropped gently on top of them, weighing them down (text-fig. 1). The
remaining distilled water floats above the glycerine and, when it is completely separated from the
graptolites, can easily be pipetted off to leave the specimens in a firm, supportive, fluid matrix.

Specimens that floated on initial extraction may sink quite quickly; if not, they will have tiny gas
bubbles lodged inside the thecae. These can be removed by pipetting the specimen directly into
alcohol, a somewhat violent process which may snap the specimens; alternatively they can be cooled

DUMICAN AND RICKARDS: GR APTOLITE ELECTRON MICROSCOPY 759

PIPETTE OFF DISTILLED WATER

ADD GLYCERINE TO SPECIMENS

t

4

usi

/\

PIPETTE TO DISTILLED WATER

ADD DISTILLED WATER

DECANT

REPEATEDLY

S E M PREPARATION TEM PREPARATION

I

PHOTOMOUNT

}

'//;

DEHYDRATION &
IMPREGNATION

4

PIPETTE
IN ALCOHOL

EMBEDDING
( Fig. 2 A )

i

COATING

OF

Au

Au / Hg

OR

c

ULTRAMICROTOME
SECTIONING
( Fig. 2B )

%

'•v :‘o 1 »>

S

T

l % / o

E PHOTOGRAPHY

E PHOTOGRAPHY

t o! • *o

1 *. 1 1 ' * * f '

v » v . i

M

M ( Figs. 4 A, B )

--

i

500cc

♦

250cc

HCI

HF

CURATION OF MICROGRAPHS,
NEGATIVES, STUBS,
SUPPLEMENTARY MATERIAL

text-fig. 1 . Flow chart summarizing preparatory stages for storage, SEM and TEM work.

760

PALAEONTOLOGY, VOLUME 28

in a refrigerator when the small gas bubbles may escape (dishes of alcohol should not be placed in
a refrigerator without proper consultation with a laboratory technician, for there is a risk of
explosion). Once the specimens have sunk, transfer to glycerine can proceed as above, and the
graptolites are now ready for the preparatory stages described below, leading to electron microscopy.

As an important aside, it should be mentioned that in preparing glycerine held specimens for light
microscopy (as, for example, in the production of slides or resin mounts; Hutt and Rickards 1967),
any chemical clearing using Schultz" Solution (KC10 3 and HN0 3 ) should be done in the absence of
glycerine since there is a slight risk of accidentally producing TNG or TNT.

SEM PREPARATION

Preparation of graptolites for use in the SEM involves mounting the specimen on a stub and
subsequently coating it, preferably with gold; other coatings may be tried, such as gold/palladium or
carbon, and occasionally it may not be necessary to coat at all if much pyrite adheres to the
specimens. The specimen is pipetted from glycerine into absolute alcohol, in which it is washed for
approximately thirty minutes. An intermediate distilled water stage has proved unnecessary. The
stub is then sprayed with a very thin film of ‘Photomount’ (see Crowther and Rickards 1977), or
double sided adhesive tape may be used, and the specimen is pipetted with a drop of alcohol to the
surface of the stub. The graptolite may be oriented at this stage by manoeuvring it with a damp hair,
although great care must be taken with very fragile specimens. The alcohol is then allowed to
evaporate. Finally, with the Sedgwick Museum material, the stub is placed in an ‘EMscope’ sputter
coater and coated with 200-500 A of gold, after which it is ready for examination in the SEM.

TEM PREPARATION

The specimen is pipetted from glycerine into a vial containing distilled water and washed in three
changes of water for fifteen minutes each. Commonly, a graded series of ethyl alcohol dilutions is then
used for dehydration of the specimen but it is possible to use three changes of absolute alcohol (about
ten to fifteen minutes each). At this stage the specimens may be stored indefinitely in the alcohol.
After dehydration the specimens are soaked in a 50/50 mixture of propylene oxide (1,2 — epoxy
propane) and absolute alcohol and then in a further two changes of 100 % propylene oxide for fifteen
minutes each. This facilitates uniform impregnation of the specimen by the epoxy.

During each of the above stages the vial is placed in a rotator so that the liquid circulates freely
inside the specimen. Whilst the specimens are soaking in the propylene oxide the epoxy resin mix may
be prepared. Propylene oxide should be used in glass rather than plastic containers and only in a fume
cupboard because of its toxicity, volatility, and flammability. After use it should be flushed away with
continuous water flow for five minutes.

Embedding

The ‘Agar 100’ embedding kit (equivalent to ‘Epon 812’) is suitable for embedding graptolite
material. The ‘Epon’ mixture is blended and accelerator added just before use. A graduated cylinder,
a small (100 ml) conical flask, and the containers of resin and hardener are warmed in an oven at
60 °C. The four components, resin, DDSA hardener (dodecynl succinic anhydride), MNA hardener
(methyl nadic anhydride), and BDMA accelerator (N benzyl N-N dimethylamine) are mixed by
pouring them in turn into the graduated cylinder. Before adding BDMA the other three components
should be stirred thoroughly since direct mixing of BDMA and MNA may be explosive. The mixture
is then poured immediately into the warm conical flask and stirred for about one minute. A few air
bubbles may develop but these will dissipate if the mixture is allowed to stand for a short time at
60 °C. Epoxy resins, hardeners, and accelerators should be handled with care in a fume cupboard.

Difficulties may be experienced in obtaining blocks of the correct hardness for ideal sectioning. The
following mixture has recently been used successfully to embed specimens of Monograptus formosus
from the Mielnik borehole in Poland: resin 25 ml; DDSA 11 ml; MNA 14 ml; BDMA 1 ml. The

DUMICAN AND RICKARDS: GRAPTOLITE ELECTRON MICROSCOPY

761

hardness of the final block may be controlled by varying the proportions of DDSA and MNA in the
resin mixture. When proportions of DDSA to MNA were tried in the quantities 12-5/12-5 ml the
resultant blocks were too soft, and at 10/15 ml they were too brittle. Hardness may also be increased
as the concentration of the accelerator (BDMA) increases, but the block may become brittle and
difficult to section as a result.

After the resin components have been stored for about five months, they begin to give inconsistent
results. It is suggested that the accelerator should be stored in a dessicator in the dark. Several
workers have found that the accelerator for ‘Spurr’ resin, for example, has a very short shelf life
(about two weeks). The liquid epoxy resin is soluble in absolute alcohol and all glassware should be
rinsed in alcohol after use.

The specimen is then transferred in propylene oxide to a small tray (a vial top or petri dish is ideal)
and an equal volume of epoxy resin mixture is added. These are left for one hour, loosely covered to
prevent evaporation. The covers may then be removed and the specimens left in a fume cupboard
overnight, during which time the propylene oxide will evaporate. They are then polymerized in a
thermostatically controlled oven in three stages (35 °C, 45 °C, and 60 °C), over a period of thirty-six
hours: overnight at 35 °C, next day at 45 C, and overnight at 60 °C. The maximum internal
temperature of the epoxy resin obtained during polymerization will affect the properties of the
resultant block. If the temperature is too low, the block will be too soft; if the temperature is too high,
bubbles may form in the epoxy and the block will be too brittle (Nye et al. 1972). Ideally the end
product should be an amber colour and of moderate hardness (i.e. will not deform when pressed with
a finger nail).

The graptolites are now ready for cutting into blocks and mounting on ‘Araldite’ stubs made
and polymerized in the same way and at the same time. The hardened resin is trimmed from the speci-
men, which is then stuck to a stub (made in polyethylene capsules known as ‘BEEM capsules’)
in the required orientation, with 2-tube ‘Araldite’. The graptolite may be trimmed down to size
with a heavy duty, backed razor blade. This step is carried out with the specimen held under
water because it is usually too small to be clamped and consequently is easily lost during trimming
in air.

Where a series of transverse sections from a specimen of three or more thecae in length is required,
it is suggested that a small hole be made in the end of a BEEM capsule and the trimmed specimen then
forced through the hole so that the end to be sectioned protrudes from the capsule (text-fig. 2a). The
capsule is then filled with liquid epoxy resin and polymerized as before. The specimen will have been
polymerized twice but this does not adversely affect the end-product. When specimens of this type
were simply stuck on stubs with 2-tube ‘Araldite’, they often snapped off during sectioning on the
ultramicrotome.

Cutting ultrathin sections

After polymerization the block is held in the clamp of the ultramicrotome and the area around the
graptolite trimmed into a pyramidal shape (text-fig. 2a). The face to be sectioned should be cut until it
reaches the specimen to save time and wear on the knife edge. The block face is shaped so that cut
sections will form a ribbon, perpendicular to the knife edge, on the surface of the water bath (text-fig.
2b). It is therefore necessary for two edges of the block to be parallel to each other and orientated in
the ultramicrotome so that they are parallel to the knife edge (text-fig. 2b). It is best if one side is
longer than the other, forming a trapezium orientated such that each new section pushes along the
whole width of the previous one, so detaching it from the knife edge. Neatly rectangular slices do not
form such a reliable and straight ‘ribbon’.

A diamond knife is held in the knife holder, tilted at approximately 4°, and the water bath is filled
with distilled water until it appears white in polarized light (i.e. maximum reflection is obtained with a
‘flat’ meniscus). The knife is advanced manually towards the block until almost touching it, as seen
through the binocular microscope, then locked in position. It is often difficult to see just how close the
block is to the knife edge. A mirror placed under the specimen will reflect bright, white light between
the knife edge and block face, making it easier to judge the gap between them. The knife should never

762

PALAEONTOLOGY, VOLUME 28

text-fig. 2. a, showing how embedded graptolite (three or more thecae in length) may be orientated for trans-
verse sectioning; the encased specimen is forced through a hole at the shaped end of the BEEM capsule which is
then filled with epoxy resin and polymerized (full explanation in text), b, sketch of trapezium-shaped block face
which causes the cut sections to form a saw-edged ribbon on the surface of the ultramicrotome’s water bath.

touch the block face at the start since this will undoubtedly damage the edge. When the knife and
block face are as close as possible the block is advanced 0-5 pm at a time until a complete section is cut
from the face, after which the block is advanced automatically by a set thickness.

Interference colours are used to estimate the thickness of cut sections. Peachey (1958) provided a
correlation between section thickness and interference colours as follows:

Colour

Thickness A (1 nm = 10 A)

grey

600

silver

600-900

gold

900-1500

purple

1500-1900

blue

1900-2400

green

2400-2800

yellow

2800-3200

Ideally sections should have grey interference colours, especially if they are to be used for high
resolution work. However, silver and occasionally pale gold sections have been used successfully for
work at low magnifications. The position and the meniscus level of the water bath may have to be
finely adjusted for optimum viewing of the interference colours. The ultramicrotome is very sensitive
to touch and atmospheric conditions. It is therefore necessary to avoid contact with it as far as
possible and also to eliminate any draughts in the room, especially those caused by doors opening; a
steady temperature of 20 °C should be maintained. A cutting speed of IT mm/s used with a knife
angle of approximately 4° provides good sections of graptolite material. In general, hard specimens

DUMICAN AND RICKARDS: GR APTOLITE ELECTRON MICROSCOPY

763

are best cut at slower speeds than soft specimens. The optimum settings for all controls can only be
realized by experimentation. Sections should float off evenly in a straight line and be of uniform
colour, flat, and with no corrugations.

Most of the difficulties in cutting good sections arise from faults with embedding and the knife
edge. For a further detailed discussion of faults observed in ultrathin sections, and their possible
causes, see Reid (1974). Generally glass knives do not provide such good sections of graptolites as
diamond knives, but they may be used to trim specimens prior to sectioning with the diamond knife.
When viewed through the binocular microscope the edge of a glass knife can often be seen to crumble
after only one cut of the material. When the embedding medium is softer, several cuts are possible but
problems with corrugations within the section and the mounting medium may be encountered (text-
fig. 3). After a few sections have been cut with a glass knife the edge becomes blunt and it is necessary
to move to an unused part of the knife. Diamond knives are more durable and may be used for
repeated sectioning over long periods of time (often several weeks or months).

text-fig. 3. TEM micrograph showing corrugations throughout the section, caused when the embedding
medium is too soft, x 8000 (Dictyonema rarum Wiman, SM XI 193).

A picking brush dipped in chloroform and held over the surface of the water bath causes the
sections to flatten, due to the heavy vapour. This eliminates, to a certain extent, deformation that
might have occurred during sectioning. An uncoated copper grid is held with fine forceps and
placed matt side down on top of the floating sections, which will then adhere to the grid. After
drying matt side up on filter paper the grids are ready for use in the electron microscope. We have
found that in general it is not necessary to coat prepared grids with carbon (but see also Crowther
and Rickards 1977).

It is more informative to study several sections cut in serial order than single sections, since any
contaminants introduced during sectioning can be recognized more easily when seen in successive
sections. If the ‘ribbon’ of cut sections is transferred to the grid intact, it may be possible during TEM
examination to obtain a three-dimensional understanding of that part of the graptolite ultra-
structure.

PHOTOGRAPHY

Initially problems were encountered with lack of contrast in the electron micrographs, a not
uncommon feature of electron microscopy; the thinner the section the less contrast there will be
between specimen and background. Since thinner sections provide better resolution, it is necessary to

764

PALAEONTOLOGY, VOLUME 28

try to increase the contrast in some other way. One such method is to take photographs slightly
underfocused. True focus on the electron microscope is found at the point of least contrast. This may
be seen by observing fresnel fringes on the edges of the specimen or around the edges of adventitiously
placed holes (text-fig. 4). (Fresnel fringes arise from the interference between scattered and
unscattered electron beams; see Agar and Chescoe (1974) for a full discussion on their formation.)
When the specimen is overfocused, the edge is outlined by a pale diffraction band with a concentric
dark one (text-fig. 4b); when underfocused, a pale band (or fresnel fringe) follows the edge (text-fig.
4a). True focus occurs where the fresnel fringe disappears but contrast appears to be minimal. Thus it
is often preferable to take photographs slightly out of focus in order to heighten contrast whilst
increasing, of course, the risk of lower resolution. Contrast may also be increased at the film
development stage. Micrographs are taken using ‘Kodalith MP11 Ortho film 2577’ on an ‘AEI
Corinth 500’ electron microscope and are developed for three minutes at 13 °C in ‘Ilford Phenisol’
developer, diluted in the proportion 1 :6.

text-fig. 4. TEM micrographs showing the pattern produced by fresnel fringes, a, overfocused; b,

underfocused. Both SM XI 193, x 5000 approx.

The micrographs are then printed on ‘Ilford Ilfoprint grade 4’ paper using a ‘Beseler MCX’
enlarger. TEM micrographs should be taken as quickly as possible, since contamination builds up
quite quickly on the specimen and may adversely affect both contrast and resolution, blurring the
section. SEM photographic techniques have been discussed comprehensively by Crowther and
Rickards (1977, pp. 11-12) and are not enlarged upon here, except to say that most of our SEM work
is now carried out using a ‘Philips 501 B’ electron microscope and ‘Ilford FP4 (70 mm)’ film processed
according to manufacturer’s instructions.

MUSEUM STORAGE

Storage of isolated reference material used under the light microscope is a relatively straightforward
matter. The authors have been working on graptolites prepared and stored in glycerine, corked, and
sealed with candlewax in 1895 and they are as satisfactory today as presumably they ever were. They
respond well to both TEM and SEM studies, and the only deterioration evident is of the glycerine,
not the graptolite. In some tubes the glycerine has gone slightly brown, though it is still transparent.
At the Sedgwick Museum, graptolites have been similarly stored in glycerine for twenty years.

DUM [CAN AND RICKARDS: GRAPTOLITE ELECTRON MICROSCOPY

765

However, in order to facilitate normal examination by research workers, they are stored in the type of
plastic container illustrated (text-fig. 1) in which very viscous glycerine only 5-10 mm deep is used,
which greatly lessens the chance of spillage. Yet the container is not sealed, has a sliding lid, and the
specimens can be easily examined under the light microscope, particularly as there is no problem with
the sphericity of the container. There is no reason to suppose that material, so stored, will not last fifty
years and normal curatorial procedures can be adopted by the museum staff.

SEM stubs pose a more difficult problem. The Sedgwick Museum has fifteen-year-old mountings
which have been re-used successfully, but deterioration is apparent, especially in the glues or gums
used to mount the specimens on the stubs but also in the coating used. The question of glues has been
examined very thoroughly by Dr Jenny Chapman (in prep, and pers. comm.), and it seems unlikely
that even well-mounted graptolites will last twenty-five years. Therefore, whilst normal curatorial
procedures can be applied to stubs, and are at the Sedgwick Museum, the following items take on
greater importance in curation: 1, the photographic negative (and prints); 2, unmounted topotypes
(preserved in glycerine); and 3, topotype rock samples known to have yielded the originals.

The most difficult preservational problem concerns TEM ultrathin sections mounted on grids.
Although Sedgwick Museum specimens have been re-used successfully after a period of five years, it
is unlikely that grids will survive a decade; re-use itself usually causes rapid deterioration so that
further use is unlikely. Therefore, the film negatives are vital from a curatorial and research
standpoint and there seems no good reason why, properly stored, they should not last for a century.
Thus the film negative becomes the ‘specimen’ for all future research reference. A supplementary part
of the specimen is that remaining in the unsectioned stub. Although this has not appeared on film, it is
a potential source of at least partial confirmatory work on the same specimen and hence should be
catalogued as a very important part of the original. The life of a specimen mounted in resin depends
upon the life of the resin, which may deteriorate in, at most, a few years, making further sectioning
difficult (at least with a valuable diamond knife), but the mounted specimens may last for several
decades as far as light microscopy is concerned.

It is clear from the authors’ work that curation must be built into the procedural system and
planned for at an early stage of the work. Much the best system is to curate all parts and products
immediately after they have been produced, any delay merely contributing to the possibility of very
small items being separated from each other and being effectively lost.

Acknowledgements. We gratefully thank Roger Northfield (Department of Zoology, University of Cambridge)
for all his help with ultramicrotomy, and David Bursill for his constant assistance during experimental
photographic work.

REFERENCES

agar, a. w. and chescoe, d. 1974. Image formation in the electron microscope. In glauert, a. m. (ed.). Practical
methods in electron microscopy, 2.

berry, w. b. n. and takagi, r. s. 1970. Electron microscope investigations of Orthograptus quadrimucronatus
from the Maquoketa Formation (Late Ordovician) in Iowa. J. Paleont. 44, 1 17-124.
crowther, p. r. 1981. The fine structure of graptolite periderm. Spec. Pap. Palaeont. 26, 119 pp.

— and rickards, r. b. 1977. Cortical bandages and the graptolite zooid. Geo/ogica Palaeont. 11, 9-46.
hutt, j. and rickards, r. b. 1967. An improved transfer technique for the preparation and preservation of

pyritized graptolites. Geol. Mag. 104, 180-181.

nye, o. b., dean, d. a. and hinds, r. w. 1972. Improved thin section techniques for fossil and recent organisms. J.
Paleont. 46, 271-275.

peachey, l. d. 1958. Thin sections. A study of section thickness and physical distortion produced during
microtomy. J. hiophys. biochem. Cytol. 4, 233-242.

REID, N. 1974. Ultramicrotomy. In glauert, a. m. (ed.). Practical methods in electron microscopy, 3.
rickards, r. b., hyde, p. j. w. and krinsley, d. h. 1971. Periderm ultrastructure of a species of Monograptus
(Phylum Hemichordata). Proc. R. Soc. B178, 347-356.

— crowther, p. r. and chapman, a. j. 1982. Ultrastructural studies of graptolites — a review. Geol. Mag. 1 19,
355-370.

766

PALAEONTOLOGY, VOLUME 28

urbanek, a. 1978. Significance of ultrastructural studies for graptolite research. Acta palaeont. pol. 23, 595-629.
williams, A. 1965. Techniques for preparation of fossil and living brachiopods, pp. H251 -H256. In moore, r. c.
(ed.). Treatise on invertebrate paleontology. Part H. Brachiopoda (1). Geological Society of America and
University of Kansas Press, New York and Lawrence, Kansas.

LORI W. DUMICAN
BARRIE RICKARDS

Sedgwick Museum
Department of Earth Sciences
University of Cambridge

Typescript received 20 October 1984 Downing Street

Revised typescript received 29 January 1985 Cambridge CB2 3EQ

ON THE IDENTITY OF THE AMPHIBIAN
HESPEROHERPETON GARNETTENSE FROM THE
UPPER PENNSYLVANIAN OF KANSAS

by ANDREW R. MILNER

Abstract. Hesperoherpeton garnettense Peabody was first described as an embolomerous antfiracosaurian
amphibian, based on a small scapulocoracoid and associated neural arch from the Upper Pennsylvanian of
Garnett, Kansas. Subsequently, on the basis of a referred specimen from the type locality, Hesperoherpeton was
claimed by Eaton and Stewart to be the most anatomically primitive tetrapod despite its Upper Pennsylvanian
age. Re-examination of the type and the referred specimen leads to the conclusion that the former is
indeterminate and the latter is a poorly preserved small specimen of a temnospondyl amphibian of the family
Trematopidae. It may be a specimen of the trematopid Actiobates peabodyi Eaton from the same locality. The
binomen H. garnettense is a nomen dubium restricted to the type specimen, and Eaton and Stewart’s Order
Plesiopoda is invalid.

The Upper Pennsylvanian tetrapod assemblage from Garnett in eastern Kansas is unique among
Carboniferous faunas in that it is made up largely of diapsid and pelycosaurian reptiles (Reisz 1981;
Reisz et al. 1982). Only three amphibian specimens from Garnett have been described. One of these,
Actiobates peabodyi Eaton 1973 is a temnospondyl of the family Trematopidae (Milner: in prep.). The
other two specimens form the subject of this study. In 1958, Peabody described a small
scapulocoracoid and associated neural arch from Garnett as the new taxon Hesperoherpeton
garnettense which he assigned to the family Cricotidae within the embolomerous anthracosaurs. In
1960, Eaton and Stewart described a second specimen from Garnett which they attributed to
Hesperoherpeton. This specimen comprised the crushed and disarticulated anterior skeleton of a
small amphibian, which Eaton and Stewart reconstructed, concluding that it was not an embolomere
but was quite unlike any other known fossil amphibian. Despite its late Pennsylvanian age, Eaton and
Stewart interpreted Hesperoherpeton as the most primitive known tetrapod, a relictual form
structurally intermediate between rhipidistian fishes and tetrapods and they made it the basis of a new
order Plesiopoda. This order was erected as a grade group characterized by a mixture of choanate fish
characters (large notochordal canal, divided braincase, postaxial processes on some fore-limb
elements, and probable articulation between tabular and pectoral girdle) and tetrapod characters
(tetrapod pectoral girdle, digits, otic notch, nostrils separate from jaw margin and rhachitomous
vertebrae). Eaton and Stewart also placed Hesperoherpeton in the monogeneric family
Hesperoherpetonidae defined by its unique specializations, notably the squamosal bordering the
orbit and the loss of most of the circumorbital series. As the skull is both disarticulated and
incomplete, it is difficult to understand how such unique characters were identified. The presence of
postaxial processes on some limb bones neatly fulfilled a prediction about primitive tetrapod limb
structure made by Eaton several years previously (Eaton 1951) and elaborated subsequently (Eaton
1960). Most later workers on Palaeozoic tetrapods who have examined this material or discussed
tetrapod origins, have either treated Eaton and Stewart’s interpretation with extreme caution (Romer
1968, p. 88; Olson 1971, p. 292) or have ignored it altogether (Thomson and Bossy 1970; Carroll and
Winer 1977; Rosen et al. 1981; Reisz et al. 1982). The only alternative systematic position suggested
for it was a doubtful placement within the Seymouriamorpha by Romer (1966, p. 363) followed by
Kuhn (1972, p. 43). However, in some more general accounts of amphibian evolution it has been
cited as a very primitive tetrapod illustrating an intermediate condition between fishes and

(Palaeontology, Vol. 28, Part 4, 1985, pp. 767-776.]

768 PALAEONTOLOGY, VOLUME 28

tetrapods (Schmalhausen 1968, pp. 54, 62; Porter 1972, p. 93, Vorob'yeva 1974, p. 222; Alexander
1975, p. 235).

During 1979 I was able to examine the two Hesperoherpeton specimens and make a number of
observations which suggest a more mundane interpretation of this material. Although precise
systematic placement of the two specimens is not possible, this reinterpretation does provide a
refutation of the status given to them by Eaton and Stewart.

MATERIAL

KUVP 9976 The holotype of H. garnettense Peabody 1958. Collected in 1954 and figured by
Peabody (1958, text-fig. 1), Eaton and Stewart (1960, fig. 8), and in this work (text-fig.

1a, B).

KUVP 10295 Referred specimen of Eaton and Stewart 1960. Collected in 1955 and figured by
Eaton and Stewart (1960, fig. 1-7 and 9-11) and in this work (text-figs. 2b-4).

Both specimens are housed in the Vertebrate Paleontology collections (KUVP) of the Museum of
Natural History at the University of Kansas, Lawrence, Kansas.

Locality and Horizon. Quarry 10 km north-west of Garnett, Putnam Township, Anderson County,
Kansas. Rock Lake Member, Stanton Formation, Lansing Group, Missourian Series, Upper
Pennsylvanian. Reisz et al. (1982) provide further information on the locality and the detailed local
stratigraphy.

DESCRIPTIONS

The holotype specimen

KUVP 9976 consists of a small right scapulocoracoid and a neural arch in close association. The
originally exposed surface has been embedded in plastic and the specimen subsequently prepared out

CM

i 1

text-fig. 1. a, b, Hesperoherpeton garnettense Peabody ( nomen dubium). KUVP 9976, the holotype and
only specimen from Garnett, Kansas. Right scapulocoracoid in a, external aspect, b, internal aspect,
together with associated neural arch, n.a. neural arch.

MILNER: PENNSYLVANIAN AMPHIBIAN

769

so that the two bones are visible from both sides. The specimen was figured by Peabody (1958, text-
fig. 1) and, less accurately, by Eaton and Stewart (1960, fig. 8) but is refigured here for ease of
reference (text-fig. 1a, b) although the figure, drawn directly from the specimen, contains no new
information.

The neural arch is not co-ossified to a centrum and has a low neural spine. The former feature
indicates that it cannot be attributed to the orders Nectridea, Ai'stopoda, Microsauria, or any amniote
group. The absence of swollen neural arches suggests, albeit from negative evidence, that it is not a
seymouriamorph. Peabody suggested that it belonged to an embolomerous form as the symmetrical
anterior and posterior faces of the pedicels implied the presence of equally large intercentra and
pleurocentra. However, although the pedicel faces are asymmetrical in some rhachitomous forms,
they are symmetrical or nearly so in others (Neldasaurus Chase 1965, fig. 9a; Eryops Moulton 1974;
Tersomius Daly 1973, text-figs. 5a and 6). Thus the neural arch does not characterize the specimen as
an embolomerous anthracosaur but simply as either a temnospondyl or an anthracosaur or a more
primitive stem-tetrapod.

The scapulocoracoid is well ossified for its size and this suggests that it belonged to a terrestrial
form, small aquatic amphibians having unossified or poorly ossified endochondral regions in the
pectoral girdle. The shape of the scapulocoracoid resembles that of labyrinthodont-grade amphibians
rather than of reptiles. The presence of supracoracoid and glenoid foramina below the glenoid
resembles the condition in both temnospondyls (Greererpeton Holmes 1980, Dissorophus DeMar
1968) and anthracosaurs (Proterogyrinus Holmes 1980). The glenoid is large and slightly triangular as
originally described by Peabody, not a small ovoid as interpreted by Eaton and Stewart, and it
resembles that in terrestrial temnospondyls such as Trematops (Olson 1941, fig. 11b).

In conclusion, the holotype specimen of H. garnettense appears to be an indeterminate
labyrinthodont-grade amphibian, possibly a terrestrial temnospondyl. The bones are entirely
consistent with attribution to a trematopid temnospondyl such as Actiobates which occurs at
Garnett, but are equally consistent with attribution to a small member of the Dissorophidae or
Eryopidae. Thus the binomen H. garnettense is a nomen dubium restricted to an indeterminate
holotype specimen.

The referred specimen

KUVP 10295 consists of a large number of small bones on a small slab of shale. The originally exposed
surface has been embedded in plastic and the specimen has been prepared from the back. As noted by
Eaton and Stewart (1960, p. 220) the bones are mostly broken or distorted by crushing and are also
scattered. All the identifiable bones are consistent in size, number, and general relative position with
the interpretation that they derive from the anterior half of a single, small amphibian skeleton. The
anterior appendicular skeleton and the anterior trunk are most readily recognized, being only slightly
disarticulated, the posterior skeleton is absent, whilst the skull is severely crushed and scattered.
Regrettably, some elements are now difficult to observe through the plastic coat. However, I have
been able to identify and draw several bones which I believe are sufficient to permit a reassessment of
this specimen. These include some elements of the skull roof, palate, mandible, axial and anterior
appendicular skeleton. Unfortunately I was unable to identify the braincase, stapes, or occiput
amongst the crushed cranial elements at one end of the slab: those skull bones which are figured and
described here being a group which have drifted back towards the pectoral region. Some of the
identified bones are figured in text-figs. 2b-4 and are described in the following paragraphs.

Skull. Five cranial elements have been confidently identified, namely a right lachrymal, a left
postfrontal, a right squamosal, a right vomer, and a left pterygoid.

The right lachrymal (text-figs. 2b and 3) is the most diagnostic element in the specimen. It is present
as a disarticulated bone and was described and figured by Eaton and Stewart (1960, fig. 3c, d) as the
right tabular. It consists of an elongate triangular plate of lightly sculptured dermal bone with a long
slender process extending ventrolaterally from the unsculptured surface along its long axis (text-fig.
2b). Eaton and Stewart interpreted this as a tabular with a small otic notch supported by a flange of the

770

PALAEONTOLOGY, VOLUME 28

text-fig. 2. a, Actiobates peabodyi Eaton. KUVP 17941, the holotype specimen. Left
anterior region of the skull based on Eaton 1973, fig. la and on first-hand observation of
the specimen. The lachrymal is heavily outlined and sculpture pits are depicted on it.
Abbreviations are: fr —frontal, ju— jugal, mx maxillary, n — nasal, pmx— premaxillary,
pr— prefrontal, b, indeterminate trematopid. KUVP 10295, isolated ossification inter-
preted as right tabular by Eaton and Stewart, but here identified as the right lachrymal of a

trematopid.

dermal surface and by the slender process. In their discussion they suggested that the slender process
may have articulated with the pectoral girdle in a rhipidistian-like manner. There is no precedent for a
tabular of this particular shape in the lower tetrapods and as the bone occurs as an isolated structure,
its identification as a tabular would appear to have been no more than a guess. However, its precise
shape and size relative to the other bones are entirely consistent with it being a lachrymal of a
temnospondyl amphibian of the Family Trematopidae. The Trematopidae are a family of
dissorophoid amphibians characterized by several features, most conspicuously the elongate external
naris (Olson 1941; Bolt 1974a). Within the Order Temnospondyli, this type of external naris, with a
posterior extension and a characteristic configuration of surrounding bones is unique to the
Trematopidae, although a similar arrangement occurs in the Upper Permian batrachosaur
Chroniosaurus from Russia (Tverdochlebova 1972). Such a narial structure has been claimed for the
dissorophid Longiscitula houghae (DeMar 1966; Boy 1 98 1 ) but Bolt (1974a) notes that this isdoubtful
and I can confirm, having examined this specimen, that the ‘elongate nostril’ is an artifact of crushing.
In the trematopids the posterior extension of the naris is a manifestation of modifications to the nasal,
prefrontal, lachrymal, maxillary, and vomer (Bolt 1974a). The primitive temnospondyl lachrymal
condition is a roughly rectangular bone extending from naris to orbit and bearing the lachrymal duct
or ducts. In most long-snouted temnospondyls it withdraws from contact with the naris and orbit and
is a rhomboidal bone bordered by nasal, maxillary, prefrontal, and jugal. In trematopids however, the
enlarged naris, combined with the retention of the lachrymal position on the orbit margin, results in
the lachrymal being preserved as a triangular plate of sculptured bone bordering the orbit
anterolaterally. This is well shown by the trematopid Actiobates, also from Garnett (text-fig. 2a).
A ventral process of bone extends anteriorly from the main body of the lachrymal along the floor of

MILNER: PENNSYLVANIAN AMPHIBIAN

771

the nasal chamber bordering the maxillary (text-fig. 2a) and this corresponds to the slender process
which Eaton and Stewart interpreted as a form of tabular horn. The resultant configuration of a
trematopid lachrymal is matched by the bone in KUVP 10295 (text-figs. 2b and 3) and this provides
compelling evidence that the bone is a lachrymal and that the specimen is a trematopid
temnospondyl.

The left postfrontal is present as an isolated element overlapping the left pterygoid (text-fig. 3). It is
anteriorly pointed and posteriorly rectangular with a small posterior lappet which would have
extended under the supratemporal. This shape of bone and the position of the lappet identify it as a
postfrontal (Boy 1972 Abb. 4k depicts a very similar postfrontal in Micromelerpeton). It has
temnospondyl-type irregular pitted sculpture and the absence of dermo-sensory pits or canals
indicates that it derives from a terrestrial temnospondyl. In these features it is consistent with
attribution to the families Eryopidae, Trematopidae, or Dissorophidae. The anterior extension
terminating in a point rather than a sutural contact indicates that there was no prefrontal-postfrontal
common suture but that both bones terminated as points over the orbit, and that the frontals entered
the orbit margin. This derived character occurs in advanced dissorophids (not Amphibamus ) and in
trematopids, but not in eryopids where there is consistently a prefrontal-postfrontal contact. No bone
of this precise shape was figured by Eaton and Stewart but it may be the ‘postorbitaF of their
description.

The right squamosal (text-fig. 3) was figured by Eaton and Stewart (1960, fig. 3c/, b) as the left
squamosal. It is a large rectangular bone, shallowly concave along one edge, which is the border of the
tympanic notch, rather than the (unprecedented) orbit margin as suggested by Eaton and Stewart.
The unsculptured ventral surface is exposed and on it, next to the otic margin, is a branched crack, the
raised edges of which appear to have been figured as flange-like structures by Eaton and Stewart. Such
flanges are not certainly present unless cracks have formed along the edges of them so that the
structures which appear as raised edges are, in fact, low flanges. Flanges on the ventral surface of the
squamosal have been reported in several Palaeozoic temnospondyls including Edops (Romer and
Witter 1942), Dendrerpeton (Watson 1956, fig. 29 as Platystegos), Tersomius (Carroll 1964, fig. 4), and
an unnamed trematopid from Fort Sill (Bolt 19746, fig. 3). The squamosal only indicates the presence
of a large otic notch.

The right vomer (text-fig. 3) is a large irregularly shaped flat plate of bone bearing a covering of tiny
denticles and what appear to be the bases of two palatal fangs on the exposed ventral surface.
A shallow concavity along one edge may be interpreted as the border of the internal naris. A slightly
larger concavity on one of the shorter edges is probably the anterior border of the left interpterygoid
vacuity. Such a large, relatively wide, vomer bearing a fang-pair and bordering a large interpterygoid
vacuity is only consistent with belonging to a temnospondyl. Other early tetrapods either have
narrow vomers, fangless vomers, or no interpterygoid vacuities.

The left pterygoid (text-fig. 3) is also denticle covered on the exposed ventral face and is of the
characteristic triradiate form associated with the presence of large interpterygoid vacuities. The
quadrate and basipterygoid rami are clearly visible and the palatine ramus is partly obscured by
the superimposed postfrontal. There is also a distinct denticle-bearing posterodistal flange. The
combination of such a flange with large interpterygoid vacuities identifies the specimen as a
dissorophoid, either belonging to the Dissorophidae or the Trematopidae.

The above described elements are the only ones which could be identified as cranial ossifications.
I could not recognize the premaxillary, maxillary, parietal, supratemporal, or any of the occipital and
braincase ossifications identified by Eaton and Stewart. It would, of course, be futile to attempt a
reconstruction of the skull based on the five elements recognized in this work, but they include most of
the bones used in Eaton and Stewart’s palatal reconstruction and some of the significant components
of their skull roof reconstruction.

Mandible. The right mandible (text-fig. 3) is crushed and lightly twisted, being visible in dorsolateral
aspect at the back and ventrolateral aspect at the front. It bears temnospondyl-type sculpturing on
one large posterior element, partly visible as impression, which appears to me to be the angular rather
than the surangular as suggested by Eaton and Stewart. Parts of the mandible may be identified as the

pt pterygoid, rad - radius, spl — splenial, sq— squamosal, uln — ulna, vom -vomer.

MILNER: PENNSYLVANIAN AMPHIBIAN

773

splenial, dentary, and articular with varying degrees of confidence but the only general conclusion
that may be drawn from the mandible is that it probably belonged to a temnospondyl.

Postcranial skeleton. There is little to add to Eaton and Stewart’s description of the vertebrae. They
comprise neural arches, some with slightly elongate neural spines, U-shaped intercentra and large
paired pleurocentra. Some neural spines are distinctly lower than depicted by Eaton and Stewart, one
is visible in text-fig. 3 between the mandible and the pterygoid. The vertebrae appear to correspond to
the typical rhachitomous condition and are consistent with attribution to many Palaeozoic
temnospondyl families. Some short straight anterior thoracic ribs are present (text-fig. 4). As Eaton
and Stewart note, they are substantially expanded proximally, but this is a common feature of the
thoracic ribs of small labyrinthodonts.

Preparation of the slab has permitted both aspects of the preserved anterior appendicular skeleton
to be recognized (text-figs. 3 and 4). The clavicle, visible in both aspects, is narrow-bladed with a long
stem. The outer face of the blade is lightly striated. The narrow blade resembles those of presumed
terrestrial temnospondyls such as eryopids, trematopids, and dissorophids and also of microsaurs and
seymouriamorphs. Part of the scapulocoracoid is visible under the clavicle on one side (text-fig. 4).
A couple of striations are visible suggesting that part of the exposed bone is the scapular blade and the
rest is coracoid. On the other face of the specimen (text-fig. 3) the crushed coracoid is visible in ventral
aspect and bears a prominent glenoid fossa. Comparison of text-fig. 3 with figure 1 1 of Eaton and
Stewart shows that their ‘humerus’ is actually the base of the coracoid with a short humerus closely
appressed to it. They interpreted the coracoid as the ridged proximal region of an ‘unusually long
humerus’ and the glenoid surface was depicted as an unfinished cartilage ridge on the proximal end of
the humerus (Eaton and Stewart 1960, fig. 11).

The genuine humerus is a short ‘propellor-blade’ shaped structure with broad ends set at 90° to
each other, the proximal end being visible as a flat blade while the distal end is visible only in section
(text-figs. 3 and 4). This shape and relative size of humerus characterizes many terrestrial
temnospondyls such as Eryops, the trematopids, and some dissorophids. In apparently aquatic
temnospondyls such as the trimerorhachoids the humerus blades are poorly ossified and less
obviously rotated at 90° to one another. Other small temnospondyls such as Amphibamus and
Branchiosaurus have longer, more slender humeri (Carroll 1964; Boy 1972; Milner 1982). The humerus
of KUVP 10295 was probably incompletely ossified at the ends and may have been slightly longer.
Eaton and Stewart interpreted the entire humerus as the distal portion of a long humerus bearing a
slender hook-like ectepicondyle, a supposed relictual fish characteristic. The radius and ulna are close
to the humerus and to each other although not in articulation (text-figs. 3 and 4). They are of typical
temnospondyl type and I cannot see the pronounced distal expansion of the ulna which Eaton and
Stewart figure as homologous to a postaxial process. Several isolated metacarpals and phalanges are
present but are not in articulation and do not permit a reconstruction of the carpus or nranus. The
only unusually shaped manus bone depicted by Eaton and Stewart which does appear to be present is
the forked bone which they describe as an ulnare with a posterior hook-like expansion. A bone of
approximately this shape is present on one face of the specimen (text-fig. 4). In view of the otherwise
orthodox nature of the appendicular skeleton, I suspect that this ossification is made up of two or
three metacarpals or phalanges superimposed and crushed across each other. Apart from this I can see
nothing in the skeleton of the fore-limb which would justify the unique reconstruction of Eaton and
Stewart.

Systematic position. The visible determinate bones of KUYP 10295 are all consistent with
attribution to the Temnospondyli. The type of dermal sculpturing on the postfrontal, the shape of the
pterygoid and the vomer, and the structure of the vertebrae in particular support this attribution. The
postfrontal, squamosal, clavicle, and humerus together specify a temnospondyl with no lateral-line
pits around the interorbital region, an otic notch, a narrow clavicle, and a short, broad-ended
humerus showing pronounced torsion. Such characteristics are consistent with the specimen being
either an eryopid, a dissorophid, or a trematopid. The pterygoid with the posterodistal flange and the
postfrontal coming to an anterior point identify it as either a dissorophid or a trematopid, while the
lachrymal is most diagnostic, permitting the specimen to be identified as a trematopid. The only other

774

PALAEONTOLOGY, VOLUME 28

text-fig. 4. Indeterminate trematopid previously referred to Hesperoherpeton garnettense
by Eaton and Stewart 1960. KUVP 10295, part of reverse face of specimen showing
disarticulated axial and anterior appendicular elements. Abbreviations as for text-fig. 3
plus: r^rib, scap-cor— scapulocoracoid.

determinate amphibian described from Garnett is the trematopid Actiolmtes peabodyi (Eaton 1973)
and it is probable though not demonstrable that KUVP 10295 is a poor specimen of Actiobates. As
Actiobates has not been fully described or comparatively diagnosed against the Texas red-bed
trematopids such as Acheloma , there is no basis for assigning KUVP 10295 to any particular genus
other than by locality and horizon so it is proposed that the specimen be considered as Trematopidae
incertae sedis.

MILNER: PENNSYLVANIAN AMPHIBIAN

775

DISCUSSION

Systematic conclusions. Re-examination of the two specimens from Garnett confirms a view, which
has been implicit in the tendency of most workers to ignore Eaton and Stewart’s paper, that their
interpretation was not justified by the material and that there is no foundation for either the Order
Plesiopoda or the Family Hesperoherpetonidae. Their limb reconstruction is based on a mis-
interpretation of the crushed anterior appendicular elements and their skull reconstructions are a
frankly imaginative assembly of a series of isolated elements which can be interpreted in a far more
justifiable manner by comparison with contemporaneous forms. As noted in the introduction, the
nature of this specimen was immediately evident to most workers on Palaeozoic tetrapods and Eaton
and Stewart’s publication has been widely ignored. However because it was ignored rather than
explicitly criticized, Hesperoherpeton found its way into several general discussions on the origin and
early evolution of tetrapods, by authors who assumed that the absence of refutation meant that the
work was accepted. It is, perhaps, a reminder of the value of published critical discussion.

Amphibians in the Garnett tetrapod fauna. The Garnett assemblage is unique among Carboniferous
tetrapod assemblages in that it is largely composed of early amniotes, namely pelycosaurs and
diapsids. Some of these have been redescribed or newly described in recent years and the known
amniote fauna currently comprises the diapsid Petrolacosaurus (Reisz 1977, 1981) and four
pelycosaurs, namely the sphenacodont Haptodus (Currie 1977, 1979), the enigmatic long-spined
Xyrospondylus and undescribed material of Edaphosaurus, and a Clepsydrops- like ophiacodont (Reisz
et al. 1982). The type specimen of the trematopid Actiobates (Eaton 1973) and the two specimens
redescribed in this paper, are the only described amphibians from Garnett. Reisz et al. ( 1 982) reported
a further four amphibian specimens recently collected and yet to be described, while Peabody (1958,
p. 571) reported an earlier discovery of a series of nectridean vertebrae which were subsequently lost
before they could be described. Thus there may be other amphibians in the Garnett assemblage, but at
present, all the described material is either trematopid temnospondyl or consistent with attribution to
that family.

Reisz et al. ( 1 982) discussed the lithology of the productive horizons at Garnett and concluded that
it represents a terrestrial environment subject to progressive incursions by brackish water, implying
an estuarine mud-flat regime. It is interesting then, that the only determinate amphibians described to
date are trematopids, the peculiar naris of which has been interpreted as modified to incorporate a
chamber for an enlarged gland, possibly a salt gland (Bolt 1974a). Possession of a salt gland permits
tetrapods to be more tolerant of both xeric and saline environments and it is possible that trematopids
had such glands and were unusual among temnospondyls in their tolerance of brackish conditions.

Acknowledgements. I should like to thank Dr Larry Martin and Dr Hans-Peter Schultze of the University of
Kansas for their hospitality and for permission to examine this material. My research travel was funded by the
University of London Central Research Fund.

REFERENCES

Alexander, R. mcn. 1975. The Chordates , 480 pp. Cambridge University Press, London.
bolt, j. r. 1974a. Osteology, function and evolution of the trematopsid (Amphibia: Labyrinthodontia) nasal
region. Fieldiana: Geol. 33, 11-30.

— 1974b. A trematopsid skull from the Lower Permian, and analysis of some characters of the dissorophoid
(Amphibia: Labyrinthodontia) otic notch. Ibid. 30, 67-79.

boy, j. a. 1972. Die Branchiosaurier (Amphibia) des saarpfalzischen Rotliegenden (Perm, SW-Deutschland).
Abh. hess. Landesamt. Bodenforsch. 65, 1 137.

— 1981. Zur Anwendung der Hennigschen Methode in der Wirbeltierpalaontologie. Palaont. Z. 55, 87-107.
carroll, r. l. 1964. Early evolution of the dissorophid amphibians. Bull. Mus. comp. Zool. Harv. 131, 163-250.

— and winer, l. 1977. Privately circulated appendix to carroll, r. l. Patterns of amphibian evolution: an
extended example of the incompleteness of the fossil record. Chapter 13 in hallam, a. (ed.). Patterns of
Evolution. Elsevier, Amsterdam.

776

PALAEONTOLOGY, VOLUME 28

chase, J. N. 1965. Neldasaurus wrightae, a new rhachitomous labyrinthodont from the Texas Lower Permian.
Bull. Mus. comp. Zool. Harv. 133 , 153-225.

currie, p. j. 1977. A new haptodontine sphenacodont (Reptilia: Pelycosauria) from the Upper Pennsylvanian
of North America. J. Paleont. 51 , 927-942.

— 1979. The osteology of haptodontine sphenacodonts (Reptilia: Pelycosauria). Palaeontographica Abt. A.
163, 130-168.

daly, E. 1973. A Lower Permian vertebrate fauna from southern Oklahoma. J . Paleont. 47, 562-589.
demar, r. e. 1966. Longiscitula houghae, a new genus of dissorophid amphibian from the Permian of Texas.
Fieldiana: Geol. 16, 45-53.

— 1968. The Permian labyrinthodont amphiban Dissorophus multicinctus and the adaptations and phylogeny
of the family Dissorophidae. J. Paleont. 42, 1210-1242.

eaton, T. H. jr. 1951. Origin of tetrapod limbs. Amer. Midi. Nat. 46, 245-251.

1960. The aquatic origin of tetrapods. Trans. Kans. Acad. Sci. 63, 115-120.

— 1973. A Pennsylvanian dissorophid amphibian from Kansas. Occas. Pap. Mus. nat. Hist. Univ. Kans. 14 ,
1 - 8 .

— and stewart, p. l. 1960. A new order of fishlike Amphibia from the Pennsylvanian of Kansas. Univ. Kans.
Pubis Mus. nat. Hist. 12, 217-240.

holmes, r. 1980. Proterogyrinus scheelei and the early evolution of the labyrinthodont pectoral limb, 351-376. In
panchen, a. l. (ed.). The terrestrial environment and the origin of land vertebrates, 633 pp. Systematics Assn
Spec. Vol. No. 15. Academic Press, London.

kuhn, o. 1972. Seymourida, 20-69. In carroll, r. l., kuhn, o. and tatarinov, l. p. Teil 5B Batrachosauria
(Anthracosauria), Gephyrostegida-Chroniosuchida. Handbuch der Palaoherpetologie. Fischer, Stuttgart.
milner, a. r. 1982. Small temnospondyl amphibians from the Middle Pennsylvanian of Illinois. Palaeontology,
25, 635-664.

moulton, j. m. 1974. A description of the vertebral column of Eryops based on the notes and drawings of
A. S. Romer. Breviora, No. 428, 1 44.
olson, e. c. 1941. The family Trematopsidae. J. Geol. 49, 149-176.

1971. Vertebrate Paleozoology, 839 pp. Wiley Interscience, New York.

peabody, F. E. 1958. An embolomerous amphibian in the Garnett fauna (Pennsylvanian) of Kansas. J. Paleont.
32, 571-573.

porter, k. r. 1972. Herpetology, 524 pp. Saunders, Philadelphia.

reisz, R. R. 1977. Petrolacosaurus kansensis Lane, the oldest known diapsid reptile. Science, 196, 1091-1093.

— 1981. A diapsid reptile from the Pennsylvanian of Kansas. Spec. Publ. Mus. nat. Hist. Univ. Kans. 7, 1-74.

— heaton, m. j. and pynn, b. r. 1982. Vertebrate fauna of late Pennsylvanian Rock Lake Shale near
Garnett, Kansas: Pelycosauria. J. Paleont. 56, 741-750.

romer, a. s. 1966. Vertebrate Paleontology (3rd edn), 468 pp. Chicago University Press, Chicago.

— 1968. Notes and Comments on Vertebrate Paleontology, 304 pp. Chicago University Press, Chicago.

— and witter, r. v. 1942. Edops, a primitive rhachitomous amphibian from the Texas red beds. J. Geol.
50, 925-960.

rosen, d. e., forey, p. L., Gardiner, b. G. and patterson, c. 1981. Lungfishes, tetrapods, paleontology and
plesiomorphy. Bull. Am. Mus. nat. Hist. 167, 159-267.

schmalhausen, i. I. 1968. The Origin of Terrestrial Vertebrates, 314 pp. Academic Press, New York and London.
Thomson, k. s. and bossy, k. h. 1970. Adaptive trends and relationships in early Amphibia. Forma et Functio, 3,
7-31.

tverdochlebova, g. i. 1972. A new batrachosaurian genus from the Upper Permian of southern Cisuralia.
Paleontol. J. 6, 84-90.

vorob'yeva, e. i. 1974. On the formation of tetrapod characters in crossopterygians. Ibid. 8, 219-224.
watson, d. m. s. 1956. The brachyopid labyrinthodonts. Bull. Br. Mus. nat. Hist. (Geol.), 2, 315-392.

ANDREW R. MILNER

Department of Biology
Birkbeck College
London WC1E 7HX

Typescript received 25 October 1984

HOMOTAXY AND BIOSTRATIGRAPHICAL

THEORY

by G. H. SCOTT

Abstract. The principal problem in biostratigraphical theory is a justification for using fossils to identify
isochronous horizons. Biostratigraphers establish the sequential order of fossil events but there is no theoretical
justification for equating constancy in stratigraphical position (homotaxy) with constancy in time of deposition.
The problem is important as the way in which it is solved greatly influences the direction of biostratigraphical
research. A partial solution is to use homotaxy as a weak test for diachroneity. Events that do not maintain
invariant stratigraphical order are regarded as diachronous. While surviving events are not shown to be
isochronous the amount of potential diachroneity throughout their individual distributions is inversely related
to their stratigraphical spacing. The closer are homotaxial events in stratigraphical space the less they have
wandered in time. Thus a major objective of biostratigraphical research should be to make tests of homotaxis
more rigorous by raising the density of events.

Since William Smith the principal role of biostratigraphers has been to provide classifications of
strata useful for estimating time of deposition. Historical geology requires a chronological framework
to study ancient geographies and the sequential classifications established with fossils have long been
used for this purpose. Indeed, the advent of radiometric and palaeomagnetic chronologies, in which
time is directly estimated, has not diminished the utility of biostratigraphical data and there are now
interesting applications of fossils as surrogate criteria for datum planes determined by quantitative
techniques for measuring age. Cenozoic biostratigraphy, for example, has progressed to the stage
where appearances and extinctions of taxa, calibrated at only a few sites with radiometric or
magnetostratigraphic chronologies, are used as datum planes expressed in years (e.g. Ryan et al. 1 974;
Poore et al. 1984) that are recognized at locations far distant from the sites of calibration. Resolutions
of less than 10000 years have been claimed for such datum planes (Thierstein et al. 1977; Berggren
and van Couvering 1978). To the bystander, these applications suggest that biostratigraphy has come
of age and that the isochronous surfaces envisaged by the International Stratigraphic Guide (Hedberg
1976) can be accurately estimated by fossils.

While utility itself may be a justification, there is little theory to underwrite biostratigraphical
practice. Further, over a long period there have been denials that fossils can accurately estimate time.
A trenchant early critic was Huxley (1862) who argued that occurrences of taxa in fixed order from
section to section (the common feature of biostratigraphical classifications since William Smith) did
not imply that each datum or unit was isochronous. As correspondence in position did not imply
contemporaneity, he alleged by way of demonstration that, Tor anything that geology or
palaeontology are able to show to the contrary, a Devonian fauna and flora in the British Islands may
have been contemporaneous with Silurian life in North America, and with a Carboniferous fauna and
flora in Africa’ (Huxley 1862, p. xliv). Modern critics (e.g. Miller 1965; Drooger 1974; Watson 1983),
while less dismissive than Huxley, have continued to wrestle with the problems of using fossils to
define isochronous surfaces. Kitts (1966) made a detailed investigation of the problem, viewing the
events of the biostratigrapher as biological signals which were transmitted from specific centres.
Because biological signal velocities (rates of dispersal) are variable and cannot be determined
intrinsically from the data, he concluded that ‘simultaneous with’ relationships cannot be derived
from fossils.

The impasse in biostratigraphical theory is that the intrinsic palaeobiological and stratigraphical
data for an individual event do not enable its isochroneity to be established. There is no rationale for

[Palaeontology, Vol. 28, Part 4, 1985, pp. 777-782.|

778

PALAEONTOLOGY, VOLUME 28

progressing from event ordering to event dating. The International Stratigraphic Guide states (p. 88)
that ‘biostratigraphic correlation is one of the most useful approaches to time correlation if used with
discretion and judgment’, repeating an earlier view (p. 63) that correlation is a matter of judgement. It
also mentions ‘subtle paleontological discrimination’ (p. 88) but in none of the discussions are clear
guidance or objective rules given for identifying the most reliable datum planes. The idealistic way in
which the Guide equates fossil datum planes with isochronous surfaces is echoed in some texts.
Krumbein and Sloss (1963, p. 370) wrote that ‘Chronospecies and “chronogenera” include the most
obvious index fossils, since the synchronism of their range zones may be established without serious
doubt’. In like vein, Donovan (1966, p. 32) found it ‘inconceivable, in view of what we now know of
evolution and dispersal, that a long and complex series of faunal changes should occur at different
times in different places’. Certainly, there are grounds for making such assertions but where is the
theory that allows them to be tested with biostratigraphical evidence? Other authors (e.g. Dunbar
and Rodgers 1957; Weller 1960; Raup and Stanley 1978) focus more on the imperfections of
palaeobiological events as time indices. Yet these ‘realists’ also fail to confront the problem: how to
identify the best events for time correlation?

In overview, there is no theory to allow a set of ordered events to be tested for isochroneity with
biostratigraphical data. The problem has existed since the advent of biostratigraphy and is ignored or
glossed over in the vast majority of the biostratigraphical literature. I believe that this is unsatisfactory
as the direction of biostratigraphical research is intimately related to the way in which the gap
between sequence and time is to be bridged. Here I outline one approach which, rather than trying to
find isochronous events directly, seeks to identify the least diachronous.

A PARTIAL SOLUTION

Perhaps ironically for Huxley, a basis for biostratigraphical theory lies in his concept of homotaxis.
To Huxley (1862) it signified similarity in orders of stratigraphical events without implication of
contemporaneity, and he used it with some relish to demolish time stratigraphical correlation with
fossils. Huxley was correct in contending that an order of events, maintained from section to section,
does not entail that each was isochronous throughout its occurrences. Nevertheless, he did not
explore some implications of his concept. Let us simply define as homotaxial those events that occur
in fixed stratigraphical order with respect to selected neighbouring events. As graphical representa-
tions show, order invariance does not necessarily identify an event as isochronous. Homotaxial events
potentially include strictly isochronous events (e.g. those caused by instantaneous global cata-
strophes) as well as diachronous events that are insufficiently time transgressive to intersect with their
immediate neighbours. Conversely, non-homotaxial events include those diachronous events that are
time transgressive to the extent that they produce inversions of order. Relative to the time
stratigraphical objective, non-homotaxial events can be rejected as certainly unsuitable. Thus the
problem focuses on homotaxial events; these may vary considerably in the amount of undetected
diachroneity. Huxley denied that this class of event was applicable in time stratigraphy. This is
substantially correct given the biogeographical model he implied in which, over a long period
(Silurian to Carboniferous), taxa originated in one region and slowly migrated in the same direction to
others. But in any model in which events may arise at various geographical loci (text-fig. 1) and
disperse at different rates in different directions, homotaxial events will include those that are
potentially the most accurate estimators of time. The operational task is to identify the least
diachronous.

While tests using various stratigraphical, physical, or chemical techniques are sometimes feasible, it
is important from the standpoint of biostratigraphical theory to identify procedures that simply
depend on stratigraphical relationships (the intrinsic data of biostratigraphy).

The primary procedure relates to the stratigraphical spacing of homotaxial events and tries to force
them into the class of non-homotaxial events. Text-fig. 1 shows a homotaxial triplet (Events 1, 2, 3).
Events 1 and 3 are well behaved in the time domain as they dispersed rapidly. Event 2 did not, but it
will still be regarded as a homotaxial event as it does not intersect with the adjacent Events 1 and 3.

SCOTT: HOMOTAXY AND BIOSTRATIGRAPHICAL THEORY

779

EVENT SPACING AND HOMOTAXY

GEOGRAPHIC DISTRIBUTION *-

text-fig. 1. Initially, only biostratigraphical Events 1, 2, 3 are known. They maintain homotaxy throughout their
joint occurrences. Event 4 is recognized subsequently and it is found that the upward sequence 1 -4-2-3 is
maintained over most of the region in which they occur jointly. However, in the vicinity of A the order of Events 4
and 2 inverts. Evidence from Event 5 helps establish that Event 2 is strongly diachronous. The slow dispersal of
Event 2 was unrecognized when the event set consisted only of Events 1, 2, 3. Decreasing the stratigraphical
spacing of events will raise the chance of detecting the most diachronous events provided that the origins of
events and their dispersal directions and rates are variable.

Consideration of an additional event (4, perhaps newly recognized) shows that while it maintains
homotaxy with Events 1 and 3 it fails with Event 2. At this stage which of Events 2 and 4 is the more
diachronous is unknown and further events, such as 5, may be needed to reject Event 2.

Interpolation of additional events in homotaxial sets should tend to eliminate the most
diachronous. Thus attempts to falsify the hypothesis that a set of events is homotaxial become
increasingly rigorous as event spacing declines. As testing relies only on hypotheses about event
orders, there is no indication of the variation in age of events that survive testing. They are not shown
to be isochronous, neither is their diachroneity definitely established. Nevertheless, for particular
sedimentary regimes, those that are closest in stratigraphical space may be expected to best
approximate isochronous horizons.

The signal model (Kitts 1966) provides another perspective on selection of events. Here, events are
considered to be transmitted away from a local geographical origin. Speciation by allopatry is an
excellent example but some extinctions (such as those caused by an environmental change moving
along a geographical gradient, progressively eliminating populations) also conform. The significance
of geographical extent is that it provides a test of signal velocity. Any event whose transmission time is
slower than those of its neighbours is liable not to maintain its order in the sequence (Event 2 in
text-fig. 1). Whether it does so depends on several factors, including its proximity (in time) to the

780

PALAEONTOLOGY, VOLUME 28

neighbouring event with faster velocity, and on variation in its own velocity. This may be
considerable. For example, zooplankton populations tend to be watermass bounded (McGowan
1971 ). Within one watermass, biological events are rapidly distributed by physical circulation. But the
spread of a taxon to a neighbouring watermass is often a trial and error process. Thus some signals
(e.g. first appearances) may terminate near the boundary (possibly indicating that founder
populations in the adjacent watermass failed to establish). On the other hand, a successful invasion
may lead to rapid expansion over the whole of the new territory.

While estimates of ancient signal velocities are speculative, it is obvious that variation in signal
velocity is the more likely to be revealed the greater the area over which homotaxis is checked. It is not
simply the global taxa that are the least diachronous, it is those that dispersed most rapidly
throughout their entire geographical range. The effectiveness of the geographical test is related to
event spacing. If events are widely spaced, velocities are only weakly testable via consideration of
geographical distributions. Rather, its effectiveness increases as event spacing decreases. Generally,
events that maintain homotaxial order over the widest area at the minimum available stratigraphical
spacing are likely to be the least diachronous.

DISCUSSION

Given suitable distributions of event origins and dispersal routes, tests of homotaxy are potentially
capable of eliminating all but strictly isochronous events. In this respect the argument simply supports
the utility of sequence classifications long used by biostratigraphers in their quest for time.
Additionally, however, the formulation raises issues in contemporary theory and practice that would
otherwise not be clearly perceived.

1. Testability. Like Donovan (1966), we may appeal to evolutionary theory to ‘guarantee’ that
certain events are isochronous or, like the International Stratigraphic Guide, we may cloak our
selections under the shrouds of ‘discretion’ and ‘judgment’. The merit of building a theory around
homotaxy is that it emphasizes practical testing procedures. Biostratigraphy is not an art, despite the
writings of the International Stratigraphic Guide.

2. Integration of data. Events used in biostratigraphy come from sources as diverse as protistans
and vertebrates. However, the potential value of this diversity is rarely realized in biostratigraphical
applications. The systematics of groups well represented in the Cenozoic record, for example, have
recently been greatly refined, often with large increases in the number of useful biostratigraphical
events recognized. Yet this research has been paralleled by the multiplication of biostratigraphical
classifications which use only events in one group (e.g. Martini 1971; Riedel and Sanfilippo 1978; Blow
1979). While all represent major advances, no single classification necessarily uses events that are the
least diachronous. In general, this will be the integrated set as it will contain the most closely spaced
events.

Integration implies more than finding the stratigraphical position of diatom species C relative to
coccolith species H in one or two sequences. To fully utilize advances in the systematics of Cenozoic
planktonic microfossils the positions of individual events need to be compared with those of their
nearest stratigraphical neighbours throughout regions of joint occurrences. While biogeographical
compatibility largely controls the severity of testing, any reduction in event spacing while maintaining
homotaxy enhances the value of the datum planes for time stratigraphy. There is an increasing effort
to correlate events in various planktonic microfossil groups (e.g. Hornibrook and Edwards 1971;
Ikebe and Chiji 1981; Abbott 1984) but is testing really rigorous?

3. Stratotypes. The International Stratigraphic Guide vigorously promoted the function of
stratotypes and type localities as standards for the definition and recognition of stratigraphical units.
It views a stratotype (p. 27) as ‘the standard of reference on which the concept of the unit is uniquely
based’. Thus biostratigraphical units (p. 63) are ‘extended away from their type localities by
biostratigraphic correlation’. Similarly, the boundaries of stages (chronostratigraphical unit, p. 71)
‘as they are extended away from the boundary stratotypes should be in principle isochronous’.
Further, the Guide states (p. 86) that 'Only after the type limits (boundary stratotypes) of a chrono-

SCOTT: HOMOTAXY AND BIOSTRATIGRAPHICAL THEORY

781

stratigraphic unit have been established can the limits be extended geographically beyond the
type section’.

These views conflict with the theory advocated here. While it is valuable to have a locality that
serves as a name bearer for a datum or unit, homotaxial theory does not require any locality to serve
as a standard endowed with special status. Homotaxis is recognizable only by occurrences of events in
the same order in several sequences. The order of events in one section without reference to the order
of the same events in other sequences is completely irrelevant in a homotaxial scheme. Rather than
initially identifying a standard and correlating outward from that site, as the Guide suggests, a
homotaxial datum, or unit, can only be recognized after inspection of events in several sequences.
From a different viewpoint Hay (1974) reached a similar conclusion. The equal status of sequences in
homotaxial schemes is implied in the test procedure for order invariance. Any sequence in which the
order of an event is not maintained becomes crucial in determining the status of the event, irrespective
of its relationships in the stratotype.

I suggest that methodological emphasis on stratotypes is misplaced. Biostratigraphy is built on
selection of events, not selection of sections.

Biostratigraphers have long advocated the utility of fossils to identify synchronous horizons and
independent methods of dating are now establishing that some palaeobiological events in fact
dispersed extremely rapidly. Such results, however encouraging, do not resolve the problem of
identifying synchronous horizons when only data on the order of events are available. While there
may not be a full solution to this problem the value of my formulation is that it points to a coherent
approach to event selection and identifies research objectives that should improve biostratigraphical
resolution. Rightly, some will say that the procedures are commonplace, used by all biostratigraphers.
Nevertheless, the way they may be used to bridge the gap between order of events and age of events is
insufficiently recognized either in practice or in the literature.

Acknowledgements. I thank Professor R. M. Carter and colleagues at N.Z. Geological Survey for many
constructive comments.

REFERENCES

Abbott, w. h. 1984. Progress in the recognition of Neogene diatom datums along the U.S. Atlantic coast.

Palaeogeography, Palaeoclimatology, Palaeoecology , 47, 5-20.
berggren, w. a. and van couvering, J. a. 1978. Biochronology, 39-55. In cohee, g. v., glaessner, m. f. and
hedberg, H. D. (eds.). The Geologic Time Scale. American Association of Petroleum Geologists, Tulsa,
Oklahoma.

blow, w. H. 1979. The Cainozoic Globigerinida , xv+1413 pp. E. J. Brill, Leiden.
donovan, D. T. 1966. Stratigraphy , 199 pp. T. Murby, London.

drooger, c. w. 1974. The boundaries and limits of stratigraphy. Proc. K. ned. Akad. Wet. B77, 159-176.
dunbar, c. o. and rodgers, J. 1957. Principles of Stratigraphy, xii + 356 pp. J. Wiley, New York.
hay, w. w. 1974. Implications of probabilistic stratigraphy for chronostratigraphy. Verb, naturf. Ges. Basel.
84, 164-171.

hedberg, H. D. (ed.). 1976. International Stratigraphic Guide, xvi + 200 pp. J. Wiley, New York.
hornibrook, n. de b. and Edwards, a. r. 1971. Integrated planktonic foraminiferal and calcareous nanno-
plankton datum levels in the New Zealand Cenozoic, 649-657. In farinacci, a. (ed.). Proceedings of the Second
Planktonic Conference, Roma 1970. Vol. 1. Edizioni Tecnoscienza, Rome.
huxley, T. h. 1862. The anniversary address. Q. Jl geol. Soc. Lond. 18, xl-liv.

ikebe, n. and chiji, M. 1981. Important datum-planes of the western Pacific Neogene (revised) with remarks on
the Neogene stages in Japan. In tsuchi, r. (ed.). Neogene of Japan. IGCP-1 14 National Working Group of
Japan, Shizuoka University, 1-14.
kitts, d. b. 1966. Geologic time. J. Geol. 74, 127-146.

krumbein, w. c. and sloss, l. l. 1963. Stratigraphy and Sedimentation, vii + 600 pp. W. H. Freeman, San
Francisco.

martini, e. 1971. Standard Tertiary and Quaternary calcareous nannoplankton zonation, 739-785. In
farinacci, A. (ed.). Proceedings of the Second Planktonic Conference, Roma 1970. Vol. 2. Edizioni Tecno-
scienza, Rome.

782

PALAEONTOLOGY, VOLUME 28

mcgowan, j. a. 1971. Oceanic Biogeography of the Pacific, 3-73. In funnell, b. m. and riedel, w. r. (eds.). The
Micropalaeontology of Oceans. University Press, Cambridge.
miller, t. G. 1965. Time in stratigraphy. Palaeontology, 8, 113-131.

POORE, R. Z., TAUXE, L., PERCIVAL, S. F. JR., LABREQUE, J. L., WRIGHT, R., PETERSEN, N. P., SMITH, C. C., TUCKER, P.

and hsu, k. j. 1984. Late Cretaceous-Cenozoic magnetostratigraphic and biostratigraphic correlations for
the South Atlantic Ocean, Deep Sea Drilling Project Leg 73, 645-655. In hsu k. j. et ah, Initial Reports of the
Deep Sea Drilling Project. Vol. 73. National Science Foundation, Washington, D. C.
raup, d. M. and Stanley, s. M. 1978. Principles of Paleontology, x + 481 pp. W. H. Freeman, San Francisco.
riedel, w. r. and sanfilippo, a. 1978. Stratigraphy and evolution of tropical Cenozoic radiolarians. Micro-
paleontology, 24 , 61-96.

ryan, w. b. f., cita, M. b., rawson, m. d., burckle, L. H. and saito, T. 1974. A paleomagnetic assignment of
Neogene stage boundaries and the development of isochronous datum planes between the Mediterranean,
the Pacific and Indian Oceans in order to investigate the response of the world ocean to the Mediterranean
salinity crisis. Riv. ital. Paleont. Stratigr. 80 , 631-687.

thierstein, h. r., geitzenauer, k. r., molfino, b. and shackleton, n. j. 1977. Global synchroneity of late
Quaternary coccolith datum levels: validation by oxygen isotopes. Geology, 5. 400-404.
watson, r. a. 1983. A critique chronostratigraphy. Am. J. Sci. 283, 173-177.
weller, j. m. 1960. Stratigraphic Principles and Practice, xvi + 725 pp. Harper, New York.

G. H. SCOTT

Typescript received 30 November 1984
Revised typescript received 4 April 1985

N.Z. Geological Survey
P.O. Box 30368
Lower Hutt, New Zealand

THE TYPE SPECIES OF CALYMENE (TRILOBITA)
FROM THE SILURIAN OF DUDLEY, ENGLAND

by DEREK J. SIVETER

Abstract. The type material of Calymene blumenbachii Brongniart in Desmarest 1817 (type species of Calymene
Brongniart, 1822) and C. tuber culata (Briinnich, 1781) has been revised, with that of the latter being figured for
the first time. The nomen dubium status accorded to C. tuber culata by Shirley (1933) can no longer be upheld as the
syntype specimens of Silurian age are, on preparation, very well preserved and identifiable, though this species is
now recognized as a senior (subjective) synonym of C. blumenbachii. In consequence application is being made to
the International Commission on Zoological Nomenclature to retain the junior name in the interest of stability.

During recent work on the Calymenidae in preparation for the second edition of the Treatise on
Invertebrate Paleontology ( Trilobita ), nomenclatural problems have resurfaced regarding the
conservation of the generic name Calymene Brongniart in Brongniart and Desmarest, 1822, type
genus for the family, and the right of Milne Edwards (1840) to stand as the author of the family name.
These uncertainties prompted Whittington (1983) to petition the International Commission on
Zoological Nomenclature (ICZN) in favour of both claims. Subsequent study by the present author
has further shown that C. blumenbachii Brongniart in Desmarest, 1817, the type species of Calymene , is
a junior (subjective) synonym of Trilobus tuberculatus Briinnich, 1781. The purpose of this paper is to
give evidence of this synonymy by providing new data on the type material of both species.
Additionally, the paper is a prerequisite to an application to the ICZN (Whittington and Siveter, in
press), by way of a rider to Whittington’s 1983 proposals, to use its plenary powers to suppress the
senior name tuberculata in favour of blumenbachii, on the grounds that only the latter name has been
used for the species by various authors during the last fifty years.

The terminology and photographic and preparation techniques are those previously employed by
Siveter (1977, 1979, 1980, 1983). Repositories holding material used in this revision are: Universite
P. et M. Curie, Laboratoire de Paleontologie des Invertebres, Paris (UPMC); Geologisk Museum,
Kobenhavns Universitet (MGUH).

THE TYPE SPECIES OF CALYMENE
The type material of Calymene blumenbachii Brongniart in Desmarest, 1817

No material of C. blumenbachii was figured by Desmarest (1817) but Brongniart in Brongniart and
Desmarest (1822) illustrated two syntype specimens, one (ibid., pi. 1, fig. 1a, b) a partially enrolled
example, the other (ibid., pi. 1, fig. lc, d) an extended incomplete individual; Dudley was indicated
(ibid., pp. 9, 1 1, and 143) as the type locality (and I can confirm that this name appears on the specimen
label accompanying the enrolled form). Shirley (1933) relocated this material in the Sorbonne, Paris,
together with a third specimen ‘which probably belonged to Brongniart’ but which Shirley left out of
his account because of its uncertain curatorial history. Shirley also selected the enrolled syntype to
stand as ‘the type’ (that is lectotype) for the species, and figured it (ibid., pi. 1, figs. 1 -3), though not the
second (now paralectotype) specimen. Since this time the type suite of specimens has remained
unstudied.

Dr Francoise Bigey (Universite P. et M. Curie, Laboratoire de Paleontologie des Invertebres,
Paris) subsequently rediscovered the lectotype in the collections of her institute, the Sorbonne

[Palaeontology, Vol. 28, Part 4, 1985, pp. 783-792, pis. 90-91. |

784

PALAEONTOLOGY, VOLUME 28

Geology Collection having been ‘removed to new buildings’ (pers. comm. 5 February 1974); I have
since fully prepared this specimen and it is refigured herein (PI. 90, figs. 1-5), but the paralectotype
was reported by Dr Bigey as missing. Further, Brongniart’s material has ‘been shared in the
past between the Museum National d’Histoire Naturelle and Paris University’ (pers. comm.
Dr Bigey, 23 November 1983), though Dr J.-C. Fischer, sub-director of this museum, informs me
(1 September 1984) that the paralectotype is not in their collections, and it does not appear in the
recently published catalogue of type and figured trilobites housed there (Carriol 1984). For the
present it is missing.

The type material o/Trilobus tuberculatus Briinnich, 1781

Four specimens comprised the syntype suite of this species, the type locality of which was indicated by
Briinnich (1781, p. 391) as ‘Rensnest’ (= Wren’s Nest), ‘Dudley’. Shirley (1933, p. 63) examined this
material of C. tuberculata and concluded that the specimens which he referred to as the ‘third’ and
‘fourth’ were, respectively, the pygidium of an undescribed Ordovician calymenid from Shropshire
and an almost complete asaphid from the Ordovician of Scandinavia or the eastern Baltic. He claimed
that the ‘first’ specimen, ‘a cranidium more than half buried in matrix and six attached thoracic
segments ... is impossible to refer to any of the species . . . occurring at Dudley’ and that the ‘second’
specimen, ‘twelve thoracic segments and an attached pygidium . . . cannot be identified with certainty’.
Shirley did not name the ‘first’ or ‘second’ specimens in his paper but in a letter (23 June 1933) to Dr
J. P. J. Ravn, the then Curator of the Geologisk Museum, Kobenhavns Universitet (where the
material is housed) he writes: ‘I cannot identify specimens one and two further than Calymene sp.
(at Dudley I have recognised at least four species). They have a matrix which is similar to other
specimens from Dudley and there seems to be no doubt that they came from that locality.’

Dr S. Floris, the present Curator, has loaned me (18 January 1984) the first two (MGUH 16.805,
16.806) and fourth (MGUH 16.807) of the syntypes, the third one (Shropshire Ordovician calymenid)
being reported as missing. The ‘Rensnest’ locality information given by Briinnich in his paper is also
written on the original labels of all these three syntypes, including that of the Ordovician asaphid.
However a Wren’s Nest origin is corroborated for at least one of the two syntypes which are of
Silurian age, that referred to by Shirley as the ‘first’ specimen (MGUH 16.805) and which is selected
below as the lectotype for the species, by evidence from ostracodes recovered from the surrounding
matrix (PI. 90, figs. 6 and 7). Dr David Siveter has kindly supplied the following comments on these:
‘The ostracodes are an almost complete female right valve of the beyrichiacean Osmotoxotis phalacra
Siveter, 1980, a left valve of the thlipsuracean Thlipsura v-scripta Jones and Holl, 1869, and two other
indeterminate non-palaeocope specimens. T. v-scripta is known from the Homerian Stage and
possibly the upper part of the Sheinwoodian Stage, Wenlock Series to the lower Gorstian Stage,
Ludlow Series of the West Midlands and the Welsh Borderland, and from at least the Wenlock Series
on Gotland (Siveter 1978). O. phalacra is unknown outside the Homerian Much Wenlock Limestone
Formation of the West Midlands and the Welsh Borderland; it has been recorded (Siveter 1980) from
Mayhill, the Malverns, Ledbury, and the type and other localities at the Wren’s Nest. The figured
O. phalacra specimen (PI. 90, fig. 6) is only the fourth female of the species known from the Wren’s Nest
and it shows exceptional preservation, having a sculpture of fine reticulo-striation and sparse
tubercles not previously observed in females from this locality.’ Ostracodes have not been recovered
from the matrix around the ‘second’ specimen of Shirley but there is no reason to doubt that it, too,
came from the Wren’s Nest.

Photographs of the two Dudley syntype trilobites were taken before (PI. 91, figs. 7, 10, 12) and after
(PI. 91, figs. 1-6, 8, 9, 11, 13) preparation, so that the nature of the specimens as seen by Shirley and
Briinnich is on record; also casts were made of the specimens before they were prepared. Both
syntypes were found to have one more thoracic segment, that is seven (PI. 91, fig. 9) and thirteen
(PI. 91, fig. 1), than the number given by Shirley. In order to fix the concept of Trilobus tuberculatus
in keeping with the previously held notion of this species as a Silurian Calymene from Dudley,
the specimen illustrated herein on Plate 91, figs. 3, 6, 8, 9, 11-13 is now selected as lectotype
(MGUH 16.805).

SIVETER: SILURIAN TRILOBITE

785

The question of synonymy

As Shirley (1933) noted, on priority grounds many Scandinavian and continental authors (e.g.
Lindstrom 1885) used the senior name C. tuherculata rather than C. blumenbachii for late Wenlock
calymenids of this type. Shirley nevertheless effectively declared C. tuherculata a nomen dubium
because of ‘inadequate description and types’, and said that ‘the name tuherculata for any species of
Calymene must be allowed to fall’. After preparing the Silurian material of Briinnich I am convinced
that there are no significant morphological differences from that of Brongniart and that they are
conspecific (see discussion below and Pis. 90 and 91). All the material is from the Dudley area, though
it is unknown in the case of C. blumenbachii whether its type locality is specifically the Wren’s Nest, as
for C. tuherculata, there being two other Silurian inliers (Dudley Castle Hill and Hurst Hill; Butler
1939) in the district. The type horizon for all the material is indisputably the Much Wenlock
Limestone Formation, of late Homerian age, considering the ostracode evidence presented above and
the fact that I have not noted any specimens of this species from Dudley, of which there are very many,
in any other formation.

The type species

The type species of Calymene Brongniart in Brongniart and Desmarest, 1822 is C. blumenbachii
Brongniart in Desmarest, 1817, by subsequent designation of Shirley (1933), and not, as stated by
Whittington (1983), Calymena blumenbachii Brongniart in Desmarest, 1817 by original designation.
Bassler’s (1915) designation of C. tuherculata to stand as the type species pre-dates that of Shirley, but
it is invalid (see Whittington and Siveter, in press).

SYSTEMATIC PALAEONTOLOGY

Family calymenidae Milne Edwards, 1840
Genus calymene Brongniart, 1822

Type species. Calymene blumenbachii Brongniart in Desmarest, 1817, from the Much Wenlock Limestone
Formation, Homerian Stage, Wenlock Series, Dudley, West Midlands, UK, by subsequent designation of Shirley
1933, p. 53.

71750

71750

1781

1816

1817

1822

71839

1851

71852

71859

1865

1868
1869
non 1872
non 1879

Calymene blumenbachii Brongniart, 1817
Plate 90, figs. 1-5; Plate 91

nondescript petrified insect; Lyttelton, p. 105, pi. 1, figs. 9-14; pi. 2 {pars). [From the figures it
is probably C. blumenbachii . ]

scolopendrae aquaticae scutatae', Mortimer,p. 106, pi. l,figs. 15-18. [From the figures it is possibly
C. blumenbachii .]

Trilobus tuberculatus Briinnich, p. 389. [Suppression requested, Whittington and Siveter, in press;
ICZN pending.]

Le Calymene de Blumenbach; Brongniart in Desmarest, p. 50.

Calymena blumenbachii, Brong.; Brongniart in Desmarest, p. 517. [Suppression of Calymena
Desmarest, 1817 requested, Whittington 1983; ICZN pending.]

Calymene blumenbachii ; Brongniart in Brongniart and Desmarest, p. II, pi. 1, fig. Ia-d.
Calymene blumenbachii Brongniart; Murchison, p. 653, pi. 7, figs. 6 and 7, non fig. 5.

Calymene blumenbachi (Brong.); M‘Coy (pars) in Sedgwick and M‘Coy, p. 165.

Calym. blumenbachi. Brongn.; Barrande, p. 566, pi. 5, fig. 8; non pi. 19, fig. 10; non pi. 43, figs. 46-48.
Calymene blumenbachii Brongniart; Murchison, p. 235, pi. 18, fig. 10.

Calymene blumenbachii, Brongn.; Salter, p. 93, pi. 8, figs. 8, 10, 12-14, ?figs. 7, 15, 16, non figs. 9 and
1 1, pi. 9, figs. 1 and 2.

Calymene ceratophthalma', Woodward, p. 489, pi. 21, fig. 1 (pars), ?fig. 2.

Calymene blumenbachii'. Woodward, p. 43.

Calym. blumenbachi. Brongn.; Barrande, p. 36, pi. 14, fig. 33.

Calymene blumenbachii, Brongniart; Nicholson and Etheridge, p. 140, pi. 10, figs. 2-6.

786

PALAEONTOLOGY, VOLUME 28

non 1906
1933
1936
non 1957

1959
1970
non 1977
71980
1980

1983

1984

Calymene blumenbachi , Brongniart, 1822; Reed, p. 133, pi. 17, figs. 12 and 13.

Calymene blumenbachi Brongniart, 1822; Shirley, pp. 52, 59, pi. 1, figs. 1-5.

Calymene lata sp. nov. Shirley, p. 414, pi. 30, figs. 11-13; pi. 31, fig. 4.

Calymene ( Calymene ) blumenbachi blumenbachi Brongniart, 1822; Tomczykowa, pp. 97, 135, pi. 3,
figs. 3 and 4; text-fig. 6 a, b.

Calymene blumenbachii Brongniart, 1822; Whittington in Moore, p. 0452, fig. 353.1a-c.
Calymene blumenbachi blumenbachi ; Schrank, p. 135, pi 9, figs. 5 and 6.

Calymene blumenbachi Brongniart, 1822; Mannil, p. 250, pi. 4, figs. 6 and 7; pi. 5, figs. 1-4.
Calymene blumenbachi Brongniart; Chatterton and Campbell, p. 95, fig. 4.

Calymene blumenbachii blumenbachii Brongniart, 1822; Siveter, p. 784, pi. 97, fig. 10; pi. 100,
figs. 9-11.

Calymene blumenbachii Brongniart in Desmarest, 1817; Whittington, p. 177.

Calymene blumenbachii Brongniart; Thomas, Owens and Rushton, fig. 23 (pars).

Lectotype. Subsequently designated Shirley 1933, p. 53. A complete, partially enrolled specimen, Collections de
Paleontologie de l’Universite P. et M. Curie, No. 3409/77; figured Brongniart in Brongniart and Desmarest
1822, pi. 1, fig. 1a, b; Shirley 1933, pi. 1, figs. 1-3; herein Plate 90, figs. 1-5.

Paralectotype. The more or less complete specimen which formed the basis for Brongniart in Brongniart and
Desmarest 1822, pi. 1, fig. lc, d. This specimen now appears to be lost (see above).

Type stratum and locality. Much Wenlock Limestone Formation, Homerian Stage, Wenlock Series, Dudley,
West Midlands. Bassett (1976, pp. 21 1 and 212) regarded at least the base of this formation at Dudley as being of
lundgreni Biozone age, with the upper part being most likely of ludensis Biozone age.

Additional material. All of the major and most provincial museums in Britain have well-preserved material of this
species and there is no attempt here to provide a complete list; the total number of specimens is probably several
hundred. It is also well represented in the collections of numerous foreign repositories.

Diagnosis. A species of Calymene with a short preglabellar area, about one-tenth as long (sag.) as
glabella; anterior border low relative to dorsal surface of frontal glabellar lobe and steeply to more or
less vertically inclined, having a fairly sharp dorsal edge; preglabellar furrow very short (sag. and
exsag.), moderately deep. Strongly inflated glabella projects well above and well in front of fixed cheek.
Pygidium with strongly convex (tr.) axis, six to eight axial rings, five pleural furrows.

Description. Cephalon is subsemicircular in outline, 2T (PI. 90, fig. 1) to 2-2 (PI. 91, fig. 6) times as wide as long.
Glabellar outline is bell-shaped, 10 (PI. 91, fig. 6) to I T (PI. 90, fig. 1) times as long as wide, projects well in front
of fixed cheeks. Occipital ring slightly narrower (tr.) than glabellar width at lp lobes, gently convex in profile, is
longest medially but gradually shortens abaxially and flexes forwards at axial furrow where it is weakly inflated.
Occipital furrow moderately long (sag.), not deeply incised behind central glabellar area, becomes slit-like
abaxially. Lobe lp is fairly large, subquadrate, has a quite strongly convex outer margin, is strongly swollen and
separated from median lobe by very shallow posterior extension of furrow lp. The latter is deep and widest
(exsag.) at axial furrow, runs inward and backward between lobes lp and 2p, bifurcates adaxially, longer
posterior branch turning first backward then inward, anterior branch directed forward and inward, continuing
very weakly anteriorly across inner side of lobe 2p. Distinct intermediate lobe within fork of furrow 1 p. Lobe 2p

EXPLANATION OF PLATE 90

Figs. 1-5. Calymene blumenbachii Brongniart in Desmarest, 1817. UPMC 3409/77, Much Wenlock Limestone
Formation, Dudley; lectotype; complete, partially enrolled specimen. 1, dorsal stereo-pair (cephalon); 2, dorsal
(pygidium); 3, oblique (pygidium); 4, lateral; 5, frontal (cephalon); all x 2. Figured Brongniart in Brongniart
and Desmarest 1822, pi. 1, fig. 1a, b; Shirley 1933, pi. 1, figs. 1-3.

Fig. 6. Osmotoxotis phalacra Siveter, 1980. MGUH 16.803, Much Wenlock Limestone Formation, Wren’s
Nest, Dudley; right valve, female, lateral view, x 33.

Fig. 7. Thlipsura v-scripta Jones and Holl, 1869. MGUH 16.804, Much Wenlock Limestone Formation, Wren’s
Nest, Dudley; left valve, tecnomorph, lateral view, x 73.

Ostracodes prepared and photographed by Dr David J. Siveter (Leicester).

PLATE 90

SIVETER, Calymene, Osmotoxotis, Thlipsura

788

PALAEONTOLOGY, VOLUME 28

rather swollen, transversely elongate, papillate. Furrow 2p directed transversely or slightly forward. Lateral lobe
3p longest (exsag.) dorsally, narrows as it runs down side of glabella, confined by shallow though distinct 3p
furrow which trends inward and forward. Small 4p lobe present. Frontal lobe with its sides vertically inclined and
directed exsagitally or slightly outward posteriorly, in dorsal view anterolateral margins are rounded, anterior
outline weakly (PI. 90, fig. 1) to strongly (PI. 91, fig. 6) convex forward. Frontal lobe is about 0-7 times as wide as
glabella at lobe lp. In lateral profile (PI. 90, fig. 4; PI. 91, fig. 8) dorsal surface of glabella projects well above fixed
cheek, is gently convex between occipital furrow and furrow 2p, thereafter moderately to strongly convex to
anterior face of frontal lobe, which falls very steeply and is undercut by preglabellar furrow.

Axial furrow very shallow at occipital ring, narrowest at base of lobe lp, becomes progressively wider (tr.) and
much deeper to furrow lp; around lobe lp it undercuts and its abaxial face curves very steeply downward and
slightly inward; it is just continuous under bridge of lobe 2p and genal buttress, from here to preglabellar furrow
it is uniformly narrow, deep, and trench like. Anterior pit is below posterior part of frontal lobe. Preglabellar
furrow very short (sag. and exsag.), moderately deep. Anterior border is a very short raised rim, in lateral view
(PI. 90, fig. 4) it curves vertically upward and a little inward, its dorsal surface is quite sharply edged and does not
reach far up anterior face of frontal lobe (PI. 90, fig. 5); opposite axial furrow it is slightly swollen (PI. 91, fig. 3).
Anterior margin moderately convex forward (PI. 90, fig. 1) and upward (PI. 90, fig. 5).

Posterior border becomes considerably wider (exsag.) and less convex abaxially from fulcrum before
narrowing (exsag.) slightly near genal angle (PI. 91, fig- 3). Posterior border furrow widest opposite fulcrum, has
shallow anterior and steep posterior slope. In lateral profile postocular part of fixed cheek slopes (exsag.) gently
to moderately to posterior border furrow, preocular part is rather narrow (tr.), curves steeply downward and
forward. Furrow between fixed cheek and anterior border is shallower than preglabellar furrow. Palpebral lobe is
moderately (PI. 91, fig. 1 1 ) to quite steeply (PI. 90, fig. 5) inclined with mid-length opposite anterior part of lobe 2p,
it is longer (exsag.) than lobe 2p but not as long as lobe lp, its outer margin is slightly pointed. At contact of
posterior facial suture and ocular suture, palpebral lobes are L6 times as wide apart as width of glabella across 2p
lobes. Posterior branch of suture runs transversely from palpebral lobe then turns obliquely backward to lateral
border where it bends more sharply backward and finally slightly outward to bisect lateral and posterior
margins; anterior branch directed forward and slightly inward to anterior border, turns sharply inward on outer
face of border to connective suture (PI. 91, fig. 11). Visual surface of eye is not preserved, reniform in outline,
supported by eye socle from which convex main field of free cheek descends very steeply to distinct, broadly
U-shaped lateral border furrow (PI. 91, figs. 3 and 8). Lateral border turns sharply over and under; doublure
acutely reflexed.

Rostral plate ( PI. 90, fig. 2; PI. 9 1 , fig. 13) composed of border and doublure sectors. Border sector and rostral
suture broadly arched; connective suture abaxially convex. Hypostoma apparently missing on figured material.

Thorax has thirteen segments. Gently convex (sag.) axial ring very gradually widens (exsag.) abaxially from
median line and is produced into gently inflated node at axial furrow (PI. 91, fig. 9). Articulating half-ring about as
wide (sag.) as axial ring sagitally; articulating furrow shallow medially, deeper and narrower (exsag.) abaxially.
Posterior pleural band strongly convex (exsag.), forms a narrow bounding rim around the flat pleural facet.
Dorsally the pleural furrow is relatively wide (exsag.), deep, and U-shaped, reduced to a rill-like slit on pleural
facet (PI. 91, fig. 1). Anterior pleural band tightly convex (exsag.), narrower, and slightly lower than posterior
band.

Pygidium is slightly less than twice as wide as long. Axis is slightly less than half the pygidial width (PI. 90,
fig. 2), reaches well above pleural regions, is strongly convex (sag.), has seven (PI. 90, fig. 2) or eight (PI. 91,
fig. 4) axial rings (other specimens have only six). All axial rings except the last defined posteriorly by complete
ring furrows which are weakest medially; last ring furrow discontinuous; terminal axial piece rounded. Axial
furrow clearly impressed, weakest posteriorly. Inner part of pleural region falls steeply from axial furrow, outer

EXPLANATION OF PLATE 91

Figs. 113. Calymene blumenbachii Brongniart in Desmarest, 1817. 1, 2, 4, 5, 7, and 10, MGUH 16.806, Much
Wenlock Limestone Formation, Wren’s Nest, Dudley; paralectotype of C. tuberculata (Briinnich, 1781) and
‘second’ specimen of Shirley (1933), pygidium and thorax. 1, lateral; 2, dorsal; 4, posterior; 5, posterior-
oblique; all views of prepared specimen, x 2. 7, posterior; 10, dorsal; both views of unprepared specimen,
x L5. 3, 6, 8, 9, and 11 13, MGUH 16.805, Much Wenlock Limestone Formation, Wren’s Nest, Dudley;
lectotype of C. tuberculata (Briinnich, 1781) and ‘first’ specimen of Shirley (1933), cephalon and partial thorax.
3, oblique; 6, dorsal (cephalon); 8, lateral; 9, dorsal (thorax); 11, frontal; 13, ventral; all views of prepared
specimen, x 2. 12, dorsal (thorax) view of unprepared specimen, x L5.

PLATE 91

SIVETER, Calymene

790

PALAEONTOLOGY, VOLUME 28

part becomes vertically inclined. There are five distinctly impressed pleural furrows which are best marked at
their mid-length and almost reach lateral margin (PI. 90, fig. 3; PI. 91, fig. 5). Interpleural furrows are slightly
longer than pleural furrows, deepest distally, become very faint more proximally though are never completely
effaced, are slightly better impressed again immediately adjacent to axial furrow. Fifth interpleural furrow runs
on outside of an exsagitally directed ridge which probably represents anterior pleural band of sixth pleura and
confines (abaxially) the postaxial sector. Inside this ridge is a very shallow furrow. In lateral profile there is a break
in slope between terminal axial piece and postaxial sector. Border rolls under at lateral margin, is widest (tr.)
anteriorly.

Sculpture on glabella and fixed and free cheek inside posterior and lateral borders consists of closely spaced
large to small granules. Abundance of granules falls off towards furrows and they are absent in deepest part of
axial furrow, preglabellar furrow, lateral and posterior border furrows, and occipital furrow. Granules are closely
packed and more uniform in size on outer side of anterior border, lateral border, rostral plate, central part of
pygidial axis, distal posterior margins of pleurae, outer pleural region, and border roll of pygidium; on posterior
part of lateral border roll they are more elongate and scale-like, on cephalic and pygidial borders more flattened.
Much more widely scattered fine- to medium-sized granules on thorax, except for pleural, axial, and articulating
furrows. Pleural facets have very fine granules. Granules are scarce on inner part of pygidial pleural region and
abaxial part of pygidial axis.

Discussion. The description above is of the three trilobites figured herein, but it can be applied to other
specimens of the species. In the lectotype of C. tuberculata the glabellar and cephalic width to length
ratio and convex outline of the frontal glabellar lobe is slightly greater than that of the C. blumenbachii
lectotype (cf. PI. 90, fig. 1; PI. 91, fig. 6), and the paralectotype of tuberculata has an extra, very weak
eighth pygidial axial ring furrow compared with the blumenbachii lectotype (cf. PI. 90, fig. 5; PI. 91,
fig. 4), but such variation is considered to be intraspecific. A description of the hypostoma and a
comparison with other closely related taxa, for example C. clavicula Campbell, 1968 from the Silurian
of Oklahoma or C. neotuberculata Schrank, 1970 from the Wenlock of the Baltic, is in preparation by
me for publication elsewhere. C. ceratophthalma Woodward, 1868 and C. lata Shirley, 1936, both from
the Much Wenlock Limestone Formation of Dudley, I consider to be junior synonyms of C.
blumenbachii.

Occurrence. Much Wenlock Limestone Formation, West Midlands inkers, England; late Wenlock mudstones
and siltstones, Penylan, Rumney inker, Cardiff area, South Wales.

Acknowledgements. I am indebted to the following who helped in the search for the material used in this paper: Dr
Francoise Bigey (Laboratoire de Paleontologie des Invertebres, Universite P. et M. Curie, Paris), Dr J.-C. Fischer
(Museum National d’Histoire Naturelle, Paris), Dr Soren Floris (Geologisk Museum, Kobenhavns Universitet),
Dr John Peel (Gronlands Geologiske Undersogelse, Kobenhavn). Professor H. B. Whittington (Cambridge)
kindly reviewed the manuscript for the author; Mr R. V. Melville (ICZN, London) and Dr J. D. D. Smith (ICZN,
London) gave advice on questions of nomenclature. The ostracode work of Dr David J. Siveter (Leicester) is
gratefully acknowledged.

REFERENCES

barrande, J. 1852. Systeme Silurien du centre de la Boheme. 1 ere partie : Recherches paleontologiques. Vol. 1.
Crustaces: Trilobites, xxx + 935 pp. Prague and Paris.

1872. Systeme Silurien du centre de la Boheme. 1 ere partie: Recherches paleontologiques. Supplement au
vol. 1. Trilobites, Crustaces divers et Poissons, xxx + 647 pp. Prague and Paris.
bassett, M. G. 1 976. A critique of diachronism, community distribution and correlation at the Wenlock-Ludlow
Boundary. Lethaia, 9, 207-218.

bassler, r. s. 1915. Bibliographic index of American Ordovician and Silurian fossils. Bull. U.S. natn. Mus. 92 ,
2 vols. 1521 pp.

brongniart, a. 1822. In brongniart, A. and desmarest, a. G. Histoire naturelle des Crustaces fossiles, sous les
rapports zoologique et geologique. Savoir: les Trilobites par A. Brongniart. Les Crustaces proprement dits par
A. G. Desmarest, vii-t- 154 pp. Paris.

brunnich, m. t. 1781. Beskrivelse over Trilobiten, en Dyreslaegt og dens Arten, med en nye Arts Aftegning.
Nye Sami. K. dansk. Vid. Selsk. Skr. 1 , 384-395.

SIVETER: SILURIAN TRILOBITE

791

butler, a. J. 1939. The stratigraphy of the Wenlock Limestone at Dudley. Q. Jl geol. Soc. Lond. 95, 37-74.
carriol, r.-p. 1984. Types et figures de trilobites du Museum national d’Histoire naturelle. Bull. Mus. natn. Hist,
nat. Paris , 4 e ser. 6, sect. C, No. 1, 3-32.

CHATTERTON, b. D. E. and Campbell, K. s. w. 1980. Silurian trilobites from near Canberra and some related
forms from the Yass Basin. Palaeontographica, AI67, 77-1 19.
desmarest, a. g. 1816. In Nouv. Diet. Hist. nat. (2nd edn.), 5, 49-50.

1817. In ibid. 8, 517-518.

lindstrom, G. 1885. Forteckning pa Gotlands Siluriska Crustaceer. Ofvers K. VetenskAkad. Fork. Stockh. 6,
37-100.

lyttelton, c. 1750. A letter from the Rev. Charles Lyttelton, Ll.D. and F.R.S., Dean of Exeter, to the
President, concerning a nondescript petrified insect. Phil. Trans. R. Soc. 46, 598-599.
mannil, R. m. 1977. East Baltic Lower Silurian Calymenidae (Trilobita). In kaljo, d. (ed.). Facies and fauna of the
Baltic Silurian. Akad. Nauk. Eston. SSR. Inst. Geol. Tallinn , 240-258. [In Russian with English and Estonian
summaries.]

m‘coy, F. 1851. In sedgwick, a. and m‘coy, f. A synopsis of the British Palaeozoic rocks , with systematic
descriptions of the Palaeozoic fossils in the geological museum of the U niversity of Cambridge , 1 84 pp. Cambridge
and London.

milne Edwards, h. 1 840. Histoire naturelle des Crustaces , comprenant I’anatomie , la physiologie et la classification
de ces animaux , 3, 638 pp. Paris.

Mortimer, c. 1750. Some further account of the before-mentioned Dudley Fossil, by the Editor of these
Transactions. Phil. Trans. R. Soc. 46, 600-602.

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.

— 1859. Siluria. The history of the oldest fossiliferous rocks and their foundations, with a brief sketch of the
distribution of gold over the earth (3rd edn.), xx + 592 pp. London.

nicholson, h. a. and etheridge, r. 1879. A monograph of the Silurian fossils of the Girvan district in
Ayrshire with special reference to those contained in the ‘Gray Collection’, 1 (2), 136-233. Edinburgh and
London.

reed, F. R. c. 1906. The lower Palaeozoic trilobites of the Girvan district, Ayrshire. Palaeontogr. Soc. [ Monogr .],
(3), 97-136.

salter, j. w. 1865. A monograph of the British trilobites from the Cambrian, Silurian and Devonian forma-
tions. Ibid. (2), 81-128.

schrank, e. 1970. Calymeniden (Trilobita) aus Silurischen Geschieben. Ber. dt. Ges. geol. kFiss., A, 15, 109-
146.

Shirley, J. 1933. A redescription of the known British Silurian species of Calymene ( s.L). Mem. Proc. Manchr lit.
phil. Soc. 77, 5-67.

- 1936. Some British trilobites of the family Calymenidae. Q. Jl geol. Soc. Lond. 92, 384 422.

siveter, david j. 1978. The Silurian. In bate, r. h. and robinson, j. e. (eds.). A stratigraphical index of
British Ostracoda. Geol. J. Special Issue, 8, 57-100.

1980. British Silurian Beyrichiacea (Ostracoda). Palaeontogr. Soc. [ Monogr .], 133, 1-76.
siveter, derek j. 1977. The middle Ordovician of the Oslo Region, Norway, 27. Trilobites of the family
Calymenidae. Norsk geol. Tiddskr. 56 (for 1976), 355-396.

— 1979. Metacalymene Kegel, 1927, a calymenid trilobite from the Kopanina Formation (Silurian) of
Bohemia. J. Paleont. 53, 367-379.

— 1980. Evolution of the Silurian trilobite Tapinocalymene from the Wenlock of the Welsh Borderlands.
Palaeontology, 23, 783-802.

1 983. Calymene lawsoni and allied species from the Silurian of Britain and their stratigraphic significance. In
briggs, d. e. G. and lane, p. d. (eds.). Trilobites and other arthropods: papers in honour of Professor H. B.
Whittington, F.R.S. Spec. Pap. Palaeont. 30, 69-88.

Thomas, a. t., Owens, r. M. and rushton, a. w. A. 1984. Trilobites in British stratigraphy. Spec. Rep. geol. soc.
Lond. 16, 78 pp.

tomczykowa, e. 1957. Trilobites from the Wenlock and lower Ludlow graptolitic shales of the Swiety Krzyz
Mountains. Biul. Inst. geol. 122, 83-143. [In Polish with Russian and English summaries.]

Whittington, H. b. 1959. In moore, R. c. (ed.). Treatise on invertebrate paleontology. Part O. Arthropoda, I,
xix + 560 pp. Geological Society of America and University of Kansas Press, New York and Lawrence,
Kansas.

792

PALAEONTOLOGY, VOLUME 28

Whittington, h. b. 1983. Calymene Brongniart, 1822 in Brongniart & Desmarest, 1822 (Trilobita): Proposed
conservation. Z.N. (S.) 637. Bull. zool. Nom. 40 , (3), 176-178.

and siveter, derek j. In press. Type species of the genus Calymene Brongniart in Brongniart & Desmarest,
1822 and proposed suppression of the name tuberculatus Briinnich, 1781: rider to Z.N. (S.) 637. Bull. zool. Nom.
woodward, H. 1868. On a newly-discovered long-eyed Calymene from the Wenlock Limestone, Dudley. Geol.
Mag. 5, 489-494.

1869. On a newly-discovered long-eyed trilobite from Dudley. Ibid. 6, 43-44.

Typescript received 13 February 1985
Revised typescript received 22 March 1985

DEREK J. SIVETER

Department of Geology
University of Hull
Cottingham Road
Hull HU6 7RX

THE PALAEONTOLOGICAL ASSOCIATION

Notes for authors submitting papers for publication in
PALAEONTOLOGY
and

SPECIAL PAPERS IN PALAEONTOLOGY
prepared by the Editors and Council of the Palaeontological Association

PUBLICATION POLICY AND PRACTICE

Scope of publications. Manuscripts on any aspect of palaeontology will be considered for publication.
Papers on Recent material may be acceptable if their palaeontological relevance is explicit. Preference
is given to manuscripts with more than local significance. Review papers are particularly welcome and
a fee may be payable by the Association. Prospective authors of such papers who wish to take
advantage of this offer should first contact the Publications Secretary. Papers whose principal
purpose is to describe one or two new species of common groups are not usually published. A high
standard of illustration is a feature of the journal. The series Special Papers in Palaeontology is for
papers longer than those normally accepted for Palaeontology (see below, p. 800) or for collections of
shorter papers with a common theme.

Palaeontology and Special Papers in Palaeontology provide opportunity for illustration using both
Plates and Text-figures, but authors are required to pay special attention to brevity in the
composition of both illustrations and text. Short papers are particularly welcome in Palaeontology.
The style should be simple. Care should be taken to avoid long complicated sentences and the
unnecessary use of unusual words or phrases, so that papers may be read easily by those
unaccustomed to the English language.

Submission. Papers should be written in English, with British rather than American spelling. Three
complete copies of the typescript (including explanatory text for illustrations and tables) should be
submitted. Three sets of good-quality copies of all text-figures and plates should accompany each
typescript; these should be reduced to publication size. The original figures (if they are small) may, if
necessary, serve as illustrations for one copy. Although failure to provide three copies will usually
result in slower handling, no paper will be rejected for this reason. Submissions should be made to
the Secretary of the Publications Committee at the address given on the inside cover of the current
issue of Palaeontology. The Secretary will assign each paper to an editor who will handle all
subsequent communication with the author. Papers will be sent to at least two referees, including at
least one specialist in the field covered by the paper. The editor will inform authors of referees’
comments and will normally require referees’ suggestions to be implemented. Changes of an editorial
nature will usually be requested at this stage, but these will be kept to a minimum if authors will follow
the instructions given below. Adherence to these rules will also improve the chances of publication
within the target time, currently less than 12 months from the date the typescript is first received.

Copyright. Authors of papers that are accepted for publication will be asked to sign over the copyright
to the Association.

794

NOTES FOR AUTHORS

THE FORMAT OF THE MANUSCRIPT

Typing. Papers should be typed on good-quality paper, preferably of International A4 size
(297 x 210 mm). All pages of typing, including references and explanations of illustrations, should be
numbered consecutively. Plate and text-figure explanations should be put on separate sheets
following the rest of the typescript. Double spacing of lines (not one and a half) is required throughout
(including references and explanations of text-figures and plates), and there should be a 3-cm left-hand
margin.

The Title should be short but informative and should normally include fossil group, age, and
general location. It should not include the names of new taxa.

A concise Abstract of not more than 200 words is required at the beginning of all papers. It should
summarize results (rather than contents) of the paper and must mention all new systematic names
erected; it should be carefully prepared so that it will serve satisfactorily for international use in
abstracting journals.

Style and arrangement. Authors should consult recently published volumes of Palaeontology and
construct their papers in accordance with the format used there. Attention is drawn to the following
practices, adherence to which will keep editorial delays to a minimum.

Three orders of heading are used. In typescript these should be represented as follows: 1st order
headings, upper-case letters centred; 2nd order headings, lower-case letters underlined for italics, flush
to the left-hand margin, on a line of their own; 3rd order headings, as 2nd order, but followed by a full
stop and with the text continuing on the same line. Nothing else should normally be underlined in the
text with the exception of words requiring italicization such as generic and specific names, and foreign
words and phrases that are not Anglicized.

In the text, references should be cited by the author’s name and the date of the publication, without
a comma in between. Brackets are used as appropriate, e.g. \ . . as discussed by Dickens (1963)’; ‘as
discussed previously (Dickens 1963)’. The page reference to any quotation must be given, e.g. Dickens
(1963, p. 20). A work with three or more authors (Squeers, Pickwick and Carton 1964), may be
shortened (Squeers et al. 1964) unless ambiguity results. Note that the ampersand (&) is not used
anywhere in the text.

Systematic work is always introduced by the 1st order heading: ‘SYSTEMATIC PALAEON-
TOLOGY’. Second order headings are often not appropriate in this section. The conventions of the
journal regarding a marginal or a central position on the page, the order of the different sections, and
the format for synonymies, references to illustrations, etc., should be followed with reference to
previous issues of Palaeontology. Note that information accessory to systematic descriptions such as
type data, etymology, stratigraphic position (which should be detailed), curatorial information,
description, etc., is set in small type. Such sections should be adjacent to avoid frequent changes
between large and small type. Care should be taken to ensure that diagnoses, descriptions, and
discussion are kept distinct. Synonymies are also set in small type. Only the most important previous
citations should be given. This style is:

1841 Ammonites Requienianus d’Orbigny, p. 315, pi. 93, figs. 1-4.

1903 Coilopoceras requienianum (d’Orbigny); Hyatt, p. 99.

1975 Coilopoceras requiem (d’Orbigny); Dassarma and Sinha, p. 70, pi. 9, fig. 6; text-fig. 8.

Note that figure and plate numbers, etc., that were originally given in Roman numerals are always
transliterated into arabic figures. This is also the case in the main text and the references.

Authors are encouraged to use symbols to indicate the degree of confidence with which particular
items in the list are referred to the species under discussion. Such symbols are given in Matthews,
s. c. 1973. Notes on open nomenclature and on synonymy lists. Palaeontology , 16 , 713-719. For
example:

vl937 = the author has seen the cited material and agrees with the reference.
v*1937 = the author has seen the type of the species.

NOTES FOR AUTHORS

795

Name(s) and correspondence address(es) should be typed on the right-hand side of the page after
the reference list.

Footnotes. These will only be allowed in exceptional circumstances. Permission to publish, for
instance, should be included with other acknowledgements.

References should be arranged in alphabetical order of authors’ names at the end of the paper. The
author’s name should be typed in capitals, with the initials after the surname. It should be followed
by the year of publication, and the title of the paper in full. In the titles of papers, capital letters
should be used only for proper nouns, and for all nouns in German. The name of the journal
(which should be underlined) should be abbreviated in the style of the fourth edition of the
World List of Scientific Periodicals , Butterworths, London, 1963-1965. New titles and abbre-
viations are listed in British Union Catalogue of Periodicals , Butterworths, London (quarterly),
1964 onwards. Volume number (part or fascicule number, in brackets, only if really necessary)
and pagination should be given in arabic figures with the items separated by commas only. Volume
numbers only should be given a wavy underline to indicate bold type. The titles of books should
be underlined for printing in italics, and the number of pages (e.g. 560 pp.), publisher, and place of
publication should be given in that order. When a reference has been translated or transliterated,
the original language should be stated in square brackets at the end. Examples:

boucot, a. j. 1981. Principles of benthic marine paleoecology, xv + 463 pp. Academic Press, New York and
London.

— and Johnson, i. G. 1979. Pentamerinae (Silurian brachiopods). Palaeontographica , A163, 87-129.
cocks, l. r. m. and mckerrow, w. s. 1978. Silurian. In mckerrow, w. s. (ed.). The ecology of fossils, 93-124.
Duckworth, London.

— 1984. Review of the distribution of the commoner animals in Lower Silurian marine benthic com-
munities. Palaeontology, 27, 663-669.

d’orbigny, a. c. v. d. 1853. Note sur le nouveau genre Hypotrema. J. Conchyliologie, 4, 432-438, pi. 10.
Holland, c. h. 1971. Some conspicuous participants in Palaeozoic symbiosis. Scient. Proc. R. Dubl. Soc. A4(2),
15-26.

jaanusson, v. 1979. Ecology and faunal dynamics, 253-294. In jaanusson, v., laufeld, s. and skoglund, r.
(eds.). Lower Wenlock faunal and floral dynamics — Vattenfallet section, Gotland. Sver. geol. Unders. Afh.
C762, 1-294.

Northrop, s. a. 1939. Paleontology and stratigraphy of the Silurian rocks of the Port Daniel— Black Cape
Region, Gaspe. Spec. Pap. geol. Soc. Am. 21, i-ix, 1-302.
obut, a. m. 1964. Podtip Stomochordata. Stomokhordovye. In orlov, y. a. (ed.). Osnovy paleontologii:
Echinodermata, Hemichordata, Pogonophora, Chaetognatha, 279-337. Nedra Press, Moscow. [In Russian.]
rasmussen, h. w. 1978. Articulata. In moore, r. c. and teichert, c. (eds.). Treatise on invertebrate paleontology.
Part T. Echinodermata 2(3), T813-T928. Geological Society of America and University of Kansas Press,
Boulder, Colorado and Lawrence, Kansas.

Sheldon, p. A. 1 979. F unctional morphology of Baltoeurypterus. Ph.D. thesis (unpubl.). University of Cambridge.
stormer, l. 1934a. Merostomata from the Downtonian sandstone of Ringerike, Norway. Skr. norske Vidensk-
Akad. Mat. -nature, kl. 1933 (10), 1 125.

19346. Downtonian Merostomata from Spitzbergen, with remarks on the suborder Synziphosura. Ibid.

1934 (3), 1-26.

watkins, R. 1979. Benthic community organization in the Ludlow Series of the Welsh Borderland. Bull. Br. Mus.
nat. Hist. (Geol.), 31, 175-280.

whittard, w. f. 1934. A revision of the trilobite genera Deiphon and Onycopyge. Ann. Mag. nat. Hist. (10), 14,
505-533.

Whittington, h. B. 1977. The Middle Cambrian trilobite Naraoia, Burgess Shale, British Columbia. Phil. Trans.
R. Soc., B 280, 409-443.

Units and symbols. As far as possible the recommendations contained in Quantities, Units and Symbols
(second edition. The Royal Society, London, 1 975) should be followed; in particular, the International
System of Units (SI) should be used whenever it is practicable to do so (e.g. /<m, mm, m, kg). Note that
no full stop is used after such symbols.

796

NOTES FOR AUTHORS

LINE ILLUSTRATIONS AND TABLES

Text-figures. Original drawings, in black on good quality white card or on good quality drawing film,
should be submitted. They should preferably be made double the final size but no larger, and when
reduced must not exceed the type area of a page 200 x 147 mm. If the caption (explanation) of a
full-page figure is long, allowance for its inclusion on the same page should be made by reducing the
height of the figure. In composing smaller text-figures, space on the page is best used if the figure is
wide rather than high, and the full width of the page should be used if possible. Individual prints and
diagrams within one text-figure should be lettered (a, b, c, etc.) and a linear scale bar added where
appropriate. All lettering should be inserted by the author, and must be readable when reduced. Good
quality photographic copies (unglazed) may be suitable for reproduction and are preferable to
excessively large originals; originals may, however, have to be requested by the editor.

Professional preparation of text-figures. The Association has allocated a limited sum of money to help
offset the cost of professional preparation of text-figures that require skills above those normally
available to authors, where such professional help is not otherwise available. Authors wishing to
apply for grants from this fund should approach the Association through the Secretary of the
Publications Committee, giving details of the proposed text-figures together with a draft manuscript.
Awards will be made at the discretion of the Association and, although funds will only be granted if
it appears likely that the manuscript will ultimately satisfy the Association’s requirements, the
awarding of financial aid will not commit the Association to publication.

Tables. Tables should preferably be carefully drafted to the same standards as text-figures;
alternatively, they may be prepared on an electric typewriter, paying special attention to compactness
and legibility, and using a carbon ribbon. Two copies at correctly reduced size must be supplied in
addition to the originals. Fold-out tables are not acceptable, because of prohibitive cost. If a table of
larger than page size is essential, it should be arranged for two facing pages.

EXPLANATION OF PLATE X

This plate has been designed to show the common defects noted in plates submitted for publication.

General comment: the plate lacks balance of contrast and is poorly laid out. If your plate shows any of the
defects illustrated here, it should not be submitted for publication. Submitted copy should be made up of
normal black and white prints, not half-tones. Otherwise, interference dots (as seen here) will result.

Top row, in general. The prints do not line up along their top and bottom edges. The gap between 1 and 2
is not parallel sided or perpendicular to the margin. Note the unacceptably variable size and shape of the
cut-off corners to allow the insertion of figure numbers.

Fig. 1. Low contrast showing loss of detail. Note that the print is not mounted perpendicular to the margin.

Fig. 2. Moderate contrast reveals greater detail.

Fig. 3. Not sharply in focus, and the appearance is therefore fuzzy.

Fig. 4. Negative scratched, causing lines on the print.

Fig. 5. SEM. The print is marred by scan lines caused by charging, the gold coating on the background is
badly cracked and the lower part of the specimen has been trimmed off.

Fig. 6. SEM. This print shows excessive contrast. Also, though the illumination should appear to come from
the top or top left, the print has been mounted at right angles to the desired orientation. The specimen is
dirty and not worth illustrating.

Fig. 7. A hair on the negative is superimposed on the glabella.

Fig. 8. The background has been inexpertly painted out with Indian ink. Note the loss of ribs below the
aperture and the crude painting-out around the ribs elsewhere. The print was damaged during mounting,
causing artefacts near the aperture.

Fig. 9. The background has been clumsily cut away using scissors. Note the damage to the outline of the ribs and
local accidental inclusion of background.

Fig. 10. The unevenly illuminated specimen merges with the background.

PLATE X

How not to make a plate

798

NOTES FOR AUTHORS

Explanations of text-figures and tables should be typed (double-spaced) and placed at the end of the
typescript. They should be numbered consecutively with the rest of the typescript.

Position in text. The approximate position desired for insertion of text-figures, tables, and plates
should be indicated on the typescript.

PLATES AND TEXT-PHOTOGRAPHS

It continues to be editorial policy to maintain a high standard of plate quality. Authors are reminded
that the published plate cannot be of better quality than the original. Plate X shows the common
defects noted in plates submitted for publication. If your plates show any of these defects, they should
be corrected before submitting them to the journal.

Size. The size of plates is 203 x 147 mm. Where authors have photographs which fill less than a full
page, they will be referred to as text-figures. These will be reproduced by the same process as plates but
to text-figure dimensions (see above). Every effort should be made to ensure that no page space is
wasted, particularly with respect to the width of text-figures composed of photographs.

Plates and text-photographs should be submitted at publication size. Large plates cause problems
of handling and offer few advantages of improved quality. Magnifications of individual figures should
be stated in the caption, not indicated by scale-bars on the photographs.

Lighting. The convention of lighting fossils from the top left should be followed where possible.

Preparation of photographic prints. Photographs should be sharply in focus and printed on glossy
paper. They should be of medium contrast, using the range of shades of grey but avoiding extremes of
black and white. Artefacts such as dirt on the negative or scratches from the enlarger should be absent
from the prints. All prints on a plate should be of even tone and contrast. For this reason, it is often
better to avoid mixing conventional photographs with photomicrographs. Where possible, remove
labels from macrofossils before photography. Avoid unsightly backgrounds.

Mounting. Mount the prints on clean, white board using dry mounting tissue or a non-aqueous glue.
Use the same method for all prints. Avoid glue on the background or on the print surface. Place prints
as close together as is reasonable, making full use of plate space. Rectangular prints should not only be
cut rectangular but also mounted parallel to each other and to the sides of the plates. Leave space for
figure numbers either directly below each print or by cutting off a bottom corner (PI. X, fig. 3).

Where it is desired to remove the background from around a fossil, the print should be carefully
trimmed to the edge of the fossil and mounted on clean white board. Black backgrounds are
permitted, but the author is responsible for accurately delimiting the edge of the fossil. Roughly
trimmed prints should be precisely outlined using black drafting ink.

Each plate should be protected by a tracing paper overlay attached along the upper edge. The
author’s name, and title of the paper, should be written on the back of the originals of each text-figure
and plate.

Numbering. Figure numbers should run consecutively from left to right from top to bottom of each
plate, la, lb, 1 c, etc., should only be used to distinguish different views of the same fossil. Numbers
should be pencilled on the overlay. Do not stencil numbers on the plates. All numbers are added by
the printers, although space for such numbers should be provided when the plate is made up by the
author.

Explanations of plates should be typed and placed at the end of the typescript. They should be brief but
adequate. The magnification of each figure should be stated. The museum number of the specimen
should be given. For example:

Fig. 1. Ammonia beccarii (Linne). Repository and catalogue number. Description, locality, horizon, mag-
nification.

NOTES FOR AUTHORS

799

Figs. 2-6. Elphidium crispum (Linne). 2, repository and catalogue number; description, locality, horizon, mag-
nification. 3 a, repository and catalogue number; description, locality, horizon, magnification. 3b, description,
magnification. 4, etc.

ADMINISTRATION

Preservation of types and other specimens. In accordance with recommendations of the International
Codes of Botanical and Zoological Nomenclature, all illustrated fossils should be registered and
deposited in an appropriate permanent institution, preferably a museum with staff and facilities
capable of ensuring their availability for future reference in perpetuity. The registered numbers should
be quoted. Papers that fail to meet this requirement will not be published.

Proofs. Authors will normally receive one proof; this proof is for the purpose of correcting printer’s
errors and not for altering the wording or substance of the paper. Authors may be charged for
excessive alterations. The editors will only be responsible for author corrections notified by return of
post. Whenever possible, plate proofs will be sent to the authors with the text proofs.

Offprints. Fifty offprints of each paper will be sent free of charge; further copies may be purchased at
prices shown on the order form which will be sent with the proofs to the author (or corresponding
author in the case of multi-author papers).

Deposition of data. The Association makes use of the scheme run by the British Library, Lending
Division, whereby publication expenses can be saved by depositing tables of data and other reference
material with the British Library rather than printing them. The deposited material is stored on
microfiches, and either microfiche or full-size copies may be obtained from the British Library by an
applicant (preferably using British Library prepaid coupons) on a standard scale of charges which
allows for postage etc.

The British Library will only accept deposited material through the Publications Committee of the
Association; all such material will be referred to as part of a published paper. The published paper will
bear a reference to deposited copy with full details of its pagination and means of acquisition, e.g. \ . .
have been deposited with the British Library, Boston Spa, Yorkshire, UK, as Supplementary
Publication No. SUP 14003 (26 pages)’. Prepaid coupons for such purposes are held by many
technical and university libraries throughout the world. They may be purchased from the British
Library, Lending Division, Boston Spa, Wetherby, Yorks. LS23 7BQ. Association policy is that
neither plates nor formal taxonomic data will be considered for deposition. Authors should indicate
and separate as appendices those parts of their papers that they propose for deposition; the
Publications Committee may also recommend that part of a paper should be deposited rather than be
printed.

Preparation of copy for deposition. Copy must be prepared by the author according to the following
specifications. Editors will not undertake the preparation of copy.

(i) Copy must be camera-ready.

(ii) Maximum page size for text or tables in typescript or computer print out is 330 mm
high x 240 mm. wide, including margins. Optimum page size is A4 size.

(iii) Tabular matter should be headed descriptively on the first page, with column headings
recurring on each page.

(iv) Prefatory text, which should contain the abstract from the parent paper, should be included.

(v) All pages must be consecutively numbered.

Authors with large sections for deposition are advised to consult the Secretary of the Publications
Committee for further information.

SPECIAL PAPERS IN PALAEONTOLOGY
Preparation of papers for this series should be in the same style as for Palaeontology.

800

NOTES FOR AUTHORS

Submission. Prospective authors should consult the Editors well in advance of submission, supplying
as much information as possible.

Cost of publication. The Association’s funds for this series are limited. Authors are asked to obtain
grants wherever possible. Manuscripts should not normally exceed the equivalent of 1 10 published
pages.

Offprints. A small number of free offprints will be supplied and further copies may be obtained at a
special reduced charge. For multiple author volumes such as conference proceedings, authors will
receive one free copy of the volume but no free reprints. Reprints of individual papers may be ordered
at the standard charge. Details will be supplied at the time by the editor concerned.

GRANTS IN AID OF PUBLICATION

Palaeontology has no compulsory page or plate charges. However, authors are requested to seek
grants in aid of publication from their institutions or from research funds, or to apply for publication
costs in research grants. Some financial support is particularly welcome, and may be necessary, for
long papers.

Although acceptance of a paper for publication will not be dependent on the receipt of such grants,
authors will appreciate that the funds available to the Association are limited. Every grant or
donation will therefore directly help the Association’s publication programme.

1985

The Palaeontological Association

THE PALAEONTOLOGICAL ASSOCIATION

Annual Report of Council for 1984

Membership and Subscriptions. Membership totalled 1,420 on 31 December 1984, an increase of 7 over the
previous year. There were 932 Ordinary Members, an increase of 22; 48 Retired Members, an increase of 5;
123 Student Members, a decrease of 20; and 317 Institutional Members (no change). The number of individuals
and institutions subscribing to Palaeontology through Marston’s agency was 435, a decrease of 7. Subscriptions
to Special Papers in Palaeontology number 123 individuals, an increase of 1, and 131 Institutions, an increase
of 8. Standing orders for Special Papers through Marston’s agency were about 100 for Special Paper 37. Sales of
back parts of Palaeontology via the Membership Treasurer realized £98-43. Sales of back parts of Palaeontology
to Institutional Members yielded £21-98 plus US $34.00. Sales of Special Papers 1-21 under the special offer
yielded £2,278, less about £700 in postage; other sales of Special Papers to individuals yielded £1,216-19, and to
Institutions, £131 plus US $1 14. Sales of 26 copies of the Atlas of the Burgess Shale through the Marketing
Manager yielded £241-50 plus US $492.94. Sales of Fossil Plants of the London Clay through the Marketing
Manager yielded £1,162- 1 5 plus US $3 1 1.50 for 281 copies.

Finance. During 1984 the Association published Volume 27 of Palaeontology at an estimated cost of £53,952
(including postage and distribution). Special Papers 31 and 32 were published at a cost of £10,305 and £13,450
(provisional) respectively. In addition the Atlas of Invertebrate Macrofossils was published jointly with
Longman, at a cost to the Association of £2,246. The Association is grateful to all those who made donations.

Publications. Volume 27 of Palaeontology , published in 4 parts during 1 984, contained 898 pages and 80 plates.
Special Papers 31: Systematic Palaeontology and Stratigraphic Distribution of the Ammonite Faunas of the
French Coniacian by W. J. Kennedy and 32: Autecology of Silurian Organisms edited by M. G. Bassett and J. D.
Lawson were published in November and December respectively.

Meetings. Eight meetings were held in 1984. The Association is indebted to the organizers, hosts, and held
leaders of these.

a. Joint Meeting with the Geological Society on ‘Fossils and Sealevel Changes', held on 8 February at
Burlington House, Piccadilly, London. More than 100 attended the meeting, which was convened by
Dr R. Riding.

b. Review Seminar on ‘Palaeobiogeography’, held on 15 February at the University of Bristol. More than
200 attended the meeting. The local secretary was Professor R. J. G. Savage.

c. Twenty-seventh Annual Genera! Meeting , held in the Lecture Theatre of the Geological Society of London
on 14 March. Professor Yves Coppens delivered the Annual Address on ‘Hominid Evolution’. The
Sylvester-Bradley Award was made to Mr P. Doyle.

d. Field Meeting, organized by the Carboniferous Group, to Fife and Lothian, Scotland led by Dr A. R. M.
Macgregor. About 50 attended the excursion which was held on 13 16 April.

e. ‘ Progressive Palaeontology ’ Meeting , aimed at younger researchers, held at Bedford College, London, on
29 May. More than 53 attended the meeting, which was convened by Dr Jane Francis.

/. Field Meeting on the ‘Lower Palaeozoic of the Lake District’, held on 14-17 September, and organized and
led by Dr A. W. A. Rushton and Dr D. E. White. More than 20 attended.

g. Review Seminar on ‘Extinctions’, held on 14 November at the Open University, Milton Keynes. Over
200 people attended. The local secretary was Dr P. W. Skelton.

h. The Annual Conference, held at Fitzwilliam College, Cambridge, on 19-22 December took the form of
a thematic session on taphonomy followed by an open meeting. About 150 attended. The President’s
Award was made to Dr M. J. Benton. An excursion was led to the Upper Jurassic of Upware by Dr S. R. A.
Kelly. The local secretary was Dr C. P. Hughes.

Council. The following members served on Council following the Annual General Meeting on 14 March 1984:
President : Professor C. Downie; Vice-Presidents: Dr J. C. W. Cope, Dr R. Riding; Treasurer: Dr M. Romano;

802

THE PALAEONTOLOGICAL ASSOCIATION

Membership Treasurer. Dr A. T. Thomas; Secretary: Dr P. W. Skelton; Marketing Manager: Dr R. J. Aldridge;
Editors: Dr D. E. G. Briggs, Dr P. R. Crowther, Dr L. B. Halstead, Dr R. Harland, Dr T. J. Palmer; Other
Members: Dr E. N. K. Clarkson, Dr D. Edwards, Dr P. D. Lane, Dr A. R. Lord (Institutional Membership
Treasurer), Dr A. W. Owen, Dr C. R. C. Paul, Dr D. J. Siveter (Circular Reporter), Dr A. B. Smith, Professor
T. N. Taylor.

Circulars. Four Circulars, numbers 115-118, were distributed to Ordinary, Student, and Retired Members,
and on request to over 90 Institutional Members.

Council Activities. Besides the now well-established Review Seminars and Field Meetings, two new kinds of
meeting were launched with great success in 1984, to be held annually thereafter: discussions with the Geological
Society have led to the institution of a meeting to be held each spring on some aspect of the broader geological
applications of palaeontology; and, as a result of requests from younger members of the Association for a regular
meeting at which they could informally present their results and discuss their work with each other, the ‘Pro-
gressive Palaeontology’ series was initiated.

The Atlas of Invertebrate Macrofossils , edited by Professor J. W. Murray and published by Longman on
behalf of the Association came out just in time to go on show at the Annual Conference. A brochure advertising
the benehts of Association membership was also published. The special offer for the sale of Special Papers 1-21
at reduced prices, in order to clear back stock, was extremely successful.

Plans for the future include the organization of a one day session at the 3rd International Conference on
Systematic and Evolutionary Biology and also the 6th Meeting of the Geological Societies of the British Isles.

BALANCE SHEET AND ACCOUNTS FOR THE
YEAR ENDING DECEMBER 1984

Balance Sheet as at 31 December 1984

1983

£

£

£

£

46,586

Investments at Cost (see schedule) .

50,459

Current Assets

1,486

Sundry Debtors ......

. 2,033

25,395

Cash at Bank ......

. 29,630

1,912

Sylvester-Bradley Fund ....

1,816

28,793

33,449

Current Liabilities

2,215

Subscriptions Received in Advance .....

Provision for cost of publication of:

2,337

13,500

Palaeontology 27 /4 ........

13,800

—

Special Paper 32 ........

13,450

9,631

Sundry Creditors ........

2,509

2,000

Loan from Royal Society .......

2,000

27,346

34,096

1447

(647)

£48,033

£49,812

Represented by:

Publications Reserve Account

43,165

Balance Brought Forward .......

Excess of Income over Expenditure for the Year Transferred

44,135

44,135

970

from Income and Expenditure Account ....

1,230

45,365

Sylvester-Bradley Fund

2,021

Balance Brought Forward .......

1,912

91

Interest ..........

104

(200)

Grant Awarded .........

(200)

1,912

Balance Carried Forward .......

1,816

1,986

Meeting Reserve

2,631

£48,033

£49,812

Income and Expenditure Account

1983

£

Subscriptions

1984

1983

40,593

Palaeontolog y

Sales

Donations ....

28,939

Special Papers
Sales .....
Donations ....

7,280

Burgess Shale Portfolio
Sales .....
Postage/Stationery

2,257

Fossil Plants of the London Clay
Sales .....
Postage/Stationery

(1,243) Offprints ....

Profit on Sales of Investments
8,43 1 Investment Income (see schedule)

330 Sundry Income

£86,587

INCOME

£

40,231

921

25,997

210

8,123

1,373

1,069

(58)

1,729

(261)

£

41,152

26,207

9,496

1,01 1

1,468

779

3,917

8,947

99

£93,076

for the Year Ended December 1984

EXPENDITURE

1983

£

Cost of Publication of Palaeontology
Volume 27— Part 1 .....

Part 2

Part 3

Part 4 (provisional)

Over Provision for Volume 26, Part 4

54,821

Cost of Publication of Special Papers

No. 31

14,729

No. 32 (provisional)
Cost of Publication of:

2,425

Fossil Plants of the London Clay

3,600

Cumulative Index

—

A tlas of Invertebrate Macrofossils

1,212

Warehousing of Publications

650

Grants

Cost of Circulars

Preparation

Postage

Credit

4,041

Administrative Costs
Postage and Stationery
Editorial and Expenses
Meeting Expenses
Audit Fee

Membership of Societies

4,139

£85,617

£

13,222

13,431

13,599

13,800

(50)

10,305

13,450

1,605

141

2,246

1,919

100

2,709

1,826

(480)

1,684

212

1,842

225

60

Excess of Income over Expenditure for the Year Transferred to
£970 Publications Reserve Account

£

54,002

23,755

3,992

4,055

4,023

£91,846

£1,230

Income and Expenditure Account

1983

£

40,593

28,939

7.280

2,257

INCOME

£

Subscriptions

1984 40,231

1983 921

Palaeontology

Sales 25,997

Donations 210

Special Papers

Sales 8,123

Donations 1,373

Burgess Shale Portfolio

Sales 1,069

Postage/Stationery (58)

Fossil Plants of the London Clay

Sales 1,729

Poslage/Stalionery (261)

(1,243) Offprints

Profit on Sales of Investments
8,43 1 Investment I ncome (see schedule)

330 Sundry Income

£

41,152

26,207

9,496

1,468

779

3,917

8,947

99

£86,587

£93,076

the Year Ended December 1984

EXPENDITURE

1983

£ t i

Cost of Publication of Palaeontology

Volume 27— Part I 13,222

Part 2 13,431

Part 3 13,599

Part 4 (provisional) 13,800

Over Provision for Volume 26, Part 4 (50)

54,821 54,002

Cost of Publication of Special Papers

No. 3 1 10,305

No. 32 (provisional) 13,450

14,729 23,755

Cost of Publication of:

2,425 Fossil Plants of the London Clay 1,605

3,600 Cumulative Index . . 141

— Atlas of Invertebrate Macrofossils 2,246

3,992

1,212 Warehousing of Publications 1,919

650 Grants 100

Cost of Circulars

Preparation 2,709

Postage 1,826

Credit (480)

4,041 4,055

Administrative Costs

Postage and Stationery 1,684

Editorial and Expenses . . 212

Meeting Expenses 1,842

Audit Fee 225

Membership of Societies . . 60

4,139 4,023

£85,617 £91,846

Excess of Income over Expenditure for the Year Transferred to
£970 Publications Reserve Account £1,230

Schedule of Investments and Investments Income as at 31 December 1984

Gross Income

Cost

£

11,520

for Year

£12,000

13 1% Exchequer Stock 1987

jb

1,590

£1,000

9% Treasury Stock 1992/1996

992

90

£1,000

9% Treasury Stock 1994

955

90

£4,000

8% Treasury Stock 2002/2006

2,192

320

£5,357

1 3 Treasury Stock 1997

5,000

710

£3,280

1 3J% Exchequer Stock 1996

3,000

435

£2,000

Agricultural Mortgage Corporation Ltd. 9J% Debenture 1980/1985

1,938

185

5,270

M. & G. Charifund Units

4,073

1,070

1,865

Imperial Group p.l.c. 25p Ordinary Shares ......

1,730

177

10,000

New Throgmorton Trust ( 1983) p.l.c. 25p Income Shares

1,706

321

700

Clarke, Nicholls & Coombs p.l.c. 25p Shares

668

61

£6,180

M.E.P.C. p.l.c. 6^% Convertible Unsecured Loan Stock 1995/2000 .

4,943

402

374

M.E.P.C. p.l.c. 25p Shares

703

45

570

National Westminster Bank p.l.c. £1 Ordinary Shares ....

2,788

£10,150

Agricultural Mortgage Corporation Ltd. 7|% Debenture Stock 1991/1993

8,251

393

1,600

Commercial Union Assurance Co. p.l.c. 25p Shares (sold August 1984)

—

159

5,000

Thorn EMI p.l.c. 7% Convertible Redeemable Second Cumulative

Preference Shares 1992/1999 (sold March 1984) .....

—

250

6,298

Bank interest . .

2,649

50,459

8,947

Market Value at 31 December 1984(1983— £68,428) ....

£71,437

Report of the Auditor to the Members of
the Palaeontological Association

In my opinion, the Accounts as set out on pages 803 to 806, give a true and fair view of the state of the affairs of
the Association at 31 December 1984 and of its income and expenditure for the year ended on that date.

March 1985

Market Harborough, Leicestershire

G. R. Powell
Chartered Accountant

INDEX

Pages 1-206 are contained in Part 1; pages 207-412 in Part 2; pages 413-618 in Part 3; pages 619-810 in Part 4. Figures in Bold

Type indicate plate numbers.

A

Aldridge, R. J. See Mabillard, J. E. and Aldridge, R. J.
Algae: mound-building coralline alga, Florida, 189
Alocopythere bhandari sp. nov., 371, 44; dhansariensis
sp. nov., 372, 44; polygona sp. nov., 374, 45; talukdari
sp. nov., 374, 44

Ammonites: Maastrichtian, Western Australia, 35
Amorophognatus ordovicicus, 691, 84, 85, 86; superbus , 692,
83; alT. A. tvaerensis , 694, 86
Amphibian: Pennsylvanian, Kansas, 767
Anagaudryceras politissimum, 46, T; subtilineatum, 45, 2
Antarctica: Lower Cretaceous bivalves, 475
Antropora tincta, 31
Aphelognathus rhodesi , 698, 84
Arsenault, M. See Schultze, H.-P. and Arsenault, M.
Aulatornoceras auris , 17; cf. A. auris , 16; ‘A.’ leoschmanni,
16

Aulichnitesl bradfordensis ichnosp. nov., 626, 75, 76
Australia: Maastrichtian ammonites from Miria Formation,
35

B

Bairdia berguaensis , 365, 42

Balticella binodis , 69; deckeri, 69

Bassett, M. G. See Savage, N. M. and Bassett, M. G.

Bavlinella faveolata , 462, 51

Beania simplex, 27

Belemnite: sexual dimorphism in Youngibelus, 133
Beltina danai , 344, 39

Benedetto, J. L. See Edwards, D. and Benedetto, J. L.
Benton, M. J. and Walker, A. D. Palaeoecology, taphonomy,
and dating of Permo-Triassic reptiles from Elgin, north-
east Scotland, 207
Bergauerial sp., 350, 38

Biorecord Cincturo-Judith, 556, 65, 66; cf. Biorecord
Cincturo-Judith, 560, 66, 67; Biorecord Cincturo-domed,
562, 66, 67

Birostrina concentrica, 506, 59; BP. cf. concentrica , 510, 59, 60,

61

Biva duncanae, 69; BP. ordoviciana, 69
Bivalvia: Lower Cretaceous, Antarctica, 475
Bosence, D. W. J. The morphology and ecology of a mound-
building coralline alga ( Neogonioliihon strictum) from the
Florida Keys, 189

Boyd, M. J. A protorothyridid captorhinomorph (Reptilia)
from the Upper Carboniferous of Newsham, Northumber-
land, 393

Brachiopoda: Taxonomy, shell structure, and palaeoecology
of Gasconia , 243; Carboniferous, Scotland, 567
Brachycycloceras koninckianum , 237, 25; obtusum, 247, 24;
lobtusum, 23; B. sp., 24

Brahmaites ( Brahmaites ) kossmati sp. nov., 68, 6, 7
Brevidorsa crassispinosa, 69; fimbriata , 69

Britain: Upper Carboniferous nautiloid Brachycycloceras,
235; Lower Palaeozoic echinoderms, 527; see also England,
Scotland, Wales, and Ireland

Bryozoa: earliest fenestrate bryozoan and a short review of
Ordovician Bryozoa, 147; zooid and colony growth of
cheilostome bryozoans, 255
Bunlonia royi sp. nov., 374, 45

C

Cainozoic: taxonomy and evolution of Protenaster, 311
Calymene blumenbachii, 783, 90, 91
Cambrian: Echinoderms and edrioasteroids, 715
Canada: mid-Proterozoic macro-biota, 331
Carboniferous: Brachycycloceras, Britain, 235; reptilia,
Northumberland, 393; brachiopoda, Scotland, 567; trace
fossils in Westphalian A, West Yorkshire, 619
Celleporaria magnified, 30
Celleporella hyalina, 29
Chaloneria cormosa , 64

Chamberlain, J. A. Jr. and Pillsbury, S. W. Fluid transport
properties of Nautilus siphuncular tube: within camera
distribution of flow rate, 121

Chisholm, J. I. Xiphosurid burrows from the Lower Coal
Measures (Westphalian A) of West Yorkshire, 619
Chuaria circularis, 342, 35, 36, 37, 38, 39; ?C. circularis, 465,

51

Chulsanbaatar vulgaris, 49, 50
Colpodexylon cachiriense sp. nov., 606, 73, 74
Communities: Estimating biomass and energy flow of
molluscs, 1

Conodont: Ordovician, Welsh Borderland, 679
Crame, J. A. Lower Cretaceous inceramid bivalves from the
Antarctic peninsula region, 475
Crassotornoceras annissi sp. nov., 168, 16 ; belgicum, 169, 16;

crassum, 165. 16; guestfalicum, 170, 16
Cretaceous: non-heteromorph ammonites, Western

Australia, 35; medusoid hydrozoan, Japan. 101; bivalves,
Antarctica, 475; Barremian dinoflagellate cysts, 555;
Isocrinus, Japan, 629
Cryptophyllus magnus, 69; CP sp., 69
Cupuladria biporosa, 27

Cupulocrona digitalis sp. nov., 541, 63; rugosa sp. nov., 538,
62, 63; salopiae sp. nov., 538, 62, 63
Curry, G. B. See Taylor, P. D. and Curry, G. B
Cushmanidea distincta sp. nov., 370, 46
Cyclodendron leslii, 114, 11, 12, 13

Cytherella antheriformis sp. nov., 362, 41; assamensis sp.
nov., 359, 40; barpatharensis sp. nov., 360, 40; deopanica
sp. nov., 358, 40; hastata sp. nov., 360, 41; ventroconcava
sp. nov., 364, 41; C. sp., 364. 42
Cytherelloidea sp. juv., 364, 42
Cytheropteron reticuloradiata sp. nov., 380, 46
Cytherura eocaenica sp. nov., 376, 46

808

INDEX

D

Daltaenia mackenziensis sp. nov., 346, 39
Desmophyllites diphylloides, 54, 4

Devonian: tornoceratid goniatite, 159; lycopods, Venezuela,
599

Distobolbina grekoffi, 70 ; warthini , 70
Donovan, S. K. and Paul, C. R. C. Coronate Echinoderms
from the Lower Palaeozoic of Britain, 527
Doyle, P. Sexual dimorphism in the belemnite Youngibelus
from the Lower Jurassic of Yorkshire, 133
Drepanoistodus suberectus, 708, 81, 82
Dumican, L. W. and Rickards, R. B. Optimum preparation,
preservation, and processing techniques for graptolite
electron microscopy, 757

E

Echinodermata: Cainozoic spatangoid Protenaster , 311;

Lower Palaeozoic, Britain, 527; Cambrian, 715
Edrioasteroids: Cambrian, 715

Edwards, D. and Benedetto, J. L. Two new species of
herbaceous lycopods from the Devonian of Venezuela with
comments on their taphonomy, 599
El-Khayal, A. A. and Romano, M. Lower Ordovician
trilobites from the Elandir Shale of Saudi Arabia, 401
Electron microscopy: preparation and processing of grapto-
lite material for electron microscopy, 757
England: type species of Calymene , Silurian, 783
Eoaquapulex sp. (= Oepikella frequens of Kraft 1962), 70 ;
frequens , 70

Eocene: ostracoda, India, 355
Eographiodaetylus eos , 70 ; sulcatus , 70
Eosynechococcus sp., 464, 51

F

Falcitornoceras gen. nov., 18, 19

Falcitornoceras falcatum , 172, 17, 18, 19, 20; falciculum sp.
nov., 178, 18; /. falciculum subsp. nov., 180, 17, 18, 19, 20;
/. constriction subsp. nov., 182, 18, 19; /. wagneri subsp.
nov., 183, 19, 20
Favososphaeridium sp., 466, 53
Fish: the panderichthyid E/pistostege , 293
Fosse, G., Kielan-Jaworowska, Z. and Skaale, S. G. The
microstructure of tooth enamel in multituberculate
mammals, 435

G

Gasconsia worsleyi sp. nov., 246
Gaudryceras kayei , 46, 1, 2

Gen. et sp. indet. A, 709, 86; Gen. et sp. indet. B, 710, 83
Goniatites: Devonian, 159

Graptolite: preparation and processing for electron micro-
scopy, 757

Grossouvrites gemmatus , 66 , 5

Grypania spiralis , 349, 39

Gunnarites kalika , 59, 5; raggatti sp. nov., 60, 6

H

Hanken, N.-M. and Harper, D. A. T. The taxonomy, shell
structure, and palaeoecology of the trimerillid brachiopod
Gasconsia Northrop, 243

Harding, I. C. See Hughes, N. F. and Harding, I. C.

Harper, D. A. T. See Hanken, N.-M. and Harper, D. A. T.
Haskinsia sagittata sp. nov., 601, 71 , 72 , 73
Henderson, R. A. and McNamara, K. J. Maastrichtian non-
heteromorph ammonites from the Miria Formation,
Western Australia, 35

Hesperidella esthonica , 68 ; michiganensis , 68
Hesperolierpeton garnettense , 767
Hippida ( Hippula ) latonoda , 68 ; (H.) varicata , 68
Hi this colonus, 68 ; hi this, 68

Hoffmann, A. Biotic diversification in the Phanerozoic:
diversity independence, 387

Hoffmann, H. J. The Mid-Proterozoic Little Dal macrobiota,
Mackenzie Mountains, north-west Canada, 331
Homo sapiens, 48

House, M. R. and Price, J. D. New late Devonian genera and
species of tornoceratid goniatites, 159
Hughes, N. F. and Harding, I. C. Wealden occurrence of an
isolated Barremian dinocyst facies, 555
Hydrozoan: first Mesozoic chondrophorine, Japan, 101

I

Icriodella superba, 706, 80 , 82 , 83 , 84 , 85
India: Eocene Ostracoda, 355

Inoceramus deltoides sp. nov., 486, 54 , 55 ; carsoni, 498, 57 ,
58 ; flemingi sp. nov., 492, 56 ; stoneleyi sp. nov., 489, 55 ;
uruis, 56 ; sp. aff. anglicus, 494, 56 ; cf. anglicus elongatus,
494, 56 ; sp. aff. comancheanus , 497, 56 ; cf. sutherlandi ,

501, 58

Instructions for Authors, 793

Isocrinus (Chaladocrinus) hanai sp. nov., 636, 77 , 78 , 79 ; (C.)
sp. cf. /.? neocomiensis ( Desor), 640, 78 , 79

J

Japan: Cretaceous medusoid hydrozoan, 101; Isocrinus,
Cretaceous, 629

Jurassic: sexual dimorphism in Youngibelus from Yorkshire,
133

K

Kanie, Y. See Stanley, G. D. Jr. and Kanie, Y.

Kerfornella sp., 405, 47

Kildinosphaera chagrinata, 466, 53 ; granulata, 466, 53
Kitchinites spathi sp. nov., 57, 4
Kloucekia sp., 408, 47

Knoll, A. H. and Swett, K. Micropalaeontology of the late
Proterozoic Veteranen Group, Spitzbergen, 451
Kossmaticeras (Natalites) brunnschweileri sp. nov., 63, 6, 7;
K. (N.) sp. juv., 64, 7

Krithe oryza sp. nov., 370. 43 , 46 ; cf. K. oryza, 371, 44

L

Laccochilina dorsoplicata, 68 ; L. (Eobromidella) eurychili-
noides, 68

Legrand-Blain, M. A. A new genus of Carboniferous spiri-
ferid brachiopod from Scotland, 567
Leiosphaeridia asperata, 466, 53

Lidgard, S. Zooid and colony growth in encrusting cheilo-
stome bryozoans, 255
Lonfengshania stipitata, 343, 38

INDEX

809

Loxochonchal sp., 375, 46

Lycopod: Permian, South Africa, 111; Devonian, Venezuela,
599

M

Mabillard, J. E. and Aldridge, R. J. Microfossil distribution
across the base of the Wenlock series in the type area, 89
McNamara, K. J. Taxonomy and evolution of the Cainozoic
spatangoid echinoid Protenaster , 3 1 1
McNamara, K. J. See Henderson, R. A. and McNamara, K. J.
Macrobiota: Proterozoic, Canada, 331
Mammals: microstructure of tooth enamel in multi-
tuberculate mammals, 435
Maorites densicostatus , 64, 4
Meniscoessus sp., 49
Mesodma thompsoni , 50; M. sp., 48
Metrarabdotos unguiculatum cookae , 26
Microfossils: distribution in Wenlock type area, 89; Protero-
zoic, Spitzbergen, 451

Milner, A. R. On the identity of amphibian Hesperoherpeton
garnettense from the Upper Pennsylvanian of Kansas, 767
Mollusca: Estimating biomass and energy flow of molluscs in
palaeo-communities, 1
Monoporella nodulifera , 28

N

Neale, J. W. and Singh, P. Ostracoda from the Middle
Eocene of Assam, 355
Neogoniolithon strictum , 22, 23
Neograhamites carvonensis sp. nov., 71, 6
Neseuretus (Neseuretus) cf. tristani, 495, 47
Nicholls, E. L. and Russell, A. P. Structure and function of
the pectoral girdle and forelimb of Struthiomimus alius
(Theropoda: Orthinomimidae), 643

O

Oji, T. Early Cretaceous Isocrinus from north-east Japan, 629
Ordovician: review of Bryozoa, 147; trilobites, Saudi Arabia,
401; conodont faunas, Wales, 679
Orthoceras anceps, 25; obusum , 24
Osmotoxotis phalacra, 90

Ostracoda: Middle Eocene, India, 355; from lapetus ocean,
577

lOzarkodina pseudofissilis, 709, 84, 85

P

Pachydiscus ( Pachydiscus ) fresvillensis, 78, 8, 9; P. (P.)
jacquoti australis subsp. nov., 76, 8; P. (P.) neubergicius
dissitus subsp. nov., 72, 7, 9

Padian, K. Palaeontology Review: The origins and aero-
dynamics of flight in extinct vertebrates, 41 3
Paijenborchella ( Eopaijenborchella ) sp., 366, 43; (E.)

assamensis sp. nov., 365, 42, 43
Palaeontology Review: The origins and aerodynamics of
flight in extinct vertebrates by Kevin Padian, 413
Palaeophacmaea annulata, 104, 10
Palaeozoic: echinoderms, Britain, 527
Panderodus cf. P. gracilis , 708, 80
Paracytherideai superdimorphica sp. nov., 375, 45
Parasmittina nitida , 27

Partschiceras {Phyllop achy c eras) forbesianum, 43, 1
Paul, C. R. C. See Donovan, S. K. and Paul, C. R. C.

Paulchojfatia sp., 49, 50
Pedomphalella intermedia , 69 ; jonesii, 69
Pennsylvanian: amphibian, Kansas, 767
Permian: lycopod. South Afruca, 111; reptiles, Scotland, 207
Phanerozoic: Biotic diversification, 387
Phillips, D. The nautiloid Brachycycloceras in the Upper
Carboniferous of Britain, 235
Phragmodus cf. P. undatus , 707, 86

Phylloceras (Neophvlloceras) ramosum , 40, 1; (N.) surya, 42,

1,2

Pigg, K. B. and Rothwell, G. W. Cortical development in
Chaloneria cormosa (Isoetales), and the biological deriva-
tion of compressed lycophyte decortication taxa, 545
Pillsbury, S. W. See Chamberlain, .1. A. Jr. and Pillsbury, S. W.
Plaesiacomia vacuvertis , 406, 47; sp. aff. P. rara, 406, 47
Platybolbina ( Rimabolbina ) omphalata, 68; {R.) rima, 68
Plectodina bergstroemi sp. nov., 700, 82; bullhillensis sp., 702,
80, 81, 82, 83; tenuis, 704, 80, 81; IP. sp., 706, 80
Polonoceras planum, 17

Powell, E. N. and Stanton, R. J. Jr. Estimating biomass and
energy flow of molluscs in palaeocommunities, 1
Price, J. D. See House, M. R. and Price, J. D.

Prioniodus sp., 709, 84, 85
Propontocypris eocaenica, sp. nov., 365, 42
Protenaster antiaustralis, sp. nov., 318, 33; australis, 313, 32,
33; philipsi sp. nov., 322, 34; preaustralis sp. nov., 321,
33, 34

Proterozoic: macrobiota, Canada, 331; microfossils from
Spitzbergen, 451

Protopanderodus liripipus, 708, 86
Pseudohippula castorensis, 70; pseudoporkornina, 70
Pseudooneotodus sp. A, 709, 80, 83; sp. B, 709, 83, 86
Pseudophy llites indra, 50, 2, 3; latus, 50, 3
Pyriporopsis(l) texana , 28

R

Rayner, R. J. The Permian lycopod Clyclodendron leslii from
South Africa, 1 1 1

Reptilia: Permo-Triassic, Scotland, 207; protorothyridid
captorhinomorph. Carboniferous, Northumberland, 393
Reticycloceras sp. juv., 25; IR. sp., 25
Rhodesognathus elegans, 696, 82, 83, 84, 85, 86
Rickards, R. B. See Dumican, L. W. and Rickards, R. B
Romano, M. See El-Khayal, A. A. and Romano, M.
Rothwell, G. W. See Pigg, K. B. and Rothwell, G. W.
Russell, A. P. See Nicholls, E. L. and Russell, A. P.

S

Satka colonialica, 468, 53
Saudi Arabia: Ordovician trilobites, 401
Savage, N. M. and Bassett, M. G. Caradoc-Ashgill conodont
faunas from Wales and the Welsh Borderland, 679
Schallreuter, R. E. L. and Siveter, D. J. Ostracodes across
the lapetus Ocean, 577
Schizocythere deopanica sp. nov., 368, 43
Schizoporella floridana, 27; floridanal, 29
Schultze, H.-P. and Arsenault, M. The panderichthyid fish
Elpistostege: a close relative of tetrapods, 293
Scotland: Permo-Triassic reptiles from Elgin, 207; Carboni-
ferous brachiopods, 567

Scott, G. H. Homotaxy and biostratigraphical theory, 777
Semicytherura indica sp. nov., 378, 45, 46; S. indica sp. nov.
?subsp. nov., 380, 46

810

INDEX

Silurian: microfossil distribution in Wenlock type area, 89;

type species of Calymene, England, 783
Singh, P. See Neale, J. W. and Singh, P.

Siveter, D. J. See Schallreuter, R. E. L. and Siveter, D. J.
Siveter, D. J. The type species of Calymene (Trilobita) from
the Silurian of Dudley, England, 783
Skaale, S. G. See Fosse, G., Kielan-Jaworowska, Z. and
Skaale, S. G.

Smith, A. B. Cambrian eleutherozoan echinoderms and the
early diversification of edrioasteroids, 715
South Africa: Permian lycopod, 1 1 1

Stanley, G. D. Jr. and Kanie, Y. The first Mesozoic chondro-
phorine (medusoid hydrozoan) from the Lower Cretaceous
of Japan, 101

Stanton, R. J. Jr. See Powell, E. N. and Stanton, R. J. Jr.
IStaufferella sp., 708, 82
Steginoporella sp. nov., 31

Stephanocrinus ramsbottomi sp. nov., 534, 62 , 63 ; IS. sensu
lato sp., 537, 62 , 63

Stromatocystites pentangularis , 738, 87 , 89 ; walcotti, 740, 88

Struthiomimus altus , 643

Stygimys kuszmauli , 49 , 50

Slylopoma spongites , 30

Swett, K. See Knoll, A. H. and Swett, K.

Synsphaeridium sp., 468, 52

T

Tasmanites rifijicus , 465, 53
Tawuia dalensis , 334, 35 , 36 , 37 , 38

Taylor, P. D. and Curry, G. B. The earliest known fenestrate

bryozoan, with a short review of Lower Ordovician
Bryozoa, 147

Tetradelli scolti , 70 ; separata , 70
Thilipsura v-scripta , 90

Trace Fossils: Westphalian A, West Yorkshire, 619
Triassic: reptilia, Scotland, 207
Trilobita: Ordovician, Saudia Arabia, 401
Tyrasotaenia sp., 349, 35 ; TP. sp., 349, 35 , 36 , 37

U

Uroleberis armeniaca sp. nov., 381, 46

V

Venezuela: Devonian lycopods, 599

Vertebrates: origins and aerodynamics of extinct vertebrates,
413

W

W alcottidiscus typicalis , 746, 89
Wales: Ordovician conodont faunas, 679
Walker, A. D. See Benton, M. J. and Walker, A. D.
Wilbertopora mutabilis , 28

X

Xestoleberisl sp., 381, 46

Y

Youngibelus levis, 138, 15 ; tubularis, 136, 14

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. D. E. G. Briggs, Department of Geology, University of
Bristol, Wills Memorial Building, Queen’s Road, Bristol BS8 1R J, England, 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 1985 is £33 (U.S. $53) for Institutional
Members, and £16-50 (U.S. $27) 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 1LB, England.

RECENT PALAEONTOLOGICAL ASSOCIATION

PUBLICATIONS

Special Papers in Palaeontology

Numbers 119 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-jigs., 30 plates. Price £16 (U.S. $24).

21 . (for 1 978): 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 and g. l. williams. 100 pp., 24 text-figs., 22 plates. Price £15 (U.S. $23).

25. (for 1980): Stereom Microstructure of the Echinoid 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, 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.
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. 202pp., 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., 16 plates. Price £10 (U.S. $15).

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 + 241 pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24.95.

© The Palaeontological Association, 1985

Palaeontology

VOLUME 28 PART 4

CONTENTS

Xiphosurid burrows from the Lower Coal Measures (West-
phalian A) of West Yorkshire

J. I. CHISHOLM 619

Early Cretaceous Isocrinus from Northeast Japan

TATSUO oji 629

Structure and function of the pectoral girdle and forelimb of
Struthiomimus altus (Theropoda: Ornithomimidae)

E. L. NICHOLLS and ANTHONY P. RUSSELL 643

Caradoc-Ashgill conodont faunas from Wales and the Welsh
Borderland

N. M. SAVAGE and M. G. BASSETT 679

Cambrian eleutherozoan echinoderms and the early diversifi-
cation of edrioasteroids

ANDREW B. SMITH 715

Optimum preparation, preservation, and processing tech-
niques for graptolite electron microscopy

L. W. DUMICAN and R. B. RICKARDS 757

On the identity of the amphibian Hesperoherpeton garnettense
from the Upper Pennsylvanian of Kansas

ANDREW R. MILNER 767

Homotaxy and biostratigraphical theory

G. H. SCOTT 777

The type species of Calymene (Trilobita) from the Silurian of
Dudley, England

DEREK J. SIVETER 783

Instructions for authors 793

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

issn 0031-0239

£ co — c o _ co £

illSNI NVINOSHilWS S3IUVHan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHilt

Z W 2 W 2 CO 2

< s < s <

5

co 2 co ^ co ' z co

ARIES SMITHSONIAN INSTITUTION NOIiniliSNI_ NV1N0SH1IWS S3IUVH8n LIBRARIES SMITHSONS
co — co z \ 1/5 „ — _ zz co

IllSNI NVIN0SH1IIAIS S3IHVHan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIII'
■ z r- z r-

m

co ^ i: co — co — co

ARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IWS S3iavaail LIBRARIES SMITHSONI/

CO z co Z CO 2 .... CO

Lvl ? ^ A ? . 5 , ,,

> X'OmS'iV 5 -V > 5 5 >

2 co ‘ Z CO 2 CO *. 2

IllSNI NVIN0SH1IIAIS S3IMVdan LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1II

co — _ co

O

„ # =

CO ' = CO _

IllSNI NVINOSHIIWS SBIHVaail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1II

2 CO _ 2 , CO Z CO 2

^r 1

> 2 s ■ w >

CO 2 CO z to 2 CO

ARIES SMITHSONIAN INSTITUTION NOIinillSNI_ NVIN0SH1IWS S3ldVaan LIBRARIES SMITHSONI)

CO ~ CO - 3 » \ co — co

CO

o

2

I1I1SNI NVIN0SH1IIAIS S3IHVaai3 LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHill
r - z r- 2 ~ ~ ,

IcTX m ,, O rr, 2 .« m /K] S\ 2 m

CO

2

0 . — Zee 1 N. O . . . ^

zo

> ^
zo

—

co tz CO 2 co £ CO

ARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3iaVBail LIBRARIES SMITHSONI,

CO z v.. CO 2 CO 2 .... CO

< v - 2 .< 2

CO - co t: CO _

y VH a n ~LI B RAR I ES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IFNS S3IHVHSn LIBRAR
z c o ^ z » co z

< 2 ^ E ~ - /M?/'

z —i z ^

I xS§ 1 s

= » 1 . s g tiy

riTUTION ° ; N0liniliSNl''NVIN0SHilkNS C/, S3 I a va a n Z LI BRAR I ES^SMITHSONIAN INSTITUTION NOlinil

— co — co — co

~z zTSo^oX u X\F£IiL ( '.

co

O

z

avaan libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavaan libraf

z

o

TITUTION NOIlfUliSNl NVIN0SH1IIAIS S3iavaail LIBRARIES SMITHSONIAN INSTITUTION NOliOl

co

1 /*y

I w

__ _ CO * Z </> V Z

avaan libraries Smithsonian institution NouniiiSNi nvinoshiiws saiavaan libraf

co Z co — co ^

^ v W CO

co

^lili^y O X LN PL ^X — O __ Q

TITUTION NOlinillSNI ^NVINOSHIIWS SBIHVaan^LIBRARI E S^ SMITHSON IAN - " INSTITUTION NOlinil

O ^ ^ O X O

33
>

-

m *’ 2 XiVA^x cn X^vosv^x 'il m

co — co _ co

BRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IFNS S 3 iavaail L
z v co z w z

^ 2 < /To^x !E -

<%v .v\ =5 mM'/s,. z Aj£Wi&\ -h WV/A//S, Z

CO

TITUTION NOlinillSNI NVIN0SH1IWS S3IHVaan LIBRARIES SMITHSONIAN INSTITUTION NOlinil

co

o

z

avaan libraries Smithsonian institution NouniiiSNi nvinoshiifns saiavaan libraf

r- Z r- z r-

m ^ ^ m X°\ dc

co t: co X ± co

TiT’ 'OlinillSNI NVIN0SH1IIAJS S3ldVaan LIBRARIES SMITHSONIAN INSTITUTION NOlini

z _ ^ co Z w . w z co

S ^ «.0s\ . ^ ><usVaT\ ^