^ -

to rn xo^otx/ ™ r m

V 5 to - to ~ ~ to _

IONIAN INSTITUTION NOlinillSNI NVINOSHiilAIS S3iaVBai1 LIBRARIES SMITHSONIAN INSTITUTI

Z V' w Z CO 2 v. W 2: ...

< .-sfeA . - s .< I S s

^ 1 'W s ^

ZZ tn V

to ':y'2s ww to IsP* ^Sq) w jjwJx 10

I ^8^* 1 I I - >

oo *■ z 00 2 <0 * z co

ihihns saiavaan libraries Smithsonian institution NouniusNi nvinoshiiins saiavai

CO ~ to 5 to — CO

\

O X^acdCX _ '<58?^ O

_J 2 _j Z

jOnian institution NouniusNi NViNOSHims S3 lava an libraries Smithsonian institut

“ 2 v 2 r* z “

m x^osv^ X^Asiy rn v ^ X^As^ix m x^st^x m

CO ± CO ‘ — CO ~ CO

jHiiims saiavaan libraries Smithsonian institution NouniusNi nvinoshiiws saiavas

to 2 v CO z > <0 Z CO

»;. < V S < \v S <

*\ x m?’W-A 9 o x 9 ^ ,9

J CO \. CO to K 33 to

p7 x o x \%^..^l 9 #<y

q A 1 5 4EV S /Srt»££\ q

cn

, -y X ' t 2 Wd&^oV F ?syKT z , v /•if/' t SCUM'S/ 2

Z W •*' Z M •*“ Z CO 2

■0NIAN INSTITUTION NQIinillSNI_ NVIN0SH1IWS SBiavaail LIBRARIES SMITHSONIAN INSTITUTI

^ *** ^ -- ^ CO

«S» H , SrnS& “ H 55

f‘A H

CO

c

tLiVASt^^ Q “ ’’ O

Z ^ _| Z

ihiiws saiavaan libraries smsthsonian institution NounisiSMi nvinoshiiins sstavai
x 1 ra z ^ ✓ y« 2 5 A i

to xctoiii^x m xo/v. pcy 5; "^g*^ m ~ m

\ 5 CO ^ = W £ ' CO . „

IONIAN INSTITUTION NOlinillSNI NVINQSHliWS SaiHVaSH LIBRARIES SMITHSONIAN INSTITUT
z iS.. CO Z «2 Z ...<•• to z

> 2 > ■•r'?rs' 2 > ^^22- s

CO '*■ Z CO Z CO ®2~co

ihiiws S3 lava an libraries Smithsonian institution NouniusNi nvinoshisws ssiavai

CO X CO S CO ™ .„ . CO

JON IAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3 IHVaa 11 LIBRARIES SMITHSONIAN INSTITUT

f” Z r; v 2 r* Z *”

<ooT?oX O _ V 9 y<S?rirSX — O

03
P3
>

m z Xiwtvgz rn X^Th^Z ^ m

JH1IWS SBiavaail LIBRARIES SMITHSONIAN^SNSTITUTION^NOlinillSNI^NVINOSHimS S3IHVHE

S=fcv 1 s W g S V S < ><^X S

-h wWm>- 2 h /MWA z <%v ^ =5 454 2 -«

- v> ± co \ t: — CO =

inmsNi nvinoshiiins saiavaan libraries Smithsonian institution NOlinillSNI nvinos

z -- 00 z: ... co 2- co z

z r* Ay z

§

(/i v/ ~ Z c /) *•. 2 CO

RARIES SMITHSON!AN_ INSTITUTION NOlinillSNI NVINOSHIIINS S 3 I a V a a 11 LIBRARIES SMITHS
c^ z ^ _ co 5 </)

UJ XW^lpvTX U) XCIsvaT's. ~ jcSsCS. LlI xT^vXr-v "4- XTT<nrr>^

o ''icCLSi^' _. XiOms^X o " „ 0

LnillSNI~NVINOSHilWS S3 laVaail^LIBRARI ES^ SMITHSONIAN-* INSTITUTION 2 NOllDlllSNI^NVINOS
n •- 2 <z 2 r- ^ I- v

ro s ^ "vj^Ast^x pri x^ostQjx 2 1X1

RAR I ES SMITHSONIAN INSTITUTION NOlinillSNI “ NVINOSHIIINS^SS 1 H V a 0 II ” L < B R A R S ES ^SMITHS

^ z \ <o z co z •, <o .

S < V s < XS^X E < >K S ytrn

OA ^ >. .V _i MWZrYsS’*. •T' /o^X’STlftXjX I rfs/s/'' tV /,. 2;

^ CO

^/; t

^ Z Vvi!^x ^ V>N^' >' S >■ 5 ^ ^

LniUSNI^NVINOSHlIIAjS^SB I HVH 8 11 U B RAR I ES^SMITHSONIAN^INSTITUTION ^NOlinillSNI NVINOS
— ~ {/> «— </> 5

^ w !f W w

z

co xToq ^

• /SsfreJ

/k

c II

H \ft

O ~ O

RARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHiilNS^ S3 1 d Vd 8 11^ LIBRARIES SMITH'
z i~ — z r~ 7 r* z

S'^c'X m X<S«y'r _

|”d| = •

/ci --

Z W rn ^L£S^ £ \gc^>y m X ^ m B

®— ^ — (/) \ Z (/) ^

miiiSNi NviNOSHiuNS saiavaan libraries Smithsonian institution NOlinillSNI nvinoi

^ <■£ 2. « v • CO -y C/5 —

l/^^l

™ 1 t 8

,1

o /?i

CO

s ^2^ > ^ s ^ i ^ I ^

co z co » z — S z ‘ w

RARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSHIIINS SBiaVHail LIBRARIES SMITHS

u> z ,S) — CO — to

CcJ XJsvaSO/v/X UJ XjwiX z x5i\ . UJ X\V«vS\ ^ x^vlso^x UJ

CO

cc

5 | XNV^ ® 5 n

LflJ,iXSNI NVINOSHillAIS S3 IBVaail LIBRARI ES*2 SMITHSONIAN'* INSTITUTION 2 NOlifUllSNI^NVINOS
_ — L, v z r~ Z f“ z r* v

03
33

> I?!

x> z NsLi'/ASygX/^ rn X^vosvg^ ^ f?T) X^O<os>AXX' n^Jvas*^X (TJ

CO — CO — CO ™ CO

RARIES SMITHSONfAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3 1 a VH 8 11 LIBRARIES SMITH'
^ ^ Z co z co z <o _

- X E S S ymr, < W E

C'^'W; o X mZ'Wk ° x

CO CO x co

Published by

The Palaeontological Association • London
Price £32-50

THE PALAEONTOLOGICAL ASSOCIATION

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

COUNCIL 1991-1992

President '. Professor J. W. Murray, Department of Geology, The University, Southampton S09 5NH
Vice-Presidents: Dr p. R. Crowther, City of Bristol Museum and Art Gallery, Queen’s Road, Bristol BS8 1RL
Dr P. A. Selden, Department of Extra-Mural Studies, University of Manchester, Manchester M13 9PL
Treasurer : Dr M. E. Collinson, Department of Biology, King’s College, London W8 7AH
Membership Treasurer : Dr H. A. Armstrong, Department of Geological Sciences, The University, South Road,

Durham DH1 3LE

lnstitutional.Memhership Treasurer : Dr A. R. I. Cruickshank, Leicestershire Museums Arts and Records Service,

96 New Walk, Leicester LEI 6TD

Secretary : Dr J. A. Crame, British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 OET
Circular Reporter : Dr D. Palmer, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Marketing Manager (Sales): Dr C. R. Hill, Department of Palaeontology, British Museum (Natural History),

London SW7 5BD

Marketing Manager (Publicity) : Dr P. B. Wignall, Department of Earth Sciences, The University, Leeds LS2 9JJ
Public Relations Officer: Dr D. M. Martill, Department of Earth Sciences, The Open University,

Milton Keynes MK7 6AA

Editors

Dr J. E. Dalingwater, Department of Environmental Biology, University of Manchester, Manchester M13 9PL
Dr P. Doyle, Department of Earth Sciences, Thames Polytechnic, London El 2NG
Dr D. Edwards, Department of Geology, University of Wales College of Cardiff, Cardiff CF1 3 YE
Dr P. D. Lane, Department of Geology, University of Keele, Keele, Staffordshire ST5 5BG (co-opted)

Dr A. R. Milner, Department of Biology, Birkbeck College, Malet Street, London WC1E 7HX
Dr P. D. Taylor, Department of Palaeontology, British Museum (Natural History), London SW7 5BD

Other Members

Dr E. A. Jarzembowski, Brighton Dr W. J. Kennedy, Oxford Dr R. B. Rickards, Cambridge

Overseas Representatives

Argentina: Dr M. O. Mancenido, Division Paleozoologia invertebrados, Facultad de Ciencias Naturales y Museo, Paseo del
Bosque, 1900 La Plata. Australia: Dr K. J. McNamara, Western Australian Museum, Francis Street, Perth, Western
Australia 6000. Canada: Professor S. H. Williams, Department of Earth Sciences, Memorial University, St John’s,
Newfoundland A1B 3X5. China: Dr Chang Mee-mann, Institute of Vertebrate Palaeontology and Paleoanthropology,
Academia Sinica, P.O. Box 643, Beijing. Dr Rong Jia-yu, Nanjing Institute of Geology and Palaeontology, Chi-Ming-Ssu,
Nanjing. France : Dr J.-L. Henry, Institut de Geologie, Universite de Rennes, Campus de Beaulieu, Avenue du General
Leclerc, 35042 Rennes Cedex. Iberia : Prof. F. Alvarez, Departamento de Geologia, Universidad de Oviedo, C/. Jesus Arias
de Velasco, s/n. 33005 Oviedo, Spain. Japan : Dr I. Hayami, University Museum, University of Tokyo, Hongo 7-3-1, Tokyo.
New Zealand: Dr R. A. Cooper, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt. Scandinavia: Dr R.
Bromley, Fredskovvej 4, 2840 Holte, Denmark. U.S.A. : Prof. A. J. Rowell, Department of Geology, University of Kansas,
Lawrence, Kansas 66044. Prof. N. M. Savage, Department of Geology, University of Oregon, Eugene, Oregon 97403. Prof.
M. A. Wilson, Department of Geology, College of Wooster, Wooster, Ohio 44961. Germany: Prof. F. T. Fursich,
Institut fur Palaontologie, Universitat, D8700 Wurzburg, Pliecherwall 1

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1991 are:
Institutional membership . . . £60 00 (U.S. $108) Student membership . . . . £1 1-50 (U.S. $20)

Ordinary membership . £28-00 (U.S. $50) Retired membership .... £14-00 (U.S. $25)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr A. R. I.
Cruikshank, 72 Thirlmere Road, Hinckley, Leicestershire LE10 0PF. 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 H. A. Armstrong, Department of Geological Sciences, The
University, South Road, Durham DH1 3LE. Subscriptions cover one calendar year and are due each January; they should
be sent to the Membership Treasurer. All members who join for 1991 will receive Palaeontology , Volume 34, Parts 1-4.
Enquiries concerning back numbers should be directed to the Marketing Manager.

Non-members may subscribe, and also obtain back issues up to 3 years old, at cover price through Basil Blackwell Ltd,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, England. For older issues
contact the Marketing Manager.

Cover: Bolboforma intermedia Daniels and Spiegler (Incertae Sedis, possibly a calcified algal cyst) from Site 552A, southwest
margin of Rockall Plateau, late Miocene NN9-10. x 800. Bolboforma was planktonic and cysts are found in epicontinental
shelf sea deposits, thus providing a useful biostratigraphic link with oceanic sequences.

THE UPPER JURASSIC DIAPSID LISBOASAURUS
ESTESI-A MANIRAPTORAN TttEWtl

by ANDREW R. MILNER and SUSAN\E.

Abstract. Lisboasaurus estesi from the Upper Jurassic of Guimarota, Port ugUwasGifstdescn heel as an
anguimorph lizard. Reexamination of the holotype and referred specimens has revealed the presence of
thecodont teeth and an antorbital fossa, leading to the conclusion that L. estesi represents a small archosaur
and not a lizard. Features of the dentition including labio-lingually compressed crowns with unserrated
carinae, waisting between the root and the crown, and expanded roots, suggest that L. estesi belonged within
the troodontid dinosaur-bird clade and, less certainly, within the Avialae (Archaeopteryx plus all later birds).
It may thus be the earliest avialan. A second species in the genus, L. mitrocostatus, is based on indeterminate
material and is a nomen dubium restricted to the type specimen.

One of the most productive localities for Jurassic continental microvertebrates is the lignite mine
of Guimarota, near the town of Leiria in Portugal (Kiilme 1968). Until recently, the Guimarota
lignites were generally interpreted as Lower Kimmeridgian (Upper Jurassic) in age on the basis of
the ostracode assemblage (Helmdach 1971a, 19716, 1973). However, the plant material suggests
a Bathonian to Oxfordian age (Brauckmann 1978), and recent palynostratigraphical studies in the
Iberian Peninsula unequivocally support an age not later than Oxfordian for the Guimarota coal
seams (Mohr and Schmidt 1988; van Erve and Mohr 1988; Mohr 1989).

The Guimarota lignites are believed to have been laid down in swamps in an upper coastal plain
with slight marine influence (Mohr 1989, p. 293). They have yielded a rich assemblage of
microvertebrates, including fishes, amphibians, reptiles, and mammals, represented by both
dissociated and associated skeletons. Among the lower tetrapods are abundant lizards or lizard-like
forms which were described by SeifTert (1973) as scincomorphs ( Saurillus , Becklesisaurus), an
anguimorph (Introrsisaurus), and a new atypical anguimorph which SeifTert called Lisboasaurus and
which he suggested might be an aigialosaur/mosasaur relative. As described, Lisboasaurus was
represented by a small number of jaw fragments and isolated teeth which SeifTert divided between
the type species L. estesi and a referred species L. mitrocostatus. Estes (1983, pp. 116, 118, 133)
reassessed all SeifTert’s taxa, transferring the Guimarota Saurillus and Becklesisaurus species to the
new genera Saurillodou and Beck/esius respectively; and transferring Introrsisaurus to the
anguimorph genus Dorsetisaurus. Estes (1983, p. 193) removed Lisboasaurus from the Anguimorpha
and categorized it as Lacertilia incertae sedis, with the comment that the holotype maxilla of L.
estesi had some similarities to those of saurischian dinosaurs.

One of the authors (S.E.E.) has recently reexamined all the material of both species of
Lisboasaurus in the collections of the Freie Universitat, Berlin (=FUB), and the following
redescription elaborates on the conclusions of Estes (1983).

SYSTEMATIC PALAEONTOLOGY

Superorder archosauria Osborn, 1902
Order saurischia Seeley, 1888
Suborder theropoda Marsh, 1881
tetanurae Gauthier, 1986
maniraptora Gauthier, 1986
?avialae Gauthier, 1986 incertae sedis
Genus Lisboasaurus SeifTert 1973

IPalaeontology, Vol. 34, Part 3, 1991, pp. 503-513.]

© The Palaeontological Association

504

PALAEONTOLOGY, VOLUME 34

Type species. L. estesi Seiffert 1973.

Diagnosis. A small relative of troodontids and Archaeopteryx with the following characters: maxilla
bearing large pit or diastema in the anterior tooth-row; maxilla with antorbital fossa, but with
accessory antorbital fenestra absent or posteriorly placed; maxilla with slender dorsal process
excluded from narial margin by tall ascending process of premaxilla; teeth with mediolaterally
compressed triangular crowns with lingual groove, expanded roots wider than crowns and with
‘waist’ between crown and root; teeth with anterior and posterior carinae but no serrations.

Lisboasaurus estesi Seiffert, 1973

Text-figures 1-3.

1973 Lisboasaurus estesi Seiffert p. 33, fig. 27.

1983 Lisboasaurus estesi Seiffert; Estes p. 193, fig. 54a.

Holotype. FUB Gui.37, a nearly complete right maxilla about 22 mm long and bearing one tooth (Text-figs
I, 2a-d). Subsequent to the reexamination by S. E.E., and prior to photography, the holotype specimen was
cleaned of residual matrix and apparently suffered damage to the wall of the antorbital fossa and the posterior
end of the tooth-row. The photographs in Text-figure 1 depict it in this condition, while Text-figure 2a-c
depicts the lost regions with broken lines.

Diagnosis. As for genus, this being the only valid species.

Locality and horizon. Guimarota lignite mine, L5 km SSE of Leiria, Portugal; Guimarota complex of lignitic
marls, Oxfordian, Upper Jurassic.

Referred material. A larger left maxilla (FUB field batch no. Jun. 81:13:1; Text-fig. 3a-c) and several isolated
teeth (including FUB Gui.L. 136, Text-fig. 2e-f) are the only specimens that can certainly be referred here.

In addition to the above jaw material of L. estesi , Seiffert referred to it two small blocks containing
associated skull and postcranial material (FUB Gui . L . 33 and L . 1 77). The preservation of this material is very
poor. On the basis of these specimens, Seiffert stated that Lisboasaurus had procoelous vertebrae and fused
frontals. However there is nothing on these blocks to link them positively with either the holotype maxilla or
the referred teeth, and they are not treated as part of the hypodigm of L. estesi in the following description
and discussion.

Description

The holotype maxilla (Text-figs 1, 2a-c) is short, with a long obliquely-sloping anterior border which is faceted
and therefore did not enter the border of the external naris. The presence of a tall ascending process on the
premaxilla is implied. The dorsal process of the maxilla is slender and probably only contacted the nasal
weakly. The anterior tip of the maxilla is eroded, but there is no evidence that it bore a premaxillary process
extending below the naris. The surface of the bone is excavated laterally to give a smooth-walled depression.
Seiffert (1973) described this as an anteriorly shifted prefrontal facet, but it is too smooth to be a facet and is
the antorbital fossa as noted by Estes (1983). The depression had a thin medial wall (lost after recent
preparation), but its borders were damaged so that the shape of the fenestra, which must have lain further back,
cannot be reconstructed. No accessory fenestrae are visible in the preserved portion of the fossa wall. Medially
this part of the maxilla bears another, shallower depression (Text-fig. 2b) which could be part of a maxillary
sinus. This sinus occurs in a comparable position in the maxilla of Troodon (Text-fig. 4).

In the holotype maxilla, the dorsal process is separated from the alveolar region by a broad palatal shelf,
widest anteriorly, although the medial edge is broken (Text-fig. 2a). Below it, the teeth sat in a deep groove,
separated from one another by low ridges of interdental bone. There are spaces for about 13-14 teeth, but the
maxilla may be incomplete posteriorly. One tooth is preserved intact (Text-fig. 2d) and its shape, unique in the
context of known Jurassic microvertebrates, can be matched by a further set of isolated teeth (e.g. FUB
Gui . L. 136 in Text-figure 2e). The teeth have mediolaterally compressed triangular crowns with anterior and
posterior carinae. The labial surface is more rounded than the lingual one so that, in cross-section (Text-fig.
2f), the carinae appear to curve lingually. There are no serrations. The lingual surface is marked by a basal

MILNER AND EVANS: THEROPOD LISBOASAURUS

505

text-fig. 1 . Lisboasaurus estesi Seiffert, holotype right maxilla FUB Gui . 37 in A, labial, and b, lingual aspects.

Scale bar, 5 mm.

pit which gives rise to a shallow groove that extends towards, but does not reach, the tooth tip. The groove
is visible only on the lingual sides of teeth in both maxillary specimens and so the lingual-labial orientation
of isolated teeth is not in doubt. The crown is supported by an expanded root which is broader than the crown
and there is a clear ‘waist’ between the two. In the holotype, the root is just visible in the sole preserved tooth
which appears to be a partly erupted replacement tooth. This tooth is held in place by a low lingual wall of
bone which may represent either a remnant of the alveolar wall or a weak development of this feature. Thus
the implantation is thecodont. The expansion of the roots suggests that the roots fully enclosed the replacing
teeth.

At the anterior end of the tooth-row of the holotype maxilla, there is a large pit, subdivided into two parts.
The smaller anterodorsal section receives the opening of the superior alveolar canal (also visible in the maxilla
of Troodon , Text-fig. 4). The larger pit is more problematical. It is too smooth to be a facet, or a socket for
a large upper caniniform tooth (as suggested by Seiffert 1973 and Estes 1983). It appears to be a true diastema,
probably a recess for reception of a lower caniniform tooth.

506

PALAEONTOLOGY, VOLUME 34

text-fig. 2. Lisboasaurus estesi Seiffert. A-c, holotype right maxilla FUB Gui.37 in a, palatal, b, lingual, and
c, labial aspects, d. Maxillary tooth from FUB Gui.37 in lingual aspect (inverted for comparison with
Gui.L. 136). e, f. Referred isolated tooth FUB Gui.L. 136 in e, lingual aspect, and f, cross-section of crown.
All scale bars, 1 mm. Abbreviations: aof, antorbital fossa: li, lingual, mx.s, possible maxillary sinus; p, anterior

pit; ps, palatal shelf.

text-fig. 3. Lisboasaurus estesi Seiffert, referred left maxilla (FUB held batch no. Jun. 81 : 13: 1) in a. lingual
aspect, b, posterior view into anterior wall of antorbital fossa, and c, labial aspect. Scale bar, I mm.

Abbreviations as in Text-hgure 2.

A search through uncatalogued FUB material from Guimarota yielded a second incomplete maxilla (held
batch no. Jun. 81 : 1 3 : 1 ) - a left element somewhat larger than the holotype specimen. This specimen (Text-hg.
3 A-c) confirms most features of the holotype - antorbital fossa (Text-hg. 3b-c), maxillary sinus (Text-hg. 3a),
anterior pit (Text-hg. 3a), tooth shape and thecodont implantation - and suggests that the holotype maxilla
belonged to a juvenile. The thin outer wall of the anterior pit has broken away to give the appearance of a large
notch in the tooth-row (Text-hg. 3c). This referred specimen differs from the holotype in one respect, namely

MILNER AND EVANS: THEROPOD LI S BO AS A U RU S

507

that there is considerably more alveolar bone, particularly along the lingual side of the jaw (Text-fig. 3a) so
that the teeth are entirely in sockets. Whether or not this lingual bone was formed from interdental plates is
impossible to determine because of the poor preservation of the specimen. The difference between the two
specimens of Lisboasaurus estesi may reflect the ontogenetic development of the alveolar wall or could simply
be due to greater post-mortem damage to the holotype.

text-fig. 4. Troodon formosus Leidy, right maxilla in A
a, reconstructed cross-section (redrawn after Currie
1985, fig. 2) and b, lingual aspect. Scale bar, 10 mm.
Abbreviations: mx.s, maxillary sinus; s.a.c., superior
alveolar canal.

Systematic position

The holotype maxilla of L. estesi , with its long ascending process, antorbital fossa and thecodont tooth
insertion, clearly belongs to a small archosaur and not a lepidosaur (see Text-figure 5a for a mosasaur maxilla).
Resemblances can be found to the maxillae of several groups of archosaurs with small Jurassic representatives,
notably pterosaurs, primitive crocodylomorphs and higher maniraptoran theropods (troodontids and birds).
Because the specimens are small and plausibly juvenile, they probably lack the allometric features of larger
archosaurs of different groups and tend to have a phenetic resemblance to primitive pterosaurs and primitive
crocodylomorphs, most of which occurred significantly earlier than Lisboasaurus.

text-fig. 5. Lateral views of the skulls of a, the mosasaur Clidastes ; b, the pterosaur Preondactylus; c, the
crocodylomorph Protosuchus\ d, the theropod Compsognathus ; e, the troodontid Saurornithoides\ and, f,
Archaeopteryx. (Drawings after a, Russell 1967; b. Wild 1983; c, Crompton and Smith 1980; d, Ostrom 1978;

e, Russell 1969; f, Wellnhofer 1974.)

508

PALAEONTOLOGY, VOLUME 34

Pterosauria. The maxilla of L. estesi resembles that of some rhamphorhynchoid pterosaurs (e.g. the Norian
Preondactylus Wild 1983) in its long anterodorsal process and the possible presence of a caniniform tooth
(Text-fig. 5b). However in rhamphorhynchoids, the caniniform tooth is not at the anterior end of the tooth-
row, but in the middle, and is separated from the tip of the bone by a long slender process bearing a further
series of small teeth. This does not appear to be the condition in L. estesi. There are further differences. In
Lisboasaurus , the dorsal process is faceted and was probably excluded from the narial margin, unlike that of
any described rhamphorhynchoid, although the character is present in pteranodonts. The rhamphorhynchoid
maxilla is slender and lightly-built and it lacks an antorbital fossa set in the front of the fenestra. Tooth shape
also differs; pterosaurs do not have the combination of waisted teeth and sharply expanded roots. One Upper
Jurassic genus, Germanodactylus, does have teeth with short triangular crowns each with a groove running
down one side (Wellnhofer 1970, pi. 10, fig. 2, and D. M. Unwin pers. comm.) but the groove is labial, not
lingual as in Lisboasaurus , and the maxilla of Germanodactylus is of very different shape. In conclusion,
Lisboasaurus might be an aberrant pterosaur, but as will be argued below, its total suite of characters suggest
a closer resemblance to another group.

Crocodylomorph and avialan teeth. Martin et al. (1980) and Martin (1983b) noted the resemblance of crocodile
and Mesozoic bird teeth and suggested that they indicated immediate relationship. Most other workers accept
the characters but treat them as convergent (e.g. Gauthier 1986). The teeth of Lisboasaurus estesi bear a distinct
resemblance to the crocodile-Mesozoic bird tooth type in that they possess expanded roots, waists, and non-
serrate carinae. The only group in which these teeth occur in combination with a large antorbital fenestra, are
the primitive crocodylomorphs of the sphenosuchian grade and the higher maniraptoran theropod dinosaurs
of the Troodontidae and Avialae (birds and relatives).

Crocodylomorpha. At least four genera of crocodile are known from Guimarota, comprising Machimosaurus ,
Goniopholis, a Theriosuchus-Mke form, and a Bernissartia- like form (Brinkmann 1989). All are typical
crocodylomorphs in that they have very reduced or no antorbital fenestrae, and are thus distinct from
Lisboasaurus. However, members of the primitive pre-crocodyliform grade of crocodiles retain a relatively
large fenestra. Most of these are Upper Triassic or Liassic in age, but one, Hallopus victor , is from the Upper
Jurassic of Colorado (Walker 1970) and hence contemporaneous with Lisboasaurus. The only specimen of
Hallopus lacks cranial material, so comparison must be made with Upper Triassic and Liassic genera.
Unfortunately, most of these are represented by specimens which are considerably larger than Lisboasaurus
and so comparisons of shape are difficult.

The recent evaluation of primitive crocodylomorph relationships by Benton and Clark (1988) places the well-
characterized primitive crocodylomorphs in two grades. The basal grade includes Saltoposuchus (= Ter-
restrisuchus of Crush 1984). The ' Terrestrisuchus' material described by Crush (1984) has a much more
elongate maxilla than Lisboasaurus with an anteroposteriorly long dorsal ramus. The teeth of Terrestrisuchus
are flattened like those of Lisboasaurus but are recurved and serrated. Neither maxilla nor teeth of
Terrestrisuchus could be confused with Lisboasaurus material.

Slightly more advanced crocodylomorphs include the sphenosuchids Sphenosuchus and Hesperosuchus and
the aberrant Platyognathus. Sphenosuchus and Hesperosuchus have teeth with compressed lanceolate crowns,
waists and cylindrical roots (Walker 1970, p. 348, fig. 12) which resemble those of Lisboasaurus. The teeth of
Sphenosuchus are serrated, while Platyognathus has unique recurved teeth with a polygonal cross-section and
a serrated posterior edge (Simmons 1965, pp. 35, 39^40). Unlike Lisboasaurus , these Triasso-Liassic forms
possess a maxilla with an anteroposteriorly long dorsal process and a posteriorly set fossa associated with the
elongate anterior muzzle. Sphenosuchus and Platyognathus both possess a large anterior mandibular fang
which fits into a deep notch in the upper jaw, but in both genera, the notch is at the premaxilla-maxilla junction
and not a pocket within the maxilla as in Lisboasaurus. This condition is shown well in the more advanced
Protosuchus (Text-fig. 5c), as described by Crompton and Smith (1980). Thus the teeth of Lisboasaurus share
some characteristics with those of early crocodylomorphs, but the maxilla shows no special similarity.

Maniraptoran Theropoda. The term Maniraptora was created by Gauthier (1986, p. 30) to define a clade of
bird-like theropod dinosaurs and birds. The group, as defined by Gauthier, comprises Compsognathus and
several minor ‘coelurosaurs’, the families Caenagnathidae, Elmisauridae, Dromaeosauridae, Troodontidae,
and the Avialae ( Archaeopteryx and all birds). Gauthier (1986) suggested that the Deinonychosauria
(Dromaeosauridae + Troodontidae) were the sister-group of the Avialae, but Currie (1987) has argued that the
Troodontidae alone are the immediate sister-group of the Avialae, and noted some derived characters which
the two groups share uniquely within the Theropoda. The characters of Lisboasaurus relate largely to the

MILNER AND EVANS: THEROPOD LISBOASAURUS

509

troodontid-avialan clade which at present has no name. For the purposes of this discussion, the Troodontidae
is taken to comprise the genera Troodon ( = Stenonychosaurus , Pectinodon), Saurornithoides, and Borogovia
(for recent reviews, see Currie 1987 and Osmolska 1987). All these genera are from the Upper Cretaceous, but
if Currie (1987) is correct in arguing that the troodontids are the sister-group to the Avialae then there must
have been at least stem-troodontids in the Upper Jurassic, contemporaneous with Archaeopteryx.

The maxilla of Lisboasaurus is very similar in shape to those of small maniraptorans, such as Compsognathus
(Text-fig. 5d) and Archaeopteryx (Text-fig. 5f). As noted in the description, the anterior maxillary teeth of
Lisboasaurus have compressed asymmetrical cross-section with rounded labial and flattened lingual surfaces
and the carinae on the lingual corners. Currie (1987, p. 77) noted this to be a character of the premaxillary teeth
of theropods, and it would not be surprising if it extended to the anterior maxillary teeth. Compressed
asymmetrical crowns of this type also occur in Archaeopteryx (Flowgate 1984ft, p. 656), but carinae are not
visible (P. Wellnhofer pers. comm.). The following characters are discussed in the context of the Maniraptora.
The condition in troodontids is taken from Currie (1987) and the condition in Archaeopteryx is taken from
Howgate 1984a, ft, and Wellnhofer (pers. comm.).

The following derived character is shared by troodontids, Archaeopteryx , and Lisboasaurus.

A. Teeth with constriction between crown and root (Text-fig. 6b-e contra the primitive theropod condition,
e.g. Megalosaurus in Text-fig. 6a).

The following derived character is shared by Archaeopteryx and Lisboasaurus , but not troodontids.

B. Teeth completely unserrated (Text-fig. 6c-d; Gauthier 1986, p. 12).

text-fig. 6. Archosaur teeth in lingual aspect, a, Megalosaurus bucklandi (original from mandible of holotype,
British Museum (Natural History) R.332; serrations present but not visible at this scale); b, Troodon (after
Currie 1987); c, Lisboasaurus ; D, Archaeopteryx (after Howgate 1984ft); e, hesperornithid (after Martin

1983a). Scale bars, 10 mm (a) and 1 mm (b-e).

Compsognathus lacks serrations on the premaxilla and anterior dentary teeth, but these teeth are recurved,
slender-crowned, and also lack carinae. The maxillary teeth of Compsognathus have carinae, the posterior
carina being serrated (Ostrom 1978). ( Archaeopteryx and Lisboasaurus also share some primitive features
contra troodontids. They appear to retain a maxillary tooth count of 12-15, whereas troodontids have 19-20
maxillary teeth, but the latter is probably the derived condition. Unlike troodontids, both Archaeopteryx and
Lisboasaurus retain the antorbital fossa, but Gauthier (1986) described this character state as primitive for
archosaurs.)

The following derived character is shared by Lisboasaurus and troodontids but is not known in
Archaeopteryx.

C. Maxillary teeth each have a depression at the base of the lingual side of crown which extends as a groove
towards the crown tip but does not reach it (Text-fig. 6b-c).

(There are other differences of uncertain polarity in the shape of the crowns, those of Archaeopteryx being more
recurved (Howgate 1984a, ft) than those of Lisboasaurus.)

The following derived characters are shared by Lisboasaurus and the Ornithurae (post -Archaeopteryx birds),
but not troodontids or Archaeopteryx.

510

PALAEONTOLOGY, VOLUME 34

Dl. Maxilla separated from naris by processes of premaxilla and nasal.

In Archaeopteryx , unlike Lisboasaurus , the anterodorsal process of the premaxilla is shorter and reconstructions
show the maxilla entering the narial margin, a typical saurischian condition (Howgate 1984a). Gauthier (1986,
p. 15) notes that exclusion of the maxilla from the naris occurs as the primitive thecodontian condition, is lost
in Saurischia, but appears as a reversal in ornithomimids and ornithurine (post -Archaeopteryx) birds.

D2. Teeth with root wider than crown (Text-fig. 6c, e; Gauthier 1986, p. 12).

The following derived character occurs in Lisboasaurus but not in troodontids or Archaeopteryx.

E. Large pocket or diastema near anterior end of maxilla suggesting presence of a lower caniniform tooth.

Relationships. If Lisboasaurus estesi is a maniraptoran theropod. then character A places it in the troodontid-
avialan clade, B places it within the Avialae, while in contradiction, C places it within the Troodontidae. Dl
and D2 are weak (because variable) characters which suggest a post- Archaeopteryx position in the Avialae and
E is a unique character in the context of a troodontid-avialan clade and serves as a defining autapomorphy for
Lisboasaurus. Text-fig. 7 depicts the alternative relationships of Lisboasaurus suggested by these characters.
This character distribution suggests that Lisboasaurus is a member of the troodontid-avialan clade and could
be a primitive avialan. The implications and qualifications to this systematic position are discussed below in
the discussion section.

text-fig. 7. Cladogram depicting the possible relationships of Lisboasaurus to troodontid dinosaurs,
Archaeopteryx , and Aves. Character-states a-e are those described in the text. Abbreviation : L, Lisboasaurus.

A NOTE ON ‘ LISBOASAURUS ’ M ITROCOSTATUS

The material which forms the basis of this species is more problematical than that of L. estesi. The holotype
dentary (FUB Gui . 34) is crushed, but appears to show a type of thecodont implantation essentially similar to
that of L. estesi. The external surface of this dentary bears a pattern of striate sculpture which is unlike that
of any lizard (Estes 1983). Some of the referred teeth resemble those of L. estesi but have a short rounded crown
(e.g. FUB Gui . L. 67). Others, however, have a longer pointed crown (e.g. FUB Gui . 18 and 24). This variation
suggests that more than one taxon may be represented - a possibility recognized by Seiffert (1970), when he

MILNER AND EVANS: THEROPOD LISBOASAURUS

511

divided the L. mitrocostatus material into two subspecies. Estes (1983) suggested that at least some of the
L. mitrocostatus material might be referrable to the genus Cteniogenys which is also known from Guimarota
(Seiffert 1973, pp. 13-17). Cteniogenys was originally described as a lizard but has recently been shown to be
an early choristodere (Evans 1989, 1990). Although superficially similar, the striations on the jaw of
Cteniogenys are created by a double row of sensory foramina, not the rather irregular sculpturing seen in the
L. mitrocostatus holotype. The teeth and their mode of implantation are also different, and reference to
Cteniogenys is highly unlikely. The L. mitrocostatus material is too limited to permit firm conclusions, but we
agree with Estes (1983) that there is no certainty that the two species are congeneric. In conclusion
‘ Lisboasaurus' mitrocostatus is, at present, a nomen dubium restricted to the holotype dentary, which is not
sufficiently determinate to be associated with L. estesi , but certainly does not belong to Cteniogenys or a true
lizard.

DISCUSSION

This study commenced as an assessment of the lacertilian status of Lisboasaurus estesi, and the
unequivocal conclusion is that the L. estesi material is not lacertilian but represents a small
archosaur. Our further conclusions must, of necessity, be circumspect for two reasons. The first is
the limited nature of the material. Maxillae and teeth can only bear a restricted set of characters and
some of these occur convergently in more than one group of small archosaurs. The resemblances
of crocodile and Mesozoic bird teeth noted by Martin et al. (1980) and Martin (1983a, b) are
generally agreed to be due to convergence (e.g. Gauthier 1986) but are clearly a source of
uncertainty when limited material such as Lisboasaurus is being studied. However, parsimonious
treatment of the available restricted suite of characters does lead to a position in the troodontid-
avialan clade with no character contradictions. A less certain parsimonious conclusion, because
there is contradiction, is the position within the avialan clade.

A second reason for circumspection is that the early evolution of birds is both poorly known and
subject to intense speculation. An Oxfordian relative of Archaeopteryx might be the earliest known
bird, and yet there is nothing in the morphology of the Lisboasaurus fragments that has any bearing
on the origin of birds or bird flight. It does, however, serve as a reminder that Archaeopteryx was
probably only one of an array of similar smaller forms, including stem-troodontids. It is likely that
most Jurassic microvertebrate assemblages will prove to contain maniraptoran material which
cannot be securely identified as avian or non-avian, the recently described material from the late
Jurassic Uncompahgre Formation of Colorado being a case in point (Jensen and Padian 1989). A
more precise systematic position for Lisboasaurus must await the identification and description of
further material from Guimarota.

It would not be too surprising if Lisboasaurus estesi were an Archaeopteryx- grade maniraptoran.
The Solnhofen Limestone represents a coral lagoon floor of Lower Tithonian age (circa 150 Ma)
in southern Germany. The Guimarota lignite represents a coastal swamp of probable Oxfordian age
(circa 160 Ma) in Portugal. The two localities are about 1500 km apart and may have been about
1200 km apart in the Jurassic. The presence of similar proto-birds in two contemporaneous
continental-coastal localities 1200 km apart within one continent could reasonably be expected.

Acknowledgements. We would like to express our thanks to Professor Bernard Krebs, Freie Universitat, Berlin,
for access to the material discussed in this paper, and for arranging for photography and further preparation
of the holotype of L. estesi. We also thank Mr David Unwin for information on pterosaur teeth, and Drs Peter
Wellnhofer, David Norman, Angela Milner, and anonymous referees for helpful criticism of the typescript.

REFERENCES

benton, m. j. and clark, j. m. 1988. Archosaur phylogeny and the relationships of the Crocodylia. 295-338.
In benton, m. J. (ed.) The Phylogeny and Classification of the Tetrapods. Volume 1 : Amphibians , Reptiles ,
Birds. Clarendon Press, Oxford.

512

PALAEONTOLOGY, VOLUME 34

brauckmann, c. 1978. Beitrag zur Flora der Grube Guimarota (Ober Jura: Mittel-Portugal). Geologic/ et
Paleontologia , 12, 213-222.

brinkmann, w. 1989. Vorlaufige Mitteilung iiber die Krokodilier-Faunen aus dem Ober-Jura (Kimmeridgium)
er Kohlegrube Guimarota, bei Leiria (Portugal) und der Unter-Kreide (Barremium) von Una (Province
Cuenca, Spanien). Documenta Naturae, 56, 1-28.

crompton, a. w. and smith, k. k. 1980. A new genus and species of crocodilian from the Kayenta Formation
(Late Triassic?) of Northern Arizona. 193-217. In Jacobs, l. (ed.) Aspects of Vertebrate History. Museum
of Northern Arizona Press, Flagstaff, Arizona.

crush, p. j. 1984. A late Upper Triassic sphenosuchid crocodilian from Wales. Palaeontology, 27, 131-157.
Currie, p. j. 1985. Cranial anatomy of Stenonychosaurus inequalis (Saurischia, Theropoda) and its bearing on
the origin of birds. Canadian Journal of Earth Sciences, 22, 1643-1658.

— 1987. Bird-like characteristics of the jaws and teeth of troodontid theropods (Dinosauria, Saurischia).
Journal of Vertebrate Paleontology, 7, 72-81.

erve, a. v. and mohr, b. 1988. Palynological investigations of the Late Jurassic microflora from the vertebrate
locality Guimarota coal mine (Leiria, Central Portugal). Neues Jahrbuch fur Geologie und Pa/donto/ogie,
Monatshefte, 1988, 246-262.

estes, R. 1983. Sauria terrestria, Amphisbaenia. 1-249. In wellnhofer, p. (ed.) Handbuch der Paldoherpetologie
Teil 10A. Gustav Fischer, Stuttgart.

evans, s. e. 1989. New material of Cteniogenys (Reptilia: Diapsida; Jurassic) and a reassessment of the
systematic position of the genus. Neues Jahrbuch fur Geologie und Palaontologie, Monatshefte, 1989,
577-589.

— 1990. The skull of Cteniogenys, a choristodere (Diapsida: Archosauromorpha) from the Middle Jurassic
of Oxfordshire, England. Zoological Journal of the Linnean Society, 98, 205-237.

Gauthier, J. 1986. Saurischian monophyly and the origin of birds. Memoirs of the California Academy of
Science , 8. 1-55.

helmdach, f. f. 1971a. Stratigraphy and ostracod-fauna from the coalmine Guimarota (Upper Jurassic).
Memorias dos Servicos Geologicos de Portugal ( Nova Serie ), 17, 41-88.

— 19716. Zur Gliederung limnisch-brackischer Sedimente des portugiesischen Ober Jura. Neues Jahrbuch
fur Geologie und Palaontologie, Monatshefte, 1971, 645-662.

— 1973. A contribution to the stratigraphical subdivision of non-marine sediments of the Portuguese Upper
Jurassic. Comunicacoes dos Servicos Geologicos de Portugal, 58, 5-21.

howgate, M. E. 1984a. The teeth of Archaeopteryx and a reinterpretation of the Eichstatt specimen. Zoological
Journal of the Linnean Society, 82, 159-175.

1984 b. On the supposed difference between the teeth of the London and Berlin specimens of
Archaeopteryx lithographica. Neues Jahrbuch fur Geologie und Palaontologie, Monatshefte, 1984, 654-660.
jensen, j. a. and padian, K. 1989. Small pterosaurs and dinosaurs from the Uncompahgre fauna (Brushy Basin
Member, Morrison Formation: ?Tithonian), late Jurassic, western Colorado. Journal of Paleontology, 63,
364-373.

kuhne, w. G. 1968. History of discovery, report on the work performed, procedure, technique and generalities.

Memorias dos Servicos Geologicos de Portugal ( Nova Serie), 14, 1—20.
martin, l. d. 1983a. The origins of birds and of avian flight. Current Ornithology, 1, 105-129.

19836. The origin and early radiation of birds. 291-338. In brush, a. l. and clark, g. a. Jr. (eds)
Perspectives in Ornithology. Cambridge University Press, Cambridge, 560 pp.

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

mohr, b. a. r. 1989. New palynological information on the age and environment of Late Jurassic and Early
Cretaceous vertebrate localities of the Iberian Peninsula (eastern Spain and Portugal). Berliner
geowissenschaftliche Abhandlung, Reihe A, 106, 291-301.

— and schmidt, D. 1988. The Oxfordian/Kimmeridgian boundary in the region of Porto de Mos (Central
Portugal): stratigraphy, facies and palynology. Neues Jahrbuch fur Geologie und Palaontologie, Abhand-
lungen, 176, 245-267.

osmolska, H. 1987. Borogovia gracilicrus gen. et sp. n., a new troodontid dinosaur from the late Cretaceous
of Mongolia. Acta Palaeontologica Polonica, 32, 133-150.
ostrom, j. h. 1978. The osteology of Compsognathus longipes Wagner. Zitteliana, 4, 73-118.
russell, d. a. 1967. Systematics and morphology of American mosasaurs. Peabody Museum of Natural
History , Bulletin, 23, 1-241.

MILNER AND EVANS: THEROPOD LISBOASAURUS

513

— 1969. A new specimen of Stenonychosaurus from the Oldman Formation (Cretaceous) of Alberta.
Canadian Journal of Earth Sciences , 6, 595-612.

seiffert, j. 1970. Oberjurassische Lazertilier aus der Kohlengrube Guimarota bei Leina (MittelPortugal).
Unpublished Inaugural-Dissertation, Freie Universitat Berlin.

— 1973. Contribuicao para o conhecimento de Fauna do Kimeridgiano do Mina de Lignito Guimarota
(Leiria Portugal). Ill parte. Upper Jurassic lizards from Central Portugal. Memorias dos Servicos Geologicos
de Portugal ( Nova Serie ), 22, 7-88.

simmons, d. j. 1965. The non-therapsid reptiles of the Lufeng Basin, Yunnan, China. Fieldiana: Geology, 15
(1), 1-93.

walker, a. d. 1970. A revision of the Jurassic reptile Hallopus victor (Marsh), with remarks on the
classification of crocodiles. Philosophical Transactions of the Royal Society of London, Series B , 257,
323-372.

wellnhofer, p. 1970. Die Pterodactyloidea (Pterosauria) der Oberjura-Plattenkalke Siiddeutschlands.
Abhandlungen, Bayerische Akademie der Wissenschafte, mathematisch-naturwissenschaftliche Klasse , 141,
1-133.

— 1974. Das fiinfte Skelettexemplar von Archaeopteryx. Paldontographica (A), 147, 169-216.

wild, r. 1983. A new pterosaur (Reptilia: Pterosauria) from the Upper Triassic (Norian) of Friuli, Italy. Atti
del Museo Friulano di Storia Nciturale , 5, 45-62.

ANDREW R. MILNER
Department of Biology
Birkbeck College
Malet Street
London WC1E 7HX

Typescript received 28 February 1990
Revised typescript received 2 October 1990

SUSAN E. EVANS

Department of Anatomy and
Developmental Biology
University College London
Rockefeller Building, University Street
London WC1E 6JJ

THE CONCHOSTRACAN FAUNA OF
THE GREAT ESTUARINE GROUP, MIDDLE
JURASSIC, SCOTLAND

by chen pei-ji and j. d. Hudson

Abstract. The Great Estuarine Group contains the most diverse conchostracan fauna so far described from
the Jurassic or Cretaceous of Europe, comprising twelve species in seven genera. Estheria murchisoniae Jones
was described in the last century; it is now referred to Pseudograpta Novojilov and is the youngest member
(latest Bathonian?) of the fauna described here. Euestheria trotternishensis and Neopolygrapta lea/tensis spp.
nov. occur near the base of the Great Estuarine Group and probably close to the Bajocian-Bathonian
boundary. Dendrostracus hebridesensis sp. nov. occurs towards the top of the Kildonnan Member. Lealt Shale
Formation. In the overlying Lonfearn Member, Skyestheria intermedia sp. nov. appears (probably descended
from Neopolygrapta) and itself gives rise to Antronestheria praecursor sp. nov. The higher parts of the Great
Estuarine Group, including most of the Kilmaluag Formation, are dominated by Antronestheria kilmaluagensis
sp. nov., with one occurrence of Fibrestheria puncta sp. nov. The Pseudograpta fauna (containing P. orbita
Chen, P. morrisi sp. nov., P.jonesi sp. nov. and P. murchisoniae (Jones)) occurs in beds probably near the top
of the Kilmaluag Formation. All the conchostracans inhabited shallow, freshwater to oligohaline, near-coastal,
lagoons. Biogeographic comparisons with China show that the earlier faunas were distinct in the two areas,
but the Pseudograpta fauna occurs in both. Details of the construction of conchostracan growth bands are
revealed by stereo scanning electron micrographs.

Conchostracans are small, bivalved branchiopod crustaceans with a non-mineralized carapace.
They occur in non-marine facies from the Devonian to the present day, sporadically but often in
great abundance; many palaeontologists still refer to them informally using the pre-occupied generic
name ‘ Estheria ’. Jurassic and Cretaceous conchostracans have been little studied in Europe, no
doubt partly because of the limited occurrence of suitable facies, but they have been extensively
studied in Siberia and China, where they are important biostratigraphically within thick successions
of non-marine strata deposited in large internal basins (e.g. Sichuan basin, Sungliao basin: Chen
and Shen 1985; general summary in English, Chen et al. 1982, pp. 1011-1014). It is therefore
important to examine conchostracan faunas in Europe that can be accurately dated by intercalation
within marine sections.

In Britain, conchostracans occur in the Triassic, including the Penarth Group (' Rhaetian ’),
sporadically in the Bathonian of central England, in the Purbeck Beds (top Jurassic-early
Cretaceous) and in the Wealden Beds (early Cretaceous). By far the most abundant Jurassic faunas,
however, are those of the Great Estuarine Group of the Inner Hebrides, NW Scotland, where
conchostracans occur profusely, associated with the non-marine end of a series of faunal
assemblages ranging from restricted-marine to freshwater (Hudson 1963a, 1963b, 1980). The age is
late Bajocian-Bathonian. This occurrence is of interest from several points of view: local and
worldwide biostratigraphical, palaeoecological, and palaeo-biogeographical. Our scanning electron
microscope (SEM) studies have also enabled us to add important morphological details to existing
knowledge of the conchostracan carapace, and we illustrate conchostracan sculpture stereo-
scopically for the first time.

THE GREAT ESTUARINE GROUP

The Great Estuarine Group is an intercalation of varied paralic facies, up to 270 m thick, in the
otherwise marine Jurassic succession of the Minch Basin (Hudson 1983). Its lithostratigraphy has

(Palaeontology, Vol. 34, Part 3, 1991, pp. 515-545, 10 pls.|

© The Palaeontological Association

516

PALAEONTOLOGY, VOLUME 34

been reviewed by Harris and Hudson (1980). The base of the Group is defined at the upward
passage from ammonite-bearing clays of the garantiana zone (late Bajocian) into dark shales, with
a restricted fauna, of the Cullaidh Shale Formation. Its top occurs where the non-marine
Skudiburgh Formation is abruptly overlain by the Upper Ostrea Member of the Staffin Bay
Formation; this is dated palynologically as no older than the discus zone (latest Bathonian), and is
overlain by the Belemnite Sands Member, macrocephalus zone, early Callovian (Bradshaw and
Fenton 1982). The Great Estuarine Group may therefore include equivalents of the latest Bajocian
parkinsoni zone, but most of it must be Bathonian. It can be divided into several distinctive
formations that can be correlated lithostratigraphically (Harris and Hudson 1980; see Text-fig. 1),

Stratigraphic Ranges of Conchostraca in the
Great Estuarine Group

SKUDIBURGH FM

KILMALUAG FM

DUNTULM FM

VALTOS FM

LONFEARN MBR

KILDONNAN MBR

ELGOL FM

CULLAIDH FM

(Not to scale)

I k '

i k
i k

...J P

2

o

DO

CL

I

CO

o

LU

= _Q

o3

ID

CD

E

t/)

CO k_

c o
(D to

o> ^

g 3

-=5 CD
03 03
£ Q.

<

CO

X

O

cc

CL

Q_

<

5

O)

>

o

Q.

O

CD

CD

O

c

CO

text-fig. 1. Stratigraphic subdivisions of the Great Estuarine Group (Harris and Hudson 1980), showing

ranges of the main conchostracan species.

CHEN AND HUDSON: JURASSIC CONCHOSTR ACANS

517

but no internal biostratigraphy is currently available. It seems likely, on grounds of regional event
stratigraphy, that the Duntulm Formation correlates with the White/Blisworth Limestones of
central England (Andrews 1985) and thus most probably with the hodsoni zone of the Bathonian
(Torrens in Cope et al. 1980).

Within the Great Estuarine Group, the predominantly shale, low-salinity formations, especially
the Lealt Shale Formation (both Kildonnan and Lonfearn Members) and the Kilmaluag
Formation, contain abundant conchostracans. There are more limited occurrences in the Cullaidh
Shale Formation, in the shale facies within the Valtos Formation, and in a freshwater intercalation
within the Duntulm Formation. A definite stratigraphical arrangement of the conchostracan taxa
emerges from our present sampling, and suggests that they will be useful in local and perhaps wider
stratigraphical correlation.

PREVIOUS WORK ON CONCHOSTRACANS FROM THE GREAT ESTUARINE

GROUP

Jones (1863) described the species Estheria murchisoniae from Skye; this was subsequently made the
type species of the genus Pseudograpta Novojilov, 1954. No systematic descriptions or illustrations
have appeared since, although conchostracans have been recorded from the Great Estuarine Group
many times, variously as Estheria, Euestheria or Cyzicus, and gave their name to the Estheria Shales
of Anderson (e.g. Anderson and Dunham 1964), now the Lealt Shale Formation of Harris and
Hudson (1980). Hudson (19636, p. 337) recorded conchostracans from several horizons within the
Group and reported that ‘each sample was reasonably homogeneous within itself, but that there
was considerable variation between samples, especially in interspace ornament and also in size’.
This suggested that several taxa were present, but in view of the then-prevailing confusion in
conchostracan taxonomy, Hudson retained all the specimens in Jones’s species murchisoniae , which
he referred to Euestheria Deperet and Mazeran.

MATERIAL STUDIED

In 1988 we made two visits to the Hebrides specifically to collect material for this paper. We have
also made extensive use of collections made by Hudson since 1963, in the collections of the Leicester
University Geology Department, and between 1956-62, in the Sedgwick Museum, Cambridge. All
material from the Great Estuarine Group used for this paper is listed in an Appendix, which is
deposited at the British Library as Supplementary Publication No. SUP 14039 (1 1 pages). We have
also re-studied the specimens described by Jones ( 1 863) at the British Geological Survey, Keyworth,
a limited amount of Geological Survey material from southern England, and some core material
loaned by British Petroleum pic from the Porcupine Basin. Chen has made extensive use of the
collections of the Nanjing Institute of Geology and Palaeontology for comparative purposes, and
some material collected jointly in China in 1989, or presented to Hudson by Chen, is now in the
Leicester University collections.

DISTRIBUTION OF CONCHOSTRACAN FAUNAS

The stratigraphy of the Great Estuarine Group and the distribution of conchostracans within it are
summarized in Text-figure 1.

Lealt Shale Formation and Cullaidh Shale Formation

Stratigraphy. It is convenient to consider the two lowest shale formations together, because from
the point of view of conchostracan distribution the unfossiliferous Elgol Sandstone Formation can
be regarded as an intercalation within them. We have sampled the Cullaidh Shale Formation and
the basal part of the Kildonnan Member of the Lealt Formation at two localities each. The middle
part of the Kildonnan Member contains few, if any, conchostracans. Its upper part contains many,

518

PALAEONTOLOGY, VOLUME 34

and they continue into the lower part of the Lonfearn Member, which we have also sampled at
several localities. The upper part of the Lonfearn Member contains conchostracans, but we have not
sampled it adequately, and the same applies to the sparse conchostracan fauna of the basal Valtos
Formation above.

In the Cullaidh Shale and basal Kildonnan Member, the genera Euestheria and Neopolygrapta
occur, each represented by one species (when specifically identifiable). Euestheria is confined to these
basal beds, except for one occurrence in the higher part of the Kildonnan Member, at which horizon
Neopolygrapta lealtensis sp. nov. is joined by Dendrostracus hehridesensis sp. nov. These two species
continue into the basal part of the Lonfearn Member, but have not been found more than about
3 m above the algal limestone that marks the contact between the members. Before the
disappearance of Neopolygrapta , Skyestheria appears. Morphological features suggest it may have
evolved from Neopolygrapta. Skyestheria is the commonest conchostracan in the middle part of the
Lonfearn Member, including the shales in the Leak River (immediately above and below a
limestone with Chara fragments) from which conchostracans have most often been collected in the
past. Neopolygrapta also occurs here. (This locality is mentioned in Hudson 1 963/?, p. 339; note that
the new road bridge is a few metres upstream of the old one and stands on what were formerly the
best exposures.)

A few metres higher, Antronestheria praecursor sp. nov. appears in abundance (noted as 'form
with punctate ornament’ by Hudson 1963 6, p. 339). It probably evolved from Skyestheria and
itself leads to Antronestheria kilmaluagensis sp. nov., which dominates the faunas of the higher
Great Estuarine Group. Our data are insufficient to show whether or not Skyestheria and
Antronestheria overlap stratigraphically. The species Skyestheria elliptica sp. nov. is only known
from one sample; its horizon within the Lonfearn Member is unknown, but likely to be near the
middle because of the occurrence of oolitic limestones nearby.

Palaeoecology. The Cullaidh Shale Formation has a sparse fauna mainly of fish scales and tiny
gastropods, with occasional occurrences of bivalves including mytilids. We do not have sufficient
information to relate the conchostracan occurrences to the rest of the fauna, which is clearly not a
normal marine one. In the Raasay section, conchostracans occur near the passage into the Elgol
Sandstone Formation and not in association with the marine-brackish faunas recorded by Forsyth
(1960) from stratigraphically lower.

In the Leak Shale Formation conchostracans occur with the freshwater gastropods Viviparus and
Valvata and the bivalve Unio, as well as commonly (from its incoming in the higher Kildonnan
Member upwards) with the probably freshwater-brackish bivalve Neomiodon. The commonest
associates of the conchostracans are freshwater to oligohaline ostracods of the genera Limnocythere ,
Darwinula and Theriosynoecum, with less commonly the euryhaline genus Micropneumatocy there.
In the type section of the Kildonnan Member, Eigg, conchostracans do not occur in Beds 3-5
(Hudson 1966 and unpublished data) which are dominated by the brackish-water bivalve
Praemytilus strathairdensis , nor in association with the marine-brackish Plaucunopsis-Cuspidaria
faunas that occur in both members of the Formation. Walton (pers. comm. 1988) finds an excellent
correlation between the proportion of the planktonic alga Botryococcus in palynological
preparations and the presence of freshwater to oligohaline indicators (including conchostracans) In
the macrofauna. In one bed in the Lonfearn Member in Trotternish (bed 1 of Hudson 19636,
p. 339) Skyestheria co-occurs with small mytilid bivalves. In general, the new data confirm that
conchostracans inhabited the freshwater-brackish end of a spectrum of coastal lagoons of varying
salinity (Hudson 1963«, 19636, 1980; see also Tasch 1987, p. 139). Details will no doubt emerge
from further collecting and from work in progress on the ostracod faunas and the palynology.

Valtos and Duntulm Formations

Few horizons in these formations yield conchostracans. We have studied only two samples, a bed
0-47 to 0 32 m below the top of the Valtos Formation on Muck, overlying a conspicuously yellow-
weathering dolomitic limestone (Andrews and Walton 1990, p. 13, fig. 10), and Beds 6-7 of the

CHEN AND HUDSON: JURASSIC CONCHOSTR ACANS

519

Duntulm Formation at Lon Ostatoin, Skye (the ‘freshwater intercalation’ of Andrews and Walton
1990, p. 8). Both yield Antronestheria kilmaluagensis , accompanied by Neomiodon and, in Muck,
Darwinula.

Kilmaluag Formation

Stratigraphy. Conchostracans occur very commonly in the Kilmaluag Formation, wherever suitable
lithologies (shales or fissile argillaceous limestones) are developed. With the exception of three
localities, all the conchostracans belong to Antronestheria kilmaluagensis gen et sp. nov. This has
been identified from the type section at Kilmaluag (beds 1, 3, 6, 7), bed 8 from Lon Ostatoin, and
bed 4 from Prince Charles’s Point; all bed numbers from Andrews (1985, p. 1122). Other Skye
localities include the shore near the mouth of Lon Sgapail (Andrews, 1985, p. 1334), cliffs near
Osmigarry (NG 390719), and slipped blocks on the shore of Lub Score (NG 404728). These
occurrences span most of the exposed section of the Kilmaluag Formation, those at Kilmaluag
being in its lower part and that at Prince Charles’s Point probably in its upper part (Andrews 1985,
p. 1134). In Strathaird, the Kilmaluag Formation is too metamorphosed to yield identifiable
conchostracans. A. kilmaluagensis is abundant in its lower part at Laig Gorge, Eigg (beds 1, 2, 3,
5, 9 of Andrews, 1985, p. 1125).

One sample from bed 7 at Kilmaluag yielded the only other conchostracan species so far known
from this horizon, Fibrestheria puncta sp. nov. Other samples from this bed yielded only
Antronestheria. Fibrestheria also occur in the Blisworth Limestone of central England (see below).

Two localities have yielded a totally different conchostracan fauna consisting of species of
Pseudograpta Novojilov. From one of these came the original material of Estheria murchisonae
Jones: ‘Mugstok’ (Monkstadt), Trotternish (a locality also referred to as ‘Icolmkill’). The
specimens were collected by Murchison in 1826 from a temporary exposure made during the
drainage of the former Loch Chaluim Cille, probably approximately NG 375695. Murchison (1829,
p. 311) recorded, in blue shale, ‘ Ammonites Koenigi , Ostraeae in masses, many belemnites, flattened
Tellinae (?), etc.’ The ‘flattened Tellinae’ are the conchostracans subsequently described by Jones
(1863), and preserved in the Geological Society Collections, now at the British Geological Survey,
Keyworth, as No. 7296. The lithology, fine-grained grey argillaceous limestone with conchostracans
almost covering one bedding plane, resembles typical Kilmaluag Formation. It is impossible to be
sure how close the association with the other fauna recorded by Murchison was, but the presence
of oysters suggests the Upper Ostrea Member, and the belemnites and ammonites the Belemnite
Sands Member, both of the Staffin Bay Formation (Sykes 1975). It therefore seems most probable
that the conchostracans occurred high in the Kilmaluag Formation or conceivably as an
intercalation of ‘Kilmaluag’ facies within the Skudiburgh Formation.

The other Pseudograpta locality is in the Kilmaluag Formation of Muck, beds 13, 14 and 15 of
Andrews (1985, p. 1125). Abundant and well preserved conchostracans occur in argillaceous
dolomites, without macrofauna or ostracods. This outcrop is tectonically separated from the nearby
Duntulm Formation and overlain by Palaeocene agglomerate. Lithological comparison with
Straithaird suggests a high horizon with the Kilmaluag Formation (Andrews, 1985, p. 1128). Our
suggestion, therefore, is that the Pseudograpta fauna is younger than the Antronestheria fauna and
both are late Bathonian. At both localities Pseudograpta murchisoniae is accompanied by the species
P. orbita Chen and P. jonesi sp. nov; the Muck locality additionally yields P. morrisi sp. nov.

Palaeoecology. As in the Lealt Shale Formation, the occurrence of conchostracans in the higher
Great Estuarine Group clearly correlates with non-marine facies, and it is likely that the Kilmaluag
Formation lagoons were only tenuously connected to the open sea (Andrews 1985). Antronestheria
kilmaluagensis is nearly always accompanied by ostracods ( Theriosynoecum , Darwinula ) that are
considered to be freshwater or oligohaline (Kilenyi in Tan andl Hudson 1974, p. 107); the
Theriosynoecum show variations in node development. These forms are often accompanied by
Klieana , which can also tolerate higher salinities. Some horizons additionally contain the brackish-
marine ostracod Progonocythere (M. Wakefield pers. comm. 1989). Other associated fossils are

520

PALAEONTOLOGY, VOLUME 34

Viviparus , Chara oogonia, fish fragments and cuticular fragments of higher plants. The occurrence
of Pseudograpta in ostracod-free dolomites from Muck may suggest tolerance of evaporated water
and perhaps elevated (but non-marine) salinities by this genus (cf. Andrews et al. 1987). The Skye
occurrence is in limestone, not dolomite (checked by X-ray diffraction), and also lacks associated
ostracods, as do the occurrences of Pseudograpta spp. in Liaoning Province, China.

Bathonian of Southern and Eastern England

We have examined conchostracans from three localities in the Bathonian of England, using material
from the BGS collections, Keyworth. Those from the top part of the Blisworth Limestone of
Newport Pagnell, Bucks., may be assigned to Fihrestheria cf. puncta sp. nov. The horizon is
probably late Bathonian and about the same age as the occurrence of F. puncta in the Kilmaluag
Formation, Skye. An occurrence in the Forest Marble near Cirencester is of poorly preserved
specimens of Dendrostracus! This horizon is also late Bathonian, and if the occurrence is confirmed
it will probably represent a later occurrence of Dendrostracus than in the Hebrides, where it is
confined to the Lealt Shale Formation of probable early to mid Bathonian age. The Blisworth Clay
of the Nettleton Bottom borehole yields Antronestheria kilmaluagensis and is probably of similar
age to the Kilmaluag Formation, late Bathonian. Locality details of the English occurrences are
given in the systematic section below.

Porcupine Basin

A particularly interesting record of Antronestheria kilmaluagensis is in material loaned to us by the
British Petroleum Company pic from a core taken in the northern part of the Porcupine Basin,
offshore SW Ireland. The fossils and lithology are indistinguishable from typical Kilmaluag
formation, and the horizon is independently dated as late Bathonian by ostracods (J. Athersuch
pers. comm. 1988).

MORPHOLOGY, TAXONOMY AND EVOLUTION

The taxonomy of fossil conchostracans is mainly based on the characteristics of their carapace and
minute ornamentation of growth bands, because their soft bodies and appendages are seldom
preserved in fossils (but see Zhang et al. 1987, for Jurassic soft-part preservation). The distinctive
leaids aside, many forms lack remarkable differences in shape of the carapaces, although some may
have exceptional features of the umbo or dorsal margin (such as large umbo, umbonal node,
reflected curved growth lines near dorsal margin). In most Jurassic and Cretaceous forms, the
carapaces vary only from elliptical, oval, or subquadrate to subcircular, and their umbones are
generally placed between the middle point and the anterior end of the dorsal margin, so the
taxonomy is only dependent upon the variety of ornamentation. Sometimes, the evolution of
ornamentation is very rapid and distinct in a particular basin or in a limited area belonging to the
same river system. The non-marine Cretaceous of NE China has been successfully subdivided in this
way (Zhang et al. 1 976, pp. 44-45, 82-85, text-fig. 39 ; Chen and Shen, 1 985, pp. 20-22, text-fig. 11).
Thus we disagree with the low taxonomic status assigned to interspace ornament by Tasch (1969).

The variously developed growth lines of conchostracan carapaces are formed from the junction
of neighbouring growth bands (Novojilov 1954, p. 96, text-fig. 69; Chen and Shen 1985, pp. 13-14;
text-figs 7 and 8). These structures are clearly shown in SEM photographs in the present paper (PI.
2, figs 8, 10, 11; PI. 3, figs 6 and 9; PI. 9, figs 9-12; pi. 10, figs 3-8). A row of tubes is situated on
the lower margin of each growth band ( = growth line), corresponding with a row of rivets on the
prolongation of the next growth band (Text-fig. 2). The basal holes of these tubes are exposed on
the interior surface of growth bands (PI. 3, fig. 6). These tubes in Dendrostracus have a cross-bar in
between (PI. 3, figs 4 and 8), but those of Neopolygrapta are shorter and without the cross-bar (PI.
2, figs 8, 10, 11; PI. 10, figs 3-8). For Skyestheria , appearing in a higher horizon of the Lealt Shale
Formation, these tubes contract and connect with each other to form beading on the lower margin
of the growth lines (PI. 4, figs 6 and 9).

CHEN AND HUDSON: JURASSIC CONCHOSTR ACANS

521

text-fig. 2. The construction of conchostracan growth bands and associated features.

Our morphological observations suggest some modifications to the classification of concho-
stracans given by Chen and Shen (1985) and to their evolutionary diagram (1985, p. 190, fig. I 13).
On the basis of the beading or serration along the lower margin of their growth lines, three genera
erected herein ( Dendrostracus , Neopolygrapta and Skyestheria ) belong to the family Afrograptidae
Novojilov, 1957. The origin of this family was regarded as uncertain by Chen and Shen (1985, p.
190, fig. 113). Our results suggest that these genera arose in the Middle Jurassic from members of
the family Euestheriidae Defretin, 1965: Dendrostracus perhaps from Euestheria and Neopolygrapta
from Polygrapta. The Afrograptidae (if monophyletic) thus originate in the Middle Jurassic and
range into the lower Tertiary; they belong to the superfamily Eosestherioidea Zhang and Chen (in
Zhang et al., 1976) and there is no need for the superfamily Afrograptioidea Novojilov, 1957.
According to interpretations in this paper, Neopolygrapta gave rise to Skyestheria in the Bathonian,
and that in turn to the genus Antronestheria of the late Bathonian. Antronestheria is the type genus
of the new family Antronestheriidae which includes other genera (see below) ranging from Middle
Jurassic to Lower Cretaceous.

Pseudograpta Novojilov appears to be transitional between the families Loxomegaglyptidae and
Eosestheriidae, a transaction already dated as middle Jurassic by Chen and Shen (1985, p. 190, fig.

1 13). However, our new material suggests that Pseudograpta should be included in the latter family
rather than in the former as in Chen and Shen (1985, p. 109). Pseudograpta is not closely related to
the group of genera belonging to the Euestheriidae or originating from that family within the
Middle Jurassic, and its incoming near the top of the Great Estuarine Group is probably the result
of immigration. It corresponds to a uniting of formerly distinct faunal provinces, as discussed
below. Text-figure 3 summarizes our present views on evolutionary relationships.

PALAEOBIOGEOGRAPHY

In this paper twelve conchostracan species in seven genera are described. Eight species in six genera
(. Skyestheria , Dendrostracus , Neopolygrapta , Euestheria , Antronestheria and Fibrestheria) form
assemblages with overlapping stratigraphical ranges, belonging to or descended from the

522

PALAEONTOLOGY, VOLUME 34

LOWER CRETACEOUS ■

Great

Estuarine

Group

text-fig. 3. Possible evolutionary relationships among some Jurassic conchostracan families (superfamily
Eosestherioidea Zhang and Chen, 1976), as illustrated by genera occurring in the Great Estuarine Group.
Based on Chen and Shen (1985, p. 120, fig. 1 13), as modified herein. Direct descent is not necessarily implied;
see discussion in text. Eu., Eustheria ; N., Neopoly grapt a ; D., Dendrostracus \ Sk., Skyestheria ; A.p.,
Antronestheria praecursor; A.k., A. kilmaluagensis ; F., Fibrestheria; Ps., Pseudograpta.

Euestheriidae. They are mostly new forms with very complex dendritic or cavernous ornamentation
in their growth bands. This fauna is here named the Skyestheria Fauna, so far known mainly from
Scotland, but with some of its members also recorded from England and from the Atlantic west of
Ireland. It occurs stratigraphically below a fauna containing four species of Pseudograpta.

The Euestheria ziliujingensis fauna is widely distributed in the Middle Jurassic deposits from E
Asia (Chen and Shen 1982). Apart from the index species, the fauna includes A. haifanggonensis
Chen, E. yanjiawanensis Chen, E. complanata Chen, E. rotunda Chen, E. jingyuanensis Chen, and
Paleoleptestheria chinensis Chen. Their carapaces are generally small, elliptical or subcircular in
outline, with simple small punctae or fine reticulate ornamentation of growth bands. P. chinensis
from the Upper Shaximiao Formation of the Sichuan basin of SW China is a late member of this
fauna and its cavernous ornamentation is similar to that of Antronestheria , described herein from
late Bathonian deposits of Western Europe.

CHEN AND HUDSON: JURASSIC CONCHOSTRACANS

523

text-fig. 4. Map showing separate distribution of the early Middle Jurassic Euestheria ziliujingensis and
Skyestheria faunas, and wide distribution of the later Middle Jurassic Pseudograpta fauna between China and
western Europe. Polar projection based on Smith and Briden 1977.

By the end of the Middle Jurassic, the distinctness of the above-mentioned two conchostracan
paleobiogeographic Provinces in the Euro-Asian continent (Text-fig. 4) disappeared, and they were
replaced by a united Pseudograpta fauna. In this genus carapaces are large in size, with coarse
hexagonal ornamentation of growth bands and very stout and convex growth lines (growth ridges).
P. orbita and P. murchisoniae have been found in the Tuchengszi Formation of W Liaoning, NE
China, as well as in Scotland (Zhang et al. 1976).

The two occurrences provide a remarkable example of the range of facies that conchostracans can
inhabit, and of their wide dispersal. The Scottish occurrences are in near-coastal lagoonal
dolomites, the Liaoning ones in thin green mudstones intercalated in thick, mainly fluvial, red-beds.
They probably represent temporary pools or abandoned channels in flood-plains. And few places
could be more different now than the shores of the Isle of Muck in the windy Hebrides, and the
semi-arid uplands of Liaoning where the authors have jointly collected these conchostracans.

SYSTEMATIC PALAEONTOLOGY
Order conchostraca Sars, 1867

Superfamily eosestherioidea Zhang and Chen, 1976 (in Zhang et al. 1976)
Family euestheriidae Defretin, 1865
Genus euestheria Deperet and Mazeran, 1912

524

PALAEONTOLOGY, VOLUME 34

Type species. Estheria minuta (Zieten, 1833)

Euestheria trotternishensis sp. nov.

Plate 1, figs 1-6

Etymology. From Trotternish, the northernmost peninsula of the Isle of Skye, Scotland.

Holotype. A left valve (BMNH In 63723, PI. 1, fig. 2) from 2-5 m above the base of the Kildonnan Member,
Lealt Shale Formation at Lonfearn Cliff, Trotternish, Skye.

Occurrence. Limited to the basal part of the Kildonnan Member of their Lealt Shale Formation (early
Bathonian) in Skye and Eigg, NW Scotland. Euestheria indet., possibly the same species, occurs also in the
Cullaidh Shale Formation in Strathaird (LEIUG 108064-6), and Euestheria ? in bed 5 of the Kildonnan
Member in Trotternish (LEIUG 108069).

Additional material. Numerous specimens are from the same locality as the holotype, where E. trotternishensis
ranges through at least 1 m of strata below the holotype horizon (BMNH In 63724, 5, LEIUG 108060-3,
108067, 108181). A few specimens are from bed 1 of the type section of the Kildonnan Member, Lealt Shale
Formation, Eigg (SM J49383-7; LEIUG 108179, 80).

Diagnosis. This new species differs from E. minuta in possessing numerous flattened growth bands,
and from E. ziliujingensis (Zhang et ctl. 1976) of the early Middle Jurassic in China in having a
subquadrate carapace outline.

Description. All specimens are well preserved carapaces or external moulds. Carapace small, subquadrate in
outline, 4-2-5-5 mm long 3-3-4-3 mm high; dorsal margin short, umbo projecting above its subcentral part;
anterior and posterior margins relatively straight, ventral margin arched downward; 23-35 growth bands,
flattened, and with many small punctae ornamenting the surface especially in lower parts of the growth bands
(PI. 1, figs 1, 5, 6) or appearing as small granules on the external mould (PI. 1, figs 3 and 4).

Genus fibrestheria gen. nov.

Etymology. From the fibrous appearance of the growth bands.

Type species. Fibrestheria puncta sp. nov. from the Kilmaluag Formation of Skye.

Occurrence. Late Bathonian ; NW Scotland and Central England.

Diagnosis. Carapace of moderate size, subcircular or elliptical in outline; growth bands flattened,
ornamented by minute and crowded punctae which form vertical lines.

Discussion. Among all forms of the family Euestheriidae, Tenuestheria from the Upper Cretaceous
is similar to this new genus in ornamentation, but differs in its thin carapace, few and broad growth
bands, and especially in its punctate sculpture without vertical alignment.

EXPLANATION OF PLATE I

Figs 1-6. Euestheria trotternishensis sp. nov., from the basal Kildonnan Member, Lealt Shale Formation,
Lonfearn Cliff, Skye. 1, 3, 4, BMNH In 63724; 1, x42; 3, x72; 4, x 120. 2, holotype, BMNH In 63723,
x 7-7. 5 and 6, BMNH In 63725; 5, x 180; 6, x 600.

Figs 7-12. Skvestheria elliptica gen. et sp. nov., from the Lonfearn Member, Lealt Shale Formation, Rigg Burn,
Skye. 7 and 8 holotype, BMNH In 63742; 7, x 12; 8, x 180. 9-11, BMNH In 63743; 9, x42; 10, x 120;
11, x 180. 12, BMNH In 63744, x240.

PLATE 1

CHEN and HUDSON, Euestheria, Skyestheria

526

PALAEONTOLOGY, VOLUME 34

Fibrestheria puncta sp. nov.

Plate 4, figs 1,8, 11

Etymology. From the punctate nature of the growth band sculpture.

Holotype. An external mould of a right valve with some pieces of surviving shell (SM J49594-604I, PI. 4, figs
1,8, 11), from the Kilmaluag Formation of Skye.

Occurrence. Kilmaluag Formation, Trotternish, Skye; so far known from one hand specimen only.
Fibrestheria cf. puncta occurs in the top 30 cm of the Blisworth Limestone (late Bajocian) on the M 1 motorway
near Newport Pagnell, Bucks (SP 8564 3840): BGS collection WA 908-910. This is probably within the
aspidioides zone, late Bathonian (Cope et al. 1980, p. 39, fig. 6b).

Additional material. Many individuals in the same slab (SM J49594-604) from Kilmaluag, Trotternish, Skye,
bed 7 of Andrews (1985, p. 1122).

Diagnosis. As for the genus.

Description. Carapace of moderate size, elliptical to subcircular in outline, 8-1 1 mm long, 5-7 mm high; dorsal
margin somewhat broken, umbo narrow and small, situated between its central and anterior end; anterior
margin straight, posteroventral margin rounded; more than 30 growth bands with overlapping junctions,
ornamented with minute and crowded punctae that are vertically aligned.

Family afrograptidae Novojilov, 1957
Genus neopolygrapta gen. nov.

Etymology. This genus resembles Polygrapta Novojilov, 1957, but is younger.

Type species. Neopolygrapta lealtensis sp. nov. from the Leak Shale Formation and Cullaidh Shale Formation
in Skye.

Occurrence. Early Bathonian; NW Scotland.

Diagnosis. Carapace small, elliptical or oval in outline; numerous growth bands with radial striae
and cross bars, forming irregular reticulations with each other; growth lines with tubiform
serrations of the lower margin.

Discussion. The new genus and Polygrapta, distributed in the Euro-Asia continent during the
Permian and Triassic, are much alike in ornamentation and growth bands, but the former differs
from the latter in its tubiform serrations of the lower margin of the growth lines.

Neopolygrapta lealtensis sp. nov.
Plate 2, figs 1-12; Plate 10, figs 1-12
Etymology. From the Leak River in Trotternish, Isle of Skye, Scotland.

EXPLANATION OF PLATE 2

Figs 1 12. Neopolygrapta lealtensis gen. et sp. nov., from the Leak Shale Formation of the Rudha nam
Brathairean area, Skye; BMNH In 63728 and 63730 from the Lonfearn Member, BMNH In 63726,7 from
the Kildonnan Member. 1, holotype, BMNH In 63726, x 12. 2 and 9, BMNH In 63727; 2, x42; 9, x 120.
3-6, 12, external mould, BMNH in 63728; 3, x42;4, xl2;5, x90;6, x90;12, x 420. 7, BMNH In 63729,
x 180. 8, 10, 11, BMNH In 63730; 8, x72; 10, x60; 11, x420.

PLATE 2

CHEN and HUDSON, Neopolygrapta lealtensis

528

PALAEONTOLOGY, VOLUME 34

Holotype. A left valve (BMNH In 63726, PI. 2, fig. I) from Upper Kildonnan Member of the Lealt Shale
Formation, Lonfearn Cliff, Skye.

Occurrence. Kildonnan Member to lower Lonfearn Member of the Lealt Shale Formation, Isle of Skye.

Additional material. Many specimens from the Kildonnan and lower Lonfearn Members of the Lealt Shale
Formation, Trotternish, Skye. Kildonnan Member: LEIUG 108082-4 from Rudha nam Brathairean; LEIUG
108085,6 from Lonfearn Cliff; LEIUG 108087-9 from South Rigg. Lonfearn Member: LEIUG 108070-81,
108184,5 from Rudha nam Braithairean; BMNH In 63728, LEIUG 108182,3 from Lonfearn Cliff; LEIUG
108194 from the Lealt River; LEIUG 108090,1, 108186-9 from North Rigg.

Diagnosis. As for the genus.

Description. Most specimens are well preserved carapaces or external moulds. Carapace small, oval or elliptical
in outline, 4-5-5 mm long, 34 mm high; dorsal margin short and straight, umbo small, situated near its
anterior end; anterior, posterior and ventral margins rounded; about 30 growth bands covered with radial
striae and cross bars, striae simple and sometimes branching, cross bars fine and separating the interspace
between striae into irregular reticulation ; growth lines with short tubiform serrations of the lower margin. In
external moulds of the carapace, growth bands with dotted ornamentation, radially arranged vertically, each
dot consisting of several granules.

Genus dendrostracus gen. nov.

Etymology. From the dendritic appearance of the growth band sculpture.

Type species. Dendrostracus hebridesensis sp. nov. from the Lealt Shale Formation in the Isle of Skye.

Occurrence. Bathonian of Britain ; the type species known only from the Great Estuarine Group of Scotland
(see below).

Dendrostracus ? also occurs in the Forest Marble, from 3^1 m beneath the Cornbrash, in the BGS Sandpool
Farm borehole (SU 0122 9427) between Cirencester and Swindon. These strata most probably belong to the
aspidioides zone, late Bathonian (Cope et al. 1980, p. 32).

Diagnosis. Carapace small, elliptical or oval in outline; growth bands flattened and ornamented by
dendritic striae and small reticulations, reticulation separated by dendritic striae into many groups
of meshwork patterns. Growth lines with tubiform serrated lower margins.

Discussion. This genus is similar in size and shape of carapace to Neopolygrapta , but differs from
the latter in broader growth bands with dendritic ornament, and in the longer tubes forming the
serrated lower margin of the growth lines, with a few cross bars between the tubes. The new genus
is also somewhat similar in ornamentation of growth bands to Jilinestheria, Plectestheria and
Glyptostracus (Zhang et al. 1976, pi. 82, fig. 2; pi. 84, figs 5 and 6; pi. 133, figs 1-5) from the Upper
Cretaceous of NE China, but differs from them chiefly in having a tubiform serrated lower margin
to its growth lines. In addition, the dendritic striae of three Upper Cretaceous forms are much
stronger and more thickened than these of Dendrostracus , and their reticulations between dendritic
striae are of clearly different origin.

EXPLANATION OF PLATE 3

Figs 1-9. Dendrostracus hebridesensis gen. et sp. nov., from the Kildonnan Member, Lealt Shale Formation,
south Rigg, Skye. 1 and 2 holotype. BMNH In 63731 ; 1, x 18; 2, x 12. 3 and 5, BMNH In 63732; 3, x 12;
5, x 1 80. 4 and 7, BMNH In 63732; 4, x240; 7, x 600. 6, BMNH In 63733, x 120. 8, BMNH In 63735,
x 420. 9, BMNH In 63736, x 200.

PLATE 3

CHEN and HUDSON, Dendrostracus hebridesensis

530

PALAEONTOLOGY, VOLUME 34

Dendrostracus hebridesensis sp. nov.

Plate 3, figs 1-9; Plate 9, figs 9-12
Etymology. From the Flebrides, islands oft' NW Scotland.

Holotype. An external mould of a left valve (BMNH In 63731, PI. 3, fig. 2) from the Kildonnan Member, Lealt
Shale Formation, in S Rigg, Skye.

Occurrence. The species has a short recorded range in the upper Kildonnan and lower Lonfearn Members of
the Lealt Shale Formation, Skye.

Additional material. At least 21 specimens from the upper Kildonnan and lower Lonfearn Members of the Lealt
Shale Formation, Trotternish, Skye. Kildonnan Member: LEIUG 108086, 108095-7 from Rudha nam
Braithairean ; LEIUG 108099, 100 from Lonfearn Cliff; BMNH In 63732-6, LEIUG 108098, 108190-3 from
South Rigg, Lonfearn Member: LEIUG 108094 from Rudha nam Brathairean.

Diagnosis. As for the genus.

Description : Most specimens are poorly preserved internal or external moulds, but a few well preserved
carapaces or external moulds show excellent ornamentation. Carapace small, oval or elliptic in outline,
3-5-4-6 mm long, 2-6-3 3 mm high; dorsal margin short and slightly arched upward; umbo small, situated
subcentrally ; anterior, posterior and ventral margins rounded, postventral margin obviously expanded; 15-30
growth bands, flattened and ornamented with dendritic striae and small reticulations which are separated by
dendritic striae into many groups of mesh (each group generally including 6-8 meshes); occasionally only small
reticulations in one or two growth bands near the ventral margin of carapace and without dendritic striae; a
row of long tubiform structures attached to the lower margin of each growth line near the ventral side of the
carapace, and a few cross bars between the tubes (PI. 3, figs 4 and 8).

Genus skyestheria gen. nov.

Etymology , From the Isle of Skye, Scotland.

Type species. Skyestheria intermedia Chen and Hudson sp. nov. from the Lonfearn Member, Lealt Shale
Formation in Skye.

Occurrence. Bathonian, Isle of Skye.

Diagnosis. Carapace small, oval, elliptical or subcircular in outline; numerous growth bands with
Neopolygrapta-type ornamentation which crosses the whole band near the dorsal side of the
carapace, but only occurs in the lower part of each band near the ventral side; growth lines with
beaded lower margin.

Discussion. This genus originates directly from Neopolygrapta. Its ornamentation of radial striae
with cross bars shrinks back to the lower part of each growth band near the ventral side of the
carapace and the upper part of the growth band is smooth. Along the lower margin of each growth

EXPLANATION OF PLATE 4

Figs 1,8, 11. Fibrestheria puncta gen. et sp. nov., holotype, SM J49594-604I, from the Kilmaluag Formation,
Kilmaluag, Skye. 1, x 5-5. 8, x42. 11, x414.

Figs 2, 3, 5, 6, 9, 12. Skyestheria intermedia gen. et sp. nov., from the Lonfearn Member, Lealt Shale Formation,
Skye. 2, BMNH In 63738, x 12. 3 and 5, holotype, BMNH In 63737; 3, x 72; 5, x 12. 6, BMNH In 63739,
x 120. 9, BMNH In 63740, x420. 12. BMNH In 63741, x 900.

Figs 4, 7, 10. Antronestheria praecursor gen. et sp. nov., from the upper part of the Lonfearn Member,
Lealt Shale Formation, Lealt River, Skye. 4, holotype, SM J49409, x4-8. 7, BMNH In 63751, x60.
10, SM J49398, x 72.

PLATE 4

mm

.-st:?,':-'"

7

10 11 12
CHEN and HUDSON, Fibrestheria , Skyestheria , Antronestheria

532

PALAEONTOLOGY, VOLUME 34

line, the tubiform structure has been replaced by beading, presumably for consolidating the
carapace, and the growth lines have thickened.

Skyestheria intermedia sp. nov.

Plate 4, figs 2, 3, 5, 6, 9, 12; Plate 9, figs 5-8

Etymology. This species is believed to occupy an intermediate position in the evolution of these conchostracans.

Holotype. An external mould of a left valve (BMNH In 63737, PI. 4, figs 3 and 5) from the middle Lonfearn
Member of the Lealt Shale Formation in the Lealt River, Skye.

Occurrence. Lonfearn Member of the Lealt Shale Formation in Skye.

Additional material. At least 32 specimens from the Lonfearn Member of the Lealt shale formation in Skye.
BMNH In 63738, LEIUG 108104,5, 108198,9, from Rudha nam Braitharean; LEIUG 108112^1 from
Lonfearn Cliff; BMNH In 63739, LEIUG 108102,3, 1081 101,1, 108195-7 from Lealt River; BMNH In 63741,
LEIUG 108106-9, 108200-5, from North Rigg; LEIUG 108115,6 from Elgol.

Diagnosis. Skyestheria with radial striae and cross bars in interspace sculpture of approximately
equal strength.

Description. Carapace small, oval, elliptical or subcircular in outline, 3 5^4-5 mm long, 2-3— 3-6 mm high; dorsal
margin short and straight, umbo small and situated between its central and anterior end; anterior and ventral
margins relatively straight, post- ventral margin rounded and slightly expanded; about 20-30 growth bands,
ornamented by radial striae with cross bars through whole intervals in dorsal side, or only occupying the lower
part of interval in ventral side; near ventral margin of carapace, thickened growth lines with beading along its
lower margin and a row of shallow cavernous sculpture along its upper margin, several minute punctae
surrounding the major structures in each case.

Skyestheria elliptica sp. nov.

Plate 1, figs 7-12

Etymology. From the shape of the carapace.

Holotype'. A left carapace (BMNH In 63742, PI. I, figs 7, 8, 10) from the Lonfearn Member of the Lealt Shale
Formation in Rigg Burn, Skye.

Occurrence. Lonfearn Member of the Lealt Shale Formation in Skye, from one bed only so far.

Additional material. Five specimens BMNH In 63743,4, LEIUG 1081 17,8, 108206 from same locality and same
horizon as the holotype.

Diagnosis. Skyestheria with radial striae markedly stronger than cross bars in interspace sculpture.

Description. This species is very similar in shape and size of carapace to S. intermedia, but differs chiefly in
ornamentation. Its radial striae are consistently stronger and cross bars are weaker than those of S. intermedia '.
compare Plate 1, figures 9, 10, 11 with Plate 4, figure 6 and Plate 9, figures 5, 6, 7, 8. The carapace of the
holotype is 3-2 mm long and 2-3 mm high. There are about 30 growth bands.

EXPLANATION OF PLATE 5

Figs 1-8. Antronestheria kilmaluagensis gen. et sp. nov., from the Kilmaluag Formation, Trotternish, Skye.
I, BMNH In 63746, x 9 6. 2, holotype, BMNH In 63745, x 13-5. 3-5 BMNH In 63747; 3, x90; 4, x 300;
5, x 300. 6, BMNH In 63748, x 600. 7, BMNH In 63749, x 600. 8, paratype, left valve BMNH 63750, x 20.

PLATE 5

CHEN and HUDSON, Antronestheria kilmaluagensis

534

PALAEONTOLOGY, VOLUME 34

Family antronestheriidae nov.

Carapace elliptical or subcircular in outline; numerous growth bands with cavernous orna-
mentation; fossae of moderate or large size, simple or consisting of small punctae. Includes
Antronestheria Chen and Hudson (gen. nov.), Pseudestherites Chen, 1976 (in Zhang et al. 1976),
Paleoleptestheria ? chinensis Chen, 1976 (in Zhang et al. 1976). Middle Jurassic to Lower Cretaceous.

Genus antronestheria gen. nov.

Etymology. From the Greek antron, a cave or cavity, in reference to the interspace sculpture.

Type species. Antronestheria kilmaluagensis sp. nov. from the Kilmaluag Formation (Bathonian) in NW
Scotland and late Bathonian strata in the Porcupine Basin, NE Atlantic off Ireland.

Occurrence. Middle to upper Bathonian ; NW Scotland, eastern England and Porcupine Basin off Ireland.

Diagnosis. Carapace moderate in size, oval to subcircular in outline; growth bands broad and
flattened, with large cavernous ornamentation; fossae, consisting of several punctae, deepened and
crowded along upper margin of each growth line; growth lines smooth and thickened.

Discussion. The new genus resembles Paleoleptestheria chinensis from the late middle Jurassic of SW
China, and Pseudestherites from early Lower Cretaceous of NE China (Zhang et al. 1976, pi. 34,
figs 1-8; pi. 39, figs 1, 2, 7, 8) in cavernous ornamentation, but differs from them in possessing
several punctae in each fossa, and a row of crowded and deepened fossae along the growth line.
Antronestheria probably evolved from Skyestheria of the Lonfearn Member, but the latter differs
from the former in having the beading along lower margin of each growth line and radial striae with
cross bars in between.

Antronestheria kilmaluagensis sp. nov.

Plate 5, figs 1-8; Plate 9. figs 1-^4

Etymology. From Kilmaluag, a village in north Trotternish, Isle of Skye, Scotland.

Holotype. An external mould of a left valve (BMNH In 63745, PI. 5, fig. 2) from the Kilmaluag Formation in
Kilmaluag Bay, Skye.

Occurrence. Commonly in the Kilmaluag Formation in Skye, NW Scotland; also in the same formation, Eigg
(beds 1-9 of Andrews, 1985, p. 1125; LEIUG 108144—6). One recorded occurrence each from the Duntulm
Formation, Lon Ostatoin, Skye, and the top of the Valtos Formation, Muck (LEIUG 108142-3). Also
recorded from the late Bathonian, Porcupine Basin, and from the Blisworth Clay of the BGS Nettleton Bottom
borehole, Lincolnshire (BGS BLJ 7762).

Additional material. At least 28 specimens from the Kilmaluag Formation, Trotternish, Skye. From Beds 1-7
of Andrews (1985, p. 1122) in Kilmaluag Bay: BMNH In 63746, LEIUG 108207-10, 108119-32. LEIUG
108133 from Lub Score. BMNH In 63747-50, LEIUG 108211-4 from Lon Sgapail.

EXPLANATION of plate 6

Figs 1, 3, 5, 6b, 8. Pseudograpta orbita Chen, 1976, from the Kilmaluag Formation!?) near Monkstadt,
Trotternish, Skye. 1 and 8, left valve, BGS, GSC 7296c; 1, x6; 8, x22. 3, 5, 6b, right valve, BGS, GSC
72966; 3, x6; 5, x22; 6b, x 2-4.

Figs 2, 4, 6a, 7. Pseudograpta murchisoniae (Jones, 1863), from the Kilmaluag Formation!?), near Monkstadt,
Trotternish, Skye. 2 and 6a, lectotype, BGS, GSC 7296a; 2, x6; 6a, x 2-4. 4, BGS, GSC 7296c, x22. 7,
BGS, GSC 7296/, x 22.

Fig. 6 d. Pseudograpta jonesi sp. nov., BGS, GSC 7296/ from the Kilmalaug Formation (?), near Monkstadt,
Trotternish, Skye, x 2-4.

PLATE 6

CHEN and HUDSON, Pseudograpta

536

PALAEONTOLOGY, VOLUME 34

Diagnosis. Antronestheria with large fossae in growth bands that are not aligned.

Description. Specimens are well preserved carapaces or external moulds. Carapace of moderate size, subcircular
in outline, 3-8-8 mm long, 3-6 mm high; dorsal margin short, small umbo located near its centre; anterior and
posterior margins relatively straight, postero-ventral margin rounded and slightly expanded; about 30 growth
bands on carapace of adult, ornamented by cavernous sculpture; fossae subcircular or irregular-elliptical ; each
big fossa including several small puntae (PI. 5, figs 4, 6, 7); a row of crowded and deepened fossae along upper
margin of each growth line (PI. 5, fig. 3); growth lines smooth and thickened, without beading or tubiform
serrated lower margin. In external moulds, fossae appear as large pits which contain small granules (PI. 5,
fig. 5).

Antronestheria praecursor sp. nov.

Plate 4, figs 4, 7, 10

Etymology. This species is a precursor of the type species stratigraphically.

Holotype. An external mould of a right valve (SM J49409, PI. 4, fig. 4) from upper Lonfearn Member of the
Lealt Shale Formation in Lealt River, Isle of Skye.

Occurrence. Upper Lonfearn Member of the Lealt Shale Formation, Skye.

Additional material. Specimens are poorly preserved carapaces or exterior moulds (SM J49397-49414 (this
series includes the holotype), BMNH In 693751, LEIUG 108136, 108215,6) from the upper Lonfearn Member
of Skye.

Diagnosis. Antronostheria with fossae in growth bands aligned in vertical rows.

Description. Carapace subcircular in outline, 6-5-8 mm long, 4-75-5-9 mm high; anterior, posterior, and ventral
margins all rounded; dorsal and umbonal area always poorly preserved, so growth lines not counted exactly;
growth bands broad and flattened, covered with cavernous ornamentation ; fossae vertically extended, crowded
and deepened along upper margin of each growth line ; growth lines thickened, without beading in their lower
margins.

Remarks. Antronestheria kilmaluagensis is the dominant conchostracan of the upper Great
Estuarine Group (see above), and A. praecursor from the upper part of the Lonfearn Member could
be the earliest species of the genus.

Family eosestheriidae Zhang and Chen, 1976 (in Zhang et al. 1976)

Genus pseudograpta Novojilov, 1954

Type species. Estheria murchisoniae (Jones, 1863), from Upper Kilmaluag Formation (?) in Skye and Muck,
NW Scotland.

Occurrence. Late Middle Jurassic; W Europe and E Asia.

EXPLANATION OF PLATE 7

Figs 1 , 4, 5, 7. Pseudograpta jonesi sp. nov. I , holotype, BGS, GSC 7296g, from the Kilmaluag Formation!?),
Skye (associated with lectotype of P. murchisoniae), x 6; 4 and 5, BGS, GSC 7296/?, fragments of same shell;
4, x%; 5, x42. 7, BMNH In 63752, from the Upper Kilmaluag Formation, Camas Mor, Muck, x4-3.
Fig. 2. Pseudograpta orbita Chen, 1976, from the Upper Kilmaluag Formation, Camas Mor, Muck, BMNH
In 63754, x 5-6.

Figs 3, 6, 8-10. Pseudograpta murchisoniae (Jones, 1863), from the Upper Kilmaluag Formation, Camas Mor,
Muck. 3 and 6, BMNH In 63753; 3, x5-6; 6, x 6. 8-10, BMNH In 63755-7, respectively; 8, x420, 9,
x 420, 10, x 42.

PLATE 7

CHEN and HUDSON, Pseudograpta

538

PALAEONTOLOGY, VOLUME 34

Diagnosis. Carapace large, elliptical or subcircular in outline; growth bands broad and flattened,
ornamented by bold polygonal reticulations in dorsal and central parts of carapace, and radial striae
in ventral or postero- ventral parts; growth lines stout and convex (growth ridges), without beading
or tubiform marginal structures.

Discussion. This genus is similar in ornament to Eosestheria , but differs in its stout and convex
growth ridges and in the coarser polygonal reticulations of its growth bands. In general, the change
from reticulation near the dorsal side into radial striae near the ventral side is abrupt for
Pseudograpta (except P. morrisi sp. nov.), but for Eosestheria there is a transitional area of two types
of ornament in the carapace. This genus has a strong resemblance to Nestoria in stout and convex
growth ridges and bold polygonal reticulation, but the latter differs from the former in not having
radial striae on its growth bands near the ventral margin of the carapace. Novojilov (1954, pp.
81-82, pi. 15, fig. 10; pi. 16, fig. 1) referred another three species to this genus: P. andrewsi (Jones),
P. orientals (Eichwald), and P. olonchurensis Novojilov, but unfortunately so far no bold polygonal
reticulations or stout convex growth ridges are to be found among them.

Pseudograpta murchisoniae (Jones, 1863)

Plate 6, figs 2, 4, 6a , 7; Plate 7, figs 3, 6, 8-10

1863 Estheria murchisoniae Jones, pp. 100-101, pi. 3, figs 1-12.

1946 Bairdestheria murchisoniae (Jones); Raymond, p. 227.

1954 Pseudograpta murchisoniae (Jones); Novojilov, p. 81.

1976 Pseudograpta aff. murchisoniae (Jones); Zhang et at., p. 173, pi. 56, figs 1-8.

Lectotype. An external mould of a left valve from the syntype (British Geological Survey, Geological Society
Collection (BGS, GSC) 7296n, PI. 6, figs 2 and 6a) figured by Jones in 1863 (pi. 3, fig. 2) from the Upper
Kilmaluag Formation (?) in the canal, Icolmkill, Skye.

Occurrence. Upper Kilmaluag Formation (?) from ‘Icolmkill ’, Skye and Camas Mor, Muck, NW Scotland.

Material. Eleven specimens: syntype (BSG, GSC 7296) from Skye; BMNF1 In 63753, 63755-7, LEIUG
108137-41, 108217 from Camas Mor, Muck. Muck specimens from Beds 13 and 15 of the Kilmaluag
Formation (Andrews 1985 p. 1125).

Original Description. This species was described by Jones (1863 p. 100) as follows: ‘carapace- valve nearly
elliptical, the straight hinge-line interfering with the symmetry of the outline. The umbo is forward, at the end
of the hinge-lme, and scarcely affects the outline. The anterior extremity has a flatter curve than the posterior.
About eighteen delicate concentric ridges are usually distinctly to be observed, with their rather wide
interspaces; but some carapaces have nearly thirty ridges, with very narrow interspaces. The ornament of the
interspaces is essentially a bold, irregularly hexagonal reticulation, like that of E. minuta ; but it generally takes
on a distinct, short, vertical wrinkling of the lower part of the interspace. This sometimes presents the
modification seen in figs 3 and 8, where the wrinkling is of a much smaller pattern, and reached half way up,
or all across the interspace. Sometimes a delicate, horizontal wrinkling, without hiding altogether the reticulate
structure of the shell. Not infrequently, both the broad and the narrow interspaces are blank (figs 6 and 10).’
Judging from this description, and new observations on the slab bearing Jones’s syntype specimens, there are
at least three different forms: the first form is elliptical in outline of carapace, with 16-32 growth bands
ornamented by bold irregular hexagons near dorsal side and radial striae near ventral side; the second form
is subcircular in outline, with 15-20 growth bands and ornamentation as the first; the third form is also

EXPLANATION OF PLATE 8

Figs 1-9. Pseudograpta morrisi sp. nov., from the Upper Kilmaluag Formation, Camas Mor, Muck. 1, BMNH
In 63758, x 8 4. 2, holotype, BMNH In 63759, x 8. 3, BMNH In 63760, x 5. 4-9, BMNH In 63751 ; 4, x24;
5, x 12; 6, x 24; 7, x60;8, x240;9, x420.

PLATE 8

CHEN and HUDSON, Pseudograpta morrisi

540

PALAEONTOLOGY, VOLUME 34

elliptical in outline, but with about 40 narrow growth bands, ornamented mostly by short radial striae. The
present authors consider that this famous conchostracan species should be represented by the first elliptical
form, and assign an exterior mould of left carapace from Jones’s slab (PI. 6, figs 2 and 6 a) as the lectotype. The
second form is referred to P. orbita Chen, and the third described herein as P. jonesi sp. nov (see below).

Revised description. Carapace large, elliptical in outline, 7-5-10-3 mm long, 5-6-2 mm high; dorsal margin long
and straight, umbo small and subanterior; anterior margin straight, the posterior rounded, the ventral arched
downward; broad growth bands ornamented by bold polygonal reticulations, mesh mostly irregular
hexagonal, mesh wall thin, mesh base shallow and flattened with short septum-like structure (PI. 7, fig. 9);
radial striae frequently developed in the lower part of each growth band near ventral side of carapace; growth
ridges stout and smooth.

Remarks. Novojilov (1954, pi. 1 1, fig. 2; pi. 15, figs 5-9) described some specimens from the Upper
Jurassic in Mongolia, referring them to this species, but they may belong to Eosestheria according
to their ornament.

Pseudograpta orbita Chen, 1976 (in Zhang et al. 1976)

Plate 6, figs 1, 3, 5, 6b, 8; Plate 7, fig. 2

1976 Pseudograpta orbita Chen; Zhang et al. pp. 173-174, pi. 55, figs 1 and 2.

Holotype. An exterior mould of a left carapace from the Middle Jurassic Tuchengzi Formation at Cajiagou
near Jinlingsi, W Liaoning, China.

Occurrence. Upper Kilmaluag Formation (?) in ‘IcolmkilF, Skye and Camas Mor, Muck.

Additional material. A right valve and another left valve from the same slab as P. murchisoniae (BGS,
GSC 1296b, c), Skye; five specimens BMNH In 63754, 63762, LEIUG 108139 (two specimens), LEIUG 108141
(M88-CPJ2) from the Upper Kilmaluag Formation at Camas Mor, Muck.

Description. Carapace large, subcircular in outline; 8-5-12 mm long, 6-5— 8-5 mm high; dorsal margin short,
small umbo located between its centre and anterior end; anterior, posterior and ventral margins rather
rounded, postero- ventral margin slight expanded; growth ridges stout and convex; 16-32 growth bands broad
and flattened, ornamented with bold irregular hexagons in dorsal and middle parts of carapace, and short
radial striae in ventral part.

Remarks. This species is frequently associated with P. murchisoniae and it is very similar to the latter
in ornamentation, but differs in outline of carapace. They might be sexual dimorphs, but so far there
is no evidence of appendages or eggs in the fossils.

Pseudograpta jonesi sp. nov.

Plate 6, fig. 6d\ Plate 7, figs 1, 4, 5, 7

Etymology. In honour of Professor T. R. Jones, author of the first monograph on fossil conchostracans.

EXPLANATION OF PLATE 9

Figs 1-4. Antr one Siberia kilmaluagensis gen. et sp. nov., BMNH In 63747, Kilmaluag Formation, Trotternish,
Skye. Stereo pairs. 1 and 2, x 90. 3 and 4, x 180.

Figs 5-8. Skyestheria intermedia gen. et sp. nov., BMNH In 63739 Lonfearn Member, Lealt Shale Formation,
Lealt River, Skye. Stereo pairs. 5 and 6, x90. 7 and 8, x 180.

Figs 9-12. Dendrostracus hebridesensis gen. et sp. nov., BMNH In 63733, Kildonnan Member, Lealt Shale
Formation, south Rigg, Skye. Stereo pairs. 9 and 10, x90. 11 and 12, x 180.

PLATE 9

CHEN and HUDSON, conchostracan sculpture

542

PALAEONTOLOGY, VOLUME 34

Holotype. An external mould of a right valve with a few pieces of surviving ‘chitinous’ shell in lower surface
of the syntype slab (BGS, GSC 7296g, PL 7, fig. 1), from the Upper Kilmaluag Formation (?) at Icolmkill, Isle
of Skye; fragments from the same shell showing ornament (BGS, GSC 7296 h, PI. 7, figs 4 and 5).

Occurrence. Upper Kilmaluag Formation (?); ‘Icolmkill’, Skye and Camas Mor, Muck.

Additional material. Apart from the holotype, there are two specimens : one from the slab’s upper surface (BGS,
GSC 1296d, PI. 6, fig. 6), another (BMNH In 63752) from upper Kilmaluag Formation at Camas Mor, Muck
(bed 15 of Andrews 1985).

Diagnosis. Pseudograpta with narrow and numerous growth bands, and dominant short radial striae
on the growth bands.

Description. Carapace moderate to large, elliptical in outline, 7-7-1 1 mm long, 4-7-7-5 mm high; dorsal margin
poorly preserved, umbo situated subcentrally or near its anterior end; anterior and posterior margins
somewhat straight, ventral margin broadly arched downward; narrow growth bands about 40 in number,
ornamented with short radial striae in lower part of each interval, and reticulation only in dorsal region of
carapace; growth ridges stout and smooth.

Pseudograpta morrisi sp. nov.

Plate 8, figs 1-9

Etymology. For Mr S. Morris of the Natural History Museum, London, who collected these conchostracans
with us in August 1988.

Holotype. An external mould of a left valve (BMNH In 63759, PI. 8, fig. 2) from the Upper Kilmaluag
Formation at Camos Mor, Muck.

Occurrence. Upper Kilmaluag Formation, Camas Mor, Muck.

Additional material. Six specimens (BMNH In 63758, 63760,1 ; LEIUG 108218-20); locality and horizon same
as holotype (beds 13-15 of Andrews 1985).

Diagnosis. Small Pseudograpta with reticulation in dorsal growth bands passing into radial striae in
ventral growth bands.

Description. Most specimens are poorly preserved internal or external moulds, and only a few with ‘chitinous’
shell and ornamentation on growth bands. Carapace of moderate size, elliptical in outline, 5-4-9-8 mm long,
3-8— 6-5 mm high; dorsal margin straight and long, small and narrow umbo situated near its anterior end;
anterior and posterior margins curved symmetrically, ventral margin arched broadly and downward ; growth
lines thickened; 15-21 growth bands, broad and flattened, ornamented by very irregular polygons near dorsal
side and gradually changed into striae near ventral or postero-ventral parts of carapace.

Discussion. This species is the smallest of the genus Pseudograpta , and might be referred to
Eosestheria based on the transitional ornament from reticulation in the dorsal side to radial striae
in the ventral side, but for the thickened growth lines. Eosestheria was widely distributed in E Asia
during the late Jurassic and it was considered to evolve from Nestoria (Zhang et al. 1976, pp. 79-80,

EXPLANATION OF PLATE 10

Figs 1-12. Neopolygrapta lealtensis gen. et sp. nov., Lealt Shale Formation, Lonfearn, Skye. Stereo pairs. 1,
2, 9-12, BMNH In 63728; 1 and 2, x90; 9 and 10, x90; II and 12, x 180. 3-8, BMNH In 63730; 3 and
4, x 180; 5 and 6, x 90; 7 and 8, x 180.

PLATE 10

CHEN and HUDSON, Neopolygrapta sculpture

544

PALAEONTOLOGY, VOLUME 34

text-fig. 38). Based on the new material from the Isle of Muck, there is a greater possibility that
Pseudograpta morrisi sp. nov. is the ancestor of Eosestheria.

Acknowledgements. We thank the Royal Society and Academia Sinica for sponsoring Chen Pei-ji’s visit to
Leicester and J. D. Hudson's visit to China, and the British Petroleum Company pic for further support for
Chen Pei- ji, as well as allowing us to examine material from their Porcupine Basin cores. We thank Sam
Morris, Matthew Wakefield and Mark Wilkinson for help in the field, and Sam Morris, Roy Clements, Mike
Dorling and Hugh Ivimey-Cook for access to material in their charge at the British Museum (Natural History),
Leicester University, the Sedgwick Museum, Cambridge, and the British Geological Survey respectively, and
for discussing it with us. Rod Branson operated the SEM and Sue Button drew the figures. Clare Stanga, Liane
Baldock and Christine Burnett processed several versions of the manuscript.

REFERENCES

anderson, F. w. and dunham, k. c. 1964. The geology of Northern Skye. Memoirs of the Geological Survey,
Scotland, Her Majesty’s Stationery Office, Edinburgh, 216 pp.

Andrews, J. e. 1985. The sedimentary facies of a late Bathonian regressive episode: the Kilmaluag and
Skudiburgh Formations of the Great Estuarine Group, Inner Hebrides, Scotland. Journal of the Geological
Society of London, 142, 1119-1137.

— Hamilton, p. j. and fallick, a. e. 1987. The geochemistry of early diagenetic dolostones from a low-
salinity Jurassic lagoon. Journal of the Geological Society of London , 144, 687-698.

— and walton, w. 1990. Depositional environments within Middle Jurassic oyster-dominated lagoons: an
integrated litho-, bio- and palynofacies study of the Duntulm Formation (Great Estuarine Group, Inner
Hebrides). Transactions of the Royal Society of Edinburgh, Earth Sciences, 81, 1-22.

bradshaw, m. j. and fenton, j. 1982. The Bajocian LCornbrash’ of Raasay, Inner Hebrides: palynology, facies
analysis and a revised geological map. Scottish Journal of Geology, 18, 131-145.
chen pei-ji and shen yan-bin 1982. Late Mesozoic conchostracans from Zhejiang, Anhui and Jiangru
Provinces. Palaeontologia Sinica, 161, (New Series B 17), 1-1 17, 47 pis. [In Chinese, with English summary].
— 1985. An introduction to fossil Conchostraca. Science Press, Beijing, 241 pp., 26 pis. [In Chinese],

— LI WENBIN, CFtEN JINHUA, YE CHUNHUI, WANG ZHEN, SHEN YAN-BIN and SUN DONGLI, 1982. Sequence of
fossil biotic groups of Jurassic and Cretaceous in China. Scientica Sinica, Series B, 25, 1011-1020.

COPE, J. C. W., DUFF, K. L., PARSONS, C. F., TORRENS, H. S., WIMBLEDON, W. A. and WRIGHT, J. K. 1980. A
correlation of the Jurassic rocks in the British Isles. Part 2. Middle and Upper Jurassic. Special Reports of
the Geological Society of London, No. 15, 109 pp.

defretin, s. 1965. Etude et revision du phyllopodes conchostraces en provenance d’U.S.S.R., Annates de la
Societe Geologique du Nord, 85, 1 5—48, pis 1-3.

deperet, c. and mazeran, p. 1912. Les Estheria du Permien d’Autun. Bulletin de la Societe d'Histoire Naturelle
d'Autun, 25, 163-167, pi. 5.

forsyth, i. h. 1960. A marine shell-bed near the base of the Estuarine Series in Raasay. Transactions of the
Edinburgh Geological Society, 17, 273-275.

Harris, j. p. and Hudson, j. d. 1980. Lithostratigraphy of the Great Estuarine Group (Middle Jurassic), Inner
Hebrides. Scottish Journal of Geology , 16, 231-250.

Hudson, j. d. 1963a. The recognition of salmity-controlled mollusc assemblages in the Great Estuarine Series
(Middle Jurassic) of the Inner Hebrides. Palaeontology, 6, 318-326.

1963fi. The ecology and stratigraphical distribution of the invertebrate fauna of the Great Estuarine
Series. Palaeontology, 6, 327-341.

1966. Hugh Miller's Reptile Bed and the Mytilus Shales, Middle Jurassic, Isle of Eigg, Scotland. Scottish
Journal of Geology, 2, 265-281.

— 1980. Aspects of brackish-water facies and faunas from the Jurassic of north-west Scotland. Proceedings
of the Geologists' Association of London, 91, 99-105.

1983. Mesozoic sedimentation and sedimentary rocks in the Inner Hebrides. Proceedings of the Royal
Society of Edinburgh, Series B, 83, 47-63.

jones, T. r. 1863. A monograph of the fossil Estheriae. Monograph of the Palaeontographical Society, 14 (62),
1-134, 5 pis.

murchinson, r. i. 1829. On the coalfield of Brora in Sutherlandshire, and some other stratified deposits in the
north of Scotland. Transactions of the Geological Society of London, Series 2, 2, 293-326.

CHEN AND HUDSON: JURASSIC CONCHOSTR ACANS

545

novojilov, N. 1954. Upper Jurassic and Cretaceous conchostracans from Mongolia. Trudy Palaeontological
Institute , Academy of Sciences, USSR , 48, 7-124, pis 1-17. [In Russian].

- 1957. Crustaces bivalves de l’ordre des conchostraces du Cretace inferieur chinois et africain. Annales de
la Societe Geologicpie du Nord, 67, 235-243, pi. 17.

Raymond, p. e. 1946. The genera of fossil Conchostraca - an order of bivalved Crustacea. Bulletin of the
Museum of Comparative Zoology , Harvard College, 96, 217-307, pis 1-6.
sars, G. o. 1867. Histoire naturelle des Crustaces d'eau douce Norvege. C. Johnson, Christiana, 145 pp., 10 pi.
smith, a. G. and briden, J. C. 1977. Mesozoic and Cenozoic palaeocontinental maps. Cambridge University
Press, Cambridge, 63 pp.

sykes, R. m. 1975. The stratigraphy of the Callovian and Oxfordian stages (Middle-Upper Jurassic) in northern
Scotland. Scottish Journal of Geology, 11, 51-78.

tan, f. c. and Hudson, j. d. 1974. Isotopic studies on the palaeoecology and diagenesis of the Great Estuarine
Series (Jurassic) of Scotland. Scottish Journal of Geology, 10, 91-128.
tasch, p. 1969. Order Conchostraca. R141-R163. In moore, r. c. (ed.). Treatise on invertebrate paleontology.
Part R. Arthropoda 4 (1). Geological Society of America and University of Kansas Press, Boulder, Colorado
and Lawrence, Kansas, 398 pp.

1987. Fossil Conchostraca of the Southern hemisphere and continental drift. Paleontology, biostrati-
graphy and dispersal. Geological Society of America, Memoir 165, 1-290, 49 pis.
zhang wen-tang, chen pei-ji and shen yan-bin, 1976. Fossil Conchostraca of China. Science Press, Beijing,
325 pp., 138 pis. [In Chinese].

— shen yan-bin and niu shao-wu 1987. Discovery of Jurassic conchostracans with well-preserved soft parts
and notes on its biological significance. Acta Palaeontologica Sinica, 26, 111-129, 5 pis.

zieten, c. h. van 1833. Die versteinerungen Wurttembergs. Schweitzerbart, Stuttgart, 453 pp.

CHEN PEI-JI

Nanjing Institute of Geology and Palaeontology
Academia Sinica, Chi-Ming Ssu
P.O. Box 210008, Nanjing, China

J. D. HUDSON

Department of Geology
University of Leicester
Leicester LEI 7RH

Typescript received 20 December 1989
Revised typescript received 30 September 1990

A NEW PRIMITIVE DINOCEPH ALI AN MAMMAL-
LIKE REPTILE FROM THE PERMIAN OF
SOUTHERN AFRICA

by BRUCE S. RUBIDGE

Abstract. A new genus of tapinocephaline dinocephalian, Tapinocaninus pamelae, is described from the
lowermost biozone of the Beaufort Group in South Africa. Tapinocaninus is older than any dinocephalian
previously discovered in Africa, being early Late Permian (Ufimian-Kazanian) in age. It is considered to
belong to the subfamily Tapinocephalinae but retains plesiomorphic features which are absent in the
Tapinocephalinae and were previously considered characteristic of the more primitive Titanosuchinae. The
Titanosuchmae are distinguished from the Tapinocephalinae in the form of the postcanine teeth (leaf-shaped
and serrated in the former and bearing a lingual heel in the latter) and in the pachyostotic thickening of the
skull roof of the latter.

Dinocephalians were a group of medium to large mammal-like-reptiles which included the
largest land-living animals to have existed up to Late Permian (Ufimian-Kazanian) times. They
were present among the earliest therapsid fauna of the San Angelo Formation of Texas (Olson 1962)
and the Russian Ocher Formation (Tchudinov 1983), and also formed an important part of the
therapsid fauna of the Tapinocephalus Zone (Dinocephalian Assemblage Zone) of the Beaufort
Group in South Africa (Boonstra 1969, 1971; Kitching 1977; Keyser and Smith 1978).

The Dinocephalia are the earliest group of therapsids for which a significant adaptive radiation
can be identified (Ruben 1986), but for all their early success, they became completely extinct by the
close of Tapinocephalus Zone (Dinocephalian Assemblage Zone) times, leaving no descendant
groups (Boonstra 1971 ; Kemp 1982). The infra-order Dinocephalia was considered to consist of six
families, namely Estemmenosuchidae, Brithopodidae, Anteosauridae, Titanosuchidae, Tapino-
cephalidae and Styracocephalidae (Boonstra 1963a; Kemp 1982). More recently. King (1988)
reclassified these as subfamilies of the three families Estemmenosuchidae, Brithopodidae, and
Titanosuchidae, and regarded Styracocephalus , the only genus in the family Styracocephalidae, as
incertae sedis. Of the three families recognized by King, only the Estemmenosuchidae have not yet
been found in Africa. The classification of King (1988) is used in this study. Apomorphic features
which characterize the Dinocephalia as a distinct infraorder are the presence of intermeshing incisor
teeth in the upper and lower jaws (Boonstra 1962, 1963a; Hopson and Barghusen 1986; King 1988),
and incisor teeth with lingual heels (Hopson and Barghusen 1986).

MATERIALS AND METHODS

Five dinocephalian skulls, which are considered to belong to the same genus, have recently been collected in
rocks of the Late Permian Eodicynodon Zone, a newly identified biozone at the base of the Beaufort Group
(Rubidge 1987). All but one of these fossils were found on Modderdrift farm, three from the same sandstone
outcrop in close proximity. These three specimens were found by Mr John Nyaphuli and are housed in the
Karoo fossil reptile collection of the National Museum, Bloemfontein (specimens NMQR 2985, 2986, 2987),
while the fourth specimen (ROZ K95), although found by the author, is housed in the private fossil collection
of Mr Roy Oosthuizen on his farm, Zwartskraal, in the district of Prince Albert. The fifth specimen (NMQR
3097), which consists only of a skull roof and some isolated teeth, was found on Swartgrond farm in the
Rietbron district and is now housed in the National Museum in Bloemfontein.

IPalaeontology, Vol. 34, Part 3, 1991, pp. 547-559.)

© The Palaeontological Association

548

PALAEONTOLOGY, VOLUME 34

The lower jaws are preserved in all but one of the specimens, but those of ROZ K95 and NMQR 2986 were
too firmly appressed to the skull prior to fossilization to prepare on the lingual side. Although the right ramus
of the lower jaw of the specimen NMQR 2985 could be dissociated from the skull, the bones of the lower jaw
on the medial side were damaged and disoriented prior to fossilization. By far the best preserved lower jaw
is that of NMQR 2987 which is undistorted and still bears several teeth in situ.

All the skulls were prepared with the aid of air-driven engravers. A hammer and chisel and small angle
grinder were used in areas where there was an excess of matrix covering the bone and where there was no
chance of damage to the bone. Glyptal 1276 Lacquer Cement, supplied by General Electric, was used as an
adhesive.

Institutional abbreviations are: NMQR, National Museum, Bloemfontein; ROZ K, Roy Oosthuizen
Collection, Klaarstroom.

SYSTEMATIC PALAEONTOLOGY

Subclass synapsida Osborn, 1903
Order therapsida Broom, 1905
Infraorder dinocephalia Seeley, 1894
Family titanosuchidae Boonstra, 1972
Subfamily tapinocephalinae Lydekker, 1890
Tapinocaninus gen. nov.

Type species. Tapinocaninus pamelae sp. nov.

Diagnosis. Large tapinocephaline dinocephalian. Moderate pachyostosis with broad postorbital
bar. Temporal opening relatively large for tapinocephaline, and hence relatively narrow
intertemporal region. Premaxilla forming anterior and antero-lateral border of the internal nares.
Maxilla with very short contact on lateral side of internal nares such that the palatine and
premaxilla almost meet. Vomer narrow in ventral view. Heterodont dentition with medium-sized
canine tooth lacking definite crushing heel in both upper and lower jaw. Lower canine passing in
front of upper canine rather than on lingual side. Five larger incisor teeth in premaxilla, all with
crushing heel. Quadrate condyles situated below posterior border of orbit. Several small palatal
teeth on palatine boss.

Etymology. Greek tapino, humble; Latin caninus , canine. The name refers to the reduced size of the canine
tooth relative to that of the Titanosuchinae and indicates that the genus belongs to the subfamily
Tapinocephalinae.

Tapinocaninus pamelae sp. nov.

Text-figures 1-3.

Diagnosis. As for genus.

Etymology. The species is named in honour of my mother Pam Rubidge who encouraged my interest in
palaeontology and was with me in the field when this fossil was discovered.

Holotype. Specimen NMQR 2987. Skull and mandible. Left side of skull well preserved with several teeth;
parietal and occipital regions on right side damaged as a result of weathering.

Paratypes. Specimens NMQR 2985 (skull and mandible), NMQR 2986 (skull and mandible), ROZ K95 (skull
and mandible).

Type horizon and locality. Eodicynodon Zone (Rubidge 1987) at the base of the Beaufort Group
(Ufimian-Kazanian age) on Modderdrift farm, Prince Albert, South Africa (map sheet : South Africa 1 : 50000,
Sheet 3322 BA Seekoegat, first edition).

RUBIDGE: DINOCEPH ALI AN FROM SOUTHERN AFRICA

549

DESCRIPTION

Skull Roof

In their general shape the newly discovered dinocephalian skulls resemble those of other tapinocephaline
dinocephalians in that the skull roof is greatly pachyostosed and the postorbital bar is extremely thick.

The anterior portion of the snout is formed by the premaxilla which also forms the anteroventral, anterior,
and dorsal border of the external nares. The premaxilla extends posterodorsally as a narrow projection on the

text-fig. 1. Tapinocaninus pamelae , lateral view, a, NMQR 2987; b, NMQR 2986; c, ROZ K95.

550

PALAEONTOLOGY, VOLUME 34

roof of the snout to a point approximately halfway between the external nares and the orbit where it is in
sutural contact with the nasals. The contact of the premaxilla with the maxilla on the lateral side of the snout
is an obliquely orientated suture which extends ventrally from approximately one-third of the way along the
antero-ventral side of the external nares to the front of the canine alveolus.

Anteriorly, the nasal is in contact with the posterior border of the external naris and the septomaxilla, while
it is in contact with the maxilla on the antero-ventral side and the prefrontal on the postero-ventral side. On
the dorsal side of the skull the nasals extend posteriorly as a narrow projection medial to the enlarged
prefrontal bone and meet the frontals in line with the anterior limit of the orbits.

The maxilla forms the greater portion of the lateral surface of the snout. Dorsally it is in sutural contact with
the nasal, and posteriorly with the prefrontal, lachrymal, and jugal in an almost straight vertical line anterior
to the orbit. The lateral surface of the maxilla, above the alveolus of the canine, is swollen laterally to
accommodate the root of the canine as in titanosuchine dinocephalians (Boonstra 1962).

As in all known South African dinocephalians, the prefrontal is a prominent bone and forms the antero-
dorsal border of the orbit. Posteriorly and posterodorsally, the prefrontal is in contact with the frontal, antero-
dorsally with the nasal and antero-ventrally with the maxilla. Almost halfway down the front of the orbit, the
prefrontal is in sutural contact with the lachrymal which extends on the lateral side of the skull anterior to the
orbit. The suture between the lachrymal and jugal is situated slightly less than one-third of the way up from
the base of the orbit and slopes anteroventrally.

The jugal is a prominent bone which forms the ventral margin of the orbit and of the lateral side of the
cranium in the suborbital region. It extends slightly anteroventrally of the orbit on the lateral surface of the
snout and has a long sutural contact with the posterior edge of the maxilla. Behind this contact, the ventral
margin of the skull curves posteroventrally until it reaches its most ventral point below the postorbital bar
lateral to the articulatory condyles of the quadrate. From here, the ventral margin is notched posterodorsally
so that the jugal forms the anterior and dorsal walls of the notch. The posterior edge of the notch is formed
by the quadratojugal.

The postorbital bar, which is broad, is made up of the postfrontal dorsally and the postorbital ventrally. The
dividing suture between the postorbital and postfrontal is orientated horizontally about halfway up the orbit.
The postorbital forms the anterodorsal and anteroventral borders of the temporal fenestra and the
posteroventral border of the orbit.

It appears that the postfrontal forms most of the dorsal margin of the temporal fenestra, with the parietal
forming the posterodorsal margin. However, the suture between the postfrontal and parietal could not be
identified. The frontal forms most of the skull roof between the orbits and temporal fenestrae. It has a laterally
tapering process which forms a small portion of the dorsal border of the orbits between the prefrontal and
postfrontal bones.

The squamosal forms the entire ventral and posterior borders of the temporal fenestra. On the ventrolateral
side of the external auditory meatus the squamosal is in contact with the quadratojugal , while on the ventral
side it meets the ascending ramus of the quadrate. Posteromedially the squamosal is in contact with the bones
of the occipital region. The quadratojugal, in posterior view, is seen to form the ventrolateral corner of the
skull. Dorsally its posterior surface is overlapped by the squamosal.

Dorsally, the squamosal is in contact with the parietal , which is badly weathered in all the specimens except
NMQR 3094. A large rounded pineal foramen is situated between the temporal fenestrae. The intertemporal
region becomes very narrow behind this foramen.

Occiput

As in most tapinocephaline dinocephalians, the occipital region of the skull slopes posterodorsally. The occiput
is roughly flat and rectangular in shape, with the ventro-lateral regions of the opisthotics curving ventrally to
meet the ventro-medial process of the squamosal and the dorsomedial portion of the quadrate medial to the
external auditory meatus. Although the lateral and dorsal contacts of the occipital and otic bones with the
tabular and postparietal bones are prominent, the sutures between the various occipital and otic bones are not
easily distinguishable. The postparietal is an unpaired element situated dorsomedially in the occipital plate and
forms a horizontal sutural contact with the parietal posterodorsally. The tabular is a roughly triangular bone
between the postparietal, squamosal, and occipital bones with its apex extending ventrally to the point of
contact between the opisthotic and squamosal. Medially it is in contact with the occipital bones.

RUBIDGE: DINOCEPHALI AN FROM SOUTHERN AFRICA

551

50 mm

text-fig. 2. Tapinocaninus pamelae. a, lateral view ofNMQR 2985; b, dorsal view of NMQR 2987; c, dorsal

view ofNMQR 2985.

552

PALAEONTOLOGY, VOLUME 34

Palate

In the palate, the premaxilla overlies the anteroventral surface of the maxilla on the ventral side. Its posterior
border extends posteromedially from a point just in front of the canine to form the anterolateral border of the
internal naris. The maxilla forms only a small portion of the lateral border of the internal nares on the medial
side of the canine and prevents contact between the palatine and premaxilla. From here it forms a thin alveolar
margin which extends posteriorly to the level of the lateral flange of the pterygoid. Medially, the maxilla forms
an extensive contact on the palatine and also meets the ectopterygoid.

The vomers are narrow vertically expanded bones which form a median trough between them. They make
up the medial and posteromedial walls of the internal nares and are in contact with the premaxillae anteriorly
and the palatines posterolaterally. Posteriorly they have a short pointed contact with the pterygoids.

An anteriorly directed process of the palatine extends almost as far as the canine tooth and forms the
posterolateral, and part of the lateral, border of the internal naris. Posteromedially, the palatine forms a
ventrally projecting boss which is transversely sutured to the medial pterygoid boss. The roots of several
palatine teeth are situated on the anterior regions of the palatine portion of this boss in specimen NMQR 2985,
and alveoli are present in specimens NMQR 2986 and NMQR 2987.

The pterygoid accounts for more than half the length of the palate, and the lateral flanges stretch ventrally
and transversely across the palate. Posterior to the lateral flanges, the pterygoid meets its mate in the midline
to form a sharp median keel. Lateral to the median keel the pterygoid is deeply excavated ventrally. On the
lateral side of this wide and posteriorly flared groove, the vertically orientated quadrate ramus of the pterygoid
extends posterolaterally to meet the medial condyle of the quadrate. The quadrate ramus does not extend as
far as the posterior edge of the quadrate. A narrow interpterygoidal vacuity is present and extends posteriorly
from the level of the lateral flanges of the pterygoid. The pterygoid meets the basisphenoid immediately behind
the interpterygoidal vacuity.

Braincase

The stapes extends medially from the posterior surface of the quadrate, immediately behind the quadrate ramus
of the pterygoid, to the fenestra ovalis. In ventral view, its distal end is expanded anteroposteriorly where it
meets the quadrate. It becomes thinner towards the medial end which extends to the fenestra ovalis. In
posterior view, the stapes is also narrower medially than laterally, and is pierced anteroposteriorly by a
stapedial foramen.

The basisphenoid is a relatively small triangular-shaped bone with its apex pointing anteriorly to meet the
pterygoid immediately posterior to the interpterygoidal vacuity. A keel extends ventrally down the midline of
this bone. On either side of the keel are two foramina for the carotid arteries. Posteriorly, the basisphenoid is
in contact with the basioccipital in the anterior region of the fenestra ovalis to form the anterior edge of the
fenestra.

Lower jaw

The lower jaw consists of the dentary, splenial, angular, surangular, prearticular, and articular bones, but no
coronoid could be recognized.

The dentaries are united by a symphysial suture in the anterior midline which is very clearly defined in
specimen NMQR 2987. The dentary forms the greater portion of the lower jaw. Posteriorly, it meets the
surangular at the top of the coronoid eminence. The surangular has its greatest surface exposure on the dorso-
medial side of the lower jaw, but also forms a laterally projecting ridge on the dorsolateral surface of the lower
jaw. This ridge continues posteriorly as far as the ventral tip of the retroarticular process of the articular bone.

The articular is situated on the posteromedial side of the surangular and angular bones, and is in sutural
contact with the posterior edge of the prearticular. The articulatory surface of the articular is orientated
posterodorsallv and has two prominent depressions which relate to the quadrate condyles. The more laterally
situated depression is slightly larger than the medial one.

The angular meets the surangular on the dorsolateral side of the lower jaw immediately ventral to its laterally
extending ridge. A prominent reflected lamina is formed by the angular and extends ventrally as far as the
ventral margin of the jaw. Anteriorly, the angular is in contact with the dentary by means of a suture which
slopes anteroventrally such that the angular forms a pointed contact on the ventral margin of the lower jaw
between the dentary on the lateral side and the splenial on the medial side. The angular is also exposed on the
ventromedial side of the posterior end of the lower jaw.

RUBIDGE: DINOCEPH ALI AN FROM SOUTHERN AFRICA

553

al can
max

earn

popar

50 mm

50 mm

for mag

text-fig. 3. Tapinocaninus pamelae. a, ventral view of NMQR 2985; b, ventral view of NMQR 2987;
c, occipital view of NMQR 2985; d, occipital view of NMQR 2987.

554

PALAEONTOLOGY, VOLUME 34

The splenial is widely exposed on the medial side of the lower jaw. It attains its greatest width near its
posterior end where it is in contact with the prearticular posterodorsally, and the angular posteroventrally. The
splenial tapers anteriorly as the dentary becomes broader.

Dentition

Teeth are generally poorly known in South African dinocephalians (Boonstra 1962), and the dentition is not
perfectly present in any of the specimens studied. In one of the skulls (NMQR 2985), all the teeth had fallen
out of the alveoli prior to fossilization.

A maximum of five incisor teeth, all of which have talons and heels, are present in the premaxilla. A single
canine , with no heel, is present as the first tooth in the maxilla (Text-fig. 3a, b). The crown of the canines curves
backwards and the posterior edge is rounded with no sharp edge or serrations.

The canine of specimen ROZ K95 is mediolaterally compressed when compared with that of specimens
NMQR 2986 and NMQR 2987 which are more rounded in section. Boonstra (1953a) records in a specimen
of Anteosaurus that a canine close to replacement is also flattened. This may well be the situation in the present
instance as the preserved canine of specimen ROZ K95 is longer than those of the other individuals. This canine
is thus possibly more mature than the others, and at the point of being replaced.

The canine is followed, on the posterior side, by at least eight postcanines (twelve in specimen NMQR 2986,
and eight in specimen NMQR 2985). The postcanine teeth have arrow-shaped crowns with small heels on the
lingual side.

In the lower jaw, four incisors with crushing heels are present anteriorly in each dentary and intermesh with
the incisors of the upper jaw when occluded. A single canine is present which passes on the anterior side of the
canine of the upper jaw when the jaws are occluded, in the same fashion as in the Titanosuchinae. At least 15
smaller postcanine teeth are also present on the dentary, and lie lingually of the upper set when the jaws are
closed.

In summary, this new dinocephalian has a heterodonl dentition, consisting of incisors, canine, and
postcanines, as in the Anteosaurinae and Titanosuchinae, and not homodont, as in the Tapinocephalinae
(Boonstra 1953a). The dental formulae of the various specimens studied are: NMQR 2985 Ijj CJ PC44; NMQR
2986 I45 CJ PC.}2; NMQR 2987 I* C{ PCjj_10; ROZ K95 If C.) PC?.

COMPARISON WITH OTHER DINOCEPHALIANS

Skull roof

In general shape, the skull of Tapinocaninus resembles that of a tapinocephaline dinocephalian as
described by Boonstra (1969) and King (1988). The skull roof is greatly pachyostosed, and the
postorbital bar is extremely thick, as in all the Tapinocephalinae. Because the postorbital and
posttemporal bars of the Anteosaurinae and Titanosuchinae are relatively slender, the temporal
fossa of these subfamilies has a relatively greater anteroposterior diameter than in the
Tapinocephalinae. In the last, the thickening of these two bars produces a narrowed dorsoventrally
elongated slit-like posttemporal fossa. As in anteosaurines and tapinocephalines, the temporal
openings of Tapinocaninus are relatively large with a resultant narrow intertemporal region when
compared with other forms of the Tapinocephalinae, although there are genera such as Avenantia
and Ulemosaurus which do have a narrow intertemporal region. In the Tapinocephalinae, a narrow
intertemporal region is considered to be primitive (Hopson and Barghusen 1986).

In dorsal view, the snout is much thinner than that of other tapinocephaline dinocephalians,
except Ulemosaurus , but is not as long as that of the Titanosuchinae.

The quadratojugal of Tapinocaninus is a relatively small bone which forms part of the
ventrolateral surface of the skull. In the early therapsids this bone has variable relations. However,
in all of them it is much reduced in size and never enters the lower temporal arch as it does in some
of the Pelycosauria. Primitively, the quadratojugal was a surface bone of the posterolateral corner
of the skull flanking the quadrate. It is reduced in size and displaced medially in some of the higher
sphenacodonts to rest on the quadrate dorsal to the lateral condyle (Boonstra 1971). This medial
displacement is continued in the Gorgonopsia, Therocephalia, Brithopodinae, and Anteosaurinae.

RUBIDGE: DINOCEPH ALI AN FROM SOUTHERN AFRICA

555

In the Titanosuchinae and Tapinocephalinae the quadratojugal, variable in size and shape, still
forms part of the lateral skull surface (Boonstra 1971).

The articulatory surface of the quadrate is situated ventrally below the posterior border of the
orbit, a relatively anterior position which corresponds more closely with the Tapinocephalinae than
the Titanosuchinae, where it is situated further posteriorly.

Occiput

In the occiput, the tabular is a triangular-shaped bone which comes to a point between the
squamosal and opisthotic on the ventral side, well below the post-temporal foramen. This
configuration is not present in any of the Titanosuchinae or Anteosauridae, while a similar
configuration is present in some of the tapinocephalines.

Braincase

A prominent stapedial foramen is present. Among the Dinocephalia, where the stapes is known, this
structure is not present in the Anteosaurinae and Brithopodinae (Boonstra 1971), but is present in
some other species of the Tapinocephalinae (Boonstra 1956, 1957, 1965).

Palate

In the palate, the premaxilla forms a large proportion of the anterolateral margin of the internal
nares as in the Titanosuchinae and the Estemmenosuchoidea. In the Tapinocephalinae and in
Anteosaurus , the premaxilla has only a small contact on the anteromedial side of the internal naris.
The vomers are thin bones, as in the Brithopodinae (Orlov 1958), and not broad and flat as in all
other South African dinocephalians, especially the Tapinocephalinae.

A single row of palatal teeth is present on the anterior margin of the palatine bosses. Small
clusters of palatal teeth are known in the Brithopodinae, Anteosaurinae, and in Styracoceplia/us ,
which is now considered incertae sedis (King 1988). Palatal teeth are normally thought not to occur
in tapinocephalines : Boonstra ( 1 936 p. 97) states ‘ in the gorgonopsians the palatines and pterygoids
carry dentigerous ridges, whereas no such teeth are known in my tapinocephalid '. Boonstra (19536)
does however mention indications of roots of small palatal teeth on a rounded mound on the
palatine of Struthiocephalus. As far as the titanosuchines are concerned, Boonstra ( 19636) could not
find any palatal teeth in the type of Jonkeria ingens.

Dentition

One of the diagnostic features of dinocephalians is the dentition, but in Tapinocaninus the
premaxillary-maxillary dentition is not perfectly preserved in any of the specimens. Boonstra (1962)
states that the entire tooth (crown plus root) falls out of the alveolus before petrification most
frequently in the Tapinocephalinae. In the Titanosuchinae the roots are generally preserved in the
alveoli even though the crowns are frequently broken off at the level of the alveolar border either
before petrification or during weathering. He considered this to be because of the difference in
implantation and the mode of replacement of teeth in these two subfamilies.

The newly-discovered dinocephalian skulls have incisors with talons and crushing heels, a single
canine with no heels, and numerous postcanine teeth; features of the Titanosuchinae (Boonstra
1963a). As in titanosuchine dinocephalians, the first pair of upper incisors of Tapinocaninus is more
lightly built than the others and they are laterally compressed, they lie close together and, in
occlusion, they passed their talons together between the pair of lower incisors (Boonstra 1953a).

In the Titanosuchinae, the fifth upper incisor has the rear face of the talon modified to receive
the lower canine with which it intermeshes. In the lower jaw, the fourth incisor is somewhat weaker
than the anterior ones (Boonstra 1962). In the Anteosaurinae, the fifth upper incisor passes labially
to the lower canine during occlusion and not anteriorly, and the two teeth consequently do not
intermesh (Boonstra 1962). In the new dinocephalian, the lower canine in specimen NMQR 2986
(the only one where it is preserved) is situated more in the titanosuchine position, passing in front
of the upper canine (Text-fig. 1b). The size of the upper canine of Tapinocaninus (diameter at

556

PALAEONTOLOGY, VOLUME 34

alveolar border, 19-21 mm anteroposteriorly, and 13-18 mm mediolaterally ; crown length, 46-
52 mm) is much smaller than that of equivalent-sized titanosuchine dinocephalians (diameter at
alveolar border, 35-50 mm anteroposteriorly, and 21-36 mm mediolaterally; crown length, 60-
110 mm) (Boonstra 1962).

As is the case in titanosuchine dinocephalians (Boonstra 1962), the postcanines of the lower jaw
of Tapinocaninus are situated on the lingual side of those of the upper jaw when the jaws are closed,
and do not intermesh as in tapinocephaline dinocephalians. The number of postcanine teeth in the
Anteosaurinae and Titanosuchinae varies greatly, from one to nineteen, and is of doubtful
systematic value as it can even vary from the left to right side of the jaw (Boonstra 1953a).

The postcanines of the lower jaw of Tapinocaninus more closely resemble those of the
Tapinocephalinae than the Titanosuchinae in that the labial portion of the crown is slightly
bulbous, and there are small heels on the lingual side.

Despite having heterodont titanosuchine-like dentition, the skulls display many tapinocephaline
features such as postcanines with talon and heel (Boonstra 1963a), the pachyostosed skull roof and
form of the postorbital bar, relatively anterior position of the quadrate (Boonstra 19636), and
orientation of the tabular bone. It is thus reasonable to consider this dinocephalian as belonging to
the Tapinocephalinae but still retaining some plesiomorphic characters which are not present in any
other known tapinocephaline dinocephalian.

TAXONOMIC IMPLICATIONS

Several palaeontologists have addressed the problem of classification of the Dinocephalia, but only
recently has a cladistic approach been used. Kemp (1982) was the first to use such an approach but
did not give characters relating to the sister groups contained in his cladogram. Hopson and
Barghusen (1986) undertook a cladistic analysis of therapsid relationships and included the various
dinocephalian infraorders, while King (1988) also drew up a cladistic analysis of the Dinocephalia.
All consider, as was previously stated by Boonstra (1965), that the Anteosaurinae are the most
primitive of the South African dinocephalians, and the Tapinocephalinae are the most derived. The
Titanosuchinae fall between the two. The scheme proposed by Hopson and Barghusen (1986) differs
from that of Kemp (1982) and King (1988) in the position of the Estemmenosuchoidea. This
however does not have any bearing on this paper and will not be discussed further.

Hopson and Barghusen (1986) use the following apomorphies of the Tapinocephalinae to
separate them from the Titanosuchinae :

1. canines incisiform;

2. pattern of interlocking extends to canines and postcanines;

3. anterior end of postcanine row does not extend lingual to the canine.

A reassessment of material of the Tapinocephalinae and Titanosuchinae housed in South African
Institutes has revealed that character 3 is shared by both of these groups and is not an apomorphy
of the Tapinocephalinae alone.

King (1988) uses the following apomorphies of the Tapinocephalinae to separate them from the
Titanosuchinae. Character 2 corresponds with that of Hopson and Barghusen (1986):

1. heels on the incisor teeth expanded;

2. all teeth interdigitating;

3. size of temporal fossa reduced due to pachyostosis.

Of all these characters, Tapinocaninus possesses only the first character of King (1988), and hence
it is considered the most primitive representative of the Tapinocephalinae known. As a result, it is
necessary to broaden the concept of what is presently considered to be a tapinocephaline
dinocephalian (Text-fig. 4).

RUBIDGE: DINOCEPH ALI AN FROM SOUTHERN AFRICA

557

Tapinocephalinae

i 1 i

Brithopodinae Anteosaurinae Titanosuchinae Tapinocaninus advanced Tapinocephalinae

text-fig. 4. Cladogram illustrating the relationships of Tapinocaninus to the Titanosuchinae and
Tapinocephalinae (modified after King 1988). Apomorphies for each node are: 1 (defining Titanosuchidae),
enlarged heel on incisor teeth; jaw hinge slightly anteriorly placed to be positioned below postorbital bar; lower
canine reduced in height; lower canine positioned anterior to upper one when jaws closed. 2 (defining
Titanosuchinae) (after King 1988), many leaf-shaped and serrated postcanine teeth. 3 (defining Tapino-
cephalinae), heel on postcanine teeth; pachyostotic thickening of postorbital bar and skull roof. 4
(autapomorphies of Tapinocaninus ), vomer narrow in ventral view; premaxilla extends far posteriorly to
almost touch palatine resulting in maxilla having very short contact on lateral border of internal nares. 5
(defining advanced Tapinocephalinae), reduced canine tooth; quadrate further anteriorly positioned to be
situated below orbit; all teeth mterdigitating; size of temporal fenestra reduced due to pachyostosis.

CONCLUSION

A new genus and species of tapinocephaline dinocephalian, Tapinocaninus pamelae, has been
discovered from the lowermost rocks of the Beaufort Group in the same zone as the primitive
dicynodont Eodicynodon (Rubidge 1987), and Patranomodon , the most primitive anomodont
known (Rubidge and Hopson 1991). Tapinocaninus pamelae is the most primitive member of the
Tapinocephalinae yet discovered and retains several plesiomorphic titanosuchine-like characters
which are not present in any other known tapinocephaline dinocephalians. This has led to
modifications of cladistic characters previously used (Hopson and Barghusen 1986; King 1988) to
separate the Tapinocephalinae from the Titanosuchinae.

Acknowledgements. I am indebted to several people for assistance with this study: John Nyaphuli who
accompanied me on all my field excursions and was responsible for finding most of the material ; Joel Mohoi,
Petrus Chalatsi, and Christian Nyaphuli for preparation of the material; Roy Oosthuizen for the loan of fossil
material from his private collection; Mike Cluver, Nick Hotton III, Gillian King, James Kitching, Izak Rust,
and Juri van den Heever for numerous discussions. Gillian King and Johann Welman are thanked for critically
reading an earlier draft of this paper.

REFERENCES

boonstra, l. d. 1936. Some features of the cranial morphology of the tapinocephalid dinocephalians. Bulletin
of the American Museum of Natural History , 72, 75-98.

— 1953a. A suggested clarification of the taxonomic status of the South African titanosuchians. Annals of
the South African Museum , 42, 19-28.

558

PALAEONTOLOGY, VOLUME 34

19536. The cranial morphology and taxonomy of the tapinocephalid genus Struthiocephalus. Annals of
the South African Museum, 42, 32-53.

1956. The skull of Tapinocephalus and its near relatives. Annals of the South African Museum, 43, 1 37-169.

1957. The moschopid skulls in the South African Museum. Annals of the South African Museum, 44,
15-38.

1962. The dentition of the titanosuchian dinocephalians. Annals of the South African Museum, 48,
233-236.

1963a. Diversity within the South African Dinocephalia. South A frican Journal of Science, 59, 196-207.

19636. Early dichotomies in the therapsids. South A frican Journal of Science , 59, 176-195.

1965. The skull of Struthiocephalus kitchingi. Annals of the South A frican Museum, 48, 251-256.

1969. The fauna of the Tapinocephalus Zone (Beaufort beds of the Karoo). Annals of the South African
Museum, 56, 1-73.

1971. The early therapsids. Annals of the South African Museum, 59, 17 46.

- 1972. Discard the names Theriodontia and Anomodontia: a new classification of the Therapsida. Annals
of the South A frican Museum, 59, 315-338.

hopson, j. a. and barghusen, h. r. 1986. An analysis of therapsid relationships. 83-106. In hotton, n.,
Maclean, p. d., roth, j. j. and roth, e. c. (eds). The ecology and biology of mammal-like reptiles. Smithsonian
Institution Press, Washington, 325 pp.

kemp, T. s. 1982. Mammal-like reptiles and the origin of mammals. Academic Press, London, 363 pp.
keyser, a. w. and smith, r. m. h. 1978. Vertebrate biozonation of the Beaufort Group with special reference
to the western Karoo Basin. Annals of the Geological Survey of South A frica, 12, 1-36.
king, G. m. 1988. Anomodontia. 1 174. In Wellnhofer, P. ( ed . ) . Encyclopedia of Paleoherpetology , 17C. Gustav
Fischer, Stuttgart.

kitching, J. w. 1977. The distribution of the Karoo vertebrate fauna. Memoirs of the Bernard Price Institute
for Palaeontological Research, University of the Witwater strand, 1, 1-133.
lydekker, R. 1890. Catalogue of the fossil Reptilia and Amphibia in the British Museum. Part IV. Containing
the orders Anomodontia, Ecaudata , Caudata , and Labyrinthodontia ; and supplement. London, British
Museum (National History).

OLSON, E. c. 1962. Late Permian terrestrial vertebrates. Transactions of the American Philosophical Society, 52,
3-224.

orlov, y. a. 1958. The carnivorous dinocephalians of the Isheevo fauna (titanosuchians). Trudy Paleonto-
logicheskogo Instituta. Akademiya Nauk SSSR , 72, 3-113. [In Russian.]
ruben, j. a. 1986. Therapsids and their environment, a summary. 307-312. In hotton, n., Maclean, p. d., roth,
j. J. and roth, E. c. (eds). The ecology and biology of mammal-like reptiles. Smithsonian Institution Press,
Washington, 325 pp.

rubidge, b. s. 1987. South Africa’s oldest land-living reptiles from the Ecca-Beaufort transition in the southern
Karoo. South African Journal of Science, 83, 165-166.
rubidge, b. s. and hopson, j. a. 1991. A new anomodont therapsid from South Africa and its bearing on
the ancestry of Dicynodontia. South A frican Journal of Science, 86, 43-45.
tchudinov, p. K. 1983. Early therapsids. Trudy Paleontologicheskogo Instituta. Akademiya Nauk SSSR , 202,
1-227. [In Russian.]

BRUCE S. RUBIDGE
National Museum
P.O. Box 266
Bloemfontein 9300
South Africa

Current address:

Bernard Price Institute (Palaeontology)
University of Witwatersrand
P.O. Wits 2050
South Africa

Typescript received 25 November 1989
Revised typescript received 4 May 1990

RUBIDGE: DINOCEPHALI AN FROM SOUTHERN AFRICA

559

ABBREVIATIONS USED IN THE TEXT FIGURES

al.can.

alveolus for canine

p.can.

postcanine

al.p.can.

alveolus for postcanine tooth

p.fr.

prefrontal

art.

articular

p.for

pineal foramen

boc.

basioccipital

pm.

premaxilla

can.

canine

po.

postorbital

den.

dentary

pof.

postfrontal

e.a.m.

external auditory meatus

po.par.

postparietal

ect.

ectopterygoid

pt.f.

posttemporal foramen

ext.nar.

external nares

q-

quadrate

for. mag.

foramen magnum

q.j.

quadratojugal

fr.

frontal

qrpt

quadrate ramus of pterygoid

i.

incisor

ref.l.

reflected lamina

jug-

jugal

sm.

septomaxilla

lac.

lachrymal

sq.

squamosal

I. pr.pt

lateral process of pterygoid

st.

stapes

max.

maxilla

st. for.

stapedial foramen

nas.

nasal

t.

tabular

pal.

palatine

VO.

vomer

par.

parietal

FISHES AND AMPHIBIANS FROM THE LATE
PERMIAN PEDRA DE FOGO FORMATION OF
NORTHERN BRAZIL

by c. barry cox and p. hutchinson

Abstract. The vertebrate fauna of the Pedra de Fogo Formation of northern Brazil includes a palaeonisciform
fish Brazilichthys macrognathus gen. et sp. nov., which is placed in a new family, the Brazilichthyidae. Other
fishes include fragments of ctenacanth and xenacanth sharks, edestid holocephalians, and dipnoans. Tetrapods
include the archegosaurid amphibian Prionosuchus plummeri, which is compared with other archegosaurs. Its
extremely long narrow snout suggests that the Pedra de Fogo Formation is of Late Permian age, rather than
Early Permian. A large specimen of Prionosuchus is probably the longest amphibian currently known.

The Pedra de Fogo Formation lies in the Maranhao-Piaui (or Parnaiba) Basin of northern Brazil,
which covers parts of the state of Maranhao and adjoining states and which originated in the Siluro-
Devonian. Permo-Carboniferous sediments deposited in this basin have, over the past thirty years,
been given various names. The most common nomenclature today recognizes a Late Carboniferous
Piaui Formation, which is overlain by the Pedra de Fogo Formation. The latter is overlain by the
Motuca Formation, which is currently regarded as of Triassic age.

The Pedra de Fogo Formation was named after the small stream that runs through the area; the
name is derived from the large amount of flint (firestone, or ‘pedra de fogo’ in Portuguese) which
is found in its bed. The Formation was first recognized and defined by Plummer (1948), following
the geological traverse of the Basin made by Plummer, Price and Gomes in 1946. In brief visits
during 1945 and 1946, Price found fish and amphibian remains south of the township of Pastos
Bons. The amphibian, which Price (1948) named Prionosuchus plummeri , was the first temnospondyl
described from South America.

As Price concluded, the very elongate snout of Prionosuchus suggested that it belonged to the
early group of long-snouted temnospondyls called archegosaurs, known from Permian sediments in
Europe. The fish remains from the Pedra de Fogo Formation included the fin-spines of ctenacanth
sharks, the teeth of xenacanth sharks, palaeoniscoid scales, coprolites and, at a lower level in the
Formation, remains of the marattialian tree-fern Psaronius. Price (1948) concluded that the Pedra
de Fogo Formation was certainly of Permo-Carboniferous age, and that it was most probably of
Early Permian age.

The suggestion that Prionosuchus was of Early Permian age was of some interest, for it was the
only Early Permian amphibian known from South America, and indeed one of the very few
tetrapods known from the Early Permian of Gondwanaland (Cox 1974).

The Pedra de Fogo Formation therefore seemed to merit further investigation, and two visits to
the Maranhao Basin were made. A reconnaissance in 1970 by L. I. Price of the Departamento
Nacional de Producao Mineral, Rio de Janeiro, and J. Attridge of Birkbeck College, London,
found a few fragments of amphibian. A larger party, consisting of the senior author and Attridge,
together with Price and his assistant D. A. Campos, visited the area in 1972. The fish and amphibian
remains described in this paper were found in the area where Price found Prionosuchus , east of the
road between Pastos Bons and Nova Iorque (lat. 6° 35' S, 44° 2' W), and mainly up to 300 m east
of those parts of the road that lie between 5-6 and 6-7 km south of the junction of that road and
the main Highway BR-230 at Pastos Bons. A few fragments were also found east of a point 8-7 km

IPalaeontology, Vol. 34, Part 3, 1991, pp. 561-573.|

© The Palaeontological Association

562

PALAEONTOLOGY, VOLUME 34

south of that junction. Geological maps today show the Pedra de Fogo Formation both to the east
and to the west of Pastos Bons. However, both these areas were unproductive, despite aerial search
for exposures, followed by ground search.

GEOLOGICAL BACKGROUND

Plummer (1948) stated that the Pedra de Logo Lormation, 10-20 m thick, was made up of alternating deposits
of silt, limestone, and chert, the last predominating and forming beds and spherical concretions. The limestones
contain oolites and pisolites, the silt layers contain the vertebrate fossils, and the chert beds contain edgewise,
intraformational, fragmental conglomerates and, in some layers, large pieces of petrified wood. The chert also
has an oolitic and pisolitic texture, and both Lisboa (1914) and Plummer (1948) stated that the rocks were
originally limestone and had later been changed to chert by chemical replacement, perhaps following
deposition of a layer of volcanic ash over the region during the Early Cretaceous, when diabase was intruded
into the earlier deposits.

Plummer (1946, fig. 2) drew a section in the Pedra de Logo Lormation 6 6 km south of the township of Pastos
Bons, near the road to Nova Iorque. He showed it there to lie upon the upper part of the Lloriano Lormation
(now included as part of the Piaui Lormation), which consisted of grey marine shales interbedded with thin-
bedded sandstones containing chert. Plummer stated that the Lloriano Lormation was of Late Pennsylvanian
or Early Permian age. His figure showed that, in this region, the Pedra de Logo Lormation was unconformably

table I Section through the upper part of the Pedra de Logo Lormation.

Height above
base of
Formation,
in metres

Thickness
in metres

14-05

0-40

Basal conglomerate of Pastos Bons Fm.

13-65

1-45

Purple shale with subspherical
concretions

12-20

Thin (0-002-0-005 m) layer of calcite

0-30

Purple siltstone, as below

1 1-90

Thin (0-0 1 m) layer of calcite

2-15

Purple siltstone, as below, including a
0-05 m layer of haematite with bone
fragments

9-75

Thin (0 01 m) layer of calcite

4-05

Finely bedded purple siltstone, with
intercalations of greenish shale and of
coarse sandstone with silica matrix.
Occasional fragments of fossil wood.

5-70

1-60

Finely stratified greenish sandstone, as
below.

4-10

0-30

Moderately stratified, light-coloured,
poorly sorted sandstone.

3-80

2-90

Very fine greenish siliceous sandstone
with cross-lamination

0-90

0

0-90

Fine siliceous purple sandstone.
Bed of large flints at bridge.

COX AND HUTCHINSON: PERMIAN FISHES AND AMPHIBIANS

563

overlain by the Melancieiras Formation (now known as the Pastos Bons Formation, and considered as Middle-
Late Jurassic in age; Pinto and Purper 1976). However, Plummer (1946, fig. 3) also described a section of the
Pedra de Fogo Formation further east, between Sao Domingos and Benedito Leite, where he found that the
Pedra de Fogo Formation was unconformably overlain by the dark brick-red sandstones of the Motuca
Formation. He stated (1948, pp. 107-108) that the Motuca Formation was separated from the Pedra de Fogo
Formation by an erosional unconformity, and that it was overlain by the Melancieiras Formation (now known
as the Sambaiba Formation; Brito 1981). Plummer regarded the age of the Motuca Formation as uncertain,
and it has been ascribed variously to the Late Permian or to the Triassic (Brito 1981 ; Medeiros, pers. comm.,
1988).

A section was taken of the upper part of the Pedra de Fogo Formation, in the bed of the Pedra de Fogo
stream where it ran under the Pastos Bons-Nova Iorque road at a point 6-6 km south of the junction of that
road with Highway BR-230 (Table 1).

The vertebrate fauna of the Pedra de Fogo Formation comprises fin-spines of ctenacanth sharks and teeth
of xenacanth (pleuracanth) sharks (Santos 1946), teeth of edestid holocephalians, dipnoan teeth, scales and
cranial remains of palaeoniscoid fish, and the remains of temnospondyl amphibians (one of very large size).
This fauna was found below a layer of fossiliferous haematite that occurred about 1 F9 m above the base of
the section. Most of the bones were free of haematite.

The levels of haematite and calcite, and the apparently lacustrine nature of the sediments, suggest a fresh-
water environment of deposition. The ctenacanth and xenacanth sharks, which are elsewhere characteristic of
fresh-water environments, is confirmatory evidence of this. The bone fragments are mostly badly worn, and
the sediments had clearly been water-sorted and transported, fragments of similar size and density being found
together. The overall impression is therefore one of deposition in the lower reaches of a river system. The large
size of some of the vertebrates indicates that the body of water inhabited by these elements of the fauna must
have been extensive, suggesting a deltaic environment including deep rivers or lakes.

The fin-spines and fish teeth from the Pedra de Fogo Formation are currently being studied by Dr Silva
Santos of the DNPM. The description of the palaeoniscoid skull (below) is the work of the late Dr P.
Hutchinson; the remainder of this paper is the work of the senior author (C. B.C.). Institutional abbreviations
are: BMNH, British Museum (Natural History), London; DNPM, Departamento National de Producao
Mineral, Rio de Janeiro.

SYSTEMATIC PALAEONTOLOGY

Class ACTINOPTERYGII
Order palaeonisciformes
Family brazilichthyidae fam. nov.

Diagnosis. As for genus.

Genus brazilichthys gen. nov.

Type species. Brazilichthys macrognathus nov.

Brazilichthys macrognathus sp. nov.

Text-fig. 1

Horizon and locality. Pedra de Fogo Formation, state of Maranhao, N. Brazil. Found about 100 m east of the
Pastos Bons-Nova Iorque road, about 6 km south of Pastos Bons.

Holotype. DNPM 1061-P, the only known specimen.

Diagnosis. Upper jaw margin strongly convex below orbit; teeth arranged in two series, the inner
row composed of long, slightly recurved, conical teeth; median and lateral gulars almost half as long
as the lower jaw. The genus is incompletely known, and this diagnosis must be regarded as
provisional.

Description. The skull of B. macrognathus is represented by a single specimen, 80 mm long and 42 mm deep.

564

PALAEONTOLOGY, VOLUME 34

The dermal bones of the skull roof and opercular region are missing, but the other bones are well preserved,
especially on the left side (Text-fig. 1a). The restoration (Text-fig. 1b) is based mainly on the left side of the
specimen, but some additional details are derived from its right side.

The postorbital part of the maxilla is long and indicates that B. macrognathus had an inclined suspensorium.
The part of the maxilla exposed below the orbit is extremely narrow, its dorsal edge being overlapped by the
infra-orbital series and its ventral edge being separated from the upper jaw margin by a long premaxilla. The
teeth of the upper jaw are arranged in two series : an inner series of eight or nine striated, conical teeth that
are up to 6 mm long, and an outer series of more numerous teeth that are about 05 mm long. The larger teeth
are terminated by small enamel caps similar to those described in Pleronisculus ( Glaucolepis ) by Nielsen (1942),
and in Nematoptychius by Gardiner (1963).

The posterior and ventral edges of the orbit are bordered by two infra-orbitals. The anterior infra-orbital
is slightly expanded at its anterior end (part of this bone was removed during preparation in order to expose
details of the snout), while the posterior infra-orbital is broad and lunate. The dorsal edge of the orbit is
bordered by a slender, curved element that extends anteriorly to meet the nasal. It is identified here as an
infraorbito-supraorbital (dermosphenotic of other authors, e.g. Moy-Thomas and Dyne 1938; Gardiner 1963).

Inf Sbo

text-fig. 1 . Brazilichthys macrognathus , gen. et sp. nov. Lateral view of skull, natural size. A, holotype, DNPM
1061-P. b, restoration of skull. Abbreviations: Ang, angular; Br, branchiostegal ray; Clav, clavicle; Den,
dentary; Inf, infra-orbital (dermosphenotic); Inf-so, infraorbito-supraorbital; L Gu, lateral gular; M Gu,
median gular; Mx, maxilla; Na, nasal; Pmx, premaxilla; R, rostral; Scr, sclerotic ring; Sbo, suborbital.

COX AND HUTCHINSON: PERMIAN FISHES AND AMPHIBIANS

565

Objections to this view have already been put forward (Hutchinson 1975). It is also uncertain whether the
infraorbito-supraorbital meets the posterior infraorbital, because the posterior edge of the latter bone is
imperfectly preserved; however, comparison with other palaeoniscoids suggests that these bones did, in fact,
meet to complete the posterior edge of the orbit.

The snout is composed of a rostral and paired nasals, premaxillae, and antorbitals. The rostral is identified
as such in view of Patterson’s (1975) conclusion that the median snout element in chondrosteans, identified as
the post-rostral by many authors (e.g. Gardiner 1963, 1967 in palaeoniscoids; Hutchinson 1973 in redfieldiids
and perleidids), is homologous with the rostral of holosteans and pholidophorids. Close to its ventral end, the
nasal bears notches for the anterior and posterior nasal apertures. The premaxillae are unique among
chondrosteans because they extend posteriorly almost to the level of the posterior orbital edges, and because
they are separated at the tip of the snout. The antorbital is large and bears the ethmoid commissure of the
infraorbital sensory canal. The antorbital branch is either absent or so poorly preserved that its course is not
visible. The posterodorsal edge of the antorbital borders the orbital edge, and its posteroventral edge either
butts against, or overlaps, the anterior ends of the anterior infra-orbital, the maxilla and the premaxilla. Its
ventral edge bears the two types of teeth identified on the maxilla and premaxilla. The extreme anterior end
of the antorbital is missing, so that the structure of the tip of the snout is unknown, but it appears likely that
the antorbitals met at the midline of the snout.

There is an incompletely preserved suborbital wedged between the posterior infra-orbital and the dorsal part
of the maxilla.

The lower jaw is long, and curves dorsally towards the symphysis. This curve is reflected in the shape of the
edge of the upper jaw and does not appear to be due to post-mortem distortion. The teeth of the lower jaw
are similar to those of the upper jaw, but are somewhat shorter. The angular extends along half the length of
the lower jaw, but is almost completely overlapped on the lateral surface by the dentary, which is extremely
thin posteriorly. There are three gular plates, which together are almost half the length of the lower jaw. The
lateral gular plates bear short pit lines, but similar pit lines are not visible on the median gular. At least nine
branchiostegal rays are present between the gular plates and the level of the posterior end of the lower jaw.

The known dermal bones of the skull bear an ornament of ridges and tubercles, and their distribution is
indicated in Text-fig. 1.

[The form of the teeth and the elongated maxilla suggest that the family Brazilichthyidae may be related to
the Permian family Acrolepidae (Gardiner and Schaeffer 1989): note suggested to senior author by Dr B.
Gardiner, 1989.]

Class AMPHIBIA
Order temnospondyli
Family archegosauridae
Genus prionosuchus Price, 1948

Prionosuchus plwnmeri Price, 1948
Text-fig. 2

Material. The genus was established by Price on the basis of a rostrum, a fragment of mandible and a femur
(DNPM 320-R), all from the Pedra de Fogo Formation south of Pastos Bons. Price noted that the elongate
rostrum resembles that of the archegosaur Platyoposaurus, but that it is proportionately much longer and more
laterally compressed than that of Platyoposaurus. (The latter genus was originally named Platyops by
Twelvetrees (1880) but, as noted by Lydekker (1889, p. xi), that name was preoccupied for a fish. Lydekker
accordingly renamed the archegosaur Platyoposaurus.) Price also noted that the nasals, maxillae and palatines
of the Brazilian specimen extend further forward than those of Platyoposaurus , and that the external nares are
smaller and more laterally placed. He therefore established the new genus Prionosuchus. The rostrum of the
holotype specimen is 345 mm long, and Price estimated that the complete skull would have been about
500 mm long.

Price collected scraps of two further rostra in the same area in a short visit in 1948 (DNPM 862-R, 863-R),
and the 1970 reconnaissance collected further scraps, including a right mandibular ramus (DNPM 864-R) and
two intercentra (DNPM 865-R). The following amphibian remains were collected during the 1972 expedition:
BMNH R12000, the mid-portion of a rostrum; BMNH R12001, the posterior part of a rostrum; BMNH
R12002, a median fragment of the frontal region of a slightly smaller skull; BMNH R12003, many fragments
of a small skull; BMNH R12004, fragments of vertebrae, including pleurocentra and intercentra, and
fragments of skull, and BMNH R12005, fragments of skull and post-cranial skeleton. The last specimen is
nearly three times the size of most of the other amphibian specimens.

566

PALAEONTOLOGY, VOLUME 34

Description. Though the large specimen provides additional information on the anatomy of the posterior
region of the skull and also contains portions of the rostrum, it is so fragmentary that the anatomy of the
anterior part of the skull can be most reliably reconstructed using the fragments of the smaller skulls, and this
is then used as a basis for estimating the size of the larger specimen.

Price’s holotype specimen (DNPM 320-R) includes an almost complete rostrum, from the anterior tip back
to a point slightly behind the internal nares. Anterior to the suture between the premaxilla and the maxilla, the
rostrum is about 20 mm thick dorso-ventrally, and about 23 mm wide. Posterior to that suture, it widens to
31 mm at its posterior end but becomes only 17 mm thick dorsoventrally. The dorsal and the ventral surface
are both flattened. The rostrum bears the anterior continuation of the supra-orbital sensory canal. The palatal
surface of the vomers and palatines is covered with a fine shagreen of conical denticles. The expanded tip of
the rostrum bears large teeth; immediately posterior to this, there is a constricted region bearing small teeth.
Behind this constricted region, and as far back as the palatal portion of the premaxilla-maxilla suture, the teeth
are of moderate size. Posterior to this suture, the teeth revert to smaller size. A pair of vomerine fangs and a
pair of palatal fangs lie respectively anterior and posterior to the internal nares.

text-fig. 2. a. Dorsal and d, ventral views of the snout of Prionosuchus plummeri x reconstructed from the
type specimen (DNPM 320-R) and from the postero-lateral fragment (DNPM 862-R). b, c. Cross-sections of
the snout at the positions shown, from specimen BMNH R 12001. Abbreviations: F, frontal; L, lacrimal; MX,
maxilla; N, nasal; PAL, palatine; PMX, premaxilla; V, vomer; ext.nar, external nares; int.nar, internal nares;
inf.orb, infra-orbital sensory canal; sup. orb, supra-orbital sensory canal.

BMNF1 R 12001 is a 105 mm long section of the middle portion of a rostrum of similar size. It differs from
this region of the holotype only in that it is less flattened, being oval in cross-section, 24 mm thick and 34 mm
wide, the palatal surface being convex (Text-fig. 2b, c). This suggests that the flattened form of the holotype
was due to post-mortem crushing which collapsed the internal narial passage.

DNPM 862-R consists of two portions of rostrum, dorsoventrally crushed like the holotype. The anterior
portion corresponds to the most posterior portion of the holotype, including the region of the internal nares
and the posterior part of the external nares. The former appear to have been elongate and about 20 mm long;
the dorsal and ventral regions of the specimen have therefore become broken apart from one another in this
area. The palatal teeth are noticeably larger anterior to the internal nares than more posteriorly. The supra-
orbital sensory canals are very well-developed, being up to 4 mm deep, and diverge posteriorly.

The more posterior part of DNPM 862-R is separated from the anterior part by a gap which ventrally
appears to have been only a few millimetres wide. Ventrally, this part shows the posterior end of the internal

COX AND HUTCHINSON: PERMIAN FISHES AND AMPHIBIANS

567

nares and, 18 mm behind this, the socket of the fang on the anterior end of the palatine bone. Dorsally, this
part shows the nasal-frontal suture, the anterior end of the lacrimal bone, and two sensory canals. The more
medial of these canals appears to be the posterior part of the supra-orbital canal. The more lateral canal is
probably the infra-orbital sensory canal which, on BMNH R 12001, can be seen to continue anteriorly and die
out c. 30 mm anterior to the external nares.

BMNH R 12000 and R 12001 are uncrushed and show a series of cavities, filled with fine yellowish matrix,
which extend through the snout. The posterior end of the rostrum fragment BMNH R 12001 shows a section
just anterior to the external nares (Text-fig. 2b); it shows a matrix-filled cavity extending from the external
nares into the centre of the rostrum. The anterior end of the same specimen (Text-fig. 2c) shows a large central
matrix-filled cavity, which presumably extended to the anterior end of the snout. Posterior to the external
nares, the cavity formed the narial passage, which presumably ended at the internal nares. However, a fragment
of the median portion of a small skull including the anterior part of the frontals (BMNH R12002) shows two
smaller pairs of matrix-filled cavities extending posteriorly into the skull. In life, all these cavities were
presumably filled with air and acted as flotation device for the head, rather like the cavities of the crocodilian
skull.

Restoration and comparison. The fragments described can be used to produce a restoration of the
anterior part of the skull of Prionosuchus plummeri (Text-fig. 2), which was about 380 mm long.
Comparison of the sutural pattern of the snout of Prionosuchus with that of its closest known
relative, Platyoposaurus (Text-fig. 3c) shows that the further elongation of the snout of the former
has been achieved primarily in the region of the external nares. The nasals of Platyoposaurus end
abruptly, at the level of the anterior end of the external nares. Those of Prionosuchus extend further
forwards, gradually tapering to a narrow point, beyond which the snout is composed exclusively of
the premaxillae. On the other hand, the lacrimals of Platyoposaurus nearly reach the external nares,
while those of Prionosuchus hardly extend beyond the anterior end of the frontals, doubtless because
of the great narrowing of the skull.

An estimate of the total skull length of Prionosuchus can be made by comparing it with that of
Platyoposaurus. It would be possible simply to assume that the position of the naso-frontal suture
was unchanged, and to enlarge a drawing of a Platyoposaurus skull until its width at this point was
equal to that of the Prionosuchus skull. However, this may be an over-simple approach. The
elongate snout of Prionosuchus , with its enlarged, spatulate tip, is identical to that found in the fish-
eating gharial crocodiles of today. The narrowing of the snout reduces water-resistance when
closing the jaws, and therefore allows the closure to take place more rapidly - an obvious
advantage, especially when snapping at fish. Some of this advantage would be lost if the posterior
end of the snout widened only gradually to match the width of the more posterior part of the skull,
and this may be the reason why the snout of the gharial widens more abruptly. It seems likely that
this would have been a factor also in the evolution of the snout of Prionosuchus , and the most
posterior preserved part of its snout does indeed appear to curve outwards quite noticeably. In the
reconstructed skull (Text-fig. 3f), it has therefore been assumed that the frontals do not project as
far anteriorly as in Platyoposaurus , and the snout is shown as widening fairly abruptly. It is also
possible that the coinciding anterior terminations of the lacrimal and of the frontal mark the level
at which the skull narrowed abruptly into a snout dominated by the nasal, maxilla, and premaxilla.
A reduction in the number of sutural joints in this latter structure would also increase its strength.
It is possible that the degree of shortening of the immediately pre-orbital region that is shown in
the reconstruction is incorrect, but there is at present no way in which this can be verified. As
reconstructed in this manner, the total skull length of Prionosuchus based on these specimens would
have been about 580 mm.

Post-cranial skeleton. A number of isolated amphibian pleurocentra and intercentra have been found in the
Pedra de Fogo Formation, and show that Prionosuchus had a rhachitomous vertebral column similar to that
known in other archegosaurs, such as Archegosaurus (Hofker 1928) and Platyoposaurus (Konzhukova 1955).
The intercentra are up to 30 mm wide, and the pleurocentra are 20 mm high.

The large specimen of Prionosuchus. During the 1972 expedition, a much larger specimen of Prionosuchus

568

PALAEONTOLOGY, VOLUME 34

(BMNH R 12005) was collected in the Pastos Bons locality. Though extremely incomplete, it included
fragments of skull, vertebrae, ribs, scapulae, cleithra, clavicles, ilium, ischium, and femur.

The skull material includes fragments of the rostrum and palatal bones, both quadrates and parts of the
margins of the orbits and otic notch. It is particularly interesting that one fragment shows the supratemporal
bone entering the otic notch and so separating the tabular from the squamosal, as in Archegosaurus. The
fragments show the expanded anterior end of the rostrum found in both Prionosuchus and Platvoposaurus. and
the shagreen of palatal teeth characteristic of the former. The fragments also include a number of small,
subcircular platelets of bone, up to 14 mm in diameter and 2 mm thick, with smooth, slightly convex under
surfaces. Each platelet bears up to 30 denticles. They were probably embedded in the soft tissues of the mouth
(for example, in the tissues overlying the interpterygoid vacuities), so that these areas, like the palatal bones,
were denticulated and could grip the prey.

Despite the difference in size, there can be little doubt that this specimen belongs to Prionosuchus.
The only part of it that is sufficiently complete to provide a basis for an estimate of its size when
complete is the tip of the rostrum. This is 100 mm across, while that of the holotype specimen is only
35 mm across. As already mentioned, the snout of Prionosuchus is very like that of the modern fish-
eating gharial crocodiles. Though the relative length of the snout of these crocodiles increases
rapidly in early ontogeny, it remains almost constant from a skull length of 400 mm up to that of
the largest specimen measured, 800 mm (data from Kalin 1933, tables II and III). It therefore seems
reasonable to assume that proportions of Prionosuchus did not change greatly between the smaller
(but definitely adult) specimens and the large specimen, in which the skull of the latter would have
been about 16m long.

Other archegosaurs, such as Archegosaurus and Platvoposaurus (Konzhukova 1955) had an
elongate body and tail but small limbs. Though the skeleton of Platvoposaurus stuckenbergi that
Konzhukova describes is incomplete, the first 9-10 vertebrae have a length approximately equal to
| the skull length. They are followed by another nine vertebrae with snout centra and ribs, with no
sign of diminution in size towards the sacral region. There therefore seems little doubt that the body
alone of such an archegosaur was considerably longer than the skull.

Prionosuchus certainly had small limbs, for the incomplete femora of the large specimen show that
the complete bone could not have been more than 15 mm in diameter and 130 mm long. This
indicates that it did not swim by means of enlarged, paddle-like limbs, but instead swam like a
crocodile, by means of lateral undulations of the body and elongate tail. Wermuth (1964) has
analysed the relationships between head, body and tail length in modern crocodilians. The largest
gharial measured by him had a skull 0-83 m long, with a body 2 m long and a tail 2-65 m long -
a total of nearly 5-5 m. Even if the snouts and body-plus-tail lengths of larger specimens of gharial
or Prionosuchus were proportionately smaller than in specimens of intermediate size, there still
seems little doubt that the large specimen of Prionosuchus , with a skull twice the length of that of
Wermuth's largest gharial, must have had a total length greater than that of the gharial. There
therefore seems little doubt that it was considerably larger than the longest amphibian presently
recorded, Eogyrinus attheyi (Panchen 1972), which had an estimated total length of 4 m.

THE RELATIONSHIPS OF PRIONOSUCHUS AND THE AGE OF THE
PEDRA DE FOGO FORMATION

As Price (1948) suggested, Prionosuchus is an archegosaurid : the great, slender elongation of the
skull and the contact between the lacrimal and the frontal agree with Romer’s (1947) definition of
that family, in which he included three genera : Archegosaurus , Melosaurus and Platvoposaurus. Two
other genera, Bashkirosaurus and Collidosuchus, were added by Gubin (1981, 1986).

The earliest member of the Archegosauridae is Archegosaurus itself (Text-fig. 3b). The type
species, A. decheni (Hofker 1928; Whittard 1928), is from the ironstone nodules of the Lebach
Group of the Saar region of Germany, of early Lower Permian age. Other specimens of
Archegosaurus , named A. ornatus (Woodward 1905) and A. kashmiriensis (Tewari 1962) have been
found in the Lower Gondwana deposits of Kashmir, which are probably also of Early Permian age.

COX AND HUTCHINSON: PERMIAN FISHES AND AMPHIBIANS

569

text-fig. 3. The skulls of archegosaurs, drawn to similar widths across posterior portion of the skull, a,
Melosawus, from Romer 1947, b, Archegosaurus , from Romer 1947. c, Collidosuchus , from Gubin 1986. d,
Platyoposaurus stuckenbergi , after Efremov 1933 (his dorsal view, fig. 1, superimposed on his ventral outline,
fig. 3). e, Platyoposaurus watsoni, after Efremov 1933. f, Prionosuchus (see text for details of reconstruction).

The other archegosaur genera, Melosawus , Bashkir osaurus, Collidosuchus , and Platyoposaurus,
are found in early Late Permian (Kazanian) deposits west of the southern Urals, in European
Russia. The archegosaurs are found in Zones I and II of the deposits (which are numbered from
the base upwards).

Melosawus (Meyer 1860) was originally placed in the Archegosauridae because of its somewhat
elongate preorbital region (Text-fig. 3a). However, this region is still comparatively short, being less
than 50% of the total skull length, and is much wider than that of Archegosaurus (Text-fig. 3b). It
is clearly not merely a juvenile archegosaur, as Hartman- Weinberg (1939) suggested, since skulls up
to 400 mm long are now known.

Collidosuchus (Gubin 1986) has a snout whose elongation is intermediate between that of
Archegosaurus and that of Platyoposaurus (Text-figs 3b D). This is due in part, at least, to the
appearance of a growth zone just anterior to the external nares, which consequently lie further back
than in Archegosaurus, though not as far back as in Platyoposaurus.

Gubin had earlier (1981) described another Late Kazanian archegosaurid, Bashkir osaurus, but
this specimen comprises only the median part of the skull from the orbits posteriorly. Rather
surprisingly, Gubin does not mention this specimen in his later (1986) paper on Collidosuchus.

The genus which Lydekker (1889) renamed Platyoposaurus (see above, p. 565) was erected by
Twelvetrees (1880) on the basis of a badly-damaged skull which he named P. rickardi. The snout
of this specimen is lacking, and it seems unlikely that it shows any characters which would allow
other specimens to be placed in this species. The name Platyoposaurus rickardi should therefore be
applied only to the type specimen.

In 1884, Trautschold described a second species, P. stuckenbergi, based on a specimen which had
an elongate, narrow snout and lower jaw. This specimen was redescribed by Efremov (1933), who
also described a third species, P. watsoni, of which ten skulls were available, all from the same
locality. Efremov stated that P. watsoni had a narrower snout than P. stuckenbergi , and this is also
shown in his figures of the skulls (Text-fig. 3d). (His figure of P. watsoni in dorsal view was later
used by Romer (1947, fig. 28) as his illustration of the genus Platyoposaurus.) However, Efremov
also figured the lower jaw of this specimen of P. stuckenbergi, and this has a degree of narrowing
and elongation very similar to that of P. watsoni. Efremov also stated that he could not locate the
position of the external nares of Platyoposaurus, even approximately - presumably because, like
those of Prionosuchus, these were narrow slits located on the sides of the snout.

570

PALAEONTOLOGY, VOLUME 34

Bystrow (1935) also gave figures of the skull of P. watsoni , but showed this as having a much
wider snout than that shown in either of Efremov’s figures, and with dorsally-directed external nares
quite unlike those of the other specimens. However, this was reconstructed from fragments in the
Paleozoological Institute in Stockholm, and is unlikely to be more reliable than Efremov’s series of
more complete specimens. Konzhukova (1955) has given the most recent account of P. stuckenbergi ,
basing this on a new specimen which includes much of the post-cranial skeleton but which
unfortunately lacks the snout. A complete skull of Platyoposaurus , identified as P. watsoni, in the
Museum of Comparative Zoology, Harvard (MCZ 1750) shows a snout shape identical to that of
Efremov’s reconstruction of that species, with no trace of the external nares in dorsal view.

There is clearly some uncertainty over the proportions of the snout in these two species of
Platyoposaurus and, in view of the variable crushing of the cavities in the snout of Prionosuchus, it
seems very likely that only one species is really present in the Russian deposits. It also seems unlikely
(though not impossible) that two similar piscivorous species could co-exist with one another in the
same environment. P. watsoni (Efremov) is therefore considered to be a junior synonym of P.
stuckenbergi.

There can nevertheless be no doubt that the snout of Prionosuchus is even more elongate than that
of the Russian forms (Text-figs 3e, f), and that it represents a new taxon of archegosaur. It is,
however, debatable whether the Brazilian form should be placed in a separate genus from
Platyoposaurus , or should instead be regarded as merely another species of that genus. Until a
complete skull of Prionosuchus is found, it is impossible to know whether its degree of
morphological difference from Platyoposaurus suggests recognition at specific or at generic level.
For the time being, therefore, there is insufficient evidence to justify an alteration in the present
taxonomic position.

Prionosuchus appears to represent a continuation of the archegosaur trend towards an elongate,
narrow snout, and this implies that it is more advanced than Platyoposaurus in this character. Price,
in his original description of Prionosuchus (1948), stated that he believed it was a primitive
platyoposaurid. He gave no reasons for this statement, but his discussion of the diagnostic features
of Prionosuchus mentions five characters in which it differs from Platyoposaurus : the greater
anterior prolongation of the nasals and premaxillae; the narrow, widely-separated internal nares;
the narrow, laterally-placed external nares; the much greater number of denticles on the palate; and
the very pronounced lateral-line canals. Barberena (1972) concurs with Price in regarding
Prionosuchus as a primitive platyoposaurid, and quotes these same characters as evidence. However,
the first two characters seem to be results of the great elongation of the snout, which is surely an
advanced character. The multiplication of the palatal teeth also seems to be an advanced character,
while the depth of the lateral-line canals is variable within Prionosuchus. There therefore appears to
be no reason for regarding Prionosuchus as a primitive platyoposaurid, but several characters
instead suggest that it is more advanced than Platyoposaurus.

Barberena and Daemon (1974) have described a long-snouted temnospondyl from the Permian
Rio do Rasto Formation of the Parana Basin, southern Brazil. Though they at first ascribed this
form to Platyoposaurus , later discovery of the more posterior parts of the skull has shown that it
is probably related to the trematosaurs rather than to the archegosaurs (Barberena pers. comm.
1976), though such a late date would be surprising for a member of that group, and the specimen
does not show the conspicuous parallel vomerine tooth-rows characteristic of the trematosaurs.

In his original description of Prionosuchus , Price (1948) suggested that the Pedra de Fogo
Formation was of Early Permian age. This was mainly because he felt that the general nature of the
sediments and fauna was very similar to that of the Early Permian Red Beds of Texas, in which he
had collected earlier. In addition. Price pointed out that the Pedra de Fogo fauna contained
ctenacanths, which were then unknown from deposits later than Early Permian age. However,
Bendix-Almgreen and Mahlzahn (1969) have since described a ctenacanth from the Kupferschiefer
of Germany, which is of Late Permian age, even later than that of the Platyoposaurus specimens
from Zone II of Russia. The other elements in the Pedra de Fogo Formation are of no use for dating
purposes: xenacanths range from the Devonian to the Late Triassic (Patterson 1967), the dipnoan

COX AND HUTCHINSON: PERMIAN FISHES AND AMPHIBIANS

571

fragments are too incomplete to allow of any identification even to family level, and Brazilichthys
is isolated in a new family of actinopterygian fish. Though the fossil wood known as Psaronius is
known only up to the Middle Permian of Europe, the Late Permian floras of Europe do not include
petrification floras in which such wood would have been preserved. However, leaves of the
Pecopteris type, considered to be those borne by Psaronius , have a much longer range in time (up
to the Triassic), but the genus Pecopteris is so poorly defined that this range cannot be considered
as reliable (W. G. Chaloner pers. comm.).

There is therefore little firm evidence on which to base an estimate of the age of the Pedra de Fogo
Formation, other than that long-snouted archegosaurs are only known elsewhere from the Late
Permian, which suggests a Late Permian age for the Formation. This in turn suggests that the
Motuca Formation, which unconformably overlies the Pedra de Fogo Formation, is of Triassic age
rather than Late Permian.

The re-assignment of the Pedra de Fogo Formation to the Late Permian is of additional
palaeobiogeographical interest, in view of the fact that Prionosuchus had previously appeared to be
one of the few tetrapods known outside Euramerica before the Late Permian. The realization of that
fact provoked the suggestion (Cox 1974) that this was because Euramerica was still isolated from
other continents until that time. However, it now appears that Euramerica had joined with
Gondwana during the Carboniferous. Parrish et al. (1986) point out that all the occurrences of
tetrapods prior to the Late Permian are found in equatorial to subequatorial palaeolatitudes, and
that cool conditions for the Late Carboniferous to Early Permian are indicated by the distribution
of coals, tillites and evaporites. They therefore suggest that climate was the principal cause of the
absence of tetrapods from the higher southern palaeolatitudes of Gondwana until the Late Permian
climatic improvement.

Acknowledgements. The senior author expresses his gratitude to the Royal Society for providing funds for the
1970 and 1972 expeditions to northern Brazil, to the Departamento National de Producao Mineral, Brazil, for
all their assistance, to L. I. Price, D. A. Campos and J. Attridge for their help during the expeditions, and to
Dr A. R. Milner (Birkbeck College, London) for his help in interpreting the remains of Prionosuchus and for
comments on temnospondyl evolution.

It is sad to have to record that Dr Peter Hutchinson died in 1977, at the early age of 33, before this paper
was completed. Its final completion took place during a year’s sabbatical leave, and I am very grateful to the
Principal of King’s College London, for permission to take that leave, and to the Fulbright Commission for
its generosity in providing a travel grant. I am also grateful to Dr Phil Hanawalt of the Department of
Biological Sciences, Stanford University, California, for providing office and library facilities, and for his
personal help and kindness.

REFERENCES

aldinger, H. 1937. Permische Ganoidfische aus Ostgronland. Meddelelsen om Gronland , 102, 31-392.
barberena, m. c. 1972. South American Late Paleozoic tetrapods. Anais Academia brasiliensis de Ciencias ,
Supplemento 44, 67-75.

— and daemon, r. f. 1974. A primeira ocorrencia de Amphibia (Labyrinthodontia) na Formacao Rio do
Rasto. Implicates geocronologicas e estratigraficas. Anais do XXVIII Congresso Brasileiro de Geologic i. 2,
251-261.

bendix-almgreen, s. e. and malzahn, e. 1969. Uber neue oder wenig bekannte Elasmobranchier aus dem
deutschen Kupferschiefer. Notizblatt des Hessisches Landesamtes fur Bodenforschung zu Wiesbaden , 97,
44-48.

boy, j. a. 1972. Palaokologischen vergleich zweien beriihmter Fossillagerstatten des deutschen Rotliegenden
(Unterperm, Saar-Nahe-Gebiet). Notizblatt des Hessisches Landesamtes fur Bodenforschung zu Wiesbaden ,
100. 46-59.

brito, j. m. 1981. Estratigrafia da Bacia do Parnaiba. Anais do Academia Brasileiro de Geologic. 53, 529-549.
bystrow, a. p. 1935. Morphologische Untersuchungen der Deckknochen des Schadels der Wirbeltiere. I.
Mitteilung. Schadel der Stegocephalen. Acta Zoologica , Stockholm , 16, 65-141.

572

PALAEONTOLOGY, VOLUME 34

cox, c. B. 1974. Vertebrate palaeodistributional patterns and continental drift. Journal of Biogeography. 1,
75-94.

efremov, j. a. 1933. Uber die Labyrinthodonten der U.d.S.S.R. II. Permische Labyrinthodonten des friiheren
Gouvernements Wjatka. 1. Neubeschreibung des Labyrinthodonten Platyops stuckenbergi, Trautschold aus
den oberpermischen Ablagerungen des Flusses Kitjak, eines Nebenflusses der Wjatka. Trudy Palaeontologii
Akademii Nauk SSSR , 2, 117-164. [In German with Russian summary.]

Gardiner, B. G. 1963. Certain palaeoniscoid fishes and the evolution of the snout in actinopterygians. Bulletin
of the British Museum ( Natural History ), (Geology), 8, 255-325.

1967. Further notes on palaeoniscoid fishes with a classification of the Chondrostei. Bulletin of the British
Museum ( Natural History ), (Geology), 14, 143-206.

- and Schaeffer, B. 1989. The interrelationships of lower actinopterygian fishes. Zoological Journal of the
Linnean Society , 97, 135-187.

gubin, y. m. 1981. A new platyoposaurid from Bashkiria. Paleontological Journal , 1981, 108-110.

- 1986. New data on the archegosaurids of the East European platform. Palaeontological Journal , 1986,
70-75.

hariman-weinberg, a. p. 1939. Melosaurus uralensis H.v.M. an Upper Permian archegosaurid. Problemy
Paleontologii, 5, 20-31 [In Russian].

hofker, j. 1928. Archegosaurus decheni Goldf. Palaeontologia hungarica, 2, 2.

hutchinson, p. 1973. A revision of the redfieldiiform and perleidiform fishes from the Triassic of Bekker’s
Kraal (South Africa) and Brookvale (New South Wales). Bulletin of the British Museum (Natural History),
(Geology), 22, 233-354.

1975. Two Triassic fish from South Africa and Australia, with comments on the evolution of the
Chondrostei. Palaeontology, 18, 613-629.

kalin, j. a. 1933. Beitrage zur vergleichenden Osteologie des Crocodilidenschadels. Zoologisches Jahrbuch, 57,
535-714.

konzhukova, E. D. 1955. Platyops stuckenbergi, Trautsch. — an archegosauroid labyrinthodont from the
lower zone of the Cis-Uralian Upper Permian. Trudy Palaeontological Institute, 49, 89-127. [In Russian.]
lisboa, M. A. 1914. The Permian geology of northern Brazil. American Journal of Science, 37, 425-443.
lydekker, r. 1889. Manual of Palaeontology, 3rd. ed. Vol. 2. Blackwood & Sons, Edinburgh and London.
meyer, h. 1860. Melosaurus uralensis aus dem permischen System des westlichen Urals. Palaeontographica , 7,
90-98.

moy-thomas, j. a. and dyne, m. b. 1938. Actinopterygian fishes from the Lower Carboniferous of
Glencartholm, Eskdale, Dumfriesshire. Transactions of the Royal Society of Edinburgh, 59, 437—480.
nielsen, e. 1942. Studies on Triassic fishes from East Greenland, 1. Glaucolepis and Boreosomus. Meddelelsen
om Gronland, 138, 1^403.

panchen, a. l. 1972. The skull and skeleton of Eogyrinus attheyi Watson (Amphibia: Labyrinthodontia).

Philosophical Transactions of the Royal Society of London , Series B, 263, 279-326.
parrish, j. m., parrish, j. T. and ziegler, a. M. 1986. Permian-Triassic paleogeography and paleoclimato-
graphy and implications for therapsid distribution. 109-131. In hotton, n. et al. (eds). The Ecology and
Biology of Mammal-like Reptiles. Smithsonian Institution Press, Washington and London.
patterson, c. 1967. Family Xenacanthidae. 667. In harland, w. b. et al. (eds). The Fossil Record. Geological
Society, London.

- 1975. The braincase of pholidophorid and leptolepid fishes, with a review of the actinopterygian
braincase. Philosophical Transactions of the Royal Society of London, Series B, 269, 275-579.

pinto, i.d. and purper, i. 1976. Observations on Mesozoic Conchostraca from the north of Brazil. Anais do
XXVIII Congresso Brasileiro de Geologia, 2, 305-316.

plummer, f. b. 1946. Report to Conselho Nacional do Petroleo, original M.S. in Departmento Nacional da
Producao Mineral, Rio de Janeiro.

plummer, F. b. 1948. Estados de Piaui e Maranhao. Relatoria Anual do Conselho Nacional de Petroleo, Rio de
Janeiro, 1946. 87-134.

price, l. j. 1948. Um anfibio labirintodonte da Formagao Pedra de Fogo, Estado do Maranhao. Boletim de
Divisao de Geologia e Mineralogia, Departamento Nacional de Producao Mineral, Rio de Janeiro, 124, 1 32.
romer, a. s. 1947. Review of the Labyrinthodontia. Bulletin of the Museum of Comparative Zoology, Harvard
University, 99. 1-368.

santos, R. s. 1946. Duas novas formas de elasmobranchios do Paleozoico do Meio Norte, Brasil. Anais do
Academia Brasileiro de Ciencias, 18, 281-285.

COX AND HUTCHINSON: PERMIAN FISHES AND AMPHIBIANS

573

tewari, a. p. 1962. On a new species of Archegosaurus , from the Lower Gondwana of Kashmir. Records of the
Geological Survey of India, 89, 427 — 4 3 4 .

trautschold, h. 1884. Die Reste Permischer Reptilien des Palaontologischen Kabmets der Universitats Kasan.

Nouveaux Memoires de la Societe Imperiale de Naturalistes de Moscou , 15, 3-39.
twelvetrees, w. h. 1880. On a labyrinthodont skull ( Platyops rickardi , Twelvetr.) from the Upper Permian
cupriferous strata of Kargalinsk near Orenburg. Bulletin de la Societe Imperiale de Naturalistes de Moscou ,
55, 117-122.

wermuth, h. 1964. Das Verhaltnis zwischen Kopf-, Rumpf- und Schwanzlange bei den rezenten Krokodilen.
Senckenbergisches Biologiae , 45, 369-385.

whittard, w. F. 1928. On the structure of the palate and mandible of Archegosaurus decheni , Goldfuss. Annals
and Magazine of Natural History , 10 (1), 255-264.

woodward, a. s. 1905. Permo-Carboniferous plants and vertebrates from Kashmir. II. Fishes and
labyrinthodonts. Memoirs of the Geological Survey of India. Palaeontologia Indica, 11, 10-1 1.

C. BARRY COX and P. HUTCHINSON
Division of Biosphere Sciences

Typescript received 4 October 1989 King’s College

Revised typescript received 2 August 1990 Campden Hill Road, London, W8 7AH

SILURIAN CRYPTOSPORES AND MIOSPORES
FROM THE TYPE LLANDOVERY AREA,
SOUTH-WEST WALES

by N. D. BURGESS

Abstract. The oldest cryptospores and miospores have great significance in studies of the evolution of land
plants: the former may represent the earliest direct evidence of such organisms and the latter may provide
evidence for rhyniophytoid land plants as they have been recovered from the sporangia of Cooksonia pertoni
Lang in the late Silurian. In the type Llandovery area, two distinct sporomorph assemblages are described from
a composite section through uppermost Ordovician, Rhuddanian, Aeronian and basal Telychian strata. The
older (latest Ordovician to late Aeronian) comprises eight genera and fourteen species of cryptospores (tetrads,
pseudodyads, true dyads and monads). The younger (late Aeronian to Telychian) is dominated by smooth-
walled trilete miospores of species of the genus Ambitisporites. Both assemblages have strong similarities
to those described from similar horizons around the world. Specimens of Ambitisporites dilutus from the
sedgwickii Graptolite Biozone of the Aeronian/Telychian type boundary section are the oldest known with
unequivocal dating. They are used to define the base of the Ambitisporites dilutus - A. avitus Sporomorph
Zone. Three new cryptospore genera ( Rimosotetras , Segestrespora and Velatitetras) are erected and
Tetrahedraletes is emended and synonomized with Nodospora. Six new cryptospore species ( Velatitetras
laevigata , V. reticulata , V. rugulata , Rimosotetras problematical Segestrespora laevigata and S. rugosa ), and two
new varieties (T. medenensis vars medenensis and parvus) are described, and two combinations are made.

Currently sporomorphs provide the most convincing evidence for the existence of land plants
in the late Ordovician and early Silurian (Gray 1985), the best being those with the greatest
similarity to ones from living plants, or fossils of proven land-plants (Gray et al. 1982; Edwards
et al. 1983; Gray 1985; Richardson 1985; Fanning et al. 1988). Spores in almost all such plants are
produced in tetrads which separate on maturity to yield four trilete miospores. There are few
records of trilete miospores in the Llandovery (Hoflfmeister 1959; Cramer 1968, 1969; Pratt et al.
1978; Aldridge et al. 1979; Hill et al. 1985; Gray 1988; Richardson 1988). However, recent work
has demonstrated the presence of permanently fused tetrads, dyads and monads, all of which can
be enclosed within an envelope, in the late Ordovician and early Llandovery (e.g. Gray and Boucot
1971 ; Strother and Traverse 1979; Gray et al. 1982; Miller and Eames 1982; Gray 1985; Johnson
1985; Vavrdova 1982, 1984, 1988). As these unusually constructed cryptospores ( sensu Richardson
et al. 1984) may represent the oldest evidence of land-plants they have become important in the
debate on the appearance and subsequent evolution of such organisms (Gray and Boucot 1977;
Taylor 1982; Richardson 1985; Gray 1985, 1988).

Despite their importance, few assemblages of Ordovician and Llandovery sporomorphs have
been described from well-dated sections, or ones covering a long stratigraphic interval (Pratt et al.
1978; Strother and Traverse 1979; Miller and Eames 1982; Johnson 1985; Richardson 1988). This
has hindered interpretation of the biostratigraphic and evolutionary significance of the
sporomorphs. This paper presents full descriptions and stratigraphic ranges for those cryptospores
and trilete miospores recovered from a composite section through the uppermost Ordovician and
the majority of the Llandovery Series in the type Llandovery area (Cocks et al. 1984).

IPalaeontology, Vol. 34, Part 3, 1991, pp. 575-599, 2 pls.|

© The Palaeontological Association

576

PALAEONTOLOGY, VOLUME 34

text-fig. I . Location of the type Llandovery area in
southern Wales.

text-fig. 2. Locations of samples collected from the
late Ordovician Scrach Formation to the early
Llandovery Bronydd and Crychan formations in the
type Llandovery area: Forestry track F.33, Crychan
Forest near Llandovery (Ordnance survey map refer-
ence SO 84583964 at sample LI = locality 24 of
Cocks et al. 1984).

text-fig. 3. Locations of samples collected from the
Rhuddanian/Aeronian type boundary section in the
type Llandovery area: Forestry track F.33, Crychan
Forest, near Llandovery (Ordnance survey map
reference SO 83913960 at sample LI I = locality 38 of
Cocks et al. 1984).

BURGESS: LLANDOVERY SPOROMORPHS

577

text-fig. 4. Locations of samples collected from the
Aeronian/Teychian type boundary section in the type
Llandovery area: roadside near Cefn Cerig farm, near
Llandovery (Ordnance survey map reference SO
77543271 at sample LI 7 = locality 1 54 of Cocks et al.
1984).

GEOLOGY, SAMPLING AND TECHNIQUE

The location of the type Llandovery area is presented in Text-figure 1 . A single sample was collected from the
uppermost Ordovician and twenty-two from the Llandovery Series. Sequences collected spanned the
Ashgill/Rhuddanian boundary exposed along Forestry Track F. 33 next to the Afon Crychan river in Crychan
Forest, c. 7 km north-east of Llandovery (Text-fig. 2), the Rhuddanian/Aeronian type boundary section
(Text-fig. 3) and the Aeronian/Telychian type boundary section (Text-fig. 4) (Cocks et al. 1984). The top
c. 100 m of the sequence in the type area were not sampled: this interval covers the crispus to crenulatus
Graptolite Biozones. The stratigraphical framework of the samples is given in Text-figure 5.

The Llandovery sediments of the type area were deposited during a transgression which affected the whole
of the Anglo-Welsh Basin (Cocks et al. 1984; Ziegler et al. 1968). As a consequence the samples collected
during this study are from increasingly distal marine facies further up the section (Table 1 ). However, there may
be slight shallowing event in the Wormwood Formation.

table 1 . Distribution of samples by geological formation with the ages and probable depositional
environment.

Samples

Formation

Age

Probable depositional environment
(Cocks et al. 1984)

L23

Cerig

early Telychian

Open marine shelf

L19-L22

Wormwood

late Aeronian

Open marine shelf

L17-L18

Rhydings

late Aeronian

Open marine shelf

LI 1 — L 1 6

Trefawr

early Aeronian

Open marine shelf

L7-L10

Crychan

late Rhuddanian

Open marine prograding delta lobe

L2-L6

Bronydd

early Rhuddanian

Mud-dominated marine shelf with storms

LI

Scrach

latest Ashgill

Shallow sub-tidal or even intertidal

578

PALAEONTOLOGY, VOLUME 34

m

uj

UJ

0

oc

<

m

H

<n

(/>

>

cc

yy

>

o

Q

>

j

UJ

h

2

<

Z

O

CC

LU

<

<

2

<

Q

Q

3

I

QC

U. ORD,

GRAPTOLITE

BSOZONE

turriculatus

sedg wickii

convolutus

magnus

triangulatus

cyphus

acmacas

acuminatus

persculptus

FORMATIONS
acrach

Telychian/Aeronian

Boundary

l2"23j x x * r®

L21 —

L 20 -
L19 -
LI 8 —

LI 7 -

+ 4 * f-®
+ + +

— ©
-©

Rhuddanian/Aeronian

Forestry

Track F.33
L10
L9-

L8 -

LI 3

L7-

L6 — .

L4L5 -

L2L33

LI

100m

©

localities of
Cocks et al 19B4

Bronydd

□

Crychan

T ref a wr

j~* *■ | Rhydings | [ Wormwood

Cerig

text-fig. 5. Biostratigraphic and lithostratigraphic position of samples from the type Llandovery area, in

relation to localities of Cocks et al. (1984).

BURGESS: LLANDOVERY SPOROMORPHS

579

Samples varied from dark mudstones in the Scrach and Bronydd formations through to muddy siltstones
and silty mudstones for the remainder of the sequence. Palynomorphs were extracted using standard
palynological methods (HC1 then HF acids followed by sieving with 10 //m filter mesh and separation of the
organic fraction using zinc bromide solution (S.G. 2 0)). Residues were oxidized in cold Schultze’s solution for
c. 48 hours. The sporomorphs remain dark, but poor preservation did not allow any further extension of the
oxidation time. When this was attempted all the sporomorphs disintegrated. After oxidation, the acidic
residues were neutralized with distilled water.

For light microscope observation, measured volumes of material were strewed on to glass coverslips, dried,
and the coverslips attached to glass slides with ‘Elvacite’ plastic mounting medium.

Photomicrographs were taken on FP4 film using a Zeiss photomicroscope III (no. 2562) housed in the
Palynology Laboratory of the British Museum (Natural History) (BM(NH)), and using Nomarski differential
interference contrast with an orange filter to reduce contrast between the palynomorphs and the background.

SYSTEMATIC PALAEONTOLOGY

Preamble. Sporomorphs are described, where possible, using the standard terminology of Grebe (1971).
Specimen dimensions are presented as the minimum and maximum of the range, with the mean in brackets.
Stratigraphic range refers to range within the type Llandovery area only. The stratigraphical occurrence of taxa
is shown in Text-figure 7 and the stratigraphical distribution and correlation of the samples in Text-figure 5.

Specimens are located by means of standard England Finder and microscope stage co-ordmates taken from
the Zeiss photomicroscope III (no. 2562) housed in the Palynology Laboratory at the BM(NH). All figured
specimens are also ringed with a red indelible pen. Slides and stubs containing figured specimens (prefixed EM)
are stored in the Palynology collection at the BM(NH).

Anteturma cryptosporites Richardson et a /., 1984

1 . Cryptospore tetrads

This group comprises tetrads and dyads which are not found separated into individual components spores and
which are believed to be ‘permanently’ fused. They may also be enclosed within a sculptured or non-sculptured
envelope.

Genus tetrahedraletes Strother and Traverse, 1979 emend.

Type species. Tetrahedraletes medinensis Strother and Traverse, 1979 from the Tuscarora Formation,
Pennsylvania, USA.

Emended diagnosis. Permanent tetrahedral cryptospore tetrads, permanently fused. Tetrads sub-
circular to circular in outline and composed of four laevigate, crassitate, sub-triangular ‘spores’.
Crassitudes of individual ‘spores' + equatorial, fused or discrete. Distal ‘spore’ walls tend to
invaginate, but can remain inflated.

Discussion. Tetrahedraletes Strother and Traverse was erected to encompass ‘tetrads of inaperturate,
sub-triangular spores or spore-like palynomorphs arranged in tightly adhering tetrahedron
configuration, with the spore walls collapsed towards the centre’ (Strother and Traverse 1979, p. 8).
Nodospora, a similar genus, was defined for ‘tetrads of inaperturate spores or spore-like
palynomorphs, spherical to sub-spherical in outline, which are arranged in a cross configuration'
Strother and Traverse (1979, p. 10). Data collected here and in other recent publications (e.g. Gray
et cd. 1985; 1986; Gray 1985; 1988) indicate that the type specimens of Tetrahedraletes (T.
medinensis) and Nodospora ( N . burnhamensis) should be placed in synonomy, the two taxa
representing different compressional morphologies of otherwise identical tetrads. Hence, in this
publication all such tetrads are placed in Tetrahedraletes and the diagnosis of this genus emended
accordingly.

580

PALAEONTOLOGY, VOLUME 34

Tetrahedraletes medinensis Strother and Traverse, 1979 emend.

This species has been subdivided into two varieties based on size because the original specimens (Strother and
Traverse 1979) are much larger than those recorded from the late Ordovician and early Llandovery (e.g. Gray
1988).

Tetrahedraletes medinensis Strother and Traverse, 1979 var. medinensis var. nov.

Holotype and type locality. Strother and Traverse, 1979, pi. 1, fig. 5: Harvard Paleobotanical Collections no.
60289, slide no. 75-4/A3, location on slide 34-8 mm x 107-9 mm, reference point 13-3 mm x 109-2 mm. Samples
75-4, Tuscarora Formation; Section - Mann Narrows along route 322, north-west of Burnham, Mifflin
County, Pennsylvania, USA.

Description. As for Strother and Traverse (1979).

Comparison. Cryptospores of Tetrahedraletes medinensis var. parvus var. nov. are < 35 pm in diameter, but
otherwise identical.

Remarks. This variety of T. medenensis is common in the basal Wenlock (Burgess and Richardson
1991) but is not known from the uppermost Ordovician or early Llandovery of the type area where
the following variety is found (also see Gray 1988).

Tetrahedraletes medinensis var. parvus var. nov.

Plate 1, figs 1-4

1985 Tetrahedraletes cf. T. medinensis Gray, fig. 5 F-H.

EXPLANATION OF PLATE 1

All figures x 1000 unless stated otherwise.

Figs 1-4. Tetrahedraletes medinensis Strother and Traverse, 1979 emend, var. parvus var. nov. 1, FM 187,
holotype (slide Llan 8B/12, 178 1170; E.F. no : S47/1 ), sample L5, Bronydd Formation, acinaces Graptolite
Biozone, x 2000. 2, FM 188 (slide Llan 8B/8, 061 1283; E.F. no: F58/2), sample L5, Bronydd Formation,
acinaces Graptolite Biozone. 3, FM 92 (slide Llan 5/2, 130 960; E.F. no: N26/1), sample L16, Trefawr
Formation, triangulatus Graptolite Biozone. 4, FM 189 (slide Llan 1/1, 112 1204; E.F. no: K50/4), sample
LI 1 , Trefawr Formation, cyphus Graptolite Biozone.

Figs 5 and 6. Velatitetras laevigata gen. et sp. nov. 5, FM 190 (slide Llan 8B/1 1, 179 1294; E.F. no: S59/S60),
sample L5, Bronydd Formation, acinaces Graptolite Biozone. 6, FM 191, holotype (slide Llan 2/1, 050
1215; E.F. no: E51/4, sample LI 2, Trefawr Formation, cyphus Graptolite Biozone.

Figs 7-9. Velatitetras reticulata gen. et sp. nov. 7, FM 91 (slide Llan 1/2, 232 1110: E.F. no: Y41/2), sample
Lll, Trefawr Formation, cyphus Graptolite Biozone. 8, FM 197 holotype (slide Llan 8B/9, 128 1253: E.F.
no: N55/2), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 9, FM 192 (slide Llan 8B/8, 106
1304; E.F. no: K60/61/L60/61 ), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Fig. 10. Velatitetras rugulata gen. et sp. nov. FM 194, holotype (slide Llan 8B/12, 128 1220; E.F. no:
M52/3/M51/4), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Fig. 11. Velatitetras sp. A. FM 195 (slide Llan 8B/15, 118 1277; E.F. no: M58/1), sample L5, Bronydd
Formation, acinaces Graptolite Biozone.

Figs 12, 14, 15. Rimosotetras problematic a gen. et sp. nov. 12, FM 196 (slide Llan 8B/12, 240 1080; E.F. no:
Y38/3), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 14, FM 198, holotype (slide Llan
8B/6, 138 1228; E.F. no: 053/1), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 15, FM 199
(slide Llan 8B/8, 147 1089; E.F. no: P38/2), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Figs 13, 16, 17. Pseudodyadospora cf. laevigata Johnson, 1985. 13, FM 197 (slide Llan 8B/26, 035 1250; E.F.
no: C55/1 /2), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 16, FM 200 (slide Llan 8B/22,
226 1112; E.F. no: X41 /I ), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 17, FM 201 (slide
Llan 8B/12, 240 1095; E.F. no: Y39/4), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

PLATE 1

fee'1

aW

BURGESS, Llandovery sporomorphs

582

PALAEONTOLOGY, VOLUME 34

1985 Tetrahedraletes cf. T. medinensis Gray et a/., p. 524, fig. 5 F-H.

1986 Tetrahedraletes cf. T. medinensis Gray et al., p. 451, fig. 7, items 1-7.

1988 ‘Smooth walled tetrads’. Gray, p. 355, pi. 1, figs 1 and 5.

1988 Tetrahedraletes sp., Richardson, pi. 19, fig. 1.

Holotype and type locality. FM 187, PI. 1, fig. I, slide Llan 8B/12, 178 1170, E.F. no: S47/1, sample L5
(Text-fig. 2), Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone, Rhuddanian.

Paratypes. FM 188, PI. 1, fig. 2, slide Llan 8B/8, 061 1283, samples L5; Forestry Track F.33, Bronydd
Formation, acinaces Graptolite Biozone; FM92, PI. I, fig. 3, slide Llan 5/2, 130 0960, sample L20,
Aeronian/Telychian boundary section. Wormwood Formation, sedgwickii Graptolite Biozone; FM 189, PI. 1,
fig. 4, slide Llan 1/1, 020 1204, sample L 1 1 , Rhuddanian/ Aeronian boundary section, Trefawr Formation,
cvphus Graptolite Biozone.

Derivation of name. Latin parvus , small.

Diagnosis. A variety of Tetrahedraletes medinensis < 35 pm in diameter with low, narrow, rounded
and unfused equatorial crassitudes.

Description. Laevigate obligate tetrads, sub-circular to circular in outline and preserved in a variety of
compressional morphologies depending on the degree of rotation of the tetrad from an apical view prior to
compression. Individual ‘spores’ laevigate, amb sub-circular to sub-triangular, joined by unfused equatorial
crassitudes 1-5 pm wide. Proximal surface not observed. Distal exine 1-2 pm thick and invaginated in most
specimens.

Dimensions. Tetrads 19 (26) 30 //m in diameter (150 specimens measured, see Text-figure 6 for range of
measurements).

MICRONS

text-fig. 6. Size frequency distribution of a hundred
Tetrahedraletes medinensis var. parvus tetrads from
sample L5, Forestry Track F.33, Bronydd Formation,
acinaces Graptolite Biozone, Rhuddanian.

Biostratigraphical range. Latest Ordovician (persculptus Graptolite Biozone), to early Telychian ( turriculatus
Graptolite Biozone; Text-fig. 7).

Comparisons. Tetrahedraletes medinensis var. medinensis is larger (> 35 pm). Specimens described as
Tetrahedraletes cf. T. medinensis (Gray 1988) from various sections spanning the Ordovician/Silurian
boundary around the world have essentially identical measurements to those described above.

Remarks. Gray (1988) has presented data showing that tetrads gradually increase in size over the
Ordovician/Silurian boundary and states that this increase continues through the Llandovery. In
the Anglo-Welsh Basin, T. medinensis var. parvus var. nov. is the dominant variety in the latest
Ordovician and early Llandovery, whereas T. medinensis var. medinensis Strother and Traverse,
1979 dominates in the basal Wenlock (Burgess and Richardson 1991).

BURGESS: LLANDOVERY SPOROMORPHS

583

Genus velatitetras gen. nov.

Type species. Velatitetras laevigata sp. nov.

This genus is erected to accommodate tightly adherent cryptospore tetrads enclosed within an envelope. Some
specimens of Nodospora Strother and Traverse are believed to possess an envelope (e.g. N. retimembrana).
However, because the type species of Nodospora (N. burnhamensis) lacks an envelope and is considered
synonomous with the type species of Tetralredraletes (T. medinensis) the genus Nodospora cannot be used for
enveloped forms. Stegambiquadrella Johnson was erected for palynomorphs with four loosely-attached
inaperturate vesicles enclosed within an envelope. This genus cannot be used to accommodate tightly-adherent
tetrahedral tetrads enclosed within an envelope.

Derivation of name. Latin velatus. covered; tetras , tetrad.

Diagnosis. Obligate cryptospore tetrads composed of tightly-adherant laevigate sub-triangular to
sub-circular ‘spores’ with low and rounded + fused equatorial crassitudes. Tetrads enclosed within
a closely adherent to completely separated, + ornamented envelope; when envelope tightly-
adpressed any ornamentation passes over ‘spore’ contacts uninterrupted.

Velatitetras laevigata sp. nov.

Plate 1, figs 5 and 6

1985 ‘Obligate tetrad with smooth perispore’, Gray, p. 177, pi. 1, figs 2 and 3.

1988 ‘Spore tetrad with smooth or possibly degraded reticulate envelope’. Gray, p. 355, pi. 1, fig. 3.
1988 Nodospora sp. B, Richardson, p. 94.

Holotvpe and type locality. FM 191, PI. I, fig. 6, slide Llan 2/1, 050 1215, E.F. no: E51.4, sample L12 (Text-
fig. 3), Rhuddanian/Aeronian type boundary section, Trefawr Formation, cyphus Graptolite Biozone, late
Rhuddanian.

Paratype. FM 190, PI. 1, fig. 5, slide Llan 8B/11, 179 1294, sample L5, Forestry Track F.33, Bronydd
Formation, acinaces Graptolite Biozone.

Derivation of name. Latin laevigatus , smooth.

Diagnosis. A Velatitetras with cryptospore tetrad enclosed within a thin, laevigate and often folded
envelope.

Description. Obligate tetrads, sub-circular to circular in outline, totally enclosed within an envelope. Tetrads
composed of laevigate, sub-triangular ‘spores’ with unfused, low and rounded equatorial crassitudes 1-3 pm
wide; distal polar exine c. 1 pm thick. Envelope laevigate, diaphanous, folded and < 1 //m thick; varies from
completely separated, to closely adpressed to tetrad.

Dimensions. Enclosed tetrads 21 (27) 34 /im in diameter (22 specimens measured).

Biostratigraphic range. Latest Ordovician (persculptus Graptolite Biozone), to late Rhuddanian (cyphus
Graptolite Biozone; Text-fig. 7).

Comparisons. Gray (1985, pi. 1, figs 2 and 3; 1988, pi. 1, fig. 3), illustrated small (c. 25 pm) and
essentially identical tetrads enclosed within smooth envelopes from the Late Ordovician of America.

Biostratigraphic and lithostratigraphic ranges of sporomorphs from the type Llandovery area.
Note that the majority of the Telychian was not sampled.

SAMPLES

584

PALAEONTOLOGY, VOLUME 34

i/i

3

>

L

(8

a

M

Ul

J

DL

I

<

01

L23

L22

L21

L20

L19

L18

L17

L16

L15

L14

L13

L12

L11

LIO

L9

L 8

L7

L 6

L 5

L4

L3

L 2

LI

text-fig. 7. Biostratigraphic and lithostratigraphic ranges of sporoniorphs from the type Llandovery area.
Note that the majority of the Telychian was not sampled.

BURGESS: LLANDOVERY SPOROMORPHS

585

Velatitetras reticulata sp. nov.

Plate 1, figs 7-9.

1985 Obligate tetrad with ‘reticulate perispore’, Gray, p. 177, pi. I, figs 4-9; pi. 2, fig. 17.

1988 Obligate tetrad with ‘reticulate envelope’. Gray, p. 355, pi. 1, figs 2, 4, 6.

Holotvpe and type locality. FM 192. PI. 1, fig. 8, slide Llan 8B/9, 128 1253, E.F. no: N55/2, sample L5 (Text-
fig. 2), Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone, Rhuddanian.

Paratypes. FM 91, PI. 1, fig. 7, slide Llan 1 /2, 232 1 1 10, sample LI 1, Rhuddanian/Aeronian boundary section,
Trefawr Formation, cyphus Graptolite Biozone; FM 193, PI. I, fig. 9, slide Llan 8B/8, 106 1304, sample L5,
Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone.

Derivation of name. Latin reticu/atus , netted, marked with a network - refers to arrangement of muri on
envelope.

Diagnosis. A Velatitetras with cryptospore tetrad enclosed within an envelope ornamented with
muri forming a reticulum with small lumen.

Description. Obligate tetrads sub-circular to circular in outline, totally enclosed within an envelope. Tetrads
composed of laevigate, inaperturate, sub-triangular to sub-circular ‘spores’ with unfused equatorial crassitudes
1-3 pm wide; exine at distal pole c. 1 pm thick. Envelope closely adpressed to widely separated from enclosed
tetrad and ornamented with low (c. 1 /mr), rounded muri 0-5(1) 1-5 pm wide, which usually form an ill defined
reticulum 1-3-5 /mi in maximum diameter, although this may be of regular size and distribution over the
envelope.

Dimensions. Enclosed tetrads 22 (27) 35 /mi in diameter (14 specimens measured).

Biostratigraphic range. Latest Ordovician ( persculptus Graptolite Biozone) to late Aeronian (sedgwickii
Graptolite Biozone; Text-fig. 7).

Comparisons. Comparable specimens have been described by Gray (1985, pi. 1, figs 4-9; pi. 2, fig.
17) and Gray (1988, pi. 1, figs 2, 4, 6) from the Ashgill of the USA. Velatitetras rugulata sp. nov.
is of similar size, but has sinuous rugulae on its envelope. ‘ Nodospora' retimembrana Miller and
Eames, 1982 is much larger (34-60 //m) and the reticulum on the envelope is both larger and more
regularly arranged.

Velatitetras rugulata sp. nov.

Plate 1, fig. 10

Holotype and type locality . FM 194, PI. 1 , fig. 10, slide Llan 8B/12, 128 1220, E.F. no : M52/3/M51 /4, sample
L5 (Text-fig. 2), Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone, Rhuddanian.

Derivation of name. Latin rugulatus , refers to rugose ornamentation on envelope.

Diagnosis. A Velatitetras with cryptospore tetrad enclosed within an envelope ornamented with
sinuous to convolute and anastomosing rugulae.

Description. Obligate tetrads sub-circular to circular in outline, totally enclosed within an envelope. Tetrad
composed of laevigate, inaperturate, sub-triangular ‘spores’ with low and rounded unfused equatorial
crassitudes 1-3 pm wide; exine at distal pole c. 1 pm thick. Envelope closely adpressed to completely separated
from enclosed tetrad and ornamented with low, rounded rugulae 0-5-1 //m wide, < I pm high and I -5 (2-5)
3 pm apart ; rugulae are closely spaced, slightly sinuous to convolute and occasionally anastomosing.

Dimensions. Enclosed tetrads 24 (28-5) 33 /mi in diameter (13 specimens measured).

586

PALAEONTOLOGY, VOLUME 34

Biostratigraphic range. Early Rhuddanian (acinaces Graptolite Biozone; Text-fig. 7).

Comparisons. Velatitetras reticulata sp. nov. is of similar size but the muri on its envelope form a
reticulum. ‘ Nodospora' rugosa Strother and Traverse, 1979, Nodospora sp. A Richardson 1988, and
N. cf. rugosa Miller and Eames 1982 are much larger (> 43 //m).

Velatitetras sp. A

Plate 1, fig. 1 1

71988 Nodospora sp. E, Richardson, p. 94.

Figured specimen. FM 195, PI. I, fig. II, slide Llan 8B/15, 118 1277, E.F. no: M58/1, sample L5, Forestry
Track F.33, Bronydd Formation, acinaces Graptolite Biozone, Rhuddanian.

Description. Obligate tetrad sub-circular in outline, totally enclosed within an envelope. Tetrad composed of
laevigate, inaperturate, sub-triangular 'spores’ with low and rounded crassitudes 1-3 pm wide. Envelope
separated from tetrad and ornamented with closely packed grana.

Dimensions. 1 specimen with enclosed tetrad of 25 pm diameter.

Biostratigraphic range. Bronydd Formation, acinaces Graptolite Biozone (Text-fig. 7).

Comparisons. The granulate ornament on the envelope separates this tetrad from all other
specimens. Segestrespora sp. A has a similarly ornamented envelope surrounding a pseudodyad.

Genus rimosotetras gen. nov.

Type species. Rimosotetras problematica sp. nov.

Derivation of name. Latin rimosus , cracked.

Diagnosis. Adherent, but usually partially separating, tetrahedral tetrads composed of alete,
laevigate, sub-triangular to circular, Tcrassitate spores or spore-like units.

Remarks. In Tetrahedraletes Strother and Traverse, 1979 emend, the spores comprising the tetrad
are much more tightly adherent. Tetrads of Ambitisporites Hoffmeister, 1959 separate more readily
and produce miospores with obvious trilete marks.

Rimosotetras problematica sp. nov.

Plate 1, figs 12, 14, 15

1971 'Spore tetrad’. Gray and Boucot, fig. 1(g).

1985 ‘Loose tetrads’, Richardson, p. 29, pi. 15, figs 5 and 6.

1988 Nodospora burnhamensis 'loose tetrad’, Richardson, pi. 19, figs II and 12.

Holotype and type locality. FM 198, PI. 1, fig. 14, slide Llan 8B/6, 138 1228, E.F. no: 053/1, sample L5 (Text-
fig. 2), Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone, Rhuddanian.

Paratypes. FM 196, PI 1, fig. 12, slide Llan 8B/12, 240 1080; FM 199, PI. 1, fig 15, slide Llan 8B/8, 147 1089,
both sample L5, Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone.

Derivation of name. Latin problematicus , problematic; refers to problematic nature of these sporomorphs which
arc similar to permanently fused cryptospore tetrads, but which appear to separate into alete ‘spores', in other
ways similar to trilete miospores.

BURGESS: LLANDOVERY SPOROMORPHS

587

Diagnosis. A Rimosotetras with sub-triangular ‘spores’ possessing rounded, non-projecting,
equatorial crassitudes.

Description. Cryptospore tetrads sub-circular to circular in outline and preserved in a variety of compressional
morphologies reflecting rotation of tetrad from an apical view prior to compression. Tetrads composed of
loosely attached, frequently partially separated, laevigate, alete or indistinctly trilete, sub-triangular to sub-
circular ‘spores’ with narrow and rounded equatorial crassitudes 0-75 (15) 2-5 //m wide. Distal exine convex
c. 1 //m thick, usually inflated, but can be invaginated. Proximal surface lacks sutures typical of trilete
miospores, although specimens may have a crack or split at the position of a trilete mark indicating a weakness
of the exine.

Dimensions. Tetrads 23 (28) 42 pm in diameter (38 specimens measured). Individual spores 15 (22) 35 pm in
diameter (67 measured).

Biostratigraphic range. Latest Ordovician ( persculptus Graptolite Biozone) to late Rhuddanian ( cyphus
Graptolite Biozone) (Text-fig. 7).

Comparisons. Comparable ‘loose’ tetrads have been recorded in the Rhuddanian of Libya by
Richardson (in Hill et al. 1 985, p. 29) and Richardson ( 1 988). The spore tetrad from North America
drawn by Gray and Boucot (1971, fig. 1(g)) also appears similar.

2. Cryptospore dyads

This group encompasses permanently fused dyads (pseudodyads) which may be enclosed within a variously
ornamented envelope and dyads which readily separate into two alete spores (true dyads).

Genus pseudodyadospora Johnson, 1985
Type species. Pseudodyadospora laevigata Johnson, 1985.

Pseudodyadospora cf. laevigata Johnson, 1985
Plate 1, figs 13, 16, 17

Figured specimens. FM 197, PI. I, fig. 13, slide Llan 8B/26, 035 1250; FM 200, PI. 1, fig. 16, slide Llan 8B/22,
226 1112; FM 201, PI. 1, fig. 17, slide Llan 8B/12, 240 1095, all specimens sample L5 (Text-fig. 2), Forestry
Track F.33, Bronydd Formation, acinaces Graptolite Biozone.

Description. Pseudodyads elliptical to sub-circular or rarely circular in outline. ‘Spores’ of pseudodyad joined
by a single encircling darkened band 1 (2) 3 /rm wide providing no evidence for ‘spore’ separation. ‘Spores’
are of unequal size in 62% of the specimens, laevigate, and usually distally convex, although they can be
invaginated. Exine often folded and 0 5—1 //m thick at distal pole.

Dimensions. Pseudodyads 22 (26) 42 pm long and 16 (22) 28 pm wide at equator (50 specimens measured).
Length to width ratio = F3.

Biostratigraphic range. Latest Ordovician (persculptus Graptolite Biozone) to late Aeronian (sedgwickii
Graptolite Biozone; Text-fig. 7).

Comparisons. Pseudodyadospora laevigata Johnson, 1985 is similar in outline, the variable size of the
‘spores’ and the fused suture between the ‘spores’, but is somewhat larger (41 (52) 60 pm long),
although there is some overlap in the sizes. Dyads in Segestrespora gen. nov. possess an envelope
surrounding an otherwise similar pseudodyad. Dyads of the genus Dyadospora Strother and
Traverse, 1979 have well developed lines of separation between the equally sized spores, and are
generally seen partially separated into two alete sporomorphs (Burgess and Richardson 1991).

588

PALAEONTOLOGY, VOLUME 34

Genus segestrespora gen. nov.

Type species. Segestrespora ( Dyadospora ) membranifera (Johnson) comb. nov.

Derivation of name. Latin segestre, covering or wrapper.

Diagnosis. Bipolar, laevigate and permanently fused pseudodyads, divided by a single central to
slightly off-centred thickened encircling band, and totally enclosed within a closely adherent to
completely separated envelope which either lacks ornamentation, or has apiculi, muri, rugulae or
verrucae.

Remarks. In some specimens an envelope can not be easily seen. However, if any sculpture passes
over the ‘spore’ contact uninterrupted, then an envelope is assumed to be present and specimens
are assigned to Segestrespora.

Comparisons. The genus Pseudodyadospora Johnson is for obligate pseudodyads with only one wall-
layer, which can be sculptured. In Dyadospora Strother and Traverse, the dyads lack an envelope
and are readily separated into two laevigate alete spores.

Segestrespora ( Dyadospora ) membranifera (Johnson, 1985) comb. nov.

Plate 2, figs 2-5

1985 Dyadospora membranifera Johnson [partim], p. 336, pi. 7, figs 1-3, 5, 6, non pi. 7, fig. 7.

1988 Pseudodyadospora cf. Dyadospora membranifera , Richardson, p. 94, pi. 15, figs 1 and 2.

Holotype and type locality. Johnson, 1985, pi. 7, figs 1-3: sample TMH81-1 1-26, location EF D35-4; c. 75 m
from base of Mill Hall section, Tuscarora Formation. Section - 200 m long and located 250 m above US Route
220, 1 km SE of Mill Hall, Clinton County, Pennsylvania, USA.

Figured specimens. FM 203, PI. 2, fig. 2, slide Flan 8B/7, 198 1020; FM 204, PI. 2, fig. 3, slide Llan 8B/7, 220
2100; FM 205, PI. 2, fig. 3, slide Flan 8B/11, 240 1170; FM 206, PI. 2, fig. 5, slide Flan 8B/10, 130 1150, all
sample F5, Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone.

Diagnosis. A Segestrespora where the envelope is ornamented with narrow muri which form a low,
regular to irregularly sized and shaped reticulum.

Description. Pseudodyads elliptical to sub-circular in lateral view and totally enclosed within an envelope.
Individual ‘spores' of pseudodyad convex in lateral compression. Exine c. 1 /an thick with rare folds. A single
darkened, possibly thickened, encircling band 1-2 //m wide joins the ‘spores’ which are anisomorphic in
c. 75% of the specimens. The envelope is diaphanous and may be tightly adherent to the pseudodyad and
difficult to resolve, or completely separated. Envelope is ornamented with muri, 0-5-1 //m wide and c. 0-5 /an
high, which form a regularly to irregularly sized and shaped reticulum with lumen 1-5 (2) 3 pm in diameter.

Dimensions. Pseudodyads 25 (29) 31 /an long, 18 (20) 22 /an wide at equator (12 specimens). Fength to width
ratio = 1 -4.

Biostratigraphic range. Late Ordovician (persculptus Graptolite Biozone) to early Rhuddanian (acinaces
Graptolite Biozone; Text-fig. 7).

Comparisons. Those specimens of ‘'Dyadospora membranifera' Johnson, 1985 which possessed a
reticulate envelope are members of this species, but those with a laevigate envelope are transferred
to another species (see below). Specimens referred to Pseudodyadospora cf. Dyadospora
membranifera by Richardson, 1988 are also members of this species. When specimens of
Segestrespora membranifera have lost their membrane they are placed in Pseudodyadospora
laevigata Johnson, 1985.

BURGESS: LLANDOVERY SPOROMORPHS

589

Segestrespora laevigata sp. nov.

Plate 2, fig. 1

71982 Dyadospora murusdensa with ‘membrane’, Miller and Eames, p. 248.

1985 ‘ Dyadospora membranifera ’ Johnson, \partim\, p. 336, pi. 7, fig. 7, non pi. 7, figs 1-3, 5, 6.

71985 Dyadospora murusdensa ‘with diaphanous sheath', Richardson in Hill et a/., p. 29.

Holotype. Johnson. 1985, pi. 7, fig. 7, Sample TMH8I-8, c. 125 m above the base of the Mill Hall section,
Tuscarora Formation. Section - 200 m long and located 250 m about US Route 220, 1 km SE of Mill Hall,
Clinton County, Pennsylvania, USA.

Figured specimen. FM 202, PI. 2, fig. 1, slide Llan 8B/2, 103 1294, sample L5, Forestry Track F.33, Bronydd
Formation, acinaces Graptolite Biozone.

Derivation of name. Latin laevigatas, smooth.

Diagnosis. A Segestrespora with a thin, folded and laevigate envelope.

Description. Pseudodyads elliptical to sub-circular in lateral view and totally enclosed within an envelope.
Individual ‘spores’ of pseudodyad laevigate, distally convex in lateral compression, occasionally folded with
walls c. 1 //m thick. A single darkened encircling band 1-2 /mi wide joins ‘spores’ which are anisomorphic in
c. 25% of observed specimens. Envelope is laevigate, < 1 /nn thick, often highly folded, and varies from being
closely adpressed to the dyad and difficult to resolve, to completely separated.

Dimensions. Pseudodyads 26 (29) 41 /nn long, 17 (23) 31 /nn wide at equator (20 specimens measured). Length
to width ratio = 1 -3.

Bio stratigraphic range. Latest Ordovician (persculptus Graptolite Biozone) to early Rhuddanian (acinaces
Graptolite Biozone; Text-fig. 7).

Comparisons. Segestrespora membranifera comb. nov. has reticulate ornamentation on its envelope
and S. rugosa comb. nov. has rugulate ornamentation. One of the specimens of ‘ Dyadospora
membranifera' Johnson possessed a smooth envelope (pi. 7, fig. 7) and this specimen is included in
this new species. Other possible records of S. laevigata, all from the Rhuddanian, are: Miller and
Eames (1982), who stated (p. 248) that membranes were occasionally found around ‘ Dyadospora
murusdensa' ; Strother and Traverse (1979, p. 7) who recovered ‘membrane enclosed bilaterally
svmetrical opaque bodies’; and Richardson in Hill et al. (1985), who recorded ‘diaphanous sheaths’
(p. 29) around ‘ Dyadospora murusdensa' .

Segestrespora ( Pseudoclyadospora) rugosa (Johnson, 1985) comb. nov.

Plate 2, figs 7-9

1985 Pseudo dyadospora rugosa Johnson, p. 337, pi. VIII, figs 2-6.

1988 Pseudodyadospora sp. C, Richardson, p. 95.

Holotype and type locality. Johnson, 1985, pi. 8, fig. 5, slide TMH8 1-3-27, location EF M39-4, c. 175 m above
the base of the Mill Hall section, Tuscarora Formation. Section - 200 m long and located 250 m above US
Route 220, I km SE of Mill Hall, Clinton County, Pennsylvania, USA.

Figured specimens. FM 208, PI. 2, fig. 7, slide Llan 8B/1 1, 077 1214; FM 209, PI. 2, fig. 8, slide Llan 8B/6, 055
1125; FM 210, PI. 2, fig. 9, slide Llan 8B/2, 086 1234; all sample L5, Forestry Track F.33, Bronydd Formation,
acinaces Graptolite Biozone.

Diagnosis. A Segestrespora with an envelope ornamented with regularly sinuous and closely spaced
rugulae.

590

PALAEONTOLOGY, VOLUME 34

Description. Pseudodyads sub-circular to elliptical in lateral view, and totally enclosed within an envelope.
Individual ‘spores' of pseudodyad distally convex with a laevigate and occasionally folded wall c. 1 //m thick.
A single darkened and possibly thickened encircling band, 1-2 /m wide, joins the ‘spores’ which are
anisomorphic in c. 33 % of the specimens. The envelope is tightly adpressed to completely separated from
pseudodyad and ornamented with closely spaced, regularly sinuous, rugulae 0-5 (1) 1-5 pm wide, < I //m high
and 0-5-1 pm apart.

Dimensions. Pseudodyads 27 (29) 31 /mi long, 15 (20) 23 pm wide at equator (12 specimens measured). Length
to width ratio = 1-4.

Biostratigraphic range. Latest Ordovician (persculptus Graptolite Biozone) to early Rhuddanian ( acinaces
Graptolite Biozone; Text-fig. 7).

Comparisons. Pseudodyadospora sp. C Richardson, 1988 appears identical to these spores.
Pseudodyadospora rugosa Johnson, 1985 is similarly ornamented, but Johnson (1985, p. 337) made
no mention of whether the rugulae were on an envelope or the dyad wall. However, Johnson now
believes (pers. comm. 1986) that there is a closely adherent envelope surrounding her pseudodyads.
The envelope surrounding Velatitetras rugulata sp. nov. is similarly sculptured, but encloses a
cryptospore tetrad.

EXPLANATION OF PLATE 2

All figures x 1000 unless stated otherwise.

Fig. 1. Segestrespora laevigata gen. et sp. nov. FM 202 (slide Llan 8B/2, co-ord 103 1294; E.F. no: K59/2/4),
sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Figs 2-5. Segestrespora membranifera (Johnson, 1985) comb. nov. 2, FM 203 (slide Llan 8B/7, co-ord 198
1020; E.F. no: T31/3), sample L5. Bronydd Formation, acinaces Graptolite Biozone, x 2000. 3, FM 204
(slide 8B/7, co-ord 220 2100; E.F. no: W50/3/W49/4), sample L5, Bronydd Formation, acinaces Graptolite
Biozone, x 2000. 4, FM 205 (slide Llan 8/11, co-ord 240 1170; E.F. no: Y47/1), sample L5, Bronydd
Formation, acinaces Graptolite Biozone, x2000. 5, FM 206 (slide Llan 8B/10, co-ord 130 1 150; E.F. no:
N45/2), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Fig. 6. Segestrespora sp.A. FM 207 (slide Llan 7B/4, co-ord 100 1015; E.F. no: K30/2), sample LI, Scrach
Formation, persculptus Graptolite Biozone.

Figs 7-9. Segestrespora rugosa gen. et sp. nov. 7, FM 208 (slide Llan 8B/1 1, co-ord 077 1214; E.F. no: G51/4),
sample L5, Bronydd Formation, acinaces Graptolite Biozone, x 2000. 8, FM 209 (Llan 8B/6, co-ord 055
1125; E.F. no: E42/1), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 9, FM 210 (slide
Llan 8B/2, co-ord 086 1234; E.F. no: H53/4), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Fig. 10. Dyadospora murusattenuata Strother and Traverse, 1979. FM 211 (slide Llan 8B/22, 138 1350; E.F.
no: N65/4), sample L5, Bronydd Formation, acinaces Graptolite Biozone.

Figs 1 1 and 12. Rugosphaera cf. R1. cerebra Miller and Eames, 1982. 1 1, FM 212 (slide Llan 8B/8, co-ord 1 10
1300; E.F. no: K60/2), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 12, FM 213 (slide
Llan 8B/24, co-ord 101 1080; E.F. no: K37/4), sample L5, Bronydd Formation, acinaces Graptolite
Biozone.

Fig. 13. Strophomorpha ovata Miller and Eames, 1982. FM 214 (slide Llan 3B/3, co-ord 035 1090; E.F. no:
B39/1/2), sample L8, Crychan Formation, cyphus Graptolite Biozone.

Fig. 14. Nematothallus cuticle (see Edwards 1986). FM 219 (slide Llan 6/2, 237 1250; E.F. no: Y55/1), sample
L21, Wormwood Formation, sedgwickii Graptolite Biozone.

Fig. 15. Ambitisporites dilutus (Hoffmeister) Richardson and Lister, 1969. FM 95 (slide Llan 6/2, co-ord 071
1160; E.F. no: G47/3), sample L21, Wormwood Formation, sedgwickii Graptolite Biozone.

Figs 16-18. Ambitisporites ? vavrdovii Richardson, 1988. 16, FM 216 (Llan 7B/6, 090 1180; E.F. no: J48),
sample LI, Scrach Formation, persculptus Graptolite Biozone. 17, FM 217 (slide Llan 8B/6, 138 1225; E.F.
no: P53/3), sample L5, Bronydd Formation, acinaces Graptolite Biozone. 18, FM 218 (slide Llan 3B/1, 130
1296; E.F. no: N59/N60), sample L8, Crychan Formation, cyphus Graptolite Biozone.

PLATE 2

BURGESS, Llandovery sporomorphs

592

PALAEONTOLOGY, VOLUME 34

Segestrespora sp. A

Plate 2, fig. 6

Figured specimen. FM 207, PI. 2, fig. 6, slide Llan 7B/4, 100 1015, sample LI, Forestry Track F.33, Scrach
Formation, persculptus Graptolite Biozone, latest Ordovician.

Description. Pseudodyads sub-circular to elliptical in equatorial view, and totally enclosed within an envelope.
Individual ‘spores' of pseudodyad distally convex, with a laevigate, occasionally folded wall c. 1 pm thick. A
single, darkened and possibly thickened encircling band, 1-2 pm wide, joins ‘spores’ which are anisomorphic
in c. 33 % of the specimens. The envelope varies from closely adpressed to the pseudodyad to completely
separated from it, and is ornamented with regularly distributed and closely spaced grana < 05 /mi high and
wide and c. 0-5 //m apart.

Dimensions. Pseudodyads 29 (305) 35 pm long, 26 (27) 29 /mi wide at equator (3 specimens only). Length to
width ratio = 1-2.

Biostratigraphic range. Latest Ordovician (persculptus Graptolite Biozone), to early Rhuddanian ( acinaces
Graptolite Biozone; Text-fig. 7).

Comparison. Velatitetras sp. A has a similarly ornamented envelope surrounding a cryptospore
tetrad.

Genus dyadospora Strother and Traverse, 1979
Type species. Dyadospora murusattenuata Strother and Traverse, 1979

(Due to the presence of naked, laevigate and permanently fused pseudodyads in the early Silurian
(Pseudodyadospora Johnson, 1985), as well as similar enveloped forms (Segestrespora gen. nov.) the genus
Dyadospora Strother and Traverse, 1979 is taken to include only non-enveloped laevigate dyads which
habitually separate to produce two laevigate alete spores (see Burgess and Richardson 1991).

Dyadospora cf. murusattenuata Strother and Traverse, 1979

Plate 2, fig. 10

Figured specimen. FM 21 1, PI. 2, fig. 10, slide Llan 8B/22, 138 1350, sample L5, Forestry Track F.33, Bronydd
Formation, acinaces Graptolite Biozone.

Dimensions. Dyad 22 (28) 39 pm long, 20 (27) 37 pm wide at equator (5 specimens measured).

Biostratigraphic range. Early Rhuddanian (acinaces Graptolite Biozone; Text-fig. 7).

Comparisons. The few specimens recovered generally conform to the description of Strother and
Traverse (1979), but preservation is too poor to be sure they are conspecific.

3. Cryptospore monads

This group encompasses monads whose wall thickness and sculpturing is comparable to that of the cryptospore
tetrads and dyads but where it is more difficult to be convinced they are derived from land plants.

Genus rugosphaera Strother and Traverse, 1979
Type species. Rugosphaera tuscarorensis Strother and Traverse, 1979

BURGESS: LLANDOVERY SPOROMORPHS

593

Rugosphaera cf. /?.? cerebra Miller and Eames, 1982
Plate 2, figs 1 1 and 1 2

Figured specimens. FM 212, PI. 2, fig. 11, slide Llan 8B/8, 1 10 1300; FM 213, PI. 2, fig. 12, slide Llan 8B/24,
101 1080, both sample L5, Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone.

Description. Cryptospore monads sub-circular to circular in outline. In some specimens a laevigate inner body
with a c. 1 pm thick wall can be discerned. Envelope always tightly adpressed to inner body and ornamented
with closely spaced regularly sinuous to angular and rarely anastomosing rugulae 1 (2)3-5 /an wide, 0-5-1 pm
high and 0-5( 1)1-5 //m apart.

Dimensions. Maximum diameter 23 (28) 37 pm, minimum diameter 16 (24) 30 pm (8 specimens measured).

Biostratigraphic range. Early Rhuddanian ( Acinaces Graptolite Biozone; Text-fig. 7).

Comparisons. Rugosphaera ? cerebra Miller and Eames is larger (38(45)55 //m in diameter), but
otherwise similar. R. tuscarorensis Strother and Traverse has broader ( 1-4 /an) more widely spaced
and less sinuous muri.

Genus strophomorpha Miller and Eames, 1982

Type species. Strophomorpha ovata Miller and Eames, 1982.

Strophomorpha ovata Miller and Eames, 1982
Plate 2, fig. 13

Figured specimen. FM 214, PI. 2, fig. 13, slide Llan 3B/3, 035 1090, sample L8, Forestry Track F.33, Crychan
Formation, cyphus Graptolite Biozone.

Description. Cryptospore monad with an elliptical to sub-rectangular outline and rounded poles. Body
ornamented with low and rounded muri 0-75 (2) 3-5 pm wide separated by striae 1 (2)3-5 pm across; muri are
spirally arranged and converge at the poles. Wall is 1-3 //m thick and a single layer.

Dimensions. Maximum diameter 29 (46) 60 pm, minimum diameter 21 (29) 36 pm (10 specimens measured).
Biostratigraphic range. Late Rhuddanian ( cyphus Graptolite Biozone; Text-fig. 7).

Comparisons. Strophomorpha ovata Miller and Eames is larger (48(56)65 //m) but in other ways
closely comparable. Moyeria cabottii Cramer, 1966 is thinner-walled with narrower muri.
Qualisaspora fragilis Richardson et a/., 1984 has two, much thinner, wall-layers.

Remarks. The presence of Strophomorpha ovata in the deeper-marine Rhuddanian aged sediments
of the type Llandovery area, rather than the near-shore strata, indicates this species may be a thick-
walled acritarch as opposed to a cryptospore. However, for the present it is described as a
cryptospore until more evidence accumulates on its facies distribution.

Anteturma sporites H. Potonie, 1893
Turma triletes Reinsch, 1891

Subturma zonotriletes Waltz, 1935 in Luber and Waltz (1938)

Infraturma crassiti Bharadwaj and Venkatachala, 1961

Genus ambitisporites Hoffmeister, 1959

Type specimens. Ambitisporites avitus Hoffmeister, 1959

594

PALAEONTOLOGY, VOLUME 34

Ambitisporites dilutus (HofTmeister) Richardson and Lister, 1969

Plate 2, fig. 15

Figured specimen. FM 95, PI. 2, fig. 15, slide Llan 6/2, 071 1160, sample L21, Aeronian/Telychian type
boundary section. Wormwood Formation, sedgwickii Graptolite Biozone.

Dimensions. Diameter 20 (25) 30 pm (7 specimens measured).

Biostratigraphic range. Late Aeronian (sedgwickii Graptolite Biozone) to early Telychian ( turriculatus
Graptolite Biozone; Text-fig. 7).

Comparisons. Ambitisporites ? vavrdovii Richardson, 1988 has triradiate splits, not sutures, in the
proximal wall and a more prominent equatorial crassitude.

Ambitisporites ? vavrdovii Richardson, 1988
Plate 2, figs 16-18

Figured specimens. FM 216, PI. 2, fig. 16, slide Llan 7B/6, 090 1180, sample LI, Forestry Track F.33, Scrach
Formation, persculptus Graptolite Biozone. FM 217, PI. 2, fig. 17, slide Llan 8B/6, 138 1225, sample L5,
Forestry Track F.33, Bronydd Formation, acinaces Graptolite Biozone; FM 218, PI. 2, fig. 18, slide Llan 3B/1,
130 1296, sample L8, Crychan Formation, cyphus Graptolite Biozone.

Description. Amb sub-triangular to triangular. Equatorial crassitude 1-3 pm wide and well defined. Proximal
surface laevigate. Trilete mark distinct, is a simple split without laesurae which extends to the equator or nearly
so. Distal exine 1-2 pm thick and laevigate.

Dimensions. Diameter 28 (24) 32 pm (8 specimens measured).

Biostratigraphic range. Latest Ordovician (persculptus Graptolite Biozone) to early Aeronian (convo/utus
Graptolite Biozone; Text-fig. 7).

Comparisons. Comparable ‘trilete’ spores have been described by Gray and Boucot (1971, lower
Llandovery of North America), Gray et al. (1982), ?Caradoc of Libya), Johnson (1985, lower
Llandovery of North America), and Richardson (1988, late Ordovician and early Llandovery of
Libya). Moreover, Gray et al. (1982) have shown that closely comparable spores can be derived
from mechanically fragmented tetrads of the genus Tetrahedraletes. Ambitisporites avitus
Hoffmeister has its earliest records in the later Llandovery of North Africa and has more prominent
trilete marks with sutures and laesurae, never a simple split.

COMPARISON WITH SIMILARLY AGED MICROFLORAS WORLDWIDE

Several continents have been recognized in the Silurian (Livermore et al., 1985). These include
Laurentia (North America and Europe north of the Iapetus suture), Baltica (Europe and Russia
south of the Iapetus suture and north of the Flercynian suture), and Gondwana (Africa, South
America and Australasia). Uppermost Ordovician and Llandovery spore microfloras from the type
Llandovery area are compared with those from these continents.

Late Ordovician and early Silurian assemblages

Baltica. Vavrdova (1982, 1984, 1988) has described a diverse assemblage of cryptospore tetrads
from the late Ordovician (bohemicus graptolite Biozone) Kosov Formation in Czechoslovakia.
Many specimens appear similar, both in size, and sculptural patterning on their envelopes, to those
from the type Llandovery area. For example, tetrad types A, B and C of Vavrdova (1984) are

BURGESS: LLANDOVERY SPOROMORPHS

595

similar to Tetrahedraletes medinensis var. parvus , Velatitetras reticulata and Velatitetras sp.A
respectively. 'Trilete spores’ ( Ambitisporitesl vavrdovii ) were also recorded in Czechoslovakia.

Laurentia. Sporomorph assemblages from the late Ordovician and early Llandovery of North
America are also similar to those from the type Llandovery area in that they are predominantly
composed of cryptospores (Gray and Boucot 1971 ; Cramer and Diez 1972; Strother and Traverse
1979; Miller and Eames 1982; Gray 1985; Johnson 1985; Gray 1988). In many of these studies
(Strother and Traverse 1979; Miller and Eames 1982; Johnson 1985) the North American
cryptospore tetrads are considerably larger (x = c. 50 //m) than those from the type Llandovery area
(x — 26 //m). However, those cryptospore tetrads obtained from upper Ordovician and basal
Llandovery strata by Gray ( 1985, 1988) are of similar size (.f = < 30 //m) and some possessed highly
comparable envelopes.

Ambitisporitesl vavrdovii has also been recorded from the Llandovery of North America (Gray
and Boucot 1971; Gray 1985).

Gondwana. A diverse assemblage of cryptospores has been recently described from subsurface
material of late Ordovician to late Llandovery age in north-east Libya (Richardson 1988). This
assemblage comprises around ten species of cryptospore tetrad, eleven species of pseudodyad, two
species of true dyad, three species of cryptospore monad and three species of trilete miospore. Some
of these are identical to those from the type Llandovery area, although eleven of the species
discovered in Libya have not been seen in Britain.

Other late Ordovician and early Llandovery records of sporomorphs (all Tetrahedraletes) from
Gondwana are from the Soom Shale Member of the Cedarberg Formation, Table Mountain
Group, South Africa (Gray et al. 1986); and from the Vila Maria Formation of the Parana Basin,
Brasil (Gray et al. 1985). Furthermore, specimens of Tetrahedraletes recovered by Gray et al. ( 1982)
from the ?Caradoc strata of the Murzurk Basin in Libya are closely comparable to uppermost
Ordovician and lower Llandovery specimens from the type Llandovery area.

Late Llandovery assemblages

Late Llandovery sporomorph assemblages are known from Baltica, Laurentia and Gondwana
(Hoffmeister 1959; Cramer 1968, 1969; Smith 1975; Emo and Smith 1978; Aldridge et al. 1979;
Pratt et al. 1978; Smith 1979, 1981 ; Gray 1985; Hill et al. 1985; Gray 1988; Richardson 1988). All
these investigations have demonstrated low spore diversity during this period with smooth-walled
trilete miospores of the genus Ambitisporites predominating. This is comparable with the late
Llandovery spore assemblage obtained from the type Llandovery area.

PALYNOFACIES

Preliminary investigation of the palynofacies in the type Llandovery area (Burgess 1987), has
demonstrated large numbers of cryptospores and reworked acritarchs (Tremadoc and Ordovician)
in the Scrach Formation and lower Bronydd Formation. This indicates deposition in nearshore
palaeoenvironments, as previously suggested following mega fossil and sedimentological investiga-
tions (Cocks et al. 1984). Higher in the Bronydd Formation and throughout the Crychan, Trefawr,
Rhydings, Wormwood, and Cerig Formations an extremely sparse microflora with few spores was
obtained. The palynological assemblage, macro-fauna and sedimentology (Cocks et al. 1984) of
these formations indicate deposition in off-shore shelf conditions at a considerable distance from
land. It is thus probable that the sporomorph assemblage obtained from the base of the Llandovery
type section is more representative of its age than that from the late Rhuddanian through to the
early Telychian (later Telychian not sampled).

596

PALAEONTOLOGY, VOLUME 34

BIOSTRATIGRAPHY

Text-figure 7 presents the stratigraphical ranges of the sporomorphs recovered from the type
Llandovery area. The two fold sub-division of the sequence is closely comparable to that recognized
by Gray (1985, 1988), Richardson and McGregor (1986) and Richardson (1988).

Gray (1985) first noted the presence of a cryptospore-dominated spore assemblage in the Late
Ordovician and Early Silurian and named this ‘Microfossil Assemblage Zone 1’. It has recently
been formally named the Nodospora sp. A ( = Velatitetras laevigata) - Dyadospora murusdensa
Assemblage Biozone (Richardson 1988), and has been further subdivided into three Assemblage
Biosubzones. Samples L1-L20, from the late persculptus to late sedgwickii Graptolite Biozones of
the type Llandovery area, fall within the Dyadospora membranifera (= Segestrespora membrani-
fera) - Pseudodyadospora sp. B Biosubzone of Richardson (1988). Ambitisporitesl vavrdovii is
present in this Biosubzone, but as these spores probably derive from mechanically fragmented
cryptospore tetrads (Gray et al. 1982) they do not represent the earliest records of trilete miospores.

Gray (1985) also recognized a later Llandovery spore zone typified by smooth-walled trilete
miospores and named it ‘Microfossil Assemblage Zone 2’. This had been named the Ambitisporites
avitus assemblage by Richardson (1974) and was later termed the Ambitisporites dilutus - A. avitus
Assemblage Zone by Richardson and McGregor (1986). In the type Llandovery area this zone is
recorded from samples L21-L23 (late Aeronian sedgwickii Graptolite Biozone to early Telychian
turriculatus Graptolite Biozone).

The oldest specimens of Ambitisporites dilutus recorded from the late Aeronian ( sedgwickii
Graptolite Biozone) Wormwood Formation are the oldest known with precise dating. Their
recovery enables the reference section for the base of the Ambitisporites avitus- A. dilutus spore
Assemblage Zone of Richardson and McGregor (1986) to be defined as sample L21 (= locality 162
of Cocks et al. 1984) in the Type Aeronian/Telychian Boundary Section of the type Llandovery
area.

A review of the literature suggests that the appearance of true Ambitisporites (not Ambitisporitesl
vavrdovii) occurs at comparable horizons globally. For instance, elsewhere in the British Isles
Ambitisporites first occurs in the Aeronian Pentamerus Beds of the Welsh Borderland (Aldridge
et al. 1979), and the Telychian griestonensis Graptolite Biozone of Ireland (Emo and Smith 1978;
Smith 1981). Furthermore, in North America the incoming of Ambitisporites is placed at around the
C2-C3 boundary (Gray 1985; 1988) which is equivalent to slightly lower in the Aeronian sedgwickii
Graptolite Biozone than recorded in the type area. In North Africa Ambitisporites is also first
recorded in the Aeronian (Hofifmeister 1959; Hill et al. 1985; Richardson and McGregor 1986;
Richardson 1988).

PALAEOBOTANICAL IMPORTANCE

That cryptospores may provide evidence of pioneering land plants has been much debated (Gray
and Boucot 1971, 1977; Gray 1985, 1988; Taylor, 1982). However, as they have yet to be recovered
from the sporangia of a land-plant megafossil their derivation from this source remains unproven.
The best evidence for a land plant source comes from their abundant recovery in fluviatile sediments
of North America (Johnson 1985; Strother and Traverse 1979), and their similarity to spores
produced by certain extant hepatics (Gray 1985).

Trilete miospores, on the other hand, have been recovered from within the sporangia of the
rhyniophytoid land plant Cooksonia pertoni in the late Silurian (Lang 1937; Fanning et al. 1988).
Consequently, the Llandovery examples may be derived from a similar source and those from the
late Aeronian may provide the oldest evidence of such organisms.

In this study the oldest cuticles probably derived from land plants (Edwards 1986) were recovered
from the late Aeronian (PI. 2, fig. 14). They are apparently absent in the Rhuddanian, even though
these samples were the most nearshore investigated and sporomorphs were at their most abundant
and diverse. Other studies have reported cuticles in the middle Caradoc and later Ordovician (Gray

BURGESS: LLANDOVERY SPOROMORPHS

597

et al. 1982; Gray 1985; Vavrdova 1988). Internally thickened tubes (Burgess and Edwards in press)
were also not recorded from the type Llandovery area, even though Pratt et al. (1978) and Gray
(1985) found them in upper Llandovery sediments of the USA, and they are present in basal
Wenlock strata of the type Wenlock area (Burgess and Richardson 1991). Their absence may be
related to the fact that extremely distal sediments were sampled in the late Llandovery of the type
area. However, the facies in the early Rhuddanian are regarded as suitable for their recovery and
their absence at this level may be genuine.

In conclusion, this study provides some evidence to support the assertions of Gray (1985) and
Edwards and Burgess (1990), that a major change in the land-flora occurred in the late Aeronian.
This change was from a flora dominated by cryptospore-producing plants, to a flora dominated by
trilete miospore-producing plants. Moreover, this change appears to have occurred at approximately
the same stratigraphical horizon worldwide.

Acknowledgements. Neil Burgess acknowledges his NERC (CASE) award held between the Department of
Palaeontology, British Museum (Natural History) and the Department of Plant Science, University College,
Cardiff.

REFERENCES

aldridge, r. j., dorning, k. j., hill, p. j., Richardson, j. b. and siveter, d. j. 1979. Microfossil distribution
in the Silurian of Britain and Ireland. 433-438. In Harris, a. l., Holland, c. h. and leake, b. e. (eds). The
Caledonides of the British Isles - reviewed. Geological Society of London. Special Publication, 8, Scottish
Academic Press, Edinburgh.

bharadwaj, d. a. and venkatachala, b. s. 1961 . Spore assemblage out of a Lower Carboniferous shale from
Spitsbergen. Palaeobotanist , 10, 18M7.

burgess, n. d. 1987. Micro- and megafossils of land plants from the Silurian and Lower Devonian of the Anglo-
Welsh Basin. Unpublished PhD. Thesis, University of Wales.

and edwards, d. 1991. Classification of uppermost Ordovician to Lower Devonian tubular and

filamentous macerals from the Anglo-Welsh Basin. Botanical Journal of the Linnean Society , 106, 41-66.

and richardson, J. b. 1991. Silurian cryptospores and miospores from the type Wenlock area,

Shropshire, England. Palaeontology , 34, 601-628.

cocks, l. r. m., woodcock, n. h., rickards, r. b., temple, j. t. and lane, p. d. 1984. The Llandovery Series
of the Type Area. Bulletin of the British Museum (Natural History ), Geology, 38, 131-182.
cramer, F. H. 1968. Palynological microfossils of the Middle Silurian Maplewood Shale in New York. Revue
M icro p a lean t o l og ie , 11, 61-70.

— 1969. Possible implications for Silurian paleogeography from phytoplankton assemblages of the Rose
Hill and Tuscarora Formations of Pennsylvania. Journal of Paleontology , 43, 485-491.

— and diez de cramer, m. d. c. r. 1972. North American palynofacies and their spatial arrangement:
acritarchs. Palaeontographica , Abteilung B. 138, 107-180.

edwards, d. 1986. Dispersed cuticles of putative non-vascular plants from the Lower Devonian of Britain.
Botanical Journal of the Linnean Society, 93, 259-275.

— and burgess, n. d. 1990. Terrestrialization : plants. 60-64. In briggs, d. e. and crowther, p. r. (eds).
Palaeobiology : a synthesis. Blackwell Scientific, Oxford.

— feehan, J. and smith, d. g. 1983. A late Wenlock flora from Co. Tipperary, Ireland. Botanical Journal of
the Linnean Society , 86, 19-36.

emo, G. t. and smith, d. g. 1978. Palynological evidence for the age of the Lower Palaeozic rocks of Slieve
Aughty, Counties Clare and Galway. Proceedings of the Royal Irish Academy, 78, 281-292.
fanning, u., richardson, j. b. and edward, d. 1988. Cryptic evolution in an early land plant, evolutionary
trends in plants, 2, 13-24.

gray, j. 1985. The microfossil record of early land plants : advances in understanding of early terrestrialization,
1970-1984. Philosophical Transactions of the Roved Society of London, Series B, 309, 167-195.

1988. Land plant spores and the Ordovician-Silurian boundary. Bulletin of the British Museum (Natural
History), Geology, 43, 351-358.

— and boucot, a. j. 1971. Early Silurian spore tetrads from New York: earliest New World evidence for
vascular plants? Science, Washington, 173, 918-921.

598

PALAEONTOLOGY, VOLUME 34

— and boucot, a. j. 1977. Early vascular land plants: proof and conjecture. Lethaia , 10, 145-174.
massa, d. and boucot, a. j. 1982. Caradocian land plant microfossils from Libya. Geology , 10, 197-201.

— theron, j. n. and boucot, a. j. 1986. Age of the Cedarberg Formation, South Africa and early land plant
evolution. Geological Magazine , 123, 445-454.

— colbath, g. k., de faria, a., boucot, a. j. and rohr, d. m. 1985. Silurian age fossils from the Paleozoic
Parana Basin, southern Brazil. Geology , 13, 521-525.

grebe, h. 1971. Terminologie morphographique recommandee et methode de description des spores. CIMP
special volume 4. Microfossiles organique du Paleozoique. 7-34.
hill, P. R., Paris, F. and Richardson, j. B. 1985. Silurian palynomorphs. Journal of Micropalaeontology, 4,
27-48.

hoffmeister, w. s. 1959. Lower Silurian plant spores from Libya. Micropaleontology, 5, 331-334.

Johnson, n. G. 1985. Early Silurian palynomorphs from the Tuscarora Formation in central Pennsylvania and
their paleobotanical and geological significance. Review of Palaeobotany and Palynology, 45, 307-360.
lang, w. H. 1937. On the plant-remains from the Downtonian of England and Wales. Philosophical
Transactions of the Royal Society of London, Series B, 227, 245-291.
livermore, r. a., smith, a. G. and briden, j. c. 1985. Palaeomagnetic constraints on the distribution of the
continents in the late Silurian and early Devonian. Philosophical Transactions of the Royal Society of London,
Series B , 309, 29-56.

luber, a. a. and waltz, i. e. 1938. Klassifikatsiya i stratigraficheskoe zhachenie spor nekotorich
kamennougol’nich mestorojdemi SSSR. Trudy TSNIGRI, Moscow, 105, 1-45.
miller, M. A. and eames, l. e. 1982. Palynomorphs from the Silurian Medina Group (Lower Llandovery) of the
Niagara Gorge, Lewiston, New York, U.S.A. Palynology, 6, 221-254.
potonie, h. 1893. Die Flora des Rotliegenden von Thuringen. Abhandlungen der Koniglich Preussischen
Geologischen- Landesanstalt , 9, 1-298.

pratt, L. m., Phillips, t. l. and dennison, j. m. 1978. Evidence of non-vascular land plants from the Early
Silurian (Llandoverian) of Virginia, U.S.A. Review of Palaeobotany and Palynology, 25, 121-149.
reinch, p. f. 1891. Neue Untersuchungen iiber die Mikrostrukt Steinkohle des Carbon, Dyas und Trias.
T O. Weigel, Leipzig. 124 pp.

richardson, j. b. 1974. The stratigraphic utilisation of some Silurian and Devonian miospore species in the
northern hemisphere: an attempt at a synthesis. International Symposium on Belgian Micropalaeontological
Limits , Namur 1974, Publication no. 9, 1-13.

1985. Lower Palaeozoic sporomorphs : their stratigraphic distribution and possible affinities. Philosophical
Transactions of the Royal Society of London, Series B, 309, 201-205.

1988. Late Ordovician and Early Silurian cryptospores and miospores from northeast Libya. 89-109. In
el-arnauti. A., Owens, b. and thusu, b. (eds). Subsurface palynostratigraphy of northeast Libya. Garyounis
University Publications, Benghazi, Libya.

— and lister, t. r. 1969. Upper Silurian and Lower Devonian spore assemblages from the Welsh
Borderland and South Wales. Palaeontology, 12, 201-252.

— and mcgregor, d. c. 1986. Silurian and Devonian spore zones of the Old Red Sandstone Continent and
adjacent regions. Geological Survey of Canada, Bulletin, 364, 1-79.

— ford, j. h. and Parker, f. 1984. Miospores, correlation and age of some Scottish Lower Old Red
Sandstone sediments from the Strathmore region (Fife and Angus). Journal of Micropalaeontology , 3,
109-124.

smith, d. g. 1975. Wenlock plant spores and tetrads from County Mayo, Ireland. Geological Magazine, 112,
411-414.

— 1979. The distribution of trilete spores in Irish Silurian rocks. 423 — 43 1 . In Harris, a. l., Holland, c. h.
and leake, B. E. (eds). The Caledonides of the British Isles - reviewed. Geological Society of London. Special
Publication, 8, Scottish Academic Press, Edinburgh.

— 1981. Progress in Irish Lower Palaeozoic Palynology. Review of Palaeobotany and Palynology , 34,
137-148.

Strother, p. k. and traverse, a. 1979. Plant microfossils from Llandoverian and Wenlockian rocks of
Pennsylvania. Palynology, 3, 1-22.

taylor, t. n. 1982. The origin of land plants: a paleobotanical perspective. Taxon, 31, 155-177.
vavrdova, m. 1982. Recycled acritarchs in the uppermost Ordovician of Bohemia. Casopis pro Mineralogii a
Geologii, 27, 337-345.

— 1984. Some plant microfossils of possible terrestrial origin from the Ordovician of Central Bohemia.
Vestmk Ustredm'ho Ustavu Geologickeho, 59, 165-170. plates I —IV.

BURGESS: LLANDOVERY SPOROMORPHS

599

— 1988. Further acritarchs and terrestrial plant remains from the Late Ordovician at Hlasna Treban
(Czechoslovakia). Casopis pro Mineralogii a Geologii , 33, 1-10.
ziegler, a. m., cocks, l. r. m. and mckerrow, w. s. 1968. The Llandovery transgression of the Welsh
Borderland. Palaeontology , 11, 736-782.

N. D. BURGESS

Department of Palaeontology
British Museum (Natural History)
Cromwell Road
South Kensington
London SW7 5BD

Typescript received 2 December 1989
Revised typescript received 22 November 1990

Current address

Royal Society for the Protection of Birds
The Lodge

Sandy, Bedfordshire SG19 2DL

SILURIAN CRYPTOSPORES AND MIOSPORES
FROM THE TYPE WENLOCK AREA, SHROPSHIRE,

ENGLAND

by n. d. burgess and j. b. richardson

Abstract. The earliest occurrence of sculptured hilate cryptospores and miospores is near the base of the cf.
protophanus-verrucatus Sporomorph Zone, now more accurately located within the lundgreni Graptolite
Biozone in the type Wenlock area. Palynofacies studies indicate that this event is unrelated to changes in the
depositional environment. All the Sheinwoodian sporomorphs are laevigate (six species) and have either a
crassitate or patinate structure. The oldest known sculptured specimens of trilete miospores (two or three
species) and hilate, crassitate/patinate cryptospores (four species) appear almost synchronously in the
Homerian (upper part of lundgreni Graptolite Biozone). Additional sculptured taxa (two species) appear in the
later Homerian but there are no innovations in structure. In North Africa (Gondwana) a closely comparable
sequence of structural and sculptural events occurs. Three groups of sporomorphs described from the
Buildwas, Coalbrookdale and lower Much Wenlock Limestone Formations are permanently fused cryptospore
tetrads (one species), hilate cryptospores derived from dyads (eight species), and trilete miospore (six species).
Systematic descriptions of seventeen species in eight genera are provided. Two genera Laevolancis and
Artemopyra , three species Laevolancis plicata , Hispanaediscus wenlockensis, and Artemopyra brevicosta , and one
combination are new, and the genera Hispanaediscus and Dyadospora are emended.

Wenlock sporomorphs although known from only a few studies provide evidence for the existence
of early land plants (Gray 1985) and hold a largely untapped potential for stratigraphical correlation
(Richardson and McGregor 1986). Many of these assemblages contain mainly laevigate trilete
miospores showing little diversity and have been obtained from poorly dated strata (Smith 1975;
Colthurst and Smith 1977; Strother and Traverse 1979). Libyan assemblages are often superbly
preserved (Richardson and Ioannides 1973) but although the sequence of spore sculptural patterns
is closely similar to that in the Welsh Borderland, and the succession is dated by graptolites, there
are samples which may be either Lower Ludlow or Upper Wenlock based on the latter. In some of
these cases sporomorph data from the present study support a Wenlock age.

In contrast to North Africa, the geology of the type Wenlock area is well known (Bassett et al.
1975) and samples from higher parts of the sequence have yielded sporomorph assemblages of
increasing diversity, although frequently spores are comparatively rare (1-20% of assemblage).
Spores have been described from the type Wenlock area by Downie (1963), Richardson and Lister
(1969), and Mabillard and Aldridge (1985) and from Dudley by Doming (1983). Most previous
works consist solely of records of the miospore genus Ambitisporites, but Richardson and Lister
(1969) described eight miospore taxa, including some questionable miospores now regarded as
cryptospores, and sculptured tetrads in the Much Wenlock Limestone Formation (Wenlock
Limestone).

Sections in the type Wenlock area have been resampled (Text-fig. 1 ) to determine the diversity and
stratigraphical ranges of the sporomorph taxa and, as precisely as possible, the horizons where
sporomorph events occur, in particular the first appearance of sculpture in ‘naked' cryptospores
(i.e. those without envelopes) and in miospores. The advent of sculptural types is considered to have
biological as well as biostratigraphical significance (Richardson and Lister 1969; Gray 1985;

IPalaeontology, Vol. 34, Part 3, 1991, pp. 601-628, 2 pls.|

© The Palaeontological Association

602

PALAEONTOLOGY, VOLUME 34

text-fig. 1 . Lithostratigraphic and biostratigraphic position of samples collected from the type Wenlock area,

in relation to localities of Bassett et a/. (1975).

Richardson 1985; Fanning et al. 1988). Such studies assist in attempts to understand the
colonization of the land by plants because rocks of this age may contain abundant plant microfossils
but few megafossils (Richardson and Ioannides 1973; Edwards and Fanning 1985; Gray 1985).

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

603

SAMPLING AND TECHNIQUES

Reconnaissance sampling at Eaton Track and Harton Hollow Wood Quarry by Richardson and Lister (1969),
and later by the authors with the Ludlow Research Group, revealed laevigate and sculptured sporomorphs at
Graptohte locality 24 of Bassett et al. (1975), near the top of the Eaton Track Section. For this study a total
of forty eight samples were collected from six sections covering a stratigraphical interval of c. 240 m (Table 1 )
from basal Sheinwoodian (basal centrifugus Graptolite Biozone) to the late Homerian (lower part of the
ludensis Graptolite Biozone) but excluding the upper c. 20 m of the Much Wenlock Limestone Formation.
However, the uppermost part of the section has been sampled independently by both authors in the type
Ludlow Area and so far has not yielded any new taxa. Details on the stratigraphy of the type Wenlock Area
sections (except for Hughley Brook) are in Bassett et al. (1975).

table 1. Locality and horizon of the sections studied

Section

Grid reference

Formation

Hughley Brook c. 400 m
SW of Gippols Farm

SO 58009920

Buildwas

Ticklerton Brook

SO 48589042

Coalbrookdale

Rushbury

SO 51329173-

Coalbrookdale

Lakehouse Brook

SO 51329168
SO 51399164

Coalbrookdale

Rushbury Pack Track

SO 51369146-

Coalbrookdale

Eaton Track

SO 51679123
SO 50049001-

Coalbrookdale (Farley

SO 50509010

Member) -Much

Harton Hollow Wood

SO 48148788

Wenlock Limestone
Much Wenlock

Quarry

Limestone

Samples were all fine-grained buff to olive-green calcareous siltstones and silty limestones. Palynomorphs
were extracted and prepared for light microscope observation using standard techniques: HC1-HF-HC1 acid
treatment followed by zinc bromide solution (S.G. 2-0) heavy mineral separation. Organic residues were
washed with distilled water between HC1 and HF treatment to remove residual calcium ions and to reduce the
formation of calcium fluoride and related compounds during HF treatment. Residues were sieved through a
10 /um sieve, dried on coverslips, mounted in ‘Elvacite’ mounting medium, and studied in normal transmitted
light and Nomarski Interference using a Zeiss Photomicroscope. Photomicrographs were taken on Ilford FP4
film.

From each sample 30 g of rock was processed and a measured amount of the residue from a known mass
of rock mounted on a slide. Palynomorphs were then counted from a proportion of the slide and from this the
absolute percentage per gram of sample was calculated for prasinophycean ‘phycomata' (Tappan 1980) and
each acritarch, cryptospore and miospore group. Palynomorph counts were made under oil at x 1000.

SYSTEMATIC PALAEONTOLOGY

The last decade has seen a number of reports showing that Upper Ordovician and Silurian strata contain spore-
like microfossils in association with trilete miospores (Strother and Traverse 1979; Miller and Eames 1982;
Gray 1985; Johnson 1985; Hill et al. 1985; Richardson 1988). We group these atypical spores (cryptospores
sensu Richardson et al. 1984; Richardson 1988) into two major categories which are morphologically distinct
and which may reflect macroplant groups and their evolution. In the descriptions below the terminology of
Grebe (1971) is used to describe all sporomorphs, included cryptospores, but we have used the term muri for
sinuous, sometimes anastomosing, ridges, whether or not they form a reticulum. The stratigraphical
distribution refers to the range within this study.

Spore dimensions are given as the maximum diameter in polar compression unless stated otherwise, with the

604

PALAEONTOLOGY, VOLUME 34

minimum and maximum of the range presented and the mean in brackets. Figured specimens are stored in the
Palynology Section, Palaeontology Department, British Museum (Natural History), London and have the
prefix FM. Specimen co-ordinates are from a Zeiss Photomicroscope III no. 2562 housed in the Department.
All specimens are also located by means of standard England Finder co-ordinates, and are ringed using a
permanent red pen.

Anteturma cryptosporites Richardson et al., 1984
1. 1 Permanently ' fused cryptospore tetrads

This group comprises tetrads and dyads which are not found separated into their individual components and
which we therefore believe are ‘permanently’ fused. Such forms are abundant in the lower Llandovery where
well-preserved specimens are enclosed within variously sculptured envelopes (Strother and Traverse 1979;
Miller and Eames 1982; Gray 1985, 1988; Johnson 1985; Burgess 1987). Occasionally individual specimens
occur with poorly preserved proximal faces which may represent fragmented tetrads (Richardson 1988).

Genus tetrahedraletes Strother and Traverse, 1979
Type species. Tetrahedraletes medinensis Strother and Traverse, 1979.

Tetrahedraletes medinensis Strother and Traverse, 1979

Plate I , figs 12 and 13

Figured specimens. FM 156, PI. 1, fig. 12, slide S24/3, 040 1210; FM 157, PI. 1, fig. 13, slide S24/6 060 1240;
both specimens from sample G 5, Hughley Brook, lower Buildwas Formation.

EXPLANATION OF PLATE 1

All figures x 1000.

Figs 1 -3. Artemopyra brevicosta sp. nov. I. Dyad; FM 147 (slide GL22/I co-ord 060 1030; England Finder
no: F32/3/4), sample ET 7a, upper Coalbrookdale Formation, lundgreni Graptolite Biozone. 2, FM 148
(slide GL24/26 co-ord 200 1110; E.F. no: U40/4 U41/3), sample ET 7c, upper Coalbrookdale Formation,
nassa Graptolite Biozone. 3, FM 146, holotype (slide ET 9A/4 co-ord 165 1300; E.F. no: R60), sample ET
10, Farley Member, nassa Graptolite Biozone.

Figs 4-9. Hispanaediscus wenlockensis sp. nov. 4, FM 150 (slide GL24/23 co-ord 030 1263; E.F. no: C56/2),
sample ET 7c, upper Coalbrookdale Formation, nassa Graptolite Biozone. 5, Proximal view showing
radial muri;FM 151 (slide GL24/5 co-ord 130 1055; E.F. no: N35/3), sample ET 7c, upper Coalbrookdale
Formation, nassa Graptolite Biozone. 6, Distal view showing verrucae; FM 149 (slide GL24/13 co-ord
100 1290; E.F. no: K59/4), sample ET 7c, upper Coalbrookdale Formation, nassa Graptolite Biozone.
7, Dyad; FM 153 (slide ET 9A/1 co-ord 090 1365; E.F. no: J67/2), sample ET 10, Farley Member, nassa
Graptolite Biozone. 8, FM 152, holotype; proximal view (slide GL 24/24 co-ord 055 1205; E.F. no: E50/4
F50/2), sample ET 7c, upper Coalbrookdale Formation, nassa Graptolite Biozone. 9, Holotype, FM
152; distal view.

Figs 10 and I 1. Artemopyra sp. A. 10, Proximal surface with radial muri and murus marking boundary of
proximal hilum; FM 154 (slide ET 16/5 co-ord 078 1088; E.F. no: H38/2), sample ET 15, lower Much
Wenlock Limestone Formation, ludensis Graptolite Biozone. 11, Specimen showing muri on proximal
surface and smooth distal surface; FM 155 (slide ET 16/8 co-ord 144 1050; E.F. no: 034/4 035/3), sample
ET 15, lower Much Wenlock Limestone Formation, ludensis Graptolite Biozone.

Figs 12 and 13. Tetrahedraletes medinensis Strother and Traverse 1979. 12, FM 156 (slide S24/3 co-ord 040
1210; E.F. no: E53), sample G 5, Buildwas Formation, centrifugus Graptolite Biozone. 13, FM 157 (slide
S24/6 co-ord 060 1240; E.F. no: G54/1/3), sample G 5, Buildwas Formation, centrifugus Graptolite
Biozone.

Figs 14 and 15. hilate cryptospore type 1. 14, Sculptured distal surface; FM 158 (slide GL24/7 co-ord 100
1070; E.F. no: K36/4), sample ET 7c, upper Coalbrookdale Formation, nassa Graptolite Biozone. 15,
Smooth ‘proximal’ surface of FM 158.

PLATE 1

BURGESS and RICHARDSON, Wenlock cryptospores

606

PALAEONTOLOGY, VOLUME 34

Description. Laevigate obligate tetrahedral tetrads, subcircular to circular in outline, preserved in many
compressional forms relating to the degree of rotation from an apical view prior to compression. Within the
tetrads individual ‘spores’ have a subtriangular to subcircular equatorial outline, and unfused equatorial
crassitudes 1-5 /;m wide. Distal exine 1-2 pm thick and usually invaginated.

Dimensions. Tetrads 30(37)52 pm in diameter (100 specimens measured from sample G 5; Text-fig. 2).

text-fig. 2. Size frequency distribution of a hundred
Tetrahedraletes medinensis (Strother and Traverse,
1979) from sample G 5, Hughley Brook ( centrifugus
Graptolite Biozone); x = 37-3 pm.

Distribution. Throughout the Wenlock, basal Sheinwoodian to late Homerian Stages, centrifugus to ludensis
Graptolite Biozones.

Comparisons. Specimens of Tetrahedraletes medinensis recovered in the Wenlock (Text-fig. 2) are
larger than those from the Rhuddanian of the type Llandovery area (Burgess 1991) but are similar
in size to those from the Llandovery of North America (Pratt et al. 1978; Strother and Traverse
1979; Miller and Eames 1982; Johnson 1985).

2. Hilate cryptospores

Sporomorphs in this group are alete, hemispherical in equatorial view, have a thin proximal surface (hilum)
and an equatorial thickening. In many cases they are seen to be derived from loosely attached dyads and as
all have the same basic structure they are probably all derived from dyads.

Genus laevolancis gen. nov.

Type species. Laevolancis ( Archaeozonotriletes ) divellomedium (Chibrickova) comb. nov.

Diagnosis. Alete proximally hilate cryptospores, originally elliptical to hemispherical in equatorial
view with an equatorial to subequatorial crassitude surrounding the hilum; exine laevigate.

Derivation of name. Latin laevigatas, smooth; lands , plate or dish.

Remarks. These cryptospores are occasionally seen as loosely attached dyads.

Discussion. The genus Gneudnaspora Balme (1988) includes similar spores but is not used here
because the Australian spores may be trilete, monolete or alete. The Wenlock spores are all alete
and many, if not all, are derived from dyads. Consequently we have erected a new genus Laevolancis

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

607

to accommodate the alete forms in Balme’s genus and other alete species of similar structure with
a laevigate exine.

Laevolancis divellomedium (Chibrickova) comb. nov.

Plate 2, figs 4 and 6.

1959 Archaeozonotriletes divellomedium Chibrickova, p. 65, pi. 9, fig. 4.

1966 Hispanciediscus bernesgae Cramer, p. 82, pi. I, fig. I ; text-fig. 2, figs 2 and 1 1.

1968 Spore no. 2651, Magloire, pi. 1, fig. 6.

1969 ? Archaeozonotriletes cf. divellomedium Chibrickova; Richardson and Lister, p. 238, pi. 43, fig.

12.

1973 ?A. cf. divellomedium Chibrickova; Richardson and Ioannides, p. 280, pi. 8, figs 10 and 1 1.

1974 Zonaletes (?) divellomedium (Chibrickova); Arkhangelskaya, pi. 6, figs 3 and 4.

1974 ?A. divellomedium Chibrickova; McGregor, pi. I, figs 35 and 40.

1978 Hispanciediscus sp., McGregor and Narbonne, p. 1296, pi. 1, figs 20-22.

1979 ‘Smooth-walled inaperturate spore’, Strother and Traverse, p. 14, pi. 12, fig. 9.

1984 ?Stenozonotriletes irregularis McGregor, p. 37, pi. 1, fig. 26.

1986 Tholisporites divellomedium Turnau, p. 349, pi. 2, fig. 12.

1980 Zonaletes (?) divellomedium (Chibrickova) Arkhangelskaya, pi. 5, fig. 34.

1988 Gneudnaspora divellomedium (Chibrickova) Balme, p. 125 [partim], pi. 3, figs 1-7.

Holotvpe and type locality. Chibrickova, 1959, prep. 977, collections of the Gorno-Geological Institute, Bashkir
Filial, Academy of Sciences, USSR; Takata Beds, Emsian.

Figured specimens. FM 163, PI. 2, fig. 4, slide S24/3, 145 1290, sample G5, Hughley Brook, Buildwas
Formation; FM 164, PI. 2, fig. 6, slide ET 8/5, 133 1265, sample ET 6, Eaton Track, upper Coalbrookdale
Formation.

Diagnosis. A Laevolancis with a rigid wall and a 1 ( 1-5)4 pm wide subequatorial crassitude.

Description. Cryptospores occasionally preserved in loosely attached dyads in a variety of compressional
morphologies but are usually separated. Separated spores: amb subcircular, or occasionally circular, specimens
frequently tipped, originally roughly hemispherical in equatorial view with + flattened proximal polar area. An
equatorial to subequatorial crassitude surrounds the proximal hilum; hilum flattened to concave in equatorial
view, laevigate, occasionally folded, or ruptured, optically appears thinner than the distal exine; distal exine
laevigate c. 2 pm thick.

Dimensions. Maximum diameter 30(38)49 pm, minimum diameter 18(30)38 /an (30 specimens measured from
sample G 5, Flughley Brook, Buildwas Formation).

Distribution. Present and often common throughout the sequence, Buildwas to Much Wenlock Limestone
Formations; basal centrifugus to lower ludensis Graptolite Biozones.

Comparisons. The size range of Chibrickova’s specimens (35-50 pm; Chibrickova 1959) overlaps
with that of the Wenlock specimens. Balme’s specimens from the Devonian Gneudna Formation
(Balme 1988) also overlap but are mostly larger (44-68 pm) ; otherwise the alete forms described by
Balme are identical (J.B.R., examination of the Australian material). Hispanciediscus bernesgae
Cramer, 1966 is regarded as synonymous but the genus Hispanaediscus is restricted herein to
verrucate hilate cryptospores similar to the type species.

Laevolancis plicata sp. nov.

Plate 2, fig. 8

Holotype and type locality. FM 167; PI. 2, fig. 8; slide Rush S14/5, 170 1150; sample R 17, Rushbury Pack
Track (Text-fig. 1; Tabic 1), Coalbrookdale Formation, late Wenlock lundgreni Graptolite Biozone.

608

PALAEONTOLOGY, VOLUME 34

Diagnosis. A Laevolancis with thin, often folded, walls.

Derivation of name. Latin plicatus, folded.

Description. Cryptospores occasionally preserved in loosely attached, often partially separated, dyads but
spores are most frequently found separated as single grains. Amb sub-circular to circular; distally slightly
convex in equatorial view. Crassitude equatorial to sub-equatorial, 05-1 -5 pm wide, surrounds a proximal
hilum; proximal hilum usually concave, laevigate and frequently folded. Distal exine c. I pm thick, laevigate.

Dimensions. Maximum diameter 22(27)35 pm, minimum diameter 14(21)27 pm (30 specimens measured from
sample ET 7c); Eaton Track Section, Coalbrookdale Formation.

Distribution. Present throughout the sequence, Buildwas to Much Wenlock Limestone Formations; basal
centrifugus to lower ludensis Graptolite Biozones.

Comparisons. Laevolancis divellomedium (Chibrickova) comb. nov. is larger, more inflated distally
and has a thicker, more rigid wall (c. 2 pm). Many of these spores are separated from dyads of the
genus Dyadospora (Strother and Traverse, 1979) emend, and possibly solely from D. murusattenuata.

Genus hispanaediscus (Cramer) emend.

Type species. Hispanaediscus verrucatus Cramer, 1966.

Emended diagnosis. Alete proximally hilate cryptospores; originally elliptical to hemispherical in
equatorial view; equatorial to sub-equatorial crassitude surrounding the hilum. Flilum laevigate, or

EXPLANATION OF PLATE 2

All figures x 1000.

Figs 1-3, 5. Dyadospora murusdensa (Strother and Traverse) emend. 1, FM 159 (slide Tick Sl/7 co-ord 063
1209; E.F. no: F51/1/3), sample T 1, lower Coalbrookdale Formation, ellesae Graptolite Biozone. 2,
Partially separated specimen with pyrite damage, FM 161 (slide Rush S 10/8 co-ord 200 1204; E.F. no:
U50/4), sample R 14, upper Coalbrookdale Formation, lundgreni Graptolite Biozone. 3, FM 160 (slide ET
5A/5 co-ord 152 1053; E.F. no: P35/3), sample ET 3, upper Coalbrookdale Formation, lundgreni Graptolite
Biozone. 5, partially separated specimen; FM 162 (slide Rush S10/1, co-ord 070 1220; E.F. no:
G52/2/G53/1), sample R 14, upper Coalbrookdale Formation, lundgreni Graptolite Biozone.

Figs 4 and 6. Laevolancis divellomedium (Chibrickova) comb. nov. 4, FM 164, holotype (slide ET 8/5 co-ord
133 1265; E.F. no: N56/4/N57/3), sample ET 6, upper Coalbrookdale Formation, lundgreni Graptolite
Biozone. 6, FM 163 (slide S24/3 co-ord 145 1290; E.F. no: P59/2/P60/1 ), sample G 5, Buildwas Formation,
centrifugus Graptolite Biozone.

Figs 7 and 9. Dyadospora murusattenuata (Strother and Traverse) emend. 7, FM 165 (slide GL24/25 co-ord
098 1202; E.F. no: K50), sample ET 7c, upper Coalbrookdale Formation, nassa Graptolite Biozone. 9,
FM 166 (slide Rush S10/8 co-ord 153, 1204: E.F. no: P50), sample R 14, upper Coalbrookdale Formation,
lundgreni Graptolite Biozone.

Fig. 8. Laevolancis plicata sp. nov. FM 167, holotype (Rush S14/5 co-ord 170 1150: E.F. no: Q44), sample
R 17, upper Coalbrookdale Formation, lundgreni Graptolite Biozone.

Figs 10 and 11. cf. Hispanaediscus sp. A. 10, FM 168 (slide ET 16/5 co-ord 090 1 140; E.F. no: J44/1/2), sample
ET 15, lower Much Wenlock Limestone Formation, ludensis Graptolite Biozone. 1 1, FM 169 (slide ET 16/2
co-ord 160 1210; E.F. no: Q51/3), sample ET 15, lower Much Wenlock Limestone Formation, ludensis
Graptolite Biozone.

Figs 12 and 13. Hispanaediscus verrucatus (Cramer) emend. 12, FM 170 (slide Rush S14/5 co-ord 065 1040:
E.F. no: F33), sample R 17, upper Coalbrookdale Formation, lundgreni Graptolite Biozone. 13, FM 171
(slide ET 8/6 co-ord 017 1350; E.F. no: A65/2), sample ET 6, upper Coalbrookdale Formation, lundgreni
Graptolite Biozone.

PLATE 2

BURGESS and RICHARDSON, Wenlock cryptospores

610

PALAEONTOLOGY, VOLUME 34

with radial and/or randomly orientated muri/folds. Distal exine ornamented with verrucae and or
muri.

Comparison. Artemopyra gen. nov. has the same structure and radial proximal muri/folds, but is
distally laevigate, apiculate or spinose.

Remarks. Hilate cryptospores in Hispanaediscus are occasionally seen as loosely attached dyads.
This genus was originally erected by Cramer (1966) for alete palynomorphs with laevigate,
verrucate/murornate and foveolate sculpture. The type species is verrucate and we have restricted
the genus to forms with verrucate distal sculpture where the proximal exine is not markedly thinner
than the distal and included species with proximal sculpture.

Hispanaediscus verrucatus (Cramer) emend.

Plate 2, figs 12 and 1 3

1966 Hispanaediscus verrucatus Cramer, p. 82, pi. 2, fig. 7 [partint]. non fig. 2.

71969 cf. Synorisporites verrucatus Richardson and Lister, pi. 40, fig. 13.

1973 cf. Synorisporites verrucatus Richardson and Lister; Richardson and Ioannides, p. 278, [partial \
pi. 6, figs 17, 18, 20, non pi. 6, fig. 19.

Holotype. Cramer, 1966, pi. 2, fig. 7; F: 22354-5-2. San Pedro Formation, near Valporquero, Leon, northwest
Spain. Probably of late Silurian age.

Figured specimens. FM 170, PI. 2, fig. 12, slide Rush 13/14, 065 1040, sample R 16, Rushbury Pack Track,
Coalbrookdale Formation; FM 171, PL 2, fig. 13, slide ET 8/6, 017 1350, sample ET 6, Eaton Track, upper
Coalbrookdale Formation.

Emended diagnosis. An Hispanaediscus with a smooth, diaphanous hilum and verrucate - murornate
distal sculpture.

Description. Hilate cryptospores probably derived from loosely attached, partially separated, dyads, but all
specimens seen are single grains. Amb sub-circular to circular, originally distally hemispherical in equatorial
view and flattened proximally. Distal exine more or less convex, c. 1 pm thick, sculpture consists mainly of low
verrucae and/or muri 1-5 pm wide, c. I pm high and 0-5-3 //m apart. The groups of fused elements may
coalesce into longer occasionally convolute and anastomosing muri.

Dimensions. Maximum diameter 23(28)32 /an, minimum diameter 22(25)30 pm (15 specimens measured).

Distribution. Upper Coalbrookdale Formation, Farley Member, and Much Wenlock Limestone Formation;
upper lundgreni to lower ludensis Graptolite Biozones.

Comparisons. Some of Cramer’s specimens appear identical to those from the Wenlock of the type
area and Wenlock-Ludlow of North Africa. The spores from England, Spain and North Africa all
have size ranges less than 35 //m, and means of between 24 and 28 pm. H. wenlockensis is larger and
has proximal radial muri.

Remarks. Cramer’s material appears to include trilete spores with a more triangular outline,
similarly the material described by Richardson and Ioannides includes tetrads with comparable
sculpture. Herein H. verrucatus is restricted to alete hilate cryptospores with a circular or subcircular
amb.

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

611

Hispanaediscus wenlockensis sp. nov.

Plate 1, figs 4-9

Holotype and type locality. FM 152, PI. 1, figs 8 and 9, slide GL 24/24, 055 1205, sample ET 7c, Eaton Track
(Text-fig. I and Table 1 ), Coalbrookdale Formation, late Wenlock nassa Graptolite Biozone.

Paratypes. FM 149, PI. 1, fig. 6, GL24/13, 100 1290, sample ET 7c; FM 150, PI. 1, fig. 4, slide GL24/23, 030
1263, sample ET 7c; FM 151, PI. 1, fig. 5, slide GL24/5. 130 1055, sample ET 7c; FM 152, PI. I, fig. 8, slide
GL24/24, 055 1205, sample ET 7c, Coalbrookdale Formation; FM 153, PI. 1. fig. 7, ET 9A/1, 090 1365,
sample ET 8, Farley Member; all specimens from Eaton Track.

Diagnosis. An Hispanaediscus with short proximal radial muri and distal sculpture dominated by
verrucae which occasionally coalesce into muri.

Description. Cryptospores derived from loosely attached, partially separated, dyads (only I found, PI. 1, fig.
7) but most spores found separated as single grains. Amb sub-circular to circular; spores weakly convex distally
in equatorial view. Crassitude equatorial to sub-equatorial, 05(1)1 -5 pm wide, surrounds a proximal hilum.
Proximal hilum sculptured by closely packed radial muri, up to 7 //m long and c. 1 /mi wide which never reach
the proximal pole. Distal exine c. 2 pm thick, sculptured with closely packed verrucae usually appearing
hexagonal in plan view, elements 1(2)3 //m wide at the base, 11-5 /an high and 1 pm or less apart ; rare elements
composed of coalescent verrucae (muri) up to 17 pm long.

Dimensions. Maximum diameter 35(39)47 pm , minimum diameter 19(34)39 mm (27 specimens measured from
sample ET 7c, Coalbrookdale Formation, Eaton Track.

Distribution. Uppermost Coalbrookdale Formation and lower Farley Member, lower nassa Graptolite
Biozone.

Comparisons. Hispanaediscus verrucatus Cramer (1966) is smaller (c. 25 pm), laevigate proximally
and has more widely spaced distal verrucae.

cf. Hispanaediscus sp. A
Plate 2, figs 10 and 1 1

Figured specimens. FM 168, PI. 2, fig. 10, slide ET 16/5, 090 1 140, sample ET 15, Eaton Track, Much Wenlock
Limestone Formation; FM 169, PI. 2, fig. II. slide ET 16/2, 160 1210, sample ET 15, as above.

Description. Cryptospores seen only as separate grains. Amb circular, rarely subcircular; originally distally
hemispherical and ± flattened proximally. Proximal hilum thin, usually absent or fragmentary; sculptured with
low convolute to straight, radial muri or folds 0-5-1-5 /mi wide and 7-10 /mr long; muri/folds taper from
margin of hilum to proximal pole. Distal exine c. 2 pm thick at the distal pole and slightly thickened, up to
3 pm thick at the equator; most specimens sculptured with closely spaced, convolute and anastomosing muri,
rounded in profile, 0-5(2)4 pm wide, < 1-3 /mi apart and < 1 pm high; elements usually composed of
coalescent verrucae but may be more regular forming narrow irregular ridges or rugulae.

Dimensions. Maximum diameter 19(25)36 /mr, minimum diameter 17(22-5)29 //m (13 specimens measured from
samples ET 15, WLI and WL2).

Distribution. Upper Coalbrookdale Formation, Farley Member and Much Wenlock Limestone Formation;
lower nassa to lower ludensis Graptolite Biozones.

Comparison. Specimens of Cymbosporites verrucosus Richardson and Lister, 1969 are similar but
larger (29-52 /mr), some specimens have trilete folds or indistinct trilete marks, and the verrucae are
larger and do not form ridges, cf. Hispanaediscus sp. A has a thin proximal hilum, and therefore is

612

PALAEONTOLOGY, VOLUME 34

not typical of the genus Hispanaediscus. Too few specimens have been found to warrant further
taxonomic separation.

Genus artemopyra gen. nov.

Type species. Artemopyra brevicosta sp. nov.

Derivation of name. Greek artema. ear-ring, pyr = fire.

Diagnosis. Alete proximally hilate cryptospores; originally elliptical to hemispherical in equatorial
view. Proximal hilum sculptured with predominantly radial muri. Distal exine laevigate or
sculptured with grana, coni, biform elements or spinae.

Remarks. Specimens have been found in loosely attached dyads.

Artemopyra brevicosta sp. nov.

Plate 1, figs 1-3

1973 ?Emphanisporites cf. protophanus Richardson and Ioannides, pi. 2, fig. 7.

1979 ? ‘Spore with radial thickenings on one surface’, Strother and Traverse, p. 14.

Holotype and type locality. FM 146, PI. 1, fig. 3, slide ET 9A/4, 165 1300; sample ET 8, Eaton Track (see Text-
fig. 1 and Table 1), Farley Member, late Wenlock nassa Graptolite Biozone.

Paratypes. FM 147, PI. 1, fig. 1, slide GL 22/1, 060 1030, sample ET 7a; FM 148, PI. 1, fig. 2, slide GL 24/26,
200 1110, sample ET 7c; both Eaton Track, upper Coalbrookdale Formation.

Derivation of name. Latin brevis , short; costa , rib.

Diagnosis. An Artemopyra with simple proximal subequatorial radial muri < half spore radius in
length; distal exine laevigate.

Description. Proximally hilate cryptospores derived from loosely attached, partially separated, dyads; dyads
frequent (PI. 1, fig. 1), but most spores found separated as single grains. Amb circular to sub-circular,
cryptospores originally hemispherical with flattened proximal pole. A more or less equatorial crassitude,
distinct to indistinct, surrounds the hilum and and is roughly circular and concentric with the amb; hilum
sculptured with straight or slightly sinuous muri < 1(3)6 /mi long, < 1(1)2 /an maximum width and < 1-2 /mi
apart; muri taper polewards. Distal exine laevigate, c. F5 pm thick.

Dimensions. Maximum diameter 22(35)49 /mi. minimum diameter 21(28)44 pm (90 specimens measured from
sample ET 7c).

Distribution. Upper Coalbrookdale Formation, Farley Member, upper lundgreni to lower nassa Graptolite
Biozones.

Comparison. ?Emphanisporites cf. protophanus Richardson and Ioannides, 1973 from the Tannezuft
and Acacus Formations of western Libya (probably upper Wenlock, Ludlow and Downton (Pridoli)
equivalents) is slightly smaller but includes synonymous spores. Wenlock(?) material of Strother
and Traverse (1979) has been examined and contains specimens of A. brevicosta , including dyads.
Artemopyra sp. A has narrower, longer and more convolute radial muri.

Remarks. This species is distinctive and abundant and, in the type Wenlock Area, its first
appearance is a good indicator of the late Homerian.

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

613

Artemopyra sp. A
Plate 1, tigs 10 and 1 1

Figured specimen. FM 154, PI. 1, fig. 10, slide ET 16/5, 078 1088; FM 155, PI. 1. fig. 11, slide ET 16/8, 144
1050; both specimens from sample ET 15, Eaton Track, lower Much Wenlock Limestone Formation.

Description. Amb circular to sub-circular, cryptospores originally elliptical with flattened proximal pole.
Crassitude equatorial to sub-equatorial, forms a narrow ring c. 1 //m wide surrounding the proximal hilum.
Proximal hilum concave, sculptured with low, more or less radial muri and/or folds 2(4)7-5 pm long, c. 1 pm
wide, < 1-1 pm apart and < I pm high; muri taper gradually to the proximal pole, becoming more convolute
and anastomosing. Distal exine c. 1 pm thick and laevigatc.

Dimensions. Maximum diameter 28 pm , minimum diameter 23 pm (two specimens measured from samples ET
15 and WL1).

Distribution. Found in two samples. Much Wenlock Limestone Formation, lower ludensis Graptolite Biozone.
Comparisons. Artemopyra brevicosta sp. nov. has shorter, more robust, radial muri.

HILATE CRYPTOSPORE type 1

Plate 1 , figs 14 and 1 5

71966 Hispanaediscus leonensis Cramer, p. 83, pi. 2, fig. 8
71979 'Perforated palynomorph ', Strother and Traverse, p. 14.

Figured specimen. FM 158, PI. 1, figs 14 and 15, slide GL 24/7, 100 1070; sample ET 7c, Eaton Track,
Coalbrookdale Formation.

Description. Proximally hilate cryptospores. Amb circular to sub-circular; cryptospores originally elliptical
with flattened proximal pole. Crassitude equatorial to sub-equatorial, forms a narrow ring c. F5 //m wide
surrounding the proximal hilum which is concave and laevigate. Distal exine c. 1 -5 pm thick, foveolate;
foveolae regularly spaced c. 1 pm x 0-5 //m and c. 1 pm apart.

Dimensions. Maximum diameter 18(21)25 pm (2 specimens recorded in sample ET 7c).

Distribution. As for figured specimen.

Comparison. Hispanaediscus leonensis Cramer, 1966 has a similar structure and may be synonymous
with this type. However, the genus Hipanaediscus is used here for verrucate/murinate hilate
cryptospores and hence could not be used for this type. Material from Pennsylvania described by
Strother and Traverse (1979) has been re-examined by N.D. B. and contains specimens which are
closely comparable.

3. Dyads of hilate cryptospores

In rocks of Llandovery age cryptospore tetrads and dyads are common and in most cases the tetrad or dyad
condition is ‘permanent’. For these propagules separate genera have been created. In younger Silurian strata
intact dyads become increasingly rare although the hilate cryptospores derived from them are common. In this
work the separated cryptospores are named whereas usually their dyad counterparts are not. This parallels the
standard treatment of dispersed miospores and their tetrads. Because this is a gradational phenomenon the
hilate cryptospore genus Dyadospora Strother and Traverse, 1979 is retained, though emended, pending further
investigations.

Genus dyadospora Strother and Traverse emend.

Type species. Dyadospora murusattenuata Strother and Traverse, 1979 emend.

614

PALAEONTOLOGY, VOLUME 34

Emended diagnosis. Dyads composed of two laevigate hilate cryptospores. Individual spores not
strongly mutually attached, a clear line of separation is always seen; in equatorial compression the
two spores are often partially separated.

Remarks. The diagnosis of the genus Dyadospora has been restricted to include only non-
‘ permanent’ dyads which separate to give two laevigate hilate cryptospores. Pseudodyads,
characteristic of the early Silurian, are retained in the genus Pseudodyadospora Johnson, 1985. With
careful observation the two genera can be distinguished and their stratigraphical distribution and
biological significance assessed.

Dyadospora murusattenuata Strother and Traverse emend.

Plate 2, figs 7 and 9

1979 Dyadospora murusattenuata Strother and Traverse, p. 15, pi. 6, figs 8 and 10.

1982 Dyadospora murusattenuata Strother and Traverse; Miller and Eames, p. 247, pi. 6, fig. 8.

1985 Dyadospora murusattenuata Strother and Traverse; Johnson, p. 334.

Figured specimens. FM 165, PI. 2, fig. 7, slide GL24/25, 098 1202, sample ET 7c, Eaton Track; FM 166, Pi.
2, fig. 9, slide Rush S 1 0/8, 153 1204, sample R 14, Rushbury Pack Track, upper Coalbrookdale Formation.

Emended diagnosis. A Dyadospora with thin folded walls.

Description. Dyads, usually isomorphic, circular to subcircular in equatorial view, usually preserved in oblique
compression, each spore is distally convex; crassitudes at the point of contact between the two cryptospores;
a cleft is always present between the crassitudes of each spore. Distal exine laevigate, c. 1 /nn thick over the
distal pole and usually folded.

Dimensions. Dyad length 22(28)39 //m, equatorial width 20(27)37 //m (35 specimens measured) from sample ET
7c, Coalbrookdale Formation, upper Wenlock nassa Graptolite Biozone.

Distribution. Throughout the sequence, Buildwas to Much Wenlock Limestone Formations; lower
centrifugus to lower ludensis Graptolite Biozones.

Comparisons. The holotype of Dyadospora murusattenuata Strother and Traverse, 1979 (? Wenlock,
pi. 3, fig. 9) is closely similar to the specimens found in this study. In both cases the specimens are
thin-walled, folded and show partially separated dyads. Slides of the material from Pennsylvania
described by Strother and Traverse have been examined for comparison and contain abundant
dyads identical to those described herein. D. murusdensa Strother and Traverse, 1979 emend, is
larger and thicker- walled. Pseudodyadospora laevigata Johnson, 1985 is more elongate, and often
anisomorphic (i.e. the two parts of the dyad are of unequal size, Richardson 1988; Burgess 1991) and
the individual ‘spores’ are joined by a single fused thickening, or crassitude.

Remarks. Separation of these dyads is believed to produce spores of the species Laevolancis plicata
gen. et sp. nov.

Dyadospora murusdensa Strother and Traverse emend.

Plate 2, figs 1-3, 5

1979 Dyadospora murusdensa Strother and Traverse, p. 15, pi. 3, figs 6 and 7.

1982 cf. Dyadospora murusdensa Strother and Traverse; Miller and Eames, p. 247, pi. 6, fig. 7.

1985 cf. Dyadospora murusdensa Strother and Traverse; Johnson, p. 334, pi. 7, fig. 9.

1985 Dyadospora murusdensa Strother and Traverse; Richardson in Hill et al., p. 29, pi. 15, figs 8
and 9.

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

615

Figured specimens. FM 159, PI. 2, fig. 1, slide Tick 1/7, 063 1209; sample T 1, Ticklerton Brook, Coalbrookdale
Formation; FM 160, PI. 2, fig. 2, slide ET 5A/5, 152 1053, sample ET 3, Eaton Track, upper Coalbrookdale
Formation; FM 161, PI. 2, fig. 3, slide Rush S 1 0/8, 200 1204, sample R 14; FM 162, PI. 2, fig. 5, slide Rush
S10/1, 070 1220, sample R 14, Rushbury Pack Track, upper Coalbrookdale Formation.

Emended diagnosis. A Dyadospora with unfolded walls.

Description. Dyads, usually isomorphic, circular to subcircular in equatorial view, usually preserved in oblique
compression, each spore is distally convex; crassitudes 2(3)4 /an thick, equatorial to sub-equatorial, the two
cryptospores are usually loosely attached or almost separated (PI. 2, figs 2 and 5). In equatorial view the length
is equal to, or slightly greater than the width at the equator; exine laevigate, 1(2)3 //m thick over the distal pole
and usually without folds.

Dimensions. Dyad length 23(39)47 //m, equatorial diameter 21(34)46 pm (13 specimens measured from various
samples).

Comparison. The holotype of Dyadospora murusdensa Strother and Traverse (1979, pi. 3, fig. 9) is
closely similar to the dyads described above. Slides of material described by Strother and Traverse
from Pennsylvania have been examined for comparison and contain dyads identical to those
described here. Dyadospora murusattenuata Strother and Traverse emend, is thinner-walled and
generally smaller.

Anteturma sporites H. Potonie, 1893
Turma triletes Reinsch, 1891

Subturma zonotriletes Waltz, 1935 in Luber and Waltz (1938)
Infraturma crassiti Bharadwaj and Venkatachala, 1961
Genus ambitisporites Hoffmeister, 1959

Type species. Ambitisporites avitus Hoffmeister, 1959

Ambitisporites dilutus (Hoffmeister) Richardson and Lister, 1969

Text-fig. 3d-h

Figured specimens. FM 175, Text-fig. 3d, slide ET 16/2, 100 1205; sample ET 15, Eaton Track, Much Wenlock
Limestone Formation; FM 176, Text-fig. 3e, slide Rush Vill Sl/3, 182 1223, sample R 2, Rushbury, lower
Coalbrookdale Formation; FM 177, Text-fig. 3e, slide ET 8/5, 130 1240, sample ET 6; FM 178, Text-fig. 3g,
slide Rush Vill Sl/3, 180 1050, sample R 2; FM 179, Text-fig. 3h, slide ET 8/5, 123 1380, sample ET 6,
Rushbury and Eaton Track; specimens FM 177 and FM 179 are from the upper Coalbrookdale Formation
and FM 178 from the lower Coalbrookdale Formation.

Dimensions. Diameter 22(28)47 pm (100 specimens measured, see Text-fig. 4 for spread of measurements),
distal exine 1-2 //m thick.

Distribution. Common throughout the sequence. Buildwas to Much Wenlock Limestone Formations; basal
centrifugus to lower ludensis Graptolite Biozones.

Comparison. Ambitisporites avitus Hoffmeister, 1959 has a well-defined crassitude.

Remarks. Ambitisporites dilutus is the most abundant spore in the samples studied, up to 10,000
specimens per gram of sediment were present in some residues. The size-distribution of these spores

PALAEONTOLOGY, VOLUME 34

text-fig. 3. All figures x 1000. a-c. trilete miospore type 1. a, FM 172 (slide Rush S10/4 co-ord 135 1090;
E.F. no: 039/1 ), sample R 14, upper Coalbrookdale Formation, lundgreni Graptolite Biozone, b, FM 173 (slide
GL22/1 co-ord 063 1048 ; E.F. no : F34), sample ET 7a, upper Coalbrookdale Formation, lundgreni Graptolite
Biozone, c, FM 174 (slide ET 9A/1 co-ord 070 1145; E.F. no: G44/4), sample ET 8, upper Coalbrookdale
Formation, nassa Graptolite Biozone, d-h, Ambitisporites dilutus (Eloffmeister) Richardson and Lister,
1969. d, FM 175 (slide ET 16/2, co-ord 100 1205; E.F. no: K50/2), sample ET 15, lower Much Wenlock

616

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

617

shows a classic normal distribution (Text-fig. 4), perhaps indicating the spores were produced by a
single population of one or only a few species.

text-fig. 4. Size frequency distribution of
a hundred Ambitisporites dilutus (Hoffmeister)
Richardson and Lister, 1969 from sample R 2,
Rushbury (ellesae Graptolite Biozone); x =
27-9 //m.

in

z

UJ

I

0

UJ

a

(D

li

0

ffi

m

i

I

□

z

25 -
20 -

15 —

no —

_j

20

25 30 35

IVlSOBOfSBS

Genus synorisporites Richardson and Lister, 1969
Type species. Synorisporites downtonensis Richardson and Lister, 1969.

Synorisporites cf. S.l libycus Richardson and Ioannides, 1973
Text-fig. 3i, j

Figured specimens. FM 180, Text-fig. 3i, slide GL 22/1, 130 1030, sample ET 7a; FM 181, Text-fig. 3j, slide
ET 8/3 180 1264, sample ET 6, both samples from the Eaton Track, upper Coalbrookdale Formation.

Description. Amb sub-triangular with convex sides and rounded apices. Distally originally elliptical in
equatorial view with a flattened pyramidal proximal surface. Equatorial crassitude narrow, 0-5( 1-5)3 pm wide,
and may be indistinct. Contact surface laevigate. Distal exine c. l/nn thick, sculptured with low, rounded and
relatively broad muri and discrete verrucae or muri often composed of coalescent verrucae, 2(4)6 pm long.

Limestone Formation, ludensis Graptolite Biozone, e, FM 176 (slide Rush Vill Sl/3 co-ord 182 1223; E.F. no:
S54/5/T54/2), sample R 2, lower Coalbrookdale Formation, ellesae Graptolite Biozone, f, FM 177 (slide ET
8/5, co-ord 130 1240; E.F. no: N54/4/054/2), sample ET 6, upper Coalbrookdale Formation, lundgreni
Graptolite Biozone. G, FM 178 (slide Rush Vill Sl/3 co-ord 180 1050; E.F. no: T35/2), sample R2, lower
Coalbrookdale Formation, ellesae Graptolite Biozone, h, FM 179 (slide ET 8/5 co-ord 123 1380; E.F. no:
N69/1), sample ET 6, upper Coalbrookdale Formation, lundgreni Graptolite Biozone. I, J. Synosporites cf. 5.7
libycus Richardson and Ioannides, 1973. i. Tetrad ; FM 180 (slide GL22/1 co-ord 130 1030; E.F. no : 032/2/4),
sample ET 7a, upper Coalbrookdale Formation, ludgreni Graptolite Biozone, j, isolated spore showing trilete
mark; FM 181 (slide ET 8/3, 180 1264; E.F. no: S56), sample ET 6, upper Coalbrookdale Formation,
lundgreni Graptolite Biozone. R. Archaeozonotriletes chulus (Cramer) var. nanus Richardson and Lister, 1969.
FM 182 (slide S29/3 co-ord 040 0950; E.F. no: C23/2/C24/1 ), sample G 9, Buildwas Formation, centrifugus
Graptolite Biozone. l. Archaeozonotriletes chulus (Cramer) var. chulus Richardson and Lister, 1969. FM 183
(slide ET 8/3 co-ord 225 1390; E.F. no: X69/X70), sample ET 6, upper Coalbrookdale Formation, lundgreni
Graptolite Biozone. m-o. murornate tetrads, m, FM 184 (slide ET 7/4 co-ord 150 1087; E.F. no: P38/4),
sample ET 5, upper Coalbrookdale Formation, lundgreni Graptolite Biozone, n, FM 185 (slide ET 15/1 157
1312; E.F. no: Q61/2), sample ET 14, Farley Member, ludensis Graptolite Biozone, o, FM 186 (slide WL 2/5
co-ord 006 1364: E.F. no: A66), sample WL 2, lower Much Wenlock Limestone Formation, ludensis

Graptolite Biozone.

618

PALAEONTOLOGY, VOLUME 34

1(2)5 //m wide, < I //m high and 0-5(1 )4 //m apart. Trilete mark distinct, with straight to slightly sinuous
sutures which extend to the equator or nearly so, and are accompanied by low lips, c. 1 pm high and wide.

Dimensions. Maximum diameter 18(26)32 //m (4 specimens measured).

Distribution. Upper Coalbrookdale Formation and lower Farley Member, upper lundgreni to lower nassa
Graptolite Biozones.

Comparisons. Synorisporites? libycus Richardson and Ioannides, 1973 is of similar size (19^12 pm)
but includes some spores with broader verrucae (2-8 /nn). S. verrucatus Richardson and Lister, 1969
from the lower Downton Group (= Pridoli age) has smaller and more prominent verrucae.

TRILETE MIOSPORE type 1
Text-fig. 3a-c

Figured specimens. FM 172, Text-fig. 3a, slide Rush S 1 0/4 135 1090, sample R 14, Rushbury Pack Track, upper
Coalbrookdale Formation; FM 173, Text-fig. 3b, slide GL22/1 063 1048, sample ET 7a, Eaton Track, upper
Coalbrookdale Formation; FM 174, Text-fig. 3c, slide ET 9A/1, 070 1 145, sample ET 8, Eaton Track, Farley
Member.

Description. Amb sub-triangular with convex sides and rounded apices. Equatorial crassitude narrow, 0 5-
3-5 pm wide. Proximal surface sculptured with convolute and anastomosing muri c. 1 pm wide and 0-5 /an
apart; muri are radially aligned near the equator but are predominantly randomly orientated over the
remainder of the proximal surface. Distal exine c. 1 //m thick; laevigate in two specimens, and with scattered
grana in one specimen.

Dimensions. Maximum diameter 28(35)40 pm (three specimens measured from samples ET 7c and ET 8).
Distribution. Upper Coalbrookdale Formation to lower Farley Member. Lower nassa Graptolite Biozone.

Comparisons. Emphanisporites McGregor, 1961 has proximal muri which are predominantly radial
whereas in Trilete Miospore type 1 the muri are mainly randomly oriented.

Infraturma patinati (Butterworth and Williams) emend. Smith and Butterworth, 1967
Genus archaeozonotriletes (Naumova) Allen, 1965

Type species. Archaeozonotriletes variabilis (Naumova) Allen, 1965.

Archaeozonotriletes chains (Cramer) var. chulus Richardson and Lister, 1969

Text-fig. 3l

Figured specimen. FM 183, Text-fig. 3l, slide ET 8/3, 225 1390, sample ET 6, Eaton Track, upper
Coalbrookdale Formation.

Dimensions. Diameter 36(39)43 /zm (17 specimens measured). Equatorial exine width 2-6 //m; exine 1-5-4 pm
thick at distal pole.

Distribution. Throughout the Wenlock, Buildwas to Much Wenlock Limestone Formations; basal centrifugus
to lower ludensis Graptolite Biozones.

Comparisons. The genus Archaeozonotriletes is preferred for these spores because the distal patina
is either the same as, or similar in thickness to, the equatorial thickness (Table 2). Tlio/isporites

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

619

table 2. Exine thickness in spores of the Archaeozonotriletes chulus var. chulus type from the Welsh Borderland
(Richardson and Lister 1969).

Distal

pole

Equator

Inter-radial

Radial

Wenlock

1-5-4 //m

2-6 //m

?

Downton

4-5 pm

4-6 //m

3-6 /<m

Ditton

2-0-2- 5 pm

2-5-5 pm (3-5— 4-5)

?

Butterworth and Williams, 1958 has a 'Patina thickest in the equatorial region, thinning slightly
towards the distal pole.’ Although Smith and Butterworth (1967) report a distal thickness for
T. scoticus (the type species) ‘usually less than the equatorial thickness’, they compare Tholisporites
with a cingulate genus Densosporites. Further, A. chulus (Cramer) is elliptical in equatorial view and
usually preserved in polar compression whereas the type species of both Tholisporites and
Archaeozonotiletes are more inflated, hemispherical in lateral view, and commonly compressed in
lateral orientation (Allen 1965, p. 721 ; Smith and Butterworth 1967, p. 268). Thus A. chulus differs
in basic shape/structure from both the genera with which it has been compared. The genus
Archaeozonotriletes has priority over Tholisporites and it remains to be seen whether these two
genera can be separated on a satisfactory basis and whether spores with a less inflated distal
hemisphere should be accommodated in a new genus.

Remarks. JBR is a member of an international working group looking at species of the chulus type
and pending its results we prefer to leave this species in the genus Archaeozonotriletes.

Archaeozonotriletes chulus (Cramer) var. nanus Richardson and Lister, 1969

Text-fig. 3k

Figured specimen. FM 182, Text-fig. 3k, slide S29/3, 040 0950, sample G 9, Hughley Brook, Buildwas
Formation.

Dimensions. Diameter 21 -5(28)35 pm (14 specimens measured from various samples); equatorial exine width
2-6 /mi; exine 1-5-4 pm at distal pole.

Distribution. Throughout the Wenlock, Buildwas to Much Wenlock Limestone Formations; basal centrifugus
to lower ludensis Graptolite Biozones.

Comparison. A bimodal size distribution was demonstrated for the two varieties of A. chulus by
Richardson and Lister (1969) for Upper Silurian (Downtonian) and Lower Devonian (Dittonian)
specimens. The present work shows that in the lower Wenlock (Buildwas Formation) A. chulus var.
nanus is dominant whereas A. chulus var. chulus dominates higher in the Wenlock.

INCERTAE SEDIS
MURORNATE TETRADS

Text-fig. 3m-o

Figured specimen. FM 184, Text-fig. 3m, slide ET 7/4, 150 1087, sample ET 5, Eaton Track, upper
Coalbrookdale Formation ; FM 185, Text-fig. 3n, slide ET 15/1, 157 1312, sample ET 14, Eaton Track, Farley
Member; FM 186, Text-fig. 3o, slide WL2/5, 006 1364, sample WL 2, Harton Hollow Wood Quarry, lower
Much Wenlock Limestone Formation.

620

PALAEONTOLOGY, VOLUME 34

Description. Tetrads of closely adherent spores with murornate sculpture. Individual spores with triangular to
subtriangular amb. Proximal surface not seen. Distal exine sculptured with elongate, narrow (05( 1)1-5 /nu
wide), sinuous to convolute and anastomosing, low (< I pm high) and rounded muri; elements discrete but
occasionally closely spaced, and fused, < 05-3 pm apart.

Dimensions. Tetrads 31(38)50 /<m in diameter; individual spores 21-5(31)36 //m in width (20 specimens
measured).

Distribution. Upper Coalbrookdale Formation, Farley Member, and Much Wenlock Limestone; nassa to
lower ludensis Graptolite Biozones.

Comparison. Nodospora rugosa Strother and Traverse, 1979 from the Llandovery of North America
is a permanent cryptospore tetrad with superficially similar sculpture. However, in N. rugosa this
sculpture is present on an envelope surrounding the cryptospore tetrad whereas in this species the
spore exine is sculptured. Tetrads of rugulate trilete miospores would also appear similar, but one
might expect to recover isolated miospores in the same samples; this has not occurred with this
species. We are therefore unsure if these tetrads are cryptospores or the tetrads of trilete miospores
and hence have included them under Incertae sedis pending further investigations.

COMPARISONS WITH OTHER ASSEMBLAGES

Most records of early sporomorphs and plant megafossils have come from Baltica (Britain,
Scandinavia and European USSR, Livermore et al. 1985). Other sporomorph records for the
Silurian come from Laurasia (USA) and Gondwana (North Africa and South America) but so far
there are none from Australia, Siberia, Antarctica or China. Even fewer of these assemblages are
represented by diverse microfloras (see Table 3).

BIOSTRATIGRAPHY

A spore zonal scheme for the Silurian and Devonian of the northern hemisphere has been proposed
by Richardson and McGregor (1986). In this inter-regional scheme Wenlock strata encompass two
assemblage biozones: the chulus-nanus (Late Llandovery to Late Wenlock, Telychian to Late
Homerian) and cf. protophanus-verrucatus (Late Wenlock and Early Ludlow, Late Homerian to
Early Gorstian). In the Wenlock of the type area characteristic species of the cf. protophanus-
verrucatus Biozone first appear in upper lundgreni Graptolite Biozone of the Eaton Track and
Rushbury Pack Track sections (Text-fig. 5). Thus all the Sheinwoodian and lowermost Homerian
probably belong to the chulus-nanus Biozone. The nominal species of the upper zone are the hilate
cryptospores, lEmphanisporites cf. protophanus ( = Artemopyra hrevicosta of this work) and cf.
Synorisporites verrucatus (= Hispanaediscus verrucatus of this work). As a result of these new data,
the base of the zone is redefined in the Eaton Track Section.

cf. protophanus-verrucatus Assemblage Biozone.

The new reference section for the base of this zone is defined in the Upper Coalbrookdale
Formation of the type area of the Wenlock Series. The base is taken as sample ET 5 (approximately
midway between localities 20 and 21 of Bassett et al. 1975) in the Eaton Track Section, of upper
lundgreni Graptolite Biozone age. The nominal species lEmphanisporites cf. protophanus ( =
Artemopyra hrevicosta ) is rare at this horizon but both nominal species occur c. 4 m above in sample
ET 6 (approximately locality 21 of Bassett et al. 1975) and become more common higher in the
lundgreni Graptolite Biozone, sample ET 7a (locality 22 of Bassett et al. 1975), and above in the
Eaton Track Section.

Remarks. In the Rushbury Pack Track, the nominal species cf. Synorisporites verrucatus ( =

North Africa North

Great Britain and Ireland (Baltica) (Gondwana) America

(Laurentia)

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

621

o

a

cn

c

.5

•3

o

o

c

'5

.c

TO

C

O

C

.2

'E

<u

£

o

X

+ +I I I I I 1+ l + l I + I+ +

+ + I I I I I I I I I I I L. I I I

+ + I I I I I I + I II I II I! I

+ + + + +II++I + II + III II

+ + I I I I I I I I I I I I

+ + + + I I II II

+ + + +I+II+IIII + III II

Z

<

5

w

S

Off I I i i i i + I I I I I I I I II

z

<

5

o

o

£

w

X

GO

+ 1111 II II

+ + I I

I I + I I I I I I I I II ++II+II II

+ + + + + + + I + + + + + + + + + + + + + + + + + + + +

O-

Czj

Co

*S ^3

^ <3

• T3

o
a ‘

<D

a

o

a

II

.2 6.

G -s

<D 2

35 "2s

i CO' 2 H fq

o

< E.
& h

X Jb

"a «

S "5

Xl 8*. | | S s

g ^ 'tj I Q’Q (5

O

52

^3

^ s

2 S

5 >

N ^

« S

"P ^

"Rs

~5

c 2
-2 «
g ^

Z

o

Si

a

~2

ca
Q

'sO ON

GO

$2^
2 Ux

Cd *“7

H ^

•- 73 ^
j cffl

Z. <u

^ b £
C QJ ■*—>

o .c —
oo c
S3 O

O i- -t
T3 5/3 'C

1— , QJ

2 o oi
— — S
.2 — £
Pi +d U

in oo >,
s — - V-0
. _ On

n , Cd

°o c a
2 .1 o

« if §
^ S-2

'c S* 03

ii 03 >

a

■§ § °
!3 -o a!

T3 ■£ £

U 11 ^

.2 £ £

— « <u

z tZ a

,2 o> >>
^fc-52

-I 2
1)

'TT — >

^ as

2 <L> +
^ .tj H
*cT J2 'O

— b/j C

oo 03 03

”, ^ (U
JZ

>DS) oo ■*->

.£ r s

£ +r55

5 r> ^

ON

cn r^

>S ’£

^ c

a> h-

g ^
£ c

o ^

(N

bXj

C

cd

<— i +

- O .QJ

o '-3

Dh Z5

& g

C/D

go cd

'O 3

3 £

<55 (55

H— •

622

i — | | Much i --

Farley

I 1 Coalbrookdale

III Limestone _ _

Member

Formation

Build was
Formation

text-fig. 5. Biostratigraphic and lithostratigraphic occurrence of sporomorphs in the type Wenlock area. The
base of the cf. protophanus-ven ucatus Zone occurs below samples R17 (Rushbury Pack Track Section) and ET 5
(Eaton Track Section). Note that the top c. 20 m of the Wenlock Limestone was not sampled. P = occurrence

of only one specimen.

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

623

Hispanaediscus verrucatus ) occurs in sample R17 (approximately locality 44 of Bassett et al. 1975).
This sample is also from the upper lundgreni Graptolite Biozone and is probably at a similar horizon
to the Eaton Track Reference Section occurrences, but may be slightly lower. The Eaton Track is
preferred as the zonal reference section because the spore record is better there. A single specimen
of Trilete Miospore type 1 was also found in sample R14, approximately equivalent to locality 38
of Bassett et al. (1975) in the upper lundgreni Graptolite Biozone. This specimen represents the
oldest sculptured miospore dated independently by graptolites.

Further zonal refinement may be possible, at least in the Anglo-Welsh Area. For example, the
presence of Artemopyra brevicosta appears to be a good, though possibly only local, marker for the
upper lundgreni to lower nassa Graptolite Biozones in that it has not yet been found from the later
nassa or ludensis Biozones. Also, cf. Hispanaediscus sp. A was not recorded in samples from the
upper lundgreni Graptolite Biozone but is increasingly common in samples from the nassa and lower
ludensis Biozones. Thus with further work this species may prove to be a useful indicator for the
later Homerian. In the Libyan borehole sequences the miospore Emphanisporites protophanus
appears higher in the sequence than ? Emphanisporites cf. protophanus (= Artemopyra brevicosta)
but has not been found in the type Wenlock sequence examined. The first appearance of E.
protophanus is placed in the overlying libycus-poecilomorphus Miospore Biozone by Richardson and
McGregor (1986) and may prove to be an indicator of the latest Homerian or earliest Gorstian.

PALYNOFACIES

The appearance of sculpture on hilate cryptospores and trilete miospores is an event of
biostratigraphical, biological and evolutionary significance. This event is useful for interregional
stratigraphical correlation, and may relate to the primary diversification of land plants
(rhyniophytoids). Consequently it is important to ascertain if the appearance of sculptured
sporomorphs in the late lundgreni Graptolite Biozone of the type Wenlock area is a true
‘evolutionary event’ representing the first appearance of sculptured miospores and hilate
cryptospores, or whether it is a preservational or depositional artifact. Several aspects of marine
environments may be determined by studying the types and abundance of marine phytoplankton
on the one hand and amounts of land-derived sporomorphs on the other. To this end organic-walled
microphytoplankton (prasinophycean cysts and acritarchs), cryptospores and miospores, and
organic-walled tubes (Burgess and Edwards in press) were counted from all samples, and used to
produce a palynofacies profile of the sequence (Text-fig. 6).

The assemblages of palynomorphs and fragments of plant megafossils recognized fall into four
distinct palynofacies, which can be related to changes in the depositional environment and the first
appearance of sculptured sporomorphs.

Palynofacies 4 was recorded from samples of the Buildwas Formation in the Hughley Brook
Section. The predominant palynomorphs are acanthomorph acritarchs (15-55 %) but there are also
significant numbers of hilate cryptospores (5-10%) and tubes (up to 15%); trilete spores are rare
( < 1 %). Using existing palynofacies models, the Buildwas Formation samples appear to have been
deposited in an offshore shelf environment with significant terrestrial input (Doming 1981u; Al-
Ameri 1983; Burgess 1987; Richardson and Rasul 1990). Sculptured palynomorphs were not
recovered from this palynofacies.

Palynofacies 3 was recorded from samples of the Coalbrookdale Formation from sections
around Rushbury. The assemblages are composed predominantly of prasinophyceae (sphaero-
morph acritarchs) (30-55%), trilete miospores (up to 20%) and, near the top of the section, hilate
cryptospores (up to 20%). Studies of Recent Sediments (Muller 1959) show that the proportion
of land derived sporomorphs normally increases shorewards. In addition the proportion of
prasinophycean ‘phycomata’ increases relative to acritarchs in inshore environments (Richardson
and Rasul 1990). Thus the palynofacies indicates a shoreward shift in the environment of
deposition. In contrast macrofossil studies (Bassett and Cocks pers. comm.) indicate that these beds
were deposited in offshore shelf environments. Despite the common occurrence of sporomorphs in

text-fig, 6. Composition of palynological assemblages (%) through the type Wenlock
stratigraphical distribution of four palynofacies.

EB3

□

624

PALAEONTOLOGY, VOLUME 34

Much

- -

Farley

Wenlock

Member

jz

Limestone

Build was
Formation

text-fig. 6. Composition of palynological assemblages (%) through the type Wenlock area and the

stratigraphical distribution of four palynofacies.

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

625

this palynofacies (up to 25% of the assemblage) sculptured specimens are rare with only two
specimens being recovered towards the top of the sequence (Text-fig. 5).

Palynofacies 2 typifies the Coalbrookdale Formation of the Eaton Track Section and is
characterized by a dominance of sphaeromorph acritarchs (50-90%). Acanthomorph acritarchs are
considerably less abundant (5-25%) and sporomorphs are relatively rare (0-2%). Using our
palynofacies model, this palynofacies represents the most distal phase of sedimentation in the type
area and probably the deepest water, but it is within this part of the succession that sculptured
sporomorphs become increasingly common (though < 1 % of the assemblage).

Palynofacies 1 typifies the upper part of the Coalbrookdale Formation, Farley Member and
lower Much Wenlock Limestone Formation of the Eaton Track Section and Harton Hollow Wood
Quarry. This palynofacies is dominated by acanthomorph acritarchs (40-60 %), with prasinphycean
‘phycomata' (20-30%) and increased percentages of sporomorphs (0-3%) and tubes (5-10%).
This palynofacies indicates an outer shelf environment and nearer to the shore than that indicated
by Palynofacies 2. Sculptured sporomorphs become increasingly diverse through this part of the
sequence but are still rare (< 1 % of the total assemblage).

These palynofacies studies show that sporomorphs are most abundant in the late Sheinwoodian
and early Homerian (ellesae to late lundgreni Graptolite Biozones) where they were all laevigate.
The earliest sculptured sporomorphs found are from the late lundgreni Zone where the palynofacies
indicates deposition in offshore, deeper water, environments and where sporomorphs are rare and
generally form < 1 % of the assemblage. Sculptured sporomorphs become increasingly diverse in
samples from the upper most lundgreni to lowermost ludensis Zones, where palynofacies studies
indicate deposition in slightly more proximal, but still offshore, shelf environments and
sporomorphs form only around 1 % of the assemblage. In conclusion, the appearance of sculptured
sporomorphs in the late lundgreni Graptolite Biozone of the Wenlock Type Area does not appear
to have been caused by a change in the depositional environment to a more proximal one, but
appears to be a true event. Sediments from inshore environments would yield larger numbers of
sporomorphs but would not be expected to affect the horizon of this event substantially.

DISCUSSION AND PALAEOBOTANICAL SIGNIFICANCE

Trilete miospores of the genera Ambitisporites and Synorisporites have been recorded in situ from
specimens of Cooksonia pertoni in the Upper Silurian (Fanning et al. 1988). These spores are
comparable to some of the dispersed miospores described above and indicate that similar
rhyniophytoid plants may have been present in the Wenlock. The increasing diversity of trilete
miospores noted from the late lundgreni Graptolite Biozone, especially the appearance of sculpture,
strongly suggests that such plants began to diversify in the Homerian. The studies of Fanning et al.
(1988) on Cooksonia pertoni indicate that some of these plants, although they produced both
laevigate and sculptured spores, may have looked remarkably similar and the main differences
between them may have related to characters of their spores.

The derivation of hilate cryptospores is poorly known. However, smooth-walled hilate
cryptospores have been recorded from Salopella-Wkt sporangia in the Lower Devonian (Fanning
et al. in press) and the distribution of these spores shows that they were undoubtedly produced by
land-plants. Hilate cryptospores are found throughout the Wenlock of the type area but the first
sculptured species occur in the upper lundgreni Graptolite Biozone, at a similar horizon to the first
sculptured trilete miospores. Thus the evidence from the hilate cryptospores also suggests that the
late Wenlock was a time of major importance in the diversification of early land-plants.

Only one species of ‘permanent ’ cryptospore tetrad ( Tetrahedraletes medinensis) was recovered
from the type Wenlock area. It ranges throughout the sequence and represents the youngest
published record of the group (Gray 1985). They are also present in strata from the Rumney Inlier
of South Wales which ranges from late Wenlock through to the early Downtonian (= Pridoli)
(Burgess 1987), and J.B.R. (unpublished) has recorded them in studies of the Downtonian and
Dittonian of the Welsh Borderland. However, higher in the sequence they are erratic in their

626

PALAEONTOLOGY, VOLUME 34

distribution, usually rare, and therefore may be reworked. The type(s) of plants producing these
enigmatic palynomorphs are unknown although Gray (1985) has suggested that they may have been
similar to some extant hepatics. The nature of the record of ‘permanent’ cryptospore tetrads in this
study indicates that their parent plants were still extant in the Wenlock although their diversity had
greatly declined when compared with early Llandovery assemblages (Burgess 1991).

Smooth-walled and internally thickened tubes (Burgess and Edwards in press), which may
represent the remains of nematophytalean land plants have also been recovered throughout the
sequence in the type Wenlock area. In addition, unornamented cuticles {sensu Edwards 1982),
believed to form part of the nematophytalean land-plant Nematothallus Lang, also occur
throughout the succession. The Wenlock records suggest to us that nematophytalean plants and
possible nematophytalean, tube-bearing plants were widespread at this time but the nature of any
contained spores remains unknown.

Acknowledgements. Neil Burgess gratefully acknowledges the receipt of an NERC (CASE) AWARD held
jointly at the British Museum (Natural History) and University College, Cardiff. Both authors thank Mrs
Caroline Bell for assistance with the diagrams and the plates. The Photographic Unit (BM[NHj) and Ms Jan
Crawley (Cardiff) are thanked for photographic assistance.

REFERENCES

al-ameri, t. K. 1983. Acid-resistant microfossils used in the determination of Palaeozoic palaeoenvironments.

Palaeogeography , Palaeoclimatology , Palaeoecology. 44, 103-116.
allen, k. c. 1965. Lower and Middle Devonian spores of north and central Vestspitzbergen. Palaeontology ,
8, 687-748.

Arkhangelskaya, A. D. 1976. Basis for Eifelian age of the Periplecotriletes tortus Zone in central districts of
the European part of the USSR. 39-66. In byvsheva, t. v. (ed.). Results of palynological research on
Precambrian , Palaeozoic and Mesozoic of the USSR. Trudy Vsesoiuznogo Nauchno-Issledovatel'skogo
Geologorazvedochnogo Neftianogo Instituta, Kama Branch, no. 192. [In Russian],

1980. Plant spores from some Lower Devonian sections of the western regions of the Russian Plate.
26-46. In byvsheva, T. V. (ed. ). Palynological research in the Proterozoic and Phanerozoic of oil and gas
bearing regions of USSR. Trudy Vsesoiuznogo Naucho-Issledovate skogo Geologoazvedochnogo
Neftianogo Instituta, Kama Branch, no. 217. [In Russian],
balme, b. 1988. Miospores from Late Devonian (early Frasnian) strata, Carnarvon Basin, Western Australia.
Palaeontographica , Abteilung B , 209, 109-166.

bassett, m. G., cocks, l. r. m„ Holland, c. h., rickards, r. b. and warren, p. t. 1975. The type Wenlock
Series. Reports , Institute of Geological Sciences , no. 75/13, 1-19.
burgess, n. d. 1987. Micro- and megafossils of land plants from the Silurian and Lower Devonian of the
Anglo-Welsh Basin. Unpublished Ph.D thesis. University of Wales (Cardiff).

1991. Silurian cryptospores and miospores from the type Llandovery area, south-west Wales.
Palaeontology , 34, 575-599.

— and edwards, d. 1991. Classification of uppermost Ordovician to Lower Devonian tubular and
filamentous macerals from the Anglo-Welsh Basin. Botanical Journal of the Linnean Society , 106, 41-66.
butterworth, m. a. and williams, r. w. 1958. The small spore floras of coals in the Limestone Coal Group
and Upper Limestone Coal Group of the Lower Carboniferous of Scotland. Transactions of the Royal
Society of Edinburgh , 63, 353-392.

chibrikova, e. v. 1959. Spores from the Devonian and older rocks of Bashkiria; Academy of Sciences of USSR,
Bashkirian Branch, Data on Palaeontology and Stratigraphy of Devonian and Older Deposits of Bashkiria, pp.
3-1 16. [In Russian].

colthurst, j. r. j. and smith, d g. 1977. Palaeontological evidence for the age of the lower Palaeozoic rocks
of the Slievenammon Inlier, County Tipperary. Proceedings of the Royal Irish Academy , 77B, 144-158.
cramer, f. H. 1966. Hoegispheres and other microfossils incertae sedis of the San Pedro Formation (Siluro-
Devonian boundary) near Valporquero, Leon, NW Spain. Notas y Comunicaciones del Instituto Geologico
y Miner o de Espana, 86, 75-94.

BURGESS AND RICHARDSON: WENLOCK SPOROMORPHS

627

dorning, K. J. 1981a. Silurian acritarch distribution in the Ludlovian shelf sea of South Wales and the Welsh
Borderland. 31—36. In neale, j. w. and brasier, m. d. (eds). Microfossils from recent and fossil shelf seas. Ellis
Horwood, Chichester.

— 19816. Silurian acritarchs from the type Wenlock and Ludlow of Shropshire, England. Review of
Pcilaeobotany and Palynology , 34, 175-203.

1983. Palynology and stratigraphy of the Much Wenlock Limestone Formation of Dudley, Central
England. Mercian Geologist , 9, 31-40.

downie, c. 1963. ‘Hystrichosphcres’ (acritarchs) and spores of the Wenlock shales (Silurian) of Wenlock,
England. Palaeontology, 6, 625-652.

edwards, d. 1982. Fragmentary non-vascular plant microfossils from the late Silurian of Wales. Botanical
Journal of the Linnean Society, 84, 223-256.

and fanning, u. 1985. Evolution and environment in the late Silurian-early Devonian: the rise of the
pteridophytes. Philosophical Transactions of the Royal Society of London, Series B , 309, 147-165.
fanning, u., Richardson, j. b. and edwards, D. 1988. Cryptic evolution in an early land plant, evolutionary
trends in plants, 2, 13-24.

, and . In press. A review of in situ spores in Silurian land plants. In barnes, s. and blackmore,

s. (eds). Pollen and spores , The Systematics Association special volume No. 44.
gray, j. 1985. The microfossil record of early land plants: advances in understanding early terrestrialization,
1970-1984. Philosophical Transactions of the Royal Society of London, Series B, 309, 167-195.

— 1988. Land plant spores and the Ordovician-Silurian boundary. Bulletin of the British Museum (Natural
History ), Geology, 43, 351-358.

grebe, h. 1971. Terminologie morphographique recommendee et methode de description des spores. CIMP
special volume 4. Microfossiles organique du Palaeozoique, pp. 7-34.
hill, p. J., PARIS, F. and richardson, j. B. 1985. Silurian palynomorphs. Journal of Micropalaeontology , 4,
27-48.

hoffmeister, w. s. 1959. Lower Silurian plant spores from Libya. Micropaleontology , 5, 331 -334.
jardine, s. and yapaudjian, l. 1968. Lithostratigraphie et palynologie du Devomen-Gothlandien greseux du
Bassin Polignac (Sahara). Revue de TInstitut Franfais du Petrol et Annates des Combustibles Liquides, 23,
439-469.

Johnson, n. g. 1985. Early Silurian palynomorphs from the Tuscarora Formation in central Pennyslvania and
their paleobotanical and geological significance. Review of Palaeobotany and Palynology, 45, 307-360.
livermore, r. a., smith, A. G. and briden, J. c. 1985. Palaeomagnetic constraints on the distribution of
continents in the late Silurian and early Devonian. Philosophical Transactions of the Royal Society of London ,
Series B, 309, 29-56.

luber, a. a. and waltz, i. e. 1938. Klassifikatsiya i stratigraficheskoe zhachenie spor nekotovich kamen-
nougol'nich mestorojdenii SSSR. Trudy TSNIGRI, Moscow. 105, 1—45.
mabillard, j. e. and aldridge, r. j. 1985. Microfossil distribution across the base of the Wenlock Series in the
type area. Palaeontology, 28, 89-100.

M rGREGOR, d. c. 1961. Spores with proximal radial pattern from the Devonian of Canada. Geological Survey
of Canada, Bulletin, 76, 1-11.

— 1974. Early Devonian spores from central Ellesmere Island, Canadian Arctic. Canadian Journal of Earth
Sciences, 11. 70-78.

— and narbonne, g. m. 1978. Upper Silurian trilete spores and other microfossils from the Read Bay
Formation, Cornwallis Island, Canadian Arctic. Canadian Journal of Earth Sciences, 15, 1292-1303.

magloire, l. 1968. Etude stratigraphi'que par la palynologie des depots argilo-greseux du Silurien et du
Devonien inferieur dans la region du Grand Erg Occidental (Sahara Algerien). International Symposium on
the Devonian System, Calgary, 2, 473-495.

miller, m. e. and eames, l. e. 1982. Palynomorphs from the Silurian Medina Group (Lower Llandovery) of the
Niagara Gorge, Lewiston, New York, U.S.A. Palynology, 6, 221-254.
muller, j. 1959. Palynology of Recent Orinoco delta and shelf sediments. Reports of the Orinoco Shelf
Expedition; vol. 5. Micropaleontology , 5, 1-32.

potonie, h. 1893. Die Flora des Rotliegenden von Thuringen. Abhandlungen der Koniglich Preussischen
Geologischen- Landesanstalt , 9, 1-298.

pratt, l. m., Phillips, t. l. and dennison, j. m. 1978. Evidence of non-vascular land plants from the early
Silurian (Llandoverian) of Virginia, U.S.A. Review of Palaeobotany and Palynology, 25, 121-149.
reinsch, p. E. 1891. Neue Unter suchungen uber die Mikrostrukt Steinkohle des Carbon, Dyas und Trias.
124 pp. T. O. Weigel, Leipzig.

628

PALAEONTOLOGY, VOLUME 34

Richardson, j. b. 1985. Lower Palaeozoic sporomorphs: their stratigraphical distribution and possible
affinities. Philosophical Transactions of the Royal Society of London , Series B. 309, 201-205.

1988. Late Ordovician and Early Silurian cryptospores and miospores from northeast Libya. 89-109. In
el-arnauti, a., Owens, b. and thusu, b. (eds). Subsurface Palynostratigraphy of northeast Libya. Garyounis
University Publications, Bengahzi, Libya.

— and ioannides, n. s. 1973. Silurian palynomorphs from the Tanezzuft and Acacus formations,
Tripolitania, North Africa. Micropaleontology , 19, 257-307.

— and lister, T. r. 1969. Upper Silurian and Lower Devonian spore assemblages from the Welsh
Borderland and South Wales. Palaeontology , 12, 201-252.

and McGregor, d. c. 1986. Silurian and Devonian spore zones of the Old Red Sandstone continent and
adjacent regions. Geological Survey of Canada , Bulletin , 364, 1-79.

and rasul, s. m. 1990. Palynofacies in a Late Silurian regressive sequence in the Welsh Borderland and
Wales. J owned of the Geological Society, London , 147, 675-686.

— , ford, j. h. and Parker, f. 1984. Miospores, correlation and age of some Scottish Lower Old Red
Sandstone sediments from the Strathmore region (Fife and Angus). Journal of Micropalaeontology, 3,
109-124.

smith, d. g. 1975. Wenlock plant spores and tetrads from County Mayo Ireland. Geological Magazine, 112,
411-414.

— 1979. The distribution of trilete spores in Irish Silurian rocks. 423-431. In Harris, a. l., Holland, c. h.
and leake, b. e. (eds). The Caledonides of the British Isles — reviewed. Geological Society of London. Special
Publications. 8, Scottish Academic Press, Edinburgh.

smith, a. h. v. and butter worth, m. a. 1967. Miospores in the coal seams of the Carboniferous of Great
Britain. Special Papers in Palaeontology, 1, 1-320.

Strother, p. k. and traverse, a. 1979. Plant microfossils from Llandoverian and Wenlockian rocks of
Pennsylvania. Palynology, 3, 1-22.

tappan, h. 1980. The paleobiology of plant protists. W. H. Freeman & Co., San Francisco.
turnau, e. 1974. Microflora from core samples of some Palaeozoic sediments from beneath the flysch
Carpathanians (Bielsko-Wadowice area, southern Poland). Annales de la Societe Geologique de Pologne, 44,
143-169.

- 1986. Fower to middle Devonian spores from the vicinity of Pionki, Central Poland. Review of
Palaeobotany and Palynology, 46, 311-353.

NEIL BURGESS
Ecology Department

Royal Society for the Protection of Birds
The Lodge, Sandy
Bedfordshire SGI 9 2DL

JOHN RICHARDSON

Palaeontology Department
British Museum (Natural History)

Typescript received 2 December 1989 Cromwell Road

Revised typescript received 22 November 1990 London SW7 5BD

A NEW M ARSUPI ATE CIDAROID ECHINOID FROM
THE MAASTRICHTIAN OF ANTARCTICA

by DANIEL B. BLAKE Cllld WILLIAM J. ZINSMEISTER

Abstract. Almucidaris durhami is a new genus and species of cidaroid echinoid from the Maastrichtian (Late
Cretaceous) of Seymour Island, Antarctic Peninsula. The species is unique in that the genital plates of the
female are expanded and hollowed to form marsupia. The incidence of brooding in invertebrates, including
echinoderms, increases with latitude and apparent stress; stressful conditions might have contributed to the
evolution of A. durhami.

Seymour Island, in the Antarctic Peninsula, has been the source of an extraordinary diversity of
superbly preserved late Mesozoic and early Cenozoic fossils (Feldmann and Woodburne 1988).
None of these many fossils is better preserved or more elegant than a specimen of the new cidaroid
echinoid described here; the new species is important for more than the aesthetic value of its
holotype, however, because the genital plates of the female were much enlarged to form deep brood
chambers. Taxonomically diverse brooding echinoids, including cidaroids, are known from
numerous other occurrences, and these shed some light on the evolution of brooding in Almucidaris
durhami gen. et sp. nov.

BROODING IN ECHINOIDS

The biology of brooding in echinoderms was reviewed by Lawrence (1987), and a taxonomic survey
of brooding echinoids was provided by Philip and Foster (1971). Lawrence (1987, p. 264) defined
brooding as ‘...the association of embryos and juveniles with an adult, usually the female parent’.
The definition is broad, and includes simple presence of the juveniles on the surface of the parent,
without structural differentiation. Brooding increases probability of offspring survival at a cost of
decreased fedundity and dispersal, and increased reproductive effort. Although brooding has been
reported from all living echinoderm classes (except the Concentricycloidea), it is not common,
indicating that costs tend to outweigh benefits. In echinoids, brooding always is external, with the
young present either in marsupia or distributed among the spines without parent structural
differentiation.

Within echinoderms, brooding is found in different trophic groups, but as with marine
invertebrates in general, it increases with latitude. Seventeen of the eighteen known species of
Antarctic echinoids are brooders (Magniez 1983), some occurring intertidally on subantarctic
islands under conditions of considerable physical stress (Lawrence 1987). Brooding also is known
from lower latitudes; Gladfelter (1978) described it in the shallow-water infaunal Caribbean
cassiduloid Cassidulus caribbearum, for example. The test of C. caribbearum is unmodified and the
young are carried among the aboral spines.

Barker (1985) studied reproductive behaviour in the modern brooding cidaroid Goniocidaris
umbraculum from the deep (95 m) continental shelf off New Zealand. Laboratory studies showed
that brooding in this species begins with the urchins partially burying themselves in bryozoan shell
debris. Oocytes are passed along the ambulacra by means of ciliary and tube foot action from the
gonopore to the peristomial region. The oocytes are buoyant, and both their buoyancy and the oral
spines help hold as many as 60 oocytes at the peristome. After 47-55 days, juveniles begin to migrate
from the peristome to the bases of the spines on the oral side of the adult. Field data suggest the

IPalaeontology, Vol. 34, Part 3, 1991, pp. 629-635, 1 pl.|

© The Palaeontological Association

630

PALAEONTOLOGY, VOLUME 34

annual breeding cycle takes place over nearly a six month period, with juveniles remaining with the
parents for approximately four months. The smallest individuals recovered with juveniles were
under 18 mm in diameter.

Associated juveniles can be quite numerous; Magniez (1980) found the mean egg and juvenile
number per female to be 27-3, with a maximum of 109, in a population of 315 females of the
schizasterid spatangoid Abatus cordatus from Kerguelen Island.

Barker (1985) found controlling factors on brooding to be uncertain in Goniocidaris , which has
a well defined annual breeding cycle. He suggested that photoperiod or nutrient accumulation might
be important but that bottom temperature showed only a relatively minor 4 °C range of variation.
This author further noted that the urchins do not feed during brooding; movement of the lantern
might be sufficient to dislodge the eggs from the peristome. Partial burial might reduce risk of egg
loss as well as provide a stable environment for the young. The range of G. umbraculum partially
overlaps that of the broadcast spawning euechinoid Pseudechinus huttoni and both graze on the
bryozoan Cinctipora elegans\ however, competition was concluded not to have been involved with
brooding in G. umbraculum because of the limited overlap of ranges. Hendler and Franz (1982)
suggested brooding in the asteroid Leptasterias is a cold adaptation and not a result of competition;
Barker (1985) preferred a temperature hypothesis for the origin of brooding in G. umbraculum.

Several living brooding cidaroids, including Goniocidaris corona (Barker 1968), G. umbraculum
(D. L. Pawson pers. comm.), and Austrocidaris canaliculata (Philip and Foster 1971) carry young
around the apical system; juveniles of other species are found at the peristome, protected by the
primary spines. In Austrocidaris , a sunken median furrow is present in both ambulacra and
interambulacra, and young can cling to the apical surface. The enlarged genital-plate marsupia of
Almucidaris gen. nov. are unique.

De Ridder and Lawrence (1982) reviewed the literature on food resources of echinoids; available
data on cidaroids were largely drawn from Mortensen (1928), whose observations were based
primarily on gut contents. Cidaroid food consists of a diversity of bottom materials, including
foraminifers, bryozoans, sponges and other generally smaller particles.

STRATIGRAPHIC OCCURRENCE OF ALMUCIDARIS DURHAMI

The holotype was collected from the Maastrichtian portion of the Lopez de Bertodano Formation
of Seymour Island, Antarctic Peninsula (Text-fig. 1). The formation consists of approximately
1200 m of sandy siltstones and siltstones with numerous concretionary horizons. Although the
invertebrate fauna is moderately diverse and well preserved, the abundance of individuals within
any given interval is not great. Fragmentary echinoid material is abundant at a number of horizons;
however, the holotype is the only known complete individual.

The specimens of Almucidaris durhami gen. et sp. nov. were collected from a 25 m interval of
medium grey siltstone 320 m below the K/T boundary. Extensive bioturbation characterized by
Bergaueria and Phycodes-like structures have destroyed all primary bedding structures. The absence
of any primary sedimentary structures and the intensely bioturbated nature of the sediments suggest
that the environment of deposition of this horizon was below wave base in a mid or outer shelf
setting.

Although the fossils from the siltstone are not abundant, the fauna is exceptionally diverse (35
taxa) for the Lopez de Bertodano Formation, and remarkably well preserved. The composition of
the invertebrate fauna from this siltstone interval differs markedly from that of the rest of the Lopez
de Bertodano Formation by the presence of a large number of small gastropods that have not been
encountered elsewhere in the formation. The molluscan fauna from the Antarctic Cretaceous
generally is characterized by relatively large, thick-shelled species; the occurrence of many small,
thin-shelled taxa indicates an unusual local facies not seen elsewhere in this unit.

BROODING IN ALMUCIDARIS DURHAMI
Although a number of aboral brooders are known among echinoids, peristomial brooding is

BLAKE AND ZINSMEISTER: ANTARCTIC ECHINOID

631

C33

p m s

text-fig. I . Geological locality map of Seymour Island, Antarctic Peninsula. The Lopez de Bertodano
Formation is Cretaceous and Paleocene; that part of it in which the echinoid material was found is
Maastrichtian. The Sobral and Cross Valley Formations are Paleocene, and the La Meseta is Eocene in age.

dominant ; the latter provides more protection, but probably interferes with the parent’s feeding. In
contrast, the genital-plate marsupia of A. durhami (Text-fig. 2) provided effective protection yet they
would not have impaired feeding. A second advantage of brood chambers in the genital plates was
that oocytes could be transferred from the interior to them with little chance of loss. Many echinoid
marsupia are formed by depressions in the test, and although all marsupia in echinoids are
topologically external, they can utilize a significant percentage of the volume of the interior; the
raised marsupia of A. durhami provided enclosed space without decreasing the parent’s internal
volume.

Almucidaris durhami, like Goniocidaris umbraculum, apparently lived below wave base, and
therefore the adaptive value of marsupia in A. durhami was not directly related to variation in
shallow-water conditions, although as in Goniocidaris , perhaps availability of nutrients was
important.

Size at which juveniles become independent varies among species, but space is quite limited within
the marsupia of A. durhami. Barker (1985) illustrated juveniles of G. umbraculum clinging to the
spines of the adult; these had a test diameter of about 3 mm, and diameter including spines was
approximately 6 mm. Perhaps juveniles of A. durhami, like those of G. umbraculum, moved onto the
spines of the adult during ontogeny, freeing marsupial space for younger individuals. Sequentially
overlapping broods are known in Cassidulus caribbearum (Gladfelter 1978).

Cidaroid taxonomy has proven difficult, in part because of individual variation and lack of
agreement on phylogenetically significant characters. Smith and Wright (1989) provided a historical
survey of cidaroid taxonomy, a revised classification, and a phylogenetic reconstruction. Their
system is followed here, along with the taxonomic treatment of Philip (1963, 1964).

632

PALAEONTOLOGY, VOLUME 34

text-fig. 2. Almucidaris durhami gen. et sp. nov., holotype USNM 446322, two marsupia with spines displaced

from scrobicular rings into the chambers, both x 6.

SYSTEMATIC PALAEONTOLOGY

Class echinoidea Leske, 1778
Order cidaroida Claus, 1880
Family cidaridae Gray, 1825
Subfamily cidarinae Gray, 1825
Tribe cidarini Gray, 1825
Genus almucidaris gen. nov.

Type species. Almucidaris durhami sp. nov.

Etymology. Almus (L.); fostering, nourishing, cherishing; in reference to the brooding behaviour.

Diagnosis. Cidarinid with large genital plates forming deep marsupia, at least in females;
ambulacral column pore rows strongly sinuous; podial pairs inclined perradially, not conjugate;
interambulacral plates of adult with densely arranged tubercles, plate sutures closed by tubercles.

Almucidaris durhami sp. nov.

Plate 1, figs 1-6; Text-fig. 2

Material. Available material consists of one complete test (holotype. United States National Museum [USNM]
446322), the base of a second, and 25 test fragments, the latter ranging from single plates to nearly complete
interambulacra with portions of adjacent ambulacra (paratypes USNM 446323-446348). Small spines are
associated with the complete specimen and the base. Four isolated marsupial plates are included among the
25 fragments. Maastrichtian, Upper Cretaceous, Seymour Island, Antarctic Peninsula.

EXPLANATION OF PLATE 1

Figs 1-6, Almucidaris durhami gen. et sp. nov., holotype, USNM 446322, Maastrictian, Seymour Island,
Antarctic Peninsula. 1, aboral view, note large facets for periproctal plates, gonopore, and spines collapsed
into marsupia, x 2. 2, ambulacral detail, x 6. 3 and 4, stereo pair, lateral views, x 2. 5, oral view, x 2. 6,
ambulacral detail, specimen is partially separated along perradial suture, x 6.

PLATE 1

BLAKE and ZINSMEISTER, Almucidaris

634

PALAEONTOLOGY, VOLUME 34

Etymology. The species is named in honour of Dr J. Wyatt Durham.

Diagnosis. As for the genus.

Description. The diameter of the holotype is 35 mm and the height is 21 mm; the height of the corona and
genital ring together are 30 mm. The corona is somewhat flattened, with the apical ring forming a prominent,
peaked aboral profile. The periproctal outline is elongate pentameral, with a length of 9 mm and a breadth of
7 mm; all periproctal plates were lost but large articular surfaces on the genital plates demonstrate that
periproctals were large. The genitals are much enlarged, approximately 14-15 mm (vertical dimension) by
13-14 mm. They contain open hemispherical marsupial chambers which are approximately 10 by 8-5 mm;
chamber openings are directed laterally and aborally. A genital pore is present at the base of each chamber;
it is approximately 2 mm in length. The outer surface of the genital plate is covered by closely spaced tubercles
which diminish in size away from the enlarged scrobicular ring; the scrobicular ring and the next lateral ring
have rounded crescentic depressions which are directed toward the chamber, thus allowing spines to be
deflected over the chamber opening. Oculars are triangular but indented at their bases; their breadth is
approximately 4 mm and height is 3-5 mm. Oculars are covered by closely spaced tubercles of uniform size.
Apical ring plate boundaries are incised and they are neither pitted nor have areas bare of tubercles. Ambulacra
are strongly sinuous throughout. Ambulacral width at the ambitus is 10-11% of test diameter, with the
combined pore zones forming nearly one-half of the ambulacral width. Ambulacral pores are uniserial and
pairs are inclined, with the perradial member of the pair lower than the adradial. Pores are not conjugate; the
interporal partition is slightly broader than the width of a single pore. A clearly defined furrow extends from
below the perradial pore to the lower edge of the ossicle; the groove is gradually lost toward the adradial pore.
The perradial end of the plate has a single enlarged tubercle occupying the full plate height; one or two
vertically aligned smaller perradial tubercles also are present, and on some ossicles, one or two additional small
miliary tubercles are fitted among the perradial series. Most tubercles other than those of the enlarged series
are absent near the peristome and periproct. Perradial sutures are not bare or pitted. Each interambulacral
column consists of four or five plates, and all but perhaps one or two of the most recently-added plates have
functional tubercles. The height and width of the plates are approximately equal, or the width is the slightly
greater dimension. Primary tubercles are large with areoles of only the smallest adoral plates being confluent.
The scrobicular series of slightly larger plates are tangent; plates ambitally and adapically in the series have
closely arranged miliary tubercles separating the neighbouring scrobicular series. Sutural boundaries are not
incised or bare, although scattered small pits are present. Areoles are deeply incised, occupying approximately
75% of the plate width at the ambitus. Mamelon diameter is 15-20% of plate width and most are perforate,
although some adoral mamelons are imperforate or their condition was not preserved; they are weakly
crenulate. Scrobicular tubercle series are clearly enlarged, with a distinct mamelon and a small crescentic
depression toward the areole. Peristomial diameter is approximately 40% of test diameter; the lantern is
unknown. Spines of genital plates are cuneate in cross-section, and therefore they were closely fitted in life; they
have longitudinal beaded ribbing.

Available isolated coronal fragments and plates (paratype specimens) add relatively little, in part because the
surfaces of most have been abraded by wind and dust. The largest interambulacral plate is 19 by 13 mm, and
that of an ambital interambulacral plate on the holotype is 1 1 by 10 mm, suggesting plate growth was
amsometric. If test diameter was proportionate to ambital plate width, a diameter of at least 60 mm was
attained. One adoral plate has a clearly crenulate tubercle. Although isolated, four marsupiate genitals were
recovered, but no dimorphic examples were found. A basal test fragment retains a number of cuneate
scrobicular spine fragments similar to those of the genitals, again indicating a closely fitted, protective
arrangement.

Remarks. Almucidaris is assigned to the Cidarinae (sensu Smith and Wright 1989) based on the
presence of well defined perforate tubercles; it is assigned to the Cidarini based on presence of well
developed adapical tubercles and the presence in adults of tubercles obscuring plate sutures, and
weakly developed sutural pits. Positive evidence on subtribal affinities is lacking, although pore
pairs are neither conjugate or subconjugate and therefore Almucidaris does not belong to the
Phyllacanthina. Smith and Wright (1989) noted that although the Cidarini is distinguished by
pedicellariae and intestinal plate characters not found in fossils, some constituent genera can be
traced back to the Lower Cretaceous; Almucidaris is similar to such genera as Prionocidaris sensu
Smith and Wright (1989), and Phyllacanthus sensu Philip (1963). It differs from both in the nature

BLAKE AND ZINSMEISTER: ANTARCTIC ECHINOID

635

of the pore pairs, and from the latter in the nature of the scrobicular spine bases (less strongly
differentiated in Almucidaris ) and the presence of marsupia. It is unlike Prionocidaris (but similar
to Phyllacanthus ) in the development of test tubercles.

Acknowledgements . The discovery of the holotype of Almucidaris would not have been possible without the
generous held support provided by Dr Carlos Rinaldi, director of the Instituto Antartico Argentino, during
the austral summer of 1988-1989. Field work was supported by Division of Polar Programs grant no. DPP
8416783 to W.J.Z. D. L. Pawson and A. B. Smith reviewed and provided numerous useful suggestions on
an earlier version of the manuscript.

REFERENCES

barker, m. f. 1968. A new cidarid echinoid from northern New Zealand. Transactions of the Royal Society of
New Zealand , Zoology, 10, 199-203.

— 1985. Reproduction and development in Goniocidaris umbraculum , a brooding echinoid. 207-214. In
keegan, b. f. and o’connor, b. d. s. (eds). Echinodermata . A. A. Balkema, Rotterdam, 662 pp.

claus, c. f. w. 1800. Grundzuge der Zoologie. Volume 2. 4th edition. Marburg and Leipzig, 522 pp.
de ridder, c. and Lawrence, j. M. 1982. Food and feeding mechanisms: Echinoidea. 57-1 16. In jangoux, m.

and Lawrence, j. m. (eds). Echinoderm Nutrition. A. A. Balkema, Rotterdam, 654 pp.
feldmann, r. m. and woodburne, m. o. (eds). 1988. Geology and paleontology of Seymour Island, Antarctic
Peninsula. Memoirs of the Geological Society of America , 169, 1-566.
gladfelter, w. b. 1978. General ecology of the cassiduloid urchin Cassidulus caribbearum. Marine Biology , 47,
149-160.

gray, j. e. 1825. An attempt to divide the Echinida, or sea eggs, into natural families. Annals of Philosophy,
10, 423-431.

hendler, g. and franz, d. r. 1982. The biology of a brooding seastar, Leptasterias tenera , in Block Island
Sound. Biological Bulletin, 162, 273-289.

Lawrence, j. 1987. A functional biology of echinoderms. Croom Helm, London, 340 pp.
leske, N. G. 1778. Jacobi Theodori Klein naturalis dispositio Echinodermatum . Edita et descriptionibus novisque
inventis et eynonymis auctorum aucta. Leipzig, 278 pp.

magniez, p. 1980. Modalites de l’incubation chez Abatus cordatus (Verrill), oursin endemique des lies
Kerguelen. 399-404. In jangoux, m. (ed.). Echinoderms : Present and Past. A. A. Balkema, Rotterdam,
428 pp.

— 1983. Reproductive cycle of the brooding echinoid Abatus cordatus (Echinodermata) in Kerguelen
(Antarctic Ocean) : changes in the organ indices, biochemical composition and caloric content of the gonads.
Marine Biology, 74, 55-64.

mortensen, T. 1928. A monograph of the Echinoidea. Volume I. C. Z. Reitzel, Copenhagen, 552 pp.

Philip, G. m. 1963. The Tertiary echinoids of south-eastern Australia. Proceedings of the Royal Society of
Victoria, 16, 181-226.

— 1964. The Tertiary echinoids of south-eastern Australia. II Cidaridae (2). Proceedings of the Royal Society
of Victoria, 77, 433-477.

— and foster, r. j. 1971. Marsupiate Tertiary echinoids from south-eastern Australia and their
zoogeographic significance. Palaeontology, 14, 666-695.

smith, a. b. and wright, c. w. 1989. British Cretaceous Echinoids. Part 1, general introduction and
Cidaroidea. Monograph of the Palaeontographical Society, 141 (for 1988), 1-101.

DANIEL B. BLAKE

Department of Geology
University of Illinois
1301 W. Green St.

Urbana, IL 61801, USA

WILLIAM J. ZINSMEISTER

Department of Earth and Atmospheric Sciences
Typescript received 2 January 1990 Purdue University

Revised typescript received 9 July 1990 West Lafayette, IN 47907, USA

A NEW SPECIES OF MACHAERIDIAN FROM THE
SILURIAN OF PODOLIA, USSR, WITH A REVIEW OF

THE TURRILEPADIDAE

by J. M. ADRAIN, B. D. E. CHATTERTON and L. R. M. COCKS

Abstract. Machaeridians are enigmatic marine fossils, known from rocks of Ordovician to Carboniferous age.
Turrilepas modzalevskae sp. nov., from Podolia, Ukraine, USSR, is only the second turrilepadid machaeridian
species to be represented by an articulated assemblage. It provides unequivocal evidence for the presence of
minute marginal spines, and demonstrates the relationship between these spines and the ornament of the outer
sclerite surface. A standardized system of terminology for turrilepadid machaeridians is presented, based on
examination of many specimens of Turrilepas wrightiana (de Koninck) from the Wenlock of Britain, together
with undescribed Canadian silicified material. The anterior structure of turrilepadids and plumulitids is very
similar, both lacking outer series sclerites on the first two segments. The Family Turrilepadidae comprises the
genera Turrilepas Woodward, Deltacoleus Withers, Clarkeolepis Elias, Spinacoleus Schallreuter, and
Mojzcalepas Dzik. All but Turrilepas are inadequately known.

During a visit to the Natural History Museum in 1988, J.M.A. discovered a specimen in the
palaeontological collections labelled Lepidocoleus sp., collected from Podolia in 1983 by L.R.M.C.
It was recognized that the specimen belonged in fact to Turrilepas Woodward, 1865, and was
important because it represented only the second unequivocal record of the genus. The type species
is from the British Wenlock.

Turrilepas modzalevskae , described herein, is only the second turrilepadid species to be
represented by an articulated specimen. Examination of the many specimens of Turrilepas
wrightiana in the Natural History Museum, together with knowledge derived from study of
undescribed North American silicified faunas by J.M.A. has resulted in a more thorough
understanding of the arrangement and structure of the turrilepadid scleritome, a subject to which
little new information has been added since T. H. Withers’s seminal work of 1926. An attempt is
made to standardize descriptive terminology, and the status of the currently recognized turrilepadid
taxa is reviewed. A historical review of machaeridian study and a discussion of recent systematic
problems is presented elsewhere (Adrain in press). Illustrations of silicified sclerites are scanning
electron micrographs. Illustrated calcareous specimens were blackened and given a light coating of
ammonium chloride sublimate prior to photography. Illustrated specimens are housed in the
Natural History Museum, prefixed BM(NH), and the palaeontological collections of the University
of Alberta, prefixed UA.

LOCALITY AND STRATIGRAPHY

Podolia lies in the centre of eastern Europe; until the Second World War it was in Poland and since
then it has formed part of the Ukraine. Access has been difficult, but in May 1983 the Silurian
Subcommission was invited to visit and collect from the sequence, which consists of nearly flat-lying
beds of mostly late Silurian and early Devonian age all unconformably overlain by the Cretaceous
(Koren et al. 1989). The locality of the new species of Turrilepas , T. modzalevskae. lies in a natural
outcrop on the left slope near the mouth of the Smotrich River, near Tsvicklevtsy village
(Tsegelnjuk et al. 1983, Stop 4, p. 75). Here the Ustje Member of the Bagovitsa Formation is
exposed to 9-5 m and overlain conformably by Konovka Formation, which is composed of a lower

IPalaeontology, Vol. 34, Part 3, 1991, pp. 637-651, 2 pls.|

© The Palaeontological Association

638

PALAEONTOLOGY, VOLUME 34

Goloskov Member (11-5 m) and an upper Shutnovsky Member (17-9 m), in turn overlain by the
Sokol Member of the Tsvicklevtsy Formation which is seen to 19 m. The new Turrilepas
undoubtedly came from the Konovka Formation, but since the single specimen was found loose,
it is uncertain whether it came from the Goloskov or the Shutnovsky Member. Both members yield
a very similar shelly fauna dominated by the brachiopods Atrypella , Microsphaeridiorhynchyus ,
Protochonetes , and Sphaerirhynchia , and in other accounts (e.g. Koren et al. 1989) the members are
not used. The age of the Konovka Formation is also securely fixed as basal Ludlow (lower Gorstian
Stage), based not only on the brachiopods, but also on the ostracods present (Koren et al. 1989, fig.
108); however, there are no useful records of graptolites from that part of the Podolian succession.

TERMINOLOGY

There is no standard system of terminology for description of machaeridian sclerites and sclerite
assemblages, although Jell (1979) and Schallreuter (1985) have made progress. Withers (1926)
provided a comprehensive set of terms. His opinion was, however, that machaeridians were sessile,
attached to the substrate by one end of their assemblage, and capable of opening and closing their
assemblage along what he termed the Tree margin’. This interpretation of function and orientation
now seems highly improbable (Jell 1979; Schallreuter 1985; Dzik 1986), and hence Withers’s
descriptive terms are largely obsolete. It is now generally accepted that machaeridians were vagile,
that the more modified end of their assemblages was anterior, with the sclerites imbricating
posteriorly, and that the median line along which opposing series sclerites meet in all known
machaeridians was dorsal.

This orientation is followed here, but problems arise when comparing isolated sclerites, as
homologous parts may be held in completely different orientations in the assemblages of different
machaeridians. Furthermore, there is often sufficient serial variation along the scleritome that
anteriorly placed, midbody, and posteriorly placed sclerites will have different orientations with
respect to the general orientation of the whole assemblage. It has therefore been found necessary
to adopt conventions when referring to particular aspects of sclerites. These conventions follow
from the most general orientation of the sclerite and facilitate comparisons between differently
arranged taxa, but do not necessarily correspond to what would have been the actual orientation
of the part in a particular living machaeridian.

Terms referring to the entire scleritome

The term scleritome (Bengtson 1985) is used to refer to the organizational plan of the entire
assemblage for a given species. It is recommended that sclerite assemblage (Jell 1979) be used to refer
to an actual set of associated sclerites recovered from the fossil record. They are together given
preference over ‘strobilus’ (Pope 1975). The division of the scleritome into the longitudinal sectors
‘head,’ ‘thorax,’ and ‘abdomen’ (Dzik 1986) should be discontinued, as it carries organizational
and phylogenetic connotations which are thus far unsupported. Likewise, the neutral term sclerite
is preferred over ‘elytrum’ (Jell 1979; Dzik 1986).

The longitudinal columns of a particular scleritome are referred to as sclerite series. Biserial
machaeridians have left and right series. Quadriserial machaeridians have left and right inner and
outer series. The rows of the scleritome may well correspond to particular body segments, if its fully
segmented nature is indicative of true, and not merely functional, metamerism. It is not possible to
evaluate this suggestion further, however, given our present state of knowledge of the group. Hence,
the term ‘segment’ is used here in a phylogenetically neutral sense. When distinguishing between
sclerites within a given series or set of series, reference should be made, where possible, to the
sclerites of the xth segment , measured from the anterior of the scleritome.

Terms referring to individual sclerites

Bengtson (1978) has determined that machaeridian sclerites are composed of two distinct layers. An
outer layer is composed of dense calcite, in which growth lines may often be visible (see PI. 2, fig. 10),

ADRAIN ET AL. . TURRILEPADID M ACH AERIDIANS

639

and from which the rugae are formed (terms defined below). The inner layer is composed of
distinct, densely packed lamellar elements which correspond closely to the granular texture of the
inner sclerite surface. The layers may vary in thickness, depending upon the taxon in question.
Bengtson (1978) noted that in the turrilepadid he investigated, two distinct layers did not appear to
occur. In many silicified turrilepadid sclerites, the outer surface between the rugae has exactly the
same granular texture as the inner surface. This indicates that the inner sclerite layer made up
almost the entire sclerite thickness, with the outer layer confined to the rugae and not deposited
between them.

Some of the following terms are illustrated in Text-figures 1-3. The outer surface of a sclerite is
defined as that which bears the rugae. Rugae are prominent, ridge-like features, evenly spaced and
developed parallel to the fine growth lines , apparently through deposition of additional material of
the outer sclerite layer at specific growth increments. Rugae terminate before they reach the sclerite
margin. Figured examples in which this appears not to be the case (e.g. Schallreuter 1985, pi. 2,
figs 1-3) represent artefacts of poor preservation. Similar preservation has been observed by the
authors in silicified material from the Canadian Arctic. Relatively coarse silicification often results
in poorly preserved margins. The rugae seem to be more amenable to silicification in these
conditions than the material of the margin and hence they are sometimes preserved while the margin
lateral to them is not. They then protrude from the margin of the fossil sclerite, creating a false
impression of having been extended into marginal spines in the living animal.

text-fig. 1. Silicified machaeridian sclerites from the Mackenzie Mountains, Northwest Territories, Canada,
to illustrate the inner groove possessed by all machaeridian sclerites. The groove runs from the apex, and its
anterior extent is indicated by an arrow in each case. All are scanning electron micrographs of the inner surface.
a, UA 8125, left sclerite belonging to a new family of the Order Lepidocoleomorpha, AV 4 126, x 37. b, UA
8126, right tongued lepidocoleid, AV 5 58-60, x 36. c, UA 8127, outer right turrilepadid, AV 5 58-60, x 37.
d, UA 8128, outer right plumulitid, AV 4 126, x 18. e, UA 8129, outer left plumulitid, AV 5 58-60, x 10. See
Over and Chatterton (1987) for locality information.

640

PALAEONTOLOGY, VOLUME 34

The inner surface has a finely granular texture, reflecting the regular arrangement of calcite
elements in the inner sclerite layer. A single muscle scar is borne on the inner surface, although in
some sclerites, particularly small ones and most Plumulitidae, it may be difficult to discern. A
narrow groove runs anteriorly from the apex on the inner surface of all machaeridian sclerites (see
Text-fig. 1). The function of this groove is not yet clear, but its universal occurrence suggests an
intimate relationship with some fundamental aspect of machaeridian soft part anatomy. It is
termed, for the time being, the inner groove. The umbo is the site of initial sclerite growth. It is
usually located at the apex, but may sometimes be set somewhat forward, creating initially
concentric growth lines (see Text-fig. 2d). The apex of turrilepadid sclerites is very often ornamented
with a bifid apical spine , although this structure is frequently broken off before or after preservation.
Some lepidocoleids also have an apical spine, but it seems never to have accessory spines developed
upon it (see below), and invariably has a simple, not bifid, tip. All machaeridian sclerites have three
margins (see Text-fig. 3). The accreting margin runs parallel to the growth lines and rugae, and is
the front of deposition of new sclerite material. It may be further subdivided by folding of the
sclerite in various types of machaeridians. The two non-accreting margins run from the apex to the
edges of the accreting margin. The medial margin is that which lies closest to the dorsal midline of
the assemblage, and its complement on the opposite side of the sclerite is the lateral margin. Many
sclerites are folded about an angle which runs anteriorly from the apex to the accreting margin. This
is the longitudinal angle (see Text-fig. 2c). When it occurs, it divides the accreting margin into medial
and lettered parts. All known turrilepadids have margined spines on the non-accreting margins. These
spines are often continued on the lateral aspects of the apical spine, and are quite independent of
the spacing of either the growth lines or the rugae, contrary to the claim of Dzik (1986, p. 1 19). In
fact, the growth lines bend anteriorly near the sclerite margin, and in Turrilepas modzalevskae, can
actually be seen to transect the marginal spines (see PI. 2, fig. 10). This indicates that the marginal
spines are formed from a different material to that which forms the outer surface. They seem
either to have been formed from an accessory material deposited along the non-accreting margins
as growth proceeded, or more likely from the material of the inner sclerite layer. Turrilepadid
sclerites often have a distinct doublure (see Text-fig. 2b, e, h) preserved around the margins on the
inner surface. The depositional and crystallographic characteristics of this feature are obscure, as
it has thus far been observed only in well preserved silicified specimens. It is not clear whether or
not it is a universal feature of turrilepadids. When it occurs, however, it is almost identical in texture
to the marginal spines, and when viewing the inner surface, these spines seem to be lateral extensions
of the doublural material. It seems likely, therefore, that the doublure is always present but often,
like the marginal spines, not well expressed.

The pattern of rugae on the outer sclerite surfaces is constant within a sclerite type of a single
species, and is therefore of much potential taxonomic value. To facilitate description and
comparison, a system of numbering the inflections in the course of the rugae has been devised (see
Text-fig. 3). This system is based upon sclerites of the genus Turrilepas , the best known turrilepadid.
The general pattern and number of inflections seems on present evidence to be universal throughout
Turrilepadidae, but this will require corroboration through further study. Numbering begins from
the medial margin. Five inflections are present on inner sclerites, labelled Ix— 15. The upper case ‘I’
is used for inner sclerites, and an R’ or ‘L’ may be appended to distinguish between right or left
sclerites. Outer sclerites feature only three inflections, labelled in lower case to distinguish them from
inner sclerites, ix— i3. Thus, for example, IXL refers to the first inflection of an inner left sclerite, while
i2 R refers to the second inflection of an outer right sclerite. Note that although I2 often coincides
with the longitudinal angle of the inner sclerites (see, for example, Text-fig. 3a), this is not always
the case, and the two terms refer to separate and distinct features.

Some combinations of features that may serve to characterize and distinguish between taxa
include the degree to which each of the inflections is expressed, and the relationship between the
longitudinal angle and the inflections. This system of numbering is adequate for reference to the
Family Turrilepadidae. Some modification will be required for the plumulitids and lepido-
coleomorphs, and will be presented elsewhere.

ADRAIN ET AL.: TURRILEPADID M ACH A E R I DI ANS

641

text-fig. 2. Silicified machaeridian sderites from the Wenlock of the Mackenzie Mountains, Northwest
Territories, Canada. All sclerites are from horizon AV 5 58-60, except I, which is from AV 2 248-8 (see Over
and Chatterton 1987 for details). All are scanning electron micrographs, a-c and e-i are turrilepadids, d is a
plumulitid. a, UA 8130, outer left sclerite, outer view, x 37. b, UA 8131, inner right sclerite, inner view, x 36.
c, UA 8132, inner left sclerite, outer view, x 40. d, UA 8133, inner right plumulitid sclerite, inner view (ridges
are not rugae, but rather the impressions of the rugae borne on the outer surface), x 40. e, f, UA 8134, outer
right sclerite, inner and outer views, x 19. G, h, UA 8135, outer left sclerite, outer and inner views, x 37. i, UA
8136, inner right sclerite, outer view, x 20. a, apex, am , accreting margin, as , apical spine, d , doublure, la ,
longitudinal angle, ms, muscle scar, msp, marginal spine, u, umbo.

642

PALAEONTOLOGY, VOLUME 34

1

a

m

i

Inm

A

B

text-fig. 3. Sclerites belonging to Turrilepas wrightiana (de Koninck), illustrating some of the terminology
introduced, a, inner right sclerite. b, outer left sclerite. a , apex; am, accreting margin; Inm, lateral nonaccreting
margin; mnm, medial nonaccreting margin; Ix, xth inflection of rugae, measured from medial margin, of inner
sclerite; zx, xth inflection of rugae, measured from medial margin, of outer sclerite; r, rugae.

Genera included. Turrilepas Woodward, 1 865 ; Deltacoleus Withers, 1926; Clarkeolepsis Elias, 1958 , Spinacoleus
( = Rugacoleus) Schallreuter, 1985; Mojczalepas Dzik, 1986.

Diagnosis. Scleritome tightly articulated; sclerites usually large, thick, coarsely rugose; most
sclerites with minute to robust bifid apical spine and multiple simple spines on non-accreting
margins; mid-scleritome inner sclerites usually with longitudinal angle of about 90 degrees,
sometimes more obtuse; medial area of inner left sclerites approximately twice that of inner right
sclerites; rugae of inner sclerites usually with five inflections; rugae of outer sclerites with three
inflections. For a given segment with four sclerites (i.e. any segment from 3 backwards), the outer
sclerites overlap the inner sclerites on each side, and the medial portion of the inner right sclerite
overlaps that of the inner left sclerite.

Remarks. It has not previously been appreciated that spinose non-accreting margins and apical
spines are features which appear to occur in most turrilepadids. This is certainly because these
features sometimes require extremely good preservation to remain evident. They are absent, for
instance, from most of the Turrilepas wrightiana material from Dudley. Knowledge, however, of
Canadian silicified material upon which detail is preserved to an exquisite degree, strongly indicates
they are a constant feature. This is corroborated by the fact that in almost all of the Dudley

SYSTEMATIC PALAEONTOLOGY

Class machaeridia Withers, 1926
Order turrilepadomorpha Pilsbry, 1916
Family turrilepadidae Clarke, 1896

ADRAIN ET AL.: TURRILEPADID M ACH AERIDI ANS

643

T. wrightiana specimens, the sclerite margins are either worn or broken off. Withers (1926, p. 41)
did note the occurrence of marginal spines on some well-preserved specimens, particularly BM(NH)
59164. Reinvestigation of this specimen (see PI. 1, figs 5, 7, 8) has shown that the spines are quite
obvious. It must be emphasized that, contrary to occasional appearance (see for example BM(NH)
59406, PI. 1, fig. 10), those Dudley specimens from which marginal spines appear to be absent have
margins which have either been poorly preserved, have been broken prior to preservation, or most
commonly have been broken and scored away during preparation in the nineteenth century. The
apices of almost all of the sclerites collected from Dudley are missing. An exception is a mid-
scleritome right inner sclerite on specimen BM(NH) 59406 (PI. 1, fig. 10), which demonstrates
unequivocally the presence of a minute, bifid apical spine.

Dzik (1986, fig. 4) has figured a poorly preserved sclerite from a Polish erratic boulder of Llanvirn
age. he has made the following claims for this specimen: (1) that it is an outer plumulitid sclerite;
and (2) that the marginal spines of the specimen coincide with the rugae. More material would be
necessary to evaluate this taxon with any confidence, but the sclerite almost certainly represents a
turrilepadid. Coincidence of the marginal spines and rugae cannot be properly evaluated on the
basis of the published evidence, since Dzik did not figure the outer sclerite surface, upon which the
rugae are borne. Faint ridges are evident on the inner surface, and these do appear broadly to match
the marginal spines. It is possible that in particular sclerites the spacing of rugae may fortuitously
approximate that of the marginal spines. The point that bears emphasis is that all other published
illustrations of turrilepadid sclerites clearly indicate that there exists no general one-to-one
relationship between the features, and their spacing, while it may occasionally match, is
independent.

Genus turrilepas Woodward, 1865

Type species. Chiton wrightiana de Koninck, 1857, from the Much Wenlock Limestone Formation of Dudley,
West Midlands. By monotypy.

Other species. T. modzalaevskae sp. nov.

Diagnosis. Turrilepadid machaeridians with minute apical spines; very small marginal spines, two
or more occurring between each ruga; large, robust sclerites; twenty-five or more rugae on a single
mature sclerite; Ij_— 15 all well expressed; Ij L sharp and angular, Ij R broadly curving; q-i3 all well
expressed, i2 generally shallow; large, deep muscle scars on inner sclerites; scleritome subrectangular
in mid-body cross section, with horizontal dorsal aspect and slightly ventromedially sloping lateral
aspects.

Remarks. Only the type species and T. modzalevskae are assigned with certainty to this genus. Hall
and Clarke (1888) erected eight new Devonian species, four of which were based upon single
sclerites. This material requires re-examination before any taxonomic conclusions can be reached.
Turrilepas witliersi Elias and T. japonicus Kobayashi and Hamada both seem closer to Deltacoleus ,
discussed below. Both species, however, are at present too poorly known for use in comparative
studies. Withers (1926) considered that material from Gotland figured by Aurivillius (1892)
belonged to T. wrightiana. Bengtson (1979) listed the occurrence of T. cf. wrightiana and T. n. sp.
from the Wenlock of Gotland, but the Gotland fauna remains to be investigated.

Turrilepas wrightiana (de Koninck, 1857)

Plate 1, figs 4, 5, 7, 8, 10-12; Plate 2, figs 6-9, 11-14
1857 Chiton wrightianus de Koninck, p. 199, pi. 1, fig. 2a-c.

1926 Turrilepas wrightiana (de Koninck); Withers, p. 37, pi. 5, figs 1-6; pi. 6, figs 1-8 (with full
synonymy).

644

PALAEONTOLOGY, VOLUME 34

71978 Turrilepas cf. wrightiana (de Koninck); Schrank, p. 10, pi. 1, fig. 3; pi. 2, figs 1 and 2; pi. 3, figs
1-3; pi. 4, fig. 1.

71979 Turrilepas cf. wrightiana (de Koninck); Bengtson, p. 211.

1986 Turrilepas wrightiana (Koninck); Dzik, p. 120, fig. Ib.

Holotype. BM(NH) 116283 (PI. 1, Fig. 11).

Localities and distribution. Known from the Much Wenlock Limestone Formation, Dudley and Malvern;
possibly occurs in the Wenlock of the island of Gotland, Sweden and an upper Wenlock erratic boulder from
the island of Riigen, Baltic area, German Democratic Republic.

Diagnosis. A species of Turrilepas with at least twenty-six segments; marginal spines on medial non-
accreting margins of inner sclerites spaced from two to four between each ruga; minute apical
spines; length of accreting margin of outer sclerites subequal to length of longitudinal angle of outer
sclerites.

Remarks. Schrank (1978) has figured some sclerites from an erratic boulder from the East German
Baltic region, assigned to the upper Wenlock on the basis of the associated trilobite fauna, as
Turrilepas cf. wrightiana. The sclerite illustrated in his plate 1, figure 3 does not belong to Turrilepas ,
but to another unnamed turrilepadid form, sclerites of which are known by the authors to occur in
the Wenlock of northern Canada. The remainder certainly belong to Turrilepas. In number and
spacing of marginal spines, the two inner sclerites figured by Schrank (1978, pi. 2, figs 1 and 2;
pi. 3, fig. 1) resemble T. wrightiana , although the outer surface of a typical inner sclerite was not
figured. The two outer sclerites, however, (Schrank 1978, pi. 3, figs 2 and 3) seem closer to
T. modzalevskae in general shape. Further specimens would be required to determine the specific
identity of this material.

Anterior structure. Withers (1926, p. 40) provided a discussion of the structure of the anterior end of the
scleritome of T. wrightiana. His examples were specimens BM(NH) 59164 (PI. 1, figs 5, 7, 8, 12) and BM(NH)
47871 (PI. 2, figs 12-14), the anterior of which he exposed through careful preparation. At the anterior of both
of these specimens there lies a small, almost flat sclerite with an odd orientation. This was interpreted by
Withers as one of a pair of ‘proximal’ sclerites, the anteriormost sclerites that were supposed to function in
attachment of the machaeridian to the substrate. We regard neither of these specimens as being true
representations of the actual scleritome arrangement. BM(NH) 47871 has very obviously been disturbed, and
although many of the modified anterior sclerites are present, the assemblage has been compacted, the

EXPLANATION OF PLATE 1

Figs 1-3. Deltacoleus crassus Withers, 1926. Medial, dorsolateral, and lateral views of holotype BM(NH)
In23708, inner right sclerite. Balclatchie Group, Balclatchie, Upper Ordovician, x 9.

Figs 4, 5, 7, 8, 10-12. Turrilepas wrightiana (de Koninck, 1857). 4, BM(NFI) 116308, left outer sclerite, view
of outer surface. Wenlock Limestone, Malvern, Middle Silurian, x 5. 5, 7, 8, 12, BM(NH) 59164, nearly
complete assemblage, Wenlock Limestone, Dudley, middle Silurian; 5, entire specimen, dorsolateral view,
x 2-5; 7, detail from mid-body, dorsal view, note marginal spines, x 5; 8, detail from anterior of assemblage,
dorsolateral view, note marginal spines, x 6; 12, detail of anterior end of assemblage, note loss of outer right
series and truncation of anterior end of assemblage, x4. 10, BM(NH) 59406, Wenlock Limestone, Dudley,
Middle Silurian, detail from midbody of articulated assemblage, right lateral view, note apical spines, x4.
11. BM(NH) 116283, holotype, Wenlock Limestone, Dudley, Middle Silurian, right inner and left inner
sclerites, Wenlock Limestone, Dudley, Middle Silurian, x 3 5.

Fig. 6. External mould, BM(NH) In23673, assigned to Deltacoleus crassus by Withers (1926), but of
indeterminate affinities. Inner left sclerite, Balclatchie Group, Dow Hill, Upper Ordovician, x4.

Fig. 9. BM(NH) In23737, assigned to Deltacoleus crassus by Withers (1926), in fact a plumulitid. Outer right
sclerite, internal mould with sclerite fragment, view of outer surface, Stinchar Limestone, Aldons, Middle
Ordovician, x 4.

PLATE 1

ADRAIN el al Deltacoleus crassus , Turrilepas wrightiana

646

PALAEONTOLOGY, VOLUME 34

relationships between sclerites distorted, and some sclerites lost. BM(NH) 59164 is likewise partially complete,
but is missing almost all of its anteriormost sclerites, in addition to the complete right outer series for the
anterior half of the sclerite assemblage.

A specimen that provides what appears to be an accurate idea of the anterior structure is known, however.
This is BM(NH) 116282 (PI. 2, figs 6-9, 1 1), a specimen mentioned but not figured by Withers (1926, p. 38).
The outer left series is concealed by the matrix in this example, but all other anterior sclerites seem to be
present. The structure of the specimen is described below.

Inner right series sclerites of segments 1-7 are present. Outer right series sclerites of segments 3-6 are present,
but some are fractured and fragmentary. Inner left series sclerites 1-5 are present, and an inner left sclerite of
an indeterminate more posteriorly placed segment is also evident, displaying a relatively long, bifid apical spine
(see especially PI. 2, fig. 9). Anteriorly in the sclerite series, the inner right sclerites tend to flatten out, and the
sclerite of segment 1 all but lacks a longitudinal angle. Conversely, the inner left sclerites become increasingly
folded, the longitudinal angle becoming acute. The inner left sclerites of segments 2 and 3 are intact on the
specimen. The sclerite of segment 1 has been broken just beneath its longitudinal angle (see PI. 2, fig. 1 1), and
the lateral portion has folded inward, giving the longitudinal angle a slightly more acute appearance than is
in fact the case. The outer right sclerites begin with segment 3, and except for a slight increase in the length
of the accreting margin posteriorly in the series, the initial segment is of very similar form to successive ones
in the series.

The sclerites of segment I are distinctive. The right sclerite has a very wide accreting margin relative to the
length of the longitudinal angle. Both sclerites have extremely closely spaced rugae. The inner sclerites of
segment 2 are intermediate in morphology between those of segments I and 3. By segment three, the sclerites
have attained the general form of successive sclerites in their respective series, and the major changes thereafter
are restricted to the longitudinal angles moving towards a stable value of approximately 90 degrees.

As regards Withers’s example, BM(NH) 47871, the ‘proximal plate’ with the anteriorly directed apex at the
anterior of the assemblage (PI. 2, fig. 12) is almost certainly the inner left sclerite of one of the anterior
segments, possibly 2 or 3, with most of its lateral portion broken off just beneath the longitudinal angle
(compare with the fractured segment 1 sclerite in PI. 2, fig. 1 1). The sclerites of segment 1 have been lost, and
the broken inner left sclerite has been displaced to overlie the accreting margin of the inner right sclerite of
segment 2. The inner left sclerite of this segment is intact and in proper spatial relation to its right side
counterpart. All of these sclerites, however, have been compacted and slid backward on the truncated
assemblage, so that they sit atop an inner right sclerite (see PI. 2, fig. 13), probably of segment 4.

Comparison of Plate 2, figure 6 with Jell’s (1979) figure 4d reveals an essential similarity in structure between
the segment 1 sclerites of turrilepadids and those of plumulitids. The sclerites are all but identical in shape and
the course of the rugae. Furthermore, both families lack outer series sclerites on the first two segments. Jell’s
(1979) material established this fact for plumulitids, while the present material provides strong evidence for this
condition in turrilepadids.

Accretionary growth. It is notable that within a given sclerite assemblage, all sclerites appear to bear
approximately the same number of rugae. On specimen BM(NH) 116282, described above, for instance, the
inner left sclerite of segment 1 has twenty-seven discernible rugae. The inner right sclerites of segments 1 and
2, both of which are incomplete, have nineteen and eighteen rugae respectively. The inner left sclerite of
segment 4 has twenty-five rugae, as does the inner right sclerite of segment 5. The outer right sclerite of segment
5 is incomplete, but has twenty-four discernible rugae. The rugae are very closely spaced on sclerites of the first

EXPLANATION OF PLATE 2

Figs 1-5, 10. Turrilepas modzalevskae sp. nov. BM(NH) In63310, Lower Ludlow, Konovka Formation,
Podolia, Ukraine, USSR, holotype. 1, dorsal view. 2, right lateral view. 3, left lateral view. 4, posterior view.
5, anterior view. All x4. 10, detail of medial aspects of inner sclerites, note rugae, growth lines, marginal
spines, and manner in which growth lines transect marginal spines, x 14.

Figs 6-9, IT 14. Turrilepas wrightiana (de Koninck). 6-9, 11, BM(NH) 116282, the anterior end of an
assemblage, Wenlock Limestone, Dudley, Middle Silurian; 6, detail of anteriormost sclerites, x4-5; 7,
dorsolateral view of medial aspects of inner sclerites, x 2-5. 8, dorsal view, x 2; 9, right lateral view, x 2-5.
11, view of lateral aspects of inner left sclerites, x4-5. 12-14, BM(NH) 47871, a disturbed anterior end of
an assemblage. Wenlock Limestone, Dudley, Middle Silurian; 12, the anteriormost sclerites, exposed by
Withers, see text for explanation, x8; 13, dorsal view, x2-5; 14, left lateral view, x2-5.

PLATE 2

ADRAIN et at. , Turrilepas modzalevskae, T. wrightiana

648

PALAEONTOLOGY, VOLUME 34

series, and there seems to be a fairly constant relationship between the size of the sclerite and spacing of the
rugae. All of this indicates that the scleritome very likely had a fixed number of segments, and that growth was
incremental, with rugae deposited one at a time, at the same time for all sclerites in the assemblage. The pattern
observed is incompatible with serial addition of segments. This information, if it can be corroborated through
further study, has important implications for reconstruction of machaeridian scleritomes when only
disarticulated sclerites are available, the general case in the palaeontological record. If this relationship holds
true, it can effectively be assumed that conspecific sclerites with a roughly equivalent number of rugae can be
accommodated together in the same scleritome reconstruction.

Turrilepas modzalevskae sp. nov.

Plate 2, figs 1-5, 10

Etymology. For Dr T. L. Modzalevskaya.

Holotype. BM(NH) In63310, a unique specimen from the Konovka Formation (lower Ludlow), Podolia (for
detailed locality see above), comprising parts of four segments from the mid-body of an articulated sclerite
assemblage.

Diagnosis. A species of Turrilepas with extremely small, closely set marginal spines, five to seven
between rugae of inner sclerites; I: crisply defined, about 100 degrees; at least twenty-eight rugae
on mature sclerite; accreting margin of outer sclerites about two-thirds to three-quarters length of
longitudinal angle.

Remarks. Turrilepas modzalevskae may be distinguished from T. wrightiana by a greater number of
smaller, more closely spaced marginal spines, rugae which tend to run straighter between inflections,
making the inflections slightly sharper, and especially by outer sclerites with accreting margins that
are much shorter relative to the length of the longitudinal angle, giving the sclerites a less erect, more
anteroposteriorly elongate appearance.

The specimen consists of an articulated portion of a sclerite assemblage, probably from some
point along the undifferentiated midbody area. Very little displacement has occurred among those
sclerites present, and the spatial relationships between sclerites appear to reflect those prevalent
during the life of the animal. The specimen comprises two nearly complete and three partially
complete inner right sclerites (the posteriormost having slipped medially to lie behind the sclerites
of the inner left series), three nearly complete outer right sclerites, one nearly complete and two
partially complete inner left sclerites, and one partially complete outer left sclerite.

Genus deltacoleus Withers, 1926

Type species. Deltacoleus crassus Withers. 1926, from the Caradoc of the Balclatchie Group, Girvan, Ayrshire.
Other species. None.

Remarks. Withers (1926) established Deltacoleus as a monotypic genus, based upon only five
disarticulated sclerites. Two were inner sclerites from the Upper Ordovician (Carodoc) of
Balclatchie, one of which he selected as the holotype. Three were outer sclerites from the Middle
Ordovician (Llandeilo) of Aldons. The holotype (PI. 1, figs 1-3) is reasonably well preserved, but
is missing most of its left non-accreting margin. It is an inner right sclerite, and has a distinctive
morphology as follows: medial area about three-quarters the size of lateral area, compared with
about one-half in Turrilepas ; longitudinal axis gently curving, compared with essentially straight in
Turrilepas ; I,R quite angular, not as broadly curving as in Turrilepas ; I2R very shallow, occurring
on medial area, while deep and placed at or lateral to the longitudinal angle in Turrilepas ; I:iR
occurring on longitudinal angle; I4R and I5R spaced respectively about one-third and two-thirds of

ADRAIN ET AL.\ TURRILEPADID M ACH AERI DI ANS

649

the way between longitudinal angle and medial non-accreting margin, both shallow; at least thirty-
six rugae; apical portion of sclerite slightly beaked.

The second inner sclerite, from the inner left series, is represented by internal and external moulds
(the external mould figured in Plate 1, figure 6, the internal figured by Withers 1926, plate 6). While
it shares a curving longitudinal axis with the first sclerite, it is otherwise much larger and of quite
different appearance, with IjL placed very near to the longitudinal angle, and I2L-I5L all occurring
on the lateral aspect of the sclerite, as in Turrilepas. This sclerite is almost certainly not conspecific,
and probably not even congeneric, with the holotype.

The outer sclerites are from much older rocks. The example figured by Withers (1926, pi. 8,
fig. 8) is an internal mould with a fragment of sclerite material preserved of a left outer plumulitid
sclerite. Withers (1926, p. 44) declined to refer it to the plumulitids because the specimen lacked a
longitudinal fold, a feature he considered diagnostic. As is evident from the rephotographed
specimen (PI. 1, fig. 9), a weak dorsal inflection is present, consisting of two ‘longitudinal folds’, one
at the midpoint between the non-accreting margins and one placed near the medial non-accreting
margin, with the thickness of the sclerite inflected slightly dorsally between them. Comparison of
the sclerite with the as yet undescribed silicified sclerite of Text-figure 1e confirms that it belongs
within the Plumulitidae.

Hence, Deltacoleus appears to have been erected as something of an odds-and-ends taxon.
Clearly, the concept of the genus must be based solely upon Withers’s holotype. This is an
unsatisfactory basis for comparison and more material would be required from the type area to
properly delineate the taxon. The sclerites figured by Schallreuter (1985, pi. 1, figs 2-5) as
Deltacoleus erraticus and Deltacoleus^. crenulatus bear little resemblance to the holotype, and should
be excluded from the genus. Schallreuter’s (1985, pi. 1, fig. 1) D. cf. crassus is very similar to the
holotype. This single sclerite certainly belongs to Deltacoleus. Whether or not it is conspecific with
D. crassus can only be determined through further collecting of both the Scottish and German
material. Dzik (1986, fig. 7) has figured three specimens from the Llanvirn and Arenig of Poland
as Deltacoleus cf. crassus. The Llanvirn specimens (Dzik 1986, fig. 7a, b) seem conspecific (they are
from the same erratic boulder), and may belong to Deltacoleus , although likely not to D. crassus.
The affinities of the Arenig fragment (Dzik 1986, fig. 7c, d) are indeterminate.

In sum, before this genus can be of any use in comparative phylogenetic studies, thorough
redescription of all its possible members will be required.

Genus clarkeolepis Elias, 1958

Type species. Designated herein as Clarkeolepis clarkei Elias, 1958, from the Mississippian Redoak Hollow
Formation of Oklahoma.

Remarks. Elias (1958) described three machaeridian species, Turrilepas whithersi (discussed above),
Clarkeolepis clarkei , and C. elegans. As he did not name a type species for the genus, C. clarkei is
selected herein by virtual tautonomy and pagination. C. clarkei is known from two figured sclerites,
and C. elegans from one. All three are likely conspecific, and hence C. elegans should likely be
considered a subjective junior synonym of C. clarkei.

The sclerites assigned to Clarkeolepis are certainly machaeridians, but are characterized by a very
low number of rugae (about ten), and an odd raised ornament apparently transecting the spaces
between the rugae. More material and more thorough treatment is needed before any taxonomic
conclusions, save to assert that the sclerites seem to belong in the Turrilepadidae, can be reached.

Genus spinacoleus Schallreuter, 1985

Nomenclatural note. Schallreuter (1985) used the genus name Spinacoleus in his text, but Rugacoleus in the
explanation of his figures. Article 24(6) of the ICZN requires that the first reviser cites the alternative names

650

PALAEONTOLOGY, VOLUME 34

together and chooses one as the correct name. Dzik’s (1986) listing of Spinacoleus did not fulfil this requirement
and is therefore not binding. In the interests of nomenclatural stability, however, Spinacoleus is selected here
as the valid name.

Type and only species. Spinacoleus breugmanni Schallreuter, 1985, from Ojlemyrflint erratic boulder No. G16.

Remarks. The holotype and paratype material (Schallreuter 1985, pi. 2, figs 1-3) is poorly preserved.
As mentioned above, the sclerite margins are missing from this material, giving the false impression
that the rugae are extended into marginal spines. A more accurate impression is given by the better
preserved sclerites questionably assigned to the species by Schallreuter (1985, pi. 2, figs 4, 6, 7), in
which the rugae clearly terminate short of the sclerite margin, and the marginal spines are distinct
and not spaced in any set relation to the rugae.

Schallreuter’s (1985) material is without question properly associated and belongs to a distinctive
turrilepadid form. Although it remains inadequately known, Spinacoleus breugmanni is sufficiently
well described to allow comparison with other machaeridians. Hence, further work on machaeridian
taxonomy may lead to a better understanding of this genus.

Genus mojczalepas Dzik, 1986

Type and only species. Mojczalepas multilamellosa Dzik, 1986, from Llanvirn strata (E. reclinatus Zone),
Mojcza Limestone, Holy Cross Mountains, Poland.

Remarks. Dzik (1986, fig. 5) figured only two fragmentary and poorly preserved inner sclerites when
he erected this taxon. Unfortunately, these sclerites give no satisfactory idea of the structure, shape,
or pattern of rugae of the originals. As such, there does not seem to exist an objective basis for
reference of further material to this taxon or for comparison with other turrilepadid forms.
Satisfactory incorporation of this taxon into a general scheme of machaeridian classification must
await adequate redescription and illustration.

REFERENCES

adrain, j. m. In press, Machaeridian classification. Alcheringa.

auri villius, c. w. s. 1892. Uber einige ober-silurische Cirripeden aus Gotland. Bihang till Kongliga Svenska
Vetenskaps-Akademiens Handlingar ( Stockholm ), 18 (4), 1-24.
bengtson, s. 1977. Aspects of problematic fossils in the Early Palaeozoic. Acta Universitatis Upsaliensis.
Abstracts of Uppsala Dissertations from the Faculty of Science , 415, 1-71.

— 1978. The Machaeridia - a square peg in a pentagonal hole. Thalassia Yugoslavia , 12, 1-10.

1979. Machaeridians. 211-212. In jaanusson, v., laufeld, s. and skoglund, r. (eds). Lower Wenlock
faunal and floral dynamics - Vattenfallet section, Gotland. Sveriges Geologiska Undersokning Arsbok , 73,
1-294.

1985. Taxonomy of disarticulated fossils. Journal of Paleontology 59, 1350-1358.

— and conway morris, s. 1984. A comparative study of Lower Cambrian Halkeria and Middle Cambrian
Wiwaxia. Lethaia , 17, 307-329.

clarke, J. m. 1896. The structure of certain Palaeozoic barnacles. American Geologist , 17, 137-143.
de koninck, L. 1857. Sur deux nouvelles especes siluriennes appartenant au genre Chiton. Bulletin de
T Academie des Sciences de Belgique , (2). 3, 190-199.

— 1860. Observations on two new species of chiton from the Upper Silurian ‘Wenlock Limestone’ of
Dudley. Annals and Magazine of Natural History , (3), 6, 91-98.

dzik, j. 1986. Turrilepadida and other Machaeridia. 116-134. In hoffman, a. and nitecki, m. h. (eds).

Problematic fossil taxa. Clarendon Press, Oxford, 267 pp.
elias, m. k. 1958. Late Mississippian fauna from the Redoak Hollow Formation of southern Oklahoma,
part 4, Gastropoda, Scaphopoda, Cephalopoda, Ostracoda, Thoracica, and Problematica. Journal of
Paleontology , 32, 1-57.

ADRAIN ET AL.: TURRILEPADID MACHAERIDIANS

651

hall, J. and clarke, J. M. 1888. Geological Survey of the State of New York. Palaeontology. Volume 7. Charles
Van Benthuysen and Sons, Albany, 236 pp.

jell, p. a. 1979. Plumulites and the machaeridian problem. Alcheringa , 3, 253-259.

kobayashi, t. and hamada, t. 1976. Occurrences of the Machaeridia in Japan and Malaysia. Proceedings of
the Japan Academy , 52, 371-374.

KOREN, T. N., ABUSHIK, A. F., MODAZLEVSKAYA, T. L. and PREDTECHENSKY, N. N. 1989. Podolia. 141-149. In
bassett, M. G. and Holland, c. H. (eds). A global standard for the Silurian System. National Museum of
Wales, Geological Series No. 9, Cardiff, 325 pp.

over, d. j. and chatterton, b. d. e. c. 1987. Silurian conodonts from the southern Mackenzie Mountains,
Northwest Territories, Canada. Geologica et Palaeontologica, 21, 1-49.

pilsbry, h. a. 1916. The sessile barnacles (Cirripedia) contained in the Collections of the U.S. National
Museum. Bulletin of the United States National Museum , 93, xii+ 1-366.

pope, j. k. 1975. Evidence for relating the Lepidocoleidae, machaeridian echinoderms, to the mitrate carpoids.
Bulletins of American Paleontology, 67, 385-406.

schallreuter, r. 1985. Mikrofossilien aus Geschieben IV. Machaeridier. Der Geschiebesammler , 18, 157-171.

schrank, e. 1978. Machaeridia as silurischen Geschieben. Der Geschiebesammler, 11, 5-22.

TSEGELNJUK, P. D., GRITSENKO, V. P., KONSTANTINENKO, L. I., ISCHENKO, A. A., ABUSHIK, A. F., BOGOYAVLENSKAYA,
O. V., DRYGANT, D. M., ZAIKA-NOVATSKY, V. S., KADLETS, N. M„ KISELEV, G. N. aild SYTOVA, V. A. 1983. The
Silurian of Podolia. Science House, Kiev, 224 pp.

withers, T. H. 1926. Catalogue of the Machaeridia (Turrilepas and its allies) in the Department of Geology.
British Museum (Natural History), London, xv + 99 pp.

- 1933. The machaeridian Lepidocoleus ketleyanus (Reed ex. Salter). Annals and Magazine of Natural
History , (10), 11, 162-163.

woodward, H. 1865. On the discovery of a new genus of Cirripedia in the Wenlock Limestone and Shale of
Dudley. Quarterly Journal of the Geological Society of London, 21, 486-489.

J. M. ADRAIN
B. D. E. CHATTERTON

Department of Geology
University of Alberta
Edmonton, Alberta, T6G-2E3
Canada

Typescript received 7 February 1990
Revised typescript received 10 August 1990

L. R. M. COCKS

Department of Palaeontology
The Natural History Museum
Cromwell Road
London SW7 5BD, UK

MOSASAURS FROM THE UPPER CRETACEOUS OF

NIGER

by T. LINGH A M-SOLI AR

Abstract. Fragmentary remains from the Dukamaje Formation (‘ Mosasaurus Shales’), Upper Maastrichtian,
S.W. Niger, reveal a diverse mosasaur fauna of at least six genera - Goronyosaurus , Igdamanosaurus nov. gen.,
Angolasaurus , Halisaurus , Plioplatecarpus , Mosasaurus , and possibly Platecarpus. This represents an
astonishingly high number of mosasaur genera from a single horizon, equalling that of the most prolific
mosasaur beds of the world, the Niobrara Formation of Kansas, U.S.A. and the Craie Phosphatee de Ciply,
Belgium. Plioplatecarpus represents the first documentation in Africa of this genus.

Mosasaurs were large marine lepidosaurian reptiles which, in a brief 25 million years, dominated
the Late Cretaceous seas. Remains are found on all continents including Antarctica (Chatterjee
et al. 1984). On the basis of skull form and, to a lesser extent, postcranial characters, they have been
divided into three subfamilies, the Mosasaurinae, Tylosaurinae, and Plioplatecarpinae. Russell
(1967) established, in addition, several tribal categories, for example, the Mosasaurini and
Globidentini.

The majority of mosasaur remains occur in North America, in particular the Niobrara Chalk of
Kansas, and the Gulf Coast, the Phosphatic Chalk of Ciply, Belgium, and the Tuffeau of Maastricht,
the Netherlands ( Russell 1 967, charts 4, 5 ; Lingham-Soliar and Nolf 1 989). Most other areas of the
world have yielded isolated finds. Flowever, two further areas, in Africa, are noteworthy, with
perhaps the best mosasaur material known outside the North American and European sites: the
‘ Mosasaurus Shales’ (Upper Maastrichtian, Dukamaje Formation) of Sokoto State, N.W. Nigeria
(Swinton 1930; Azzaroli et al. 1972, 1975; Halstead 1979; Soliar 1988) and the grey limestone
coastal cliffs (Upper Turonian, ‘Camadas do Tadi’) of Iembe, Angola (Antunes 1964; Lingham-
Soliar, in prep.).

In 1988 a joint expedition to Niger by the British Museum (Natural History), London, and
Kingston Polytechnic recovered mosasaur remains, which Cyril Walker of the BMNH, the leader
of the expedition, and responsible for collecting some significant specimens, placed at my disposal.
The material, although fragmentary, was the first record of reptiles from the Maastrichtian of Niger.
The only previous documentation of vertebrates from this horizon is of fishes (Tabaste, 1963;
Cappetta 1972).

Repository abbreviations are: BMNH, British Museum (Natural History); GSN Geological Society of
Nigeria; IRSNB, Institut Royal des Sciences Naturelles de Belgique; IGF, Institute of Geology and
Paleontology of the University of Florence; YPM, Yale Peabody Museum. All the material described, unless
otherwise stated, is housed in the BMNH.

GEOLOGY

The following outline of the geology of Niger is my interpretation of information from David Ward and
Richard Moody. The three principal localities. In Tahout, Mt Igdaman, and Kehehe, lie in the south-western
corner of Niger, in an area bordering Dahomey in the south west, Nigeria in the south and Mali in the west
(Text-fig. 1). This region was once part of the large West African intracontinental sedimentary basin, the
Iullemmedan Basin. According to Petters (1977, 1979a), the subsidence of the South Central Saharan platform
during Maastrichtian times created a marine embayment of the Tethys sea with its southern limits in the south-

IPalaeontology, Vol. 34, Part 3, 1991, pp. 653-670.|

© The Palaeontological Association

654

PALAEONTOLOGY, VOLUME 34

Ward pers. comm.). The Iullemmedin Basin (shaded) is figured after Petters (1977).

western part of Niger Republic and extending into the north-western tip of Nigeria. Other authors (Reyment
1965; Kogbe 1973; Adeleye 1975; Howarth 1 98 1 ), in contrast, take the view that the trans-Saharan seaway was
open-ended, joining the Mediterranean with the Gulf of Guinea.

The Maastrichtian horizon of N.W. Nigeria has been described in detail (reviewed Soliar 1988). Kogbe
( 1 979) reported that ‘ . . . similar fossil beds also occur in Niger Republic . . . ’ where they ‘ . . . outcrop extensively ’.
However, there is no record, to my knowledge, of detailed stratigraphical work on this horizon in Niger besides
brief mention by Greigert (1966) and Krasheninnikov and Trofimov (1969). The three localities represent
exposures of the Late Cretaceous (Maastrichtian) Dukamaje Formation (David Ward, Richard Moody, pers.
comm.). A detailed stratigraphic section is available for Igdaman and tentative results for the other two areas
(Text-fig. 2).

In Tahout (5° 52' 15° 22'), near Kao, a previously unrecorded locality, shows exposures of a complete series
of sections of the Dukamaje Formation, with the mosasaurs coming from the lower levels on or very near the
Cretaceous/Tertiary boundary. Above this lie the Palaeocene phosphate beds. Kehehe (5° 36' 15° 2')
represents an outcrop of Late Cretaceous ‘ Mosasaurus Shales’, the beds resembling those described at Gilbedi
and Kaffe (Reyment 1965). Of the three localities, Mt Igdaman (5° 50' 15° 23'; Greigert 1966) yielded the
largest amount of mosasaur material. Earlier work by Cappetta (1972) has shown the horizon to be abundant
in inshore and marine fish remains which were apparently collected from a single thick phosphatic bed (David
Ward, pers. comm.). Outcropping all the way around the hill are a series of bone beds from which the reptile
and fish remains were obtained. This ‘ Mosasaurus Shales’, or Dukamaje Formation, is both overlain and
underlain by loose sandstones equivalent to the Nigerian Wurno and Taloka Formations respectively (Text-
fig. 2).

In addition to mosasaur material, the above horizons yielded the remains of a number of vertebrates, some
similar to those found in the ‘ Mosasaurus Shales’ of Nigeria: the sea snake Palaeophis , pelomedusid turtles,
sharks (including sawfish sharks and rays), cat fish, and the remains of the marine teleost, Stratodus (David
Ward, pers. comm.)

LINGHAM-SOL1AR: MOSASAURS FROM NIGER

655

text-fig. 2 Stratigraphic section at Igdaman (David Ward pers. comm.) with approximate extent of sections

at In Tahout and Kehehe.

SYSTEMATIC PALAEONTOLOGY

Order squamata
Family mosasauridae
Subfamily Incertae Sedis
Genus Goronyosaurus Azzaroli et al., 1972

Goronyosaurus sp.

Text-figures 3, 4

Referred material. Premaxilla and two fragments of maxilla ( R 1 1 909), probably associated. Poorly preserved
dentary ( R 1 1947) with tooth sockets and remains of teeth, probably from a different individual. Mt Igdaman.

Revised diagnosis. Premaxilla lacking rostrum extends on to wide internarial bar; closed canal for
olfactory lobes; ectopterygoid processes dorsoventrally flattened, two fork-like processes; snout
approximately 55% total length of skull; maxillary teeth no more than 1 1, extend close to posterior

656

PALAEONTOLOGY, VOLUME 34

border of the orbit; deep lateral interdental pits; little tapering of the dentary anteriorly; very high
number of large foramina on premaxilla; elongated anterior tooth bases.

Description. A series of unique interdental pits on the jaw rami of specimens R11909 and R1 1947, and on the
dental rami of the holotype of G. nigeriensis (IGF 14750, Azzaroli et al. 1975; Soliar 1988) characterise the
genus Goronyosaurus.

Premaxilla. The premaxilla R11909 (Text-fig. 3a, b) is broken at a point just anterior to its union with the
maxilla. No teeth are present, only four large sockets. The most striking condition of the premaxilla is the
presence of a large deep interdental pit located laterally between the first and second premaxillary teeth (Text-
fig. 3b). The premaxilla also shows an unusually high number of large foramina for the exits of cranial nerve
V (ophthalmic branch), approximately 15 on either side, situated on the lateral and anterodorsal surfaces and
extending close to the midline. The premaxilla is blunt, similar to that of Prognathodon (Russell 1967;
Lingham-Soliar and Nolf 1989).

Despite poor surface preservation, the right premaxilla of G. nigeriensis (IGF 14750) shows approximately
10 foramina in a small area. In other mosasaurs, the number of large foramina generally does not exceed 6-7
on either side. In tylosaurs such as Tylosaurus ( Liodon ) dyspelor (Cope 1875, pi. 28), there are few large
foramina, the majority being small, situated in a cluster on the lower lateral and ventral surface of the
premaxilla.

Maxilla. Two well preserved fragments of maxilla, probably part of the same specimen R11909 (Text-fig.
3c, d), reveal the characteristic deep interdental pitting. Tooth crowns are not preserved, only an unusually
long and well preserved tooth base (Text-fig. 3d) reminiscent of the one noted in the maxilla of the holotype
of Goronyosaurus (Soliar 1988, fig. 3a; Text-fig. 10a). There are no more than eleven maxillary teeth in IGF
14750 although arguably there may be as few as nine if the two teeth at the end of the ramus are in fact
displaced posterior pterygoid teeth (Text-fig. 10a).

Dentary. The left dentary ( R 1 1 947) is complete, although fractured into four separate fragments (Text-fig.
3f, g), and represents the first complete dentary of Goronyosaurus to be reported. It is fairly slender and
scarcely tapers anteriorly, contrasting with the usual description of the element in mosasaurs : ‘ . . . an anteriorly
narrowing girder of bone...’ (Russell 1967, p. 49). Medially, the excavation for Meckel’s cartilage extends up
to the second dentary tooth. All that remains of the teeth are six tooth bases and six poorly preserved relatively
straight tooth crowns (Text-fig. 3f, g). The strongest indication that the dentary is referable to Goronyosaurus
lies in the deep lateral interdental pits, extending to about the seventh tooth (Text-fig. 3g), which accords
closely with those of the maxilla and the condition seen in the holotype material. Marginal teeth of R1 1909,
R1 1947, and of the holotype are widely spaced.

Discussion. In occlusion, the jaws fitted tightly together, with the teeth overlapping their opposite
numbers and fitting, uniquely for mosasaurs, into the deep interdental excavations (Text-fig. 4a).
Plotosaunis (Camp 1942, pi. 1) shows a form of dental interdigitation, although there are no
excavations for the opposing dentition - the battery of teeth (the highest number found in
mosasaurs, 18 on the maxilla and 17 on the dentary) presumably formed a fish trap resembling that
seen in plesiosaurs (McFarland et al. 1979; Brown 1981). In certain smaller specimens of
Mosasaurus hoffmanni (e.g. IRSNB 1559, ‘ R 1 2 ’) a form of shallow interdental pitting is observed,
but it is scarcely comparable with that of Goronyosaurus. However, Marsh (1869, p. 394) described
'the unusual depth of pits’ on the outer superior edge of the lower jaw of Mosasaurus copeanus
(synonymised Plioplatecarpus depressus; Russell 1967), the only record of this character in the
literature, to my knowledge. Further study of this material is essential as it is probable that it
belongs to a form related to Goronyosaurus.

The long snout of Goronyosaurus indicates that the adductor muscles originated relatively far
back on the robust temporal arcade (Text-figs 4, 10). It is probable that static or bite forces were
consequently reduced (Olson 1961; Greaves 1983), but that speed of jaw closure was conversely
increased. A similar form of ‘snapping’ jaw mechanism was apparently prevalent in long-snouted
marine crocodiles such as the Palaeocene Rhabdognathus and Dyrosaurus of the Sokoto area in
N.W. Nigeria (Buflfetaut 1976, 1979, p. 35). This technique of rapid impaling of prey was

LINGH AM-SOLI A R: MOSASAURS FROM NIGER

657

6

text-fig. 3. Goronyosaurus (A-E, R11909). a, b, dorsal and lateral views of premaxilla, c, d, maxilla fragments, e, ventral view of c. x 1.

R1 1947, medial and lateral views of left dentary. xO-5.

658

PALAEONTOLOGY, VOLUME 34

100mm

text-fig. 4. Reconstruction of the skull of Goronyosaurus nigeriensis based on specimens IGF 14750, BMNH
R1 1909, and R1 1947. a, jaws in occlusion, b, jaws open, x 0 2.

undoubtedly effective in capturing fairly evasive, moderately small, reptiles and fishes. The long
pointed, relatively straight teeth in the dentary ( R 1 1903). long tooth base in the maxilla ( R 1 1909)
and in the holotype (Soliar 1988; Text-fig. 10a), are consistent with such an hypothesis (Text-fig.
4).

The condition of interdental pits in Goronyosaurus , first figured in the holotype (Soliar 1988, fig.
3), represents an autapomorphy of the genus. The condition has also enabled a re-evaluation of
previously mis-identified mosasaur material. For instance, the dentary of ‘G. ’ nigeriensis, figured by
Azzaroli et al. (1975, pi. la, b), lacks interdental pits, suggesting that it is not referable to the genus,
whereas a previously undescribed fragment of dentary (IGF 14751/2), revealing a series of deep
external interdental pits, clearly is. Furthermore, two highly gypsiferous dentary fragments
(R5682, R5683) referred to Mosasaurus nigeriensis (Swinton 1930), demonstrate the characteristic
interdental pits of Goronyosaurus.

LINGHAM-SOLIAR: MOSASAURS FROM NIGER

659

Subfamily ?plioplatecarpinae
Igdamanosaurus gen. nov.

Type species Igdamanosaurus aegyptiacus Zdansky, 1935
Text-figure 5

Etymology. Genus named after the village of Igdaman, near to which the specimen was found.

Holotype. BMNH R1 1898, consisting of three poorly preserved fragments of jaw, probably comprising a single
specimen, which include two almost complete teeth and the remains of five more, and three tooth bases.

Horizon and locality. Mt Igdaman, near the village of Igdaman (sometimes In Dama), 5° 50', 15° 22'.

Diagnosis. Massive dentary with dome-shaped unwaisted very finely striated teeth; ?anterior and
posterior carina on teeth, sub-circular cross-sections.

Description. R11898, a right dentary, is massively proportioned (Text-fig. 5). Medially, a shallow recess
indicates the groove for the splenial. There are seven teeth in various states of preservation (plus three tooth
sockets) progressively increasing in size in an anteroposterior direction, except for the last, which is slightly
smaller than the penultimate of the preserved teeth. The teeth are straight broad cones with rounded domed
tips. The signs are that carinae are present but poor preservation makes this hard to confirm. In cross section
the teeth are subcircular. Tooth crowns have unwaisted bases and are covered by fine parallel ribbing or striae,
approximately 65-70 per tooth. The largest tooth is approximately 27-2 mm high and 22-5 mm wide at the
base. The tooth form is clearly uniform along the entire segment of jaw (Text-fig. 5).

text-fig. 5. Right dentary of Igdamanosaurus. A, dorsal view, b, medial view.

Discussion. Two unusually small foramina for the exits of the mandibular nerve are distinguishable
on the lower lateral surface of the dentary. Absence of dentary material in Globidens precludes any
comparison, although there are similar, relatively small foramina in a maxilla of the holotype of G.
cilabamaensis (Gilmore 1912, pi. 39).

The broad domed teeth, unconstricted at the base, and parallel fine striations of R 1 1 898 are
unusual for the Globidentini (cf. G. alabamanensis Gilmore 1912, G. dakotaensis Russell 1975). The

660

PALAEONTOLOGY. VOLUME 34

teeth of R1 1898 on the other hand show some resemblance to those of "G timorensis (Huene 1935),
but even more to the teeth of ‘G. ’ aegyptiacus (Zdanski 1935, pi. 2). The sharing of such an unusual
tooth morphology by R1 1898 and ‘G. ’ aegyptiacus suggests that they should be included in the same
genus. The status of ‘ G. ’ aegyptiacus , however, was questioned by Russell (1975, p. 240). ‘These
teeth are quite unlike those in the type skull of G. alabamaensis and G. dakotaensis , and could belong
to an unrecognized genus of durophagous mosasaur’. In view of this, it seems appropriate now to
erect a new genus, Igdamanosaurus , for the reception of R 11898 and ‘ G. ’ aegyptiacus.

Russell (1967, p. 144) considered Globidens a derivative form of C/idastes but ‘because of the
highly peculiar nature of its spherical teeth’, separated the genus into a new tribe of the
Mosasaurinae, the Globidentini (Russell 1975). In contrast, vertical striae in I. aegyptiacus suggest
that it represents an entirely new form of durophagous mosasaur derived from a Platecarpus- like
ancestor and the genus is consequently assigned to the subfamily Plioplatecarpinae.

Subfamily plioplatecarpinae
cf. ? Angolasaurus Antunes, 1964

Test-figure 6

Referred material. Four vertebrae, BMNH R 1 1 90 1 , R11902, R 1 1 903 , R 1 1 904, of varying sizes, the smallest
representing an immature individual. Mt Igdaman.

Description. R 11901 is a well preserved almost complete cervical vertebra (Text-fig. 6a-c) which in general
configuration compares with Angolasaurus bocagei (Antunes 1964, pi. 22). The neural spine is very narrow,
with well preserved anterior and posterior zygapophyses. The lateral edges of the anterior zygapophyses extend
smoothly on to the transverse processes. Distally, the synapophyses are not preserved, but the bases indicate
that they were dorsoventrally compressed. Condyles and cotyles are shallow, approximately heart-shaped with
the dorsal surfaces depressed slightly, as in Platecarpus. The lateral configuration of the hypapophyseal
peduncle is similar to that of A. bocagei and Platecarpus , with the ventral surface sub-circular (Antunes 1964,
pi. 22, fig. lb, c). However, unlike A. bocagei , zygosphenes are not observed in R1 1901, although relatively poor
preservation in this part of the vertebra makes identification inconclusive.

R11902 (Text-fig. 6d) is a small, fragmentary cervical vertebra probably belonging to an immature
individual; it is in general similar to R11901. The main difference (which may be ontogenetic) lies in the
hypapophyseal peduncle, which is not as deep as in R1 1901 but much broader laterally, spanning almost the
entire length of the centrum. The ventral surface of the peduncle unlike that of R1 1901 is concave rather than
convex. The cotyles seem to be fairly flat, but may be exaggerated because of weathering. Specimens R11903
and R1 1904 (Text-fig. 6e, f) are posterior cervical or early dorsal vertebrae. The main differences from R1 1901
and R 11902 lie in the anteroposterior rather that dorso ventral flattening of the synapophyses. All the above
vertebrae show strong cartilaginous articulations (indicated by pitting), suggesting a possible immature state
of development.

Discussion. The vertebrae of Angolasaurus share broad similarities with those of Platecarpus ,
consistent with other characters that the two genera have in common : very large suprastapedial
process of the quadrate, deep groove or foramen for basal artery on basioccipital, long striated
teeth, and elliptical or heart-shaped articulations of centrum and free haemal spines. For the
present, the above characters clearly suggest that Angolasaurus should be referred to the subfamily
Plioplatecarpinae and not, as indicated by Antunes (1964, p. 164), to the Mosasaurinae.

Subfamily plioplatecarpinae
Genus Halisaurus Marsh, 1869

Halisaurus sp.

Text-figure 7a-d

Referred material. Vertebrae, BMNH R11983, R I 1906, R1195I and R 1 1 952. Mt. Igdaman.

LINGH AM-SOLI A R: MOSASAURS FROM NIGER

661

text-fig. 6. Angolasaurus anterior cervical vertebrae, a, b, c, R1 1901. posterior, lateral, and anterior views, d
R 1 1 902, small anterior cervical vertebra. E, F R11903, 11904 posterior cervical or early thoracic vertebrae.
d, e, f in anterior (top) and lateral (bottom) views, x 1.

Description. R1 1923 (Text-fig. 7a-d) is an anterior dorsal, or possibly posterior cervical, vertebra. Much of the
neural spine and left synapophysis are absent. The right synapophysis is stout, considerably longer than in
other mosasaurs, and situated posteriorly on the centrum. The only preserved posterior zygapophysis is
moderately weak. Characteristically, the neural canal is large and the condyle and cotyle nearly twice as wide
as deep, nearly kidney bean-shaped with a slight emargination dorsally. There is no evidence of zygosphenes
or zygantra.

PALAEONTOLOGY, VOLUME 34

text-fig. 7. a-d, Halisaurus vertebra (BMNH R1 1983). a, lateral view; b, ventral view; c, anterior view, d,
posterior view, x 0-5. e-g, Plioplatecarpus sp. (BMNH R1 1984). e, anterior view; f, lateral view; G, posterior

view.

Discussion. R 1 1951, 1 1952, and 1 1906 are the smallest known Halisaurus vertebrae (cf. Bukowski
1984; table 1 ). They are placed in the genus Halisaurus on the basis of the characteristically flattened
condyles and cotyles, although the problem of identification is augmented here by small size coupled
with the generally primitive nature of Halisaurus vertebrae (Bukowski 1984). For instance, the
relatively large neural canal and relatively broad span across the zygapophyses generally accord
with conditions seen in less highly modified squamate families. Flowever, a somewhat less

LINGHAM-SOLIAR: MOSASAURS FROM NIGER

663

pronounced state of these features in R 1 1923, R 1 1951, R 1 1952 suggests that the vertebrae belong
to Halisaurus.

Halisaurus was poorly known, being represented by isolated remains, mainly vertebrae and a few
fragments of skull, until Russell (1970) assigned ‘ Clidastes" sternbergi , a complete skeleton in the
University of Uppsala) to the genus. Until it is studied in detail, ‘C.’ sternbergi is excluded from
consideration of Halisaurus.

In a label in the IRSNB, Russell referred the Belgian mosasaur Phosphor osaurus ortliebi
(IRSNB R34, formerly 4671; Dollo 1889), in my view correctly, to Halisaurus. It seems that this
assignment was based on the similarity of the frontal of IRSNB R34 to that described by Baird and
Case (1966, p. 1212), which they indicated had been identified by Russell as H. platyspondylus. In
contrast, IRSNB R34 and ‘C. ’ sternbergi differ in the size and the position of the parietal foramen.
In IRSNB R34 the parietal foramen is relatively large, approximately 24 mm in diameter (Dollo
18896, pi. 9, fig. 6), one of the largest in the Mosasauridae, and it is situated on the fronto-parietal
suture and partly on the frontal. On the other hand, in C. sternbergi , ‘the foramen parietale is
small and lies on the boundary between the first and second third of the parietale, thus considerably
further back than in Clidastes velox and other mosasaurians’ (Wiman 1920, p. 14, fig. 3). These
differences in the size and position of the parietal foramen suggest that Halisaurus and ‘ C. sternbergi
may not belong in the same genus. Estes et al. (1988, pp. 148-149) and Lingham-Soliar and Nolf
(1989) discuss the apomorphic nature of these characters in squamates and mosasaurs respectively.
Provisional assignment of Halisaurus to the subfamily Plioplatecarpinae, indicated in a previous
cladistic analysis (Soliar 1988), is retained.

Subfamily plioplatecarpinae
Plioplatecarpus Dollo, 1882
Plioplatecarpus sp.

Text-figures 7e-g, 8a-d

Referred material. Fairly well preserved mid-cervical vertebra, BMNH R I 1924. Mt Igdaman. A posterior
caudal vertebra, BMNH R 1 1 982. Numerous vertebrae probably belonging to a single specimen, BMNH
R 1 1 950. In Tahout, top half of ' Mosasaurus Shales’.

Description. R11924 (Text-fig. 7e-g) consists of a complete fairly well preserved vertebra probably from the
mid-cervical vertebral region. The anterior and posterior zygapophyses are well developed. The neural canal
is large, tall, and moderately broad and the neural spine posteriorly directed. A fairly slender hypapophyseal
peduncle is situated on the posterior ventral surface of the centrum. The peduncle is sub-circular in cross-
section. Synapophyses are quite slight and short with the dorsal surface continuing smoothly onto the anterior
zygapophyses. The articulating surfaces of the centrum show the characteristic elliptical shape of cervical and
early trunk vertebrae found in members of the genus Plioplatecarpus , the width approximately one and a half
times the depth.

The caudal vertebra (R 11922, Text-fig. 8a) is poorly preserved, with most of the neural spine, the left
transverse process, and the distal tip of the right transverse process absent. The shape of the condyle is
markedly hexagonal. Haemal peduncles indicate the presence of free haemal spines which characterise
plioplatecarpines and tylosaurines. The size of the specimen (unless it represents an immature individual) and
the sharply defined hexagonal configuration of the condyle, confirm assignment to Plioplatecarpus rather than
to Tylosaurus. R1 1950 (Text-fig. 8b-d) represents a series of partly articulated vertebrae from different parts
of the vertebral column (Text-fig. 8b-d). All show the characteristic vertebral conditions described above.

Discussion. R1 1924, R1 1922 and R1 1950 are to my knowledge the first record of Plioplatecarpus
from the African continent. However, among poorly preserved undescribed material (GSN
1928-1929) from the Dukamaje Formation of Gilbedi, Nigeria), a cervical vertebra Rl 1954 and a
well preserved caudal vertebra showing free haemal peduncles, probably belong to a plioplatecarpine
mosasaur. A part of the lower jaw and several fragments of vertebrae may also be plioplatecarpine
but are too poorly preserved and gypsiferous to permit assignment.

664

PALAEONTOLOGY, VOLUME 34

text-fig. 8. Plioplatecarpus vertebrae (top to bottom), anterior, posterior, and ventral views, a, caudal
( R 1 1 922). b, anterior cervical ( R 1 1 950). c, d, lumbar and posterior lumbar respectively (11950). e, f,
? Platecarpus lumbar vertebra ( R 1 1899); E, posterior view; F, anterior view; G, lateral view. xO-5.

LINGHAM-SOLIAR: MOSASAURS FROM NIGER

665

Subfamily plioplatecarpinae
Genus Platecarpus Cope, 1869

? Platecarpus sp.

Text-figure 8e-g

Referred material. BMNH R1 1 899, a fairly well preserved dorsal vertebra with most of the neural spine absent.
Mt Igdaman.

Description. R1 1899 (Text-fig. 8e-g), although otherwise well preserved, lacks its neural spine from a point a
few centimetres above the zygapophyses. The span and robustness of the anterior and posterior zygapophyses,
suggest that the vertebra belongs to the anterior trunk series. The inflated heart-shaped condyles and cotyles,
and the general configuration of the vertebra, suggest that it may be referable to Platecarpus.

Subfamily mosasaurinae

Mosasaurus Conybeare, 1822
cf. M. hojfmanni Mantell, 1829

Text-figure 9e

Referred material. BMNH R 1 1913, a fragment of a large tooth crown. Mt Igdaman.

Description. The distal tip and part of the anteroventral surface of the tooth crown are absent (Text-fig. 9e).
The specimen is abraded, but shows distinct signs of enamel. The buccal face is divided into five vertical prisms.
On the lingual surface, however, owing to weathering, it is not possible to discern prisms with confidence,
although this is difficult even in well preserved specimens of M. hojfmanni (i.e. the holotype skull and IRSNB
26 and ‘RI2’). In cross section the tooth shows a fairly deep lingual and slightly rounded buccal surface.
Posterior and anterior carinae are present. The tip of the tooth appears to have been gently posteriorly
recurved.

Discussion. The tooth probably belongs to M. hojfmanni, representing a middle-posterior tooth on
the dental ramus, although there is some resemblance to teeth found in larger examples of M.
lemonnieri , especially in the slight beading present between a few of the facets. At any rate the tooth
is referable to Mosasaurus judging from the external facets, the cross section, and the shape of the
tooth in general.

R 1 1913 represents an important find because it is the first tooth from Niger or Nigeria that can
be positively identified as Mosasaurus. Swinton’s (1930) referral of a tooth from the Dukamaje
Formation of Gilbedi to M. nigeriensis, is based on insufficient diagnostic characters for the genus
(there are no descriptions of facetting or provision of cross-sections).

The literature on mosasaur dentition is sparse, the most significant papers being Edmund (1960)
and Russell (1967). A problem in mosasaur classification is the erection of species based on single
tooth crowns. Mosasaur teeth manifest certain changes along the tooth row, simple variations of
shape and in the nature of tooth facets. This was noted by Leidy (1865, p. 56; pi. 9. fig. 3) and
Thevenin (1896, fig. 3, p. 904). Despite this, variation seems to have been ignored by certain authors.
For instance, Persson (1959) identified a tooth from Scania as M. hojfmanni and subsequently (1963)
re-identified it as a new species, M. ivoensis, based on what appears to have been a larger number
of tooth facets with slightly more concave rather than convex surfaces (conceivably M. lemonnieri).
Another species, M. beaugei , was diagnosed by Arambourg (1952, p. 282-283, pi. 31) on similarly
dubious grounds. His description included a variation in tooth shape in an anteroposterior
direction. Following this, M. beaugei was identified by Price (1957) from the Gramame Formation,
in Brazil.

666

PALAEONTOLOGY, VOLUME 34

text-fig. 9. Mosasaur teeth, buccal and lingual views, a (R 1 1896), b ( R 1 1 949), c (R 11894), d (R 1 1985), E
( R 1 191 3), Mosasaurus cf. hoffmanni. x 1. f, mosasaur vertebra ( R 1 1 907), dorsal and lateral views, x 2. G,

mosasaur coracoid, lateral and medial views, x 05.

Mosasaurinae or Tylosaurinae indet.

Text-figures 9a-d, f, g

Referred material. Three jaw fragments probably representing the same specimen, and including a large tooth

LINGH AM-SOLIAR: MOSASAURS FROM NIGER

667

base, BMNH R11914 (Kehehe); teeth Rl 1986, R11949, R11894, R 1 1 895 (Igdaman); two large disassociated
vertebrae, BMNH R 11917 and R 1 1918 (Kehehe); a very small vertebra R11907; a coracoid R 11919 and a
radius R 1 193 1 (Igdaman).

Description. R 1 1914 consists of three poorly preserved fragments of jaw, probably from a maxilla, which
appear to represent a single specimen. Associated with one of the fragments is a large tooth base including the
jagged remains of the basal part of a tooth crown. This specimen represents the largest mosasaur jaw found
in Africa. A BMNH cast of part of the maxilla of the huge holotype skull of M. hoffmanni (original in the Paris
Museum), described by Cuvier (1834, 1836), is of almost identical proportions, and it permits an estimate of
the length of the entire skull of R 1 1914 as approximately 14 m.

The largest of the teeth, Rl 1896 (Text-fig. 9a), is approximately 43 mm long and 19 7 mm wide at the base.
It is posteriorly recurved with sharp anterior and posterior carinae, extending almost the entire length of the
tooth. The surface is enamelled and bears no evidence of striae or facets. The size of the tooth, its general
configuration, and lack of strong prisms or facets indicate that it might be tylosaurine. The rest of the teeth,
R11949, Rl 1894 and R 1 1 895 (Text-fig. 9b-d) are not as well preserved, but are probably mosasaurian.

Two very large vertebrae, R 1 1917 and Rl 1918 from the lower levels of the Dukamaje Formation, clearly
come from a very large mosasaur with the proportions of R 1 1914. Large vertebral material mentioned
previously in the literature by various authors (Swinton 1930; Azzaroli et al. 1975; Soliar 1988), as well as
previously undescribed material from Kaffe and Gilbedi, is similarly impregnated with a matrix of yellow marl
as is Rl 1914, possibly indicating a similar horizon for all specimens.

R11907 (Text-fig. 9f) represents probably the smallest mosasaur vertebra known, clearly belonging to an
immature individual or hatchling. It belongs to the lumbar series, and the general configuration suggests that
it is mosasaurian. The neural spine is absent. Articulating surfaces are heart-shaped, deeper than wide. Slight
pinched depressions on the anterodorsal surfaces of the vertebra on either side of the neural spine may indicate
the inception of zygosphenes in a very young individual.

Rl 1919 (Text-fig. 9g) is a fairly well preserved coracoid, the first record of a mosasaur pectoral element in
Africa. It lacks an emargination of the distal border and the coracoid foramen is moderate in size. The small
size of the specimen indicates that it may belong to an immature individual or a small taxon such as Clidastes.
R11931 is a large fragment of a radius, but no further distinction is possible.

PALAEOECOLOGY OF MOSASAURS OF THE IULLEMMEDIN BASIN OF NIGER
In the south-western part of the Iullemmedin Basin, Petters (19796, p. 950) defined four marginal
environments, marshes, lagoons, estuaries, and epeiric seas. These were based upon computed alpha
diversity values for the Dukamaje benthic foraminiferan assemblages and on the preponderance of
the arenaceous foraminifera Miliammina, Haplophragmoides , Trochamina, and Ammobaculites ,
which indicates adaptation to stress and marginal environments.

The environments suggested for particular mosasaurs are: Mosasaurus (coastal waters);
? Platecarpus (open sea); Plioplatecarpus ? (deep water) (Russell 1967); Halisaurus (unknown);
Goronyosaurus , Angolasaurus, and Igdamanosaurus (coastal waters and lagoons).

If Petter’s interpretation of the environment of the south-western Iullemmedin Basin is correct,
then the presence of suggested open-sea forms of mosasaurs such as Platecarpus and deep-water
divers such as Plioplatecarpus may appear inconsistent. There is, however, very little evidence in the
literature to corroborate these interpretations of the lifestyle of Platecarpus and Plioplatecarpus.
Russell (1967) has shown there to be no especial significance in the tympanic membrane for example
in diving. A calcified tympanic membrane in fact occurs widely in most mosasaurs, being present
in members of all three subfamilies (Williston 1898; Russell 1967).

Among mosasaurs, the unique presence of enclosed olfactory lobes in Goronyosaurus is puzzling.
The enclosed condition of the lobes in the Varanoidea (Mertens 1942) is regarded as a
synapomorphy by Estes et cd. (1988), but is not normally present in mosasaurs. The enclosed
condition in Goronyosaurus may be associated with more highly developed olfactory lobes and sense
of olfaction. This is consistent with other features of the genus that suggest an efficient detection of
hidden prey; for example, the high number of large foramina on the premaxilla which suggest an
array of large nerve endings to an apparently highly sensitized snout. Either condition would have
been particularly useful for hunting in the quiet murky waters of sheltered bays and estuaries and

668

PALAEONTOLOGY, VOLUME 34

text-fig. 10. Goronyosaurus nigeriensis (IGF 14750). a, lateral view, b, dorsolateral view. x0-3.

m preying upon hidden young of marine fauna including perhaps those of mosasaurs (Russell 1967,
pp. 1, 65). Further, the very small eyes of Goronyosaurus (relatively perhaps the smallest in the
Mosasauridae) suggest that Goronyosaurus did not rely heavily on vision.

Modern-day analogues of this kind of predation are seen in varanids which search out hidden
prey by ‘specialized chemoreception ’ followed by ‘rapid, skilful capture' (Losos and Greene 1988,
p. 379). These authors suggest that this is a derived condition in varanids, a view that may hold
equally true for mosasaurs.

Igdamanosaurus is the first record in the Mosasauridae of a predator complying with Massare’s
(1987) tooth morphotype in the ‘crunch’ guild. This contrasts with the ‘crush’ guild in which she
placed G/obidens. From an ecological perspective, Igdamanosaurus probably occupied a feeding
mode somewhere between Prognathodon (Lingham-Soliar and Nolf 1989) and G/obidens (Russell
1967), subsisting presumably on a diet of moderately soft-shelled invertebrates.

Acknowledgements. I thank the Niger Government for generously allowing material to be collected. Special
thanks go to Dr Oumarou Hamadou for arranging export permits, and to Professor Boureima Ousmane, Mr
Ahmed Akassa, and many other who facilitated work in Niger. I am grateful to Mr Cyril Walker, Mr David
Ward, and Mrs Alison Ward for the mosasaur material collected. Thanks go to the following members of the
BM(NH): Mr Cyril Walker and Dr Angela Milner for providing facilities at the BM(NH), Ms Sandra
Chapman for help in many ways, and Mr Gary Summons for the photographs. I thank Mr David Ward and
Dr Richard Moody for provision of geological and geographical information. Professor Richard Estes and Dr
Beverly Halstead for discussions on the material and critical reviews of earlier versions of the manuscript. I also
thank an anonymous reviewer for helpful criticism.

Dedication. This paper is dedicated to the memory of Beverly Halstead who made the dreams of dragons a

wonderful reality for me.

LINGHAM-SOLIAR: MOSASAURS FROM NIGER

669

REFERENCES

adeleye, D. R. 1975. Nigerian Late Cretaceous stratigraphy and paleogeography. Bulletin of the American
Association of Petroleum Geologists , 59, 119-121.

antunes, m. t. 1964. O Neocretacico eo Cenozoico do litoral de Angola: 1 Estratigrafia ; Repteis. Junta de
Investigaco do Ultramar , Lisbon, 257 pp.

arambourg, c. 1952. Les vertebres fossiles des gisements de phosphates (Maroc Algerie Tunisie). Notes et
Memoir es du Service geologicpie du Maroc, 92, 1-372.

azzaroli, a., de guili, c. and torre, d. 1972. An aberrant mosasaur from the Upper Cretaceous of North
Western Nigeria. Accademia nazionale dei Lincei, Rendiconti. C/asse di Science ftsiche matematiche e naturali ,
(8) 52, 53-56.

— 1975. Late Cretaceous mosasaurs from the Sokoto District, Nigeria. Accademia nazionale dei Lincei,
Memorie. C/asse di Science fisiche matematiche e naturali, 13, 21-34.

baird, d. and case, G. r. 1966. Rare marine reptiles from the Cretaceous of New Jersey. J owned of
Paleontology, 40. 1211-1215.

brown, d. s. 1981. The English Upper Jurassic Plesiosauroidea (Reptilia) and a review of the phylogeny
and classification of the Plesiosauria. Bulletin of the British Museum (Natural History), ( Geology ) 35,
253-347.

buffetaut, e. 1976. Une nouvelle definition de la famille des Dyrosaundae De Stefano, 1903 (Crocodylia,
Mesosuchia) et ses consequences: inclusion de genres Hyposaurus et Sokotosuchus dans le Dyrosauridae,
Geobios, 9, 333-336.

— 1979. Sokotosuchus ianwilsoni and the evolution of the dyrosaurid crocodiles. Nigerian Field Monograph ,
1, 31-41.

bukowski, f. 1984. Halisaurus platyspondylus ; the third reported occurrence of this mosasaur in New Jersey.
Mosasaur , 1 , 119-121.

camp, c. l. 1942. California mosasaurs. Memoirs of the University of California, 13, 1-68.
cappetta, h. 1972. Les poissons cretaces et tertiaires du Bassin des lullemmeden. Palaeovertebrata, 5, 179-251.
chatterjee, s. small, b. j. and nickell, m. w. 1985. Late Cretaceous marine reptiles from Antarctica.
Antarctic Journal, 19, 7-8.

cope, e. d. 1875. The Vertebrata of the Cretaceous formations of the West. Report of the United States
Geological Survey of the Territories, 2, 1-303.
cuvier, G. f. 1834 1836. Recherches sur les ossemens fossiles. 4th ed. Paris. 10 vols.

dollo, l. 1889. Nouvelle note sur les vertebres fossiles recement offerts au Musee de Bruxelles par M. Alfred
Lemonnier. Bulletin de la Societe Beige de Geologie, de Pa/eontologie et d Hvdrologie, proces-verbaux, 3,
271-304.

EDMUND, a. G. 1960. Tooth replacement phenomena in the lower vertebrates. Contribution of the Royal Ontario
Museum of Life Sciences Division , 52, 1-190.

estes, r., de queiroz, k. and Gauthier, k. 1988. Phylogenetic relationships within Squamata. 119-281. In
estes, R. and pregill, G. (eds). Phylogenetic Relationships of the Lizard Families. Stanford University Press,
Stanford, California.

Gilmore, c. w. 1912. A new mosasauroid reptile from the Cretaceous of Alabama. Proceedings of the United
States National Museum, 41, 479 484.

greaves, w. s. 1983. A functional analysis of carnassial biting. Biological Journal of the Linnean Society, 20,
253-263.

greigert, j. 1966. Description des Formations Cretacees et Tertiaires du Bassm des lullemmeden (Afrique
Occidentale) (France). Memoires de Bureau de Recherches Geologiques et Minieres, 36, 1-234.

Halstead, l. b. 1979. The International Palaeontological Expedition to Sokoto State 1977-1978. Nigerian Field
Monograph, 1, 3-30.

howarth, m. k. 1981 . Palaeogeography of the Mesozoic. 197-220. In cocks, l. r. m. (ed.). The Evolving Earth.
British Museum (Natural History), London.

huene, E. von 1935. Mosasaurier Zahne von Timor. Zentralblatt fur Miner alogie, Geologie und Palaontologie,
B, 10. 412-416.

kogbe, c. a. 1973. Geology of the Upper Cretaceous and Tertiary sediments of the Nigerian sector of the
Iullemmedin Basin (West Africa). Geologische Rundschau, 62, 197-211.

— 1979. Geology of the south eastern (Sokoto) sector of the Iullemmedin Basin. Geology Monograph,
Ahmadu Bello University Zaria, Nigeria. 2 (1), 1^J20.

670

PALAEONTOLOGY, VOLUME 34

krasheninnikov, v. a. and trofimov, d. m. 1969. Stravitel'nyy analiz bcnthosnykh foraminifer datsko-
paleotsenovykh otlozhneniy Mali, oblasti tetisa i severo zapadnoy Evropy. Akademia Nauk USSR. Voprosy
Micropalaeontologie , 12, 108-144.

leidy, j. 1865. Memoir on the extinct reptiles of the Cretaceous formations of the United States. Smithsonian
Contributions to Knowledge , 14 (6), 1-65.

lingham-soliar (soliar), t. and nolf, d. 1990. The mosasaur Prognathodon from the Upper Cretaceous of
Belgium and a review of the genus. Bulletin de I Institut Roval des Sciences naturelles de Belgique, 59,
137-190.

losos, J. B. and Greene, h. w. 1988. Ecological and evolutionary implications of diet in monitor lizards.

Biological Journal of the Linnean Society, 35, 379-407.
massare, j. a. 1987. Tooth morphology and prey preference of Mesozoic marine reptiles. Journal of Vertebrate
Paleontology, 7, 121-137.

McFarland, w. n., pough, f. h., cade, t. j. and heiser, j. b. 1979. Vertebrate Life. MacMillan, New York,
875 pp.

mertens, R. 1942. Die Familie der Varane (Varanidae). Zweiter Teil: Der Schadel. Abhandlungen
Senckenbergischen Naturforschenden Gesellschaft, 465, 117-234.
olson, e. c. 1961. Jaw mechanisms: rhipidistians, amphibians, reptiles. American Zoologist , 1, 205-215.
persson, p. o. 1959. Reptiles from the Senonian of Scania. Arkiv for mineralogi och geologi, Stockholm, 2,
431-480.

1963. Studies on Mesozoic marine reptile faunas with particular regard to the Plesiosauria. Publications
of the University of Lund. Institute of Mineralogy and Paleontology, Quaternary Geology, 118, 1-15.
petters, s. w. 1977. Ancient seaway across the Sahara. Nigerian Field, 42, 22-30.

1979a. Stratigraphic history of the south-central Saharan region. Bulletin of the Geological Society of
America, 90, 753-760.

19796. Maastrichtian arenaceous foraminifera from north and western Nigeria. Paleontology, 22,
947-963.

price, l. i. 1957. A presenca de Globidens no Cretacico superior do Brasil. Boletim Divisdo de Geologia e
Miner alogia, Rio de Janeiro. 169, 1-24.

reyment, r. a. 1965. Aspects of the Geology of Nigeria. Ibadan University Press, 145 pp.
russell, d. a. 1967. Systematics and morphology of American mosasaurs. Bulletin of the Peabody Museum of
Natural History , Yale University, 23, 1-237.

1970. The vertebrate fauna of the Selma Formation of Alabama. Part VII. The mosasaurs. Fieldiana ,
Geology Memoirs, 3, 365-380.

1975. A new species of Globidens from South Dakota and a review of globidentine mosasaurs. Fieldiana ,
Geology, 3, 235-256.

soliar (LINGHAM-), t. 1988. The mosasaur Goronyosaurus from the Upper Cretaceous of Sokoto State, Nigeria.
Paleontology, 31, 747-762.

swinton, w. E. 1930. On fossil Reptilia from Sokoto Province. Bulletin of the Geological Survey of Nigeria, 13,
1-56.

tabaste, n. 1963. Etude de restes de poissons du Cretace Saharien. Melanges Ichthyologiques, 68, 438-485.
thevenin, a. 1896. Mosasauriens de la Craie Grise de Vaux Eclusier pres Peronne (Somme). Bulletin de la
Societe geologique de la France, Paris, (3), 24, 900-916.
vaughn, p. p. and dawson, m. r. 1956. On the occurrence of calcified tympanic membranes in the mosasaur
Platecarpus. Transactions of the Kansas Academy of Science, 59, 382-384.

1898. Mosasaurs, Kansas University Geological Survey, 4, 83-221.

wiman, c. 1920. Some reptiles from the Niobrara group in Kansas. Bulletin of the Geological Institution of the
University of Upsala, 18, 9-18.

zdansky, o. 1935. The occurrence of mosasaurs in Egypt and Africa in general. Bulletin de /'Institut cf Egvpte,
17, 83-94.

T. LINGHAM-SOLIAR

Department of Animal and Microbial Sciences

Typescript received 15 August 1989 University of Reading

Revised typescript received 9 June 1990 Reading RG6 2AJ UK

ISOLATED GRAPTOLITES FROM THE
LLANDOVERY OF K A LLHOLEN, SWEDEN

by D. K. LOYDELL

Abstract. A graptolite fauna of Monograptus argenteus Biozone age (Aeronian = Middle Llandovery),
chemically isolated from limestone nodules formerly thought to be of M. turriculatus Biozone age (Telychian
= Upper Llandovery), is described. The total fauna comprises Metaclimacograptus hughesi (Nicholson),
Glyptograptus aff. incertus Rickards, G. sinuatus sinuatus (Nicholson), Clinoclimacograptus retroversus Bulman
and Rickards, Agetograptus primus Obut and Sobolevskaya, Agetograptus sp., ‘ Orthograptus ’ cyperoides
(Tornquist), ' insectiformis (Nicholson), Monoclimacis sp., Pristiograptus concinnus (Lapworth), Pribylo-
graptus leptotheca (Lapworth), Monograptus communis communis Lapworth, M. denticulatus sensu Sudbury,
M. millepeda M'Coy and Rastrites peregrinus Barrande. Spines are recognized for the first time on the thecae
of M. c. communis and M. millepeda. It is suggested that M. lobiferus evolved from M. millepeda which had
itself evolved from M. c. communis. The generic diagnosis for Pribylograptus Obut and Sobolevskaya is
emended, as the existing diagnosis for the genus is based upon a misinterpretation of the thecal structure
derived from examination of pyrite internal moulds.

In 1970, Hutt, Rickards and Skevington described a number of graptolites (eleven species in total)
chemically isolated from limestone nodules from the turriculatus Biozone of Osmundsberget,
Dalarna, Sweden. When similar material, collected by Drs M. G. Bassett, R. B. Rickards and
J. Gluyas became available, it seemed an ideal opportunity to compare uncrushed, isolated material
with the common low-relief internal moulds and rare flattened rhabdosomes found in Wales. Such
study would aid interpretation of, in particular, the thecal structures of such imperfect material.

Dr Bassett also provided limestone nodules from Kallholen (also in Dalarna), which were
considered (pers. comm.) to come from a similar horizon to the Osmundsberget nodules. However,
the recognition of species such as Monograptus millepeda (IVLCoy, 1850) and Pribylograptus
leptotheca (Lapworth, 1876) in the residues of some of the first samples processed indicated that this
was not the case, and that the Kallholen nodules contained a fauna from rather earlier in the
Llandovery, from the Aeronian M. argenteus Biozone.

The vast majority of the species extracted from these nodules have never before been described
in isolated condition and thus a large amount of new (and often surprising) morphological
information, in some cases of considerable taxonomic importance, has been derived.

TECHNIQUES

The isolation techniques used are fully described in Dumican and Rickards (1985) and in Bates
et al. (1988). The nodules were dissolved in acetic acid as this has a less effervescent reaction with
carbonate than hydrochloric acid, and thus causes less damage to emerging specimens. Some
specimens were cleared (i.e. rendered transparent) using the technique described by Berry (pp.
106-107, in Kummel and Raup 1965). Initial examination under a light microscope was followed
by more detailed study with a scanning electron microscope (SEM) again using the techniques
described in Dumican and Rickards (1985).

IPalaeontology, Vol. 34, Part 3, 1991, pp. 671-693, 3 pls.|

© The Palaeontological Association

672

PALAEONTOLOGY, VOLUME 34

PREVIOUS WORK ON ISOLATED NON-RETIOLITE LLANDOVERY GR APTOLITES

Any retiolite graptolites isolated from the nodules were passed to Drs Bates and Kirk. The following
account of previous research therefore considers only monograptids and ‘normal’ diplograptids.

Other than Hutt, Rickards and Skevington's (1970) work, which also included descriptions of
seven species from the Coronograptus gregarius Biozone (Aeronian) of Silvberg, also in Dalarna,
Sweden, little work has been published on isolated non-retiolite Llandovery graptolites, although
undoubtedly a wealth of material exists, particularly from the Canadian Arctic. This dearth of
studies is surprising, for as Fortey (1989) states: ‘One good isolated fossil fauna is worth any
amount of speculation on structure and phylogenetics based on flattened specimens.’

Bulman (1932) described three monograptid species from the Retioliles- Shales of Stygforsen,
Dalarna, Sweden, which had been prepared by Holm in the latter part of the nineteenth century.
Two of these he assigned to Monograptus priodon (Bronn, 1835) and M. spiralis (Geinitz, 1842). The
third he was unable to identify. It is herein recognized as Streptograptus grayae (Lapworth, 1876).

Lenz (1974) analysed evolutionary trends within M. priodon from the uppermost Llandovery to
lower Wenlock, using isolated material from Cornwallis Island, Arctic Canada.

Paskevicius (1976) erected the genus Lithuanograptus on the basis of examination of isolated
Rhuddanian and Aeronian diplograptaceans from Lithuania, USSR. It is probable that many of his
specimens are simply flattened metaclimacograptids.

In 1977, Rickards, Hutt and Berry illustrated thecae considered by them to be from M. delicatulus
Elies and Wood, 1913. Crowther (1981) described Paraclimacogr aphis innotatus innotatus
(Nicholson, 1869) from the Dergaish River, south Ural Mountains, USSR, and Dimorphograptus ?
sp. from Cornwallis Island, Canada.

Obut and Sennikov (1980) described a number of Monograptus triangulatus Biozone (Aeronian)
monograptid and diplograptid taxa from the Siberian Platform, USSR. Their fine material would
certainly benefit from further examination with an SEM.

Bates and Kirk (1984) gave a description and discussed the probable function of ancoras from
three different diplograptid species. Their suggestion of horizon as ? M. turriculatus Zone for this
Kallholen material is, as mentioned above, now known to be incorrect.

Melchin and Lenz (1986) described M. turriculatus (Barrande, 1850) from Cornwallis Island, and
discussed its possible derivation from M. sedgwickii (Portlock, 1843). Chen (1986a) described seven
generally poorly preserved species from the Aeronian of Yichang, W Hubei, China. This material
was extracted from siliceous marl concretions by dissolution first in hydrochloric and then
hydrofluoric acid. Chen (19866) has also described isolated material of Streptograptus plumosus
(Baily, 1871) (misidentified by him as S. nodifer (Tornquist, 1881)) from north-east Guizhou, and
further work on isolated faunas is in progress.

Most recently, Lenz and Melchin (1989) have described uncompressed specimens of M. spiralis
again from Cornwallis Island, and suggest that this species is unlikely to be ancestral to the genus
Cyrtograptus Carruthers, 1 867, three isolated species of which they also describe - C. sakmaricus
Koren', 1968, C. sp. and C. cf. C. laqueus Jackson and Etherington, 1969.

LOCALITY INFORMATION

As Kallholen has not been visited by the author, the information below is derived entirely from
Bassett (pers. comm, to D. Bates).

The graptolites were isolated from nodules from the lowest band exposed (within basal 2 m) in
the Llandovery sequence unconformably overlying the Upper Ordovician Boda Limestone. The
nodules were collected from sections on both sides of the western entrance to the quarry on the east
side of road 296, directly opposite Kallholen limestone works, 7-5 km NE of Orsa Church, Siljan
district, Dalarna (UTM Grid reference VH 8440 8153).

LOYDELL: GRAPTOLITES FROM SWEDEN

673

SYSTEMATIC PALAEONTOLOGY

The terminology used is mainly that of Bulman (1970). Thecal spacing is expressed in terms of a two
theca repeat distance (2TRD), as defined by Howe (1983).

Synonymies are brief for the well-known taxa, as more complete synonymies may be found in,
for example, Rickards (1970) and Hutt (1974, 1975). They are annotated with the symbols proposed
by Richter (1948) and Rabien (1954), summarized in Matthews (1973).

Suprageneric classification is based on Mitchell (1987), with reservation, particularly with regard
to the lowly status conferred by Mitchell (op. cit.) upon the ‘monograptids’.

All material figured in the plates is housed in the Institute of Earth Studies, University College
of Wales, Aberystwyth.

Repositories of type material are abbreviated as follows: BM (NH) - Natural History Museum;
BU - Birmingham University; SM - Sedgwick Museum, Cambridge.

Order graptoloidea Lapworth, 1873
Suborder virgellina Fortey and Cooper, 1986
Superfamily diplograptacea Lapworth, 1873 (emend. Mitchell, 1987)

Family glyptograptidae Fortey and Cooper, 1986
Subfamily glyptograptinae Fortey and Cooper, 1986
Genus glyptograptus Lapworth, 1873 (emend. Mitchell, 1987)

Type species. Original designation; Diplograpsus tamariscus Nicholson, 1868, p. 526, pi. 19, figs 10-13.

Glyptograptus aflf. incertus Rickards, 1970
Plate 1 , fig. 1

1970 Glyptograptus aff. incertus (Elies and Wood, 1907); Rickards pp. 40-41, text-fig. 14, fig. 11.

pl974 Glyptograptus (G.) incertus (Elies and Wood, 1907); Hutt, p. 25. pi. 3, fig. 3; pi. 4, fig. 13 (non

12); text-fig. 8, fig. 11.

Material. One specimen.

Description. The 4-5 mm long specimen has been laterally compressed producing a corrugated appearance
along its midline. The thecae have sharp genicula on thF-32. There is then an abrupt change to sigmoidally
curved thecae and flowing, rounded genicula. Thecal apertures are horizontal. Dorso-ventral width increases
from 0-5 mm at th 1 1 to 1 mm at th5E 2TRD ranges from T3 to 1-5 mm.

Remarks. The specimen agrees very well with the material illustrated by Rickards (1970) and Hutt
(1974). G. incertus Elies and Wood, 1907 differs in having a broader, more rounded proximal end
and also in that the geniculum becomes less angular more gradually along the length of the
rhabdosome. The latter species is fully described by Packham (1962).

Glyptograptus sinuatus sinuatus (Nicholson, 1869)

Plate 1, fig. 2

*.1869 Diplograpsus sinuatus Nicholson, p. 235, pi. 11, fig 11.

1907 Diplograptus ( Glyptograptus ) sinuatus Nicholson; Elies and Wood, pp. 255-257, pi. 31, fig. 6;
text-fig. 175.

.1970 Glyptograptus sinuatus sinuatus (Nicholson); Rickards, pp. 41^12, pi. 4, fig. 1.

.1974 Glyptograptus (G.) sinuatus sinuatus (Nicholson); Hutt, p. 28, pi. 4, figs l^t, 10.

.1975 Glyptograptus sinuatus sinuatus (Nicholson); Bjerreskov, p. 31, pi. 4, fig. d.

674

PALAEONTOLOGY, VOLUME 34

Holotype. By monotypy; specimen figured Nicholson 1869, pi. 1 1, fig. 11; from the Llandovery of Skelgill, the
English Lake District.

Material. Several proximal ends, and a few more fully developed rhabdosomes.

Description. The rhabdosome increases in width fairly rapidly from 05-06 mm at th 1 1 to L3 mm at t h 5 1 .
2TRD is L35-L5 mm at th2\ and L5-L55 mm at th5L The first three thecal pairs have sharp genicula and
straight supragenicular walls inclined at a very low angle to the rhabdosome axis. An abrupt change in thecal
morphology then occurs, all further thecae being sigmoidally curved with flowing genicula.

Remarks. The abrupt change in the thecal morphology has been widely illustrated. This species,
perhaps more than any other, illustrates the futility of the traditional diplograptacean generic
classification (see Bulman 1970) which was based largely on thecal morphology. The character of
the proximal thecae of this species would place it in Climacograptus Hall, 1865, whilst that of the
distal thecae would place it in Glyptograptus Lapworth, 1873. As a result of the work of Fortey and
Cooper (1986) and Mitchell (1987) this traditional classification has now been abandoned in favour
of one which is phylogenetically based, concentrating on early astogenetic development.

Genus clinoclimacograptus Bulman and Rickards, 1968.

Type species. Original designation; Pseudoclimacograptus ( Clinoclimacograptus ) retroversus Bulman and
Rickards, 1968, p. 8; from the Llandovery of Sweden and Britain.

Clinoclimacograptus retroversus Bulman and Rickards, 1968

Plate I, figs 5 and 14

1893 Climacograptus scalaris Lin.; Tornquist, pp. 2-6, figs 1-3, 5-8, 11-15, ? 4, 9, 10, 16-22.

*.1968 P. ( Clinoclimacograptus ) retroversus Bulman and Rickards, pp. 8-12, text-figs 3-5.

.1970 Pseudoclimacograptus ( Clinoclimacograptus ) retroversus Bulman and Rickards; Rickards, pp.
34-35, text-fig. 14, figs 1—4.

.1974 Pseudoclimacograptus ( Clinoclimacograptus ) retroversus Bulman and Rickards; Hutt, p. 23, pi.
2, figs 10-12.

.1975 Pseudoclimacograptus retroversus (Bulman and Rickards); Bjerreskov, pp. 25-26, text-fig. 9d, e.

Holotype. Original designation; figured Bulman and Rickards 1968, fig. 46; SM A52951; from Tomarp
Sweden, probably M. sedgwickii Biozone.

Material. Approximately 200 specimens.

Description. The rhabdosome is essentially parallel-sided, increasing in dorso-ventral width only very gradually
from 0 6-0-8 mm at th 1 1 to a distal maximum of 1 -1—1-35 mm, although some specimens exhibit a slight
subsequent reduction in width (as shown by specimen UCWG961 B, figured PI. 1, fig- 14). 2TRD increases from
1-2 1-4 mm at th2x to 1 -45—1 -6 mm distally. The thecae have sharp genicula and characteristic slightly concave
supragenicular walls. The apertures alternate and are horizontal or very slightly introverted. Their excavations
each comprise c. one-tenth of the dorso-ventral width for the first thecal pair, increasing to c. one-fifth distally.
The median septum (seen on cleared specimens) is gently undulating proximally, becoming almost straight
distally.

Remarks. The material agrees well with Bulman and Rickards’s (1968) description other than in its
attainment of a greater distal width. In this respect it resembles Pseudoclimacograptus
( Clinoclimacograptus )? washing toni Bjerreskov, 1981, but this latter species has a narrower proximal
end and more widely spaced thecae. Bulman and Rickards (1968) noted the long range of Cl.
retroversus ( M . triangulatus to M. sedgwickii Biozones) and suggested that it might be possible to
subdivide it into a number of stratigraphically useful forms.

LOYDELL: GRAPTOLITES FROM SWEDEN

675

Genus metaclimacograptus Bulman and Rickards, 1968

Type species. Original designation; Diplograpsus Hughesi Nicholson, 1869, p. 235, pi. 11, figs 9 and 10; from
the Llandovery of the Lake District, England.

Metaclimacograptus hughesi (Nicholson, 1869)

Plate 1, figs 3, 4, 6, 9, 12

*.1869 Diplograpsus Hughesi Nicholson, p. 235, pi. 11, figs 9 and 10.

1906 Climacograptus Hughesi Nicholson; Elies and Wood, pp. 208-210, pi. 27, fig. II; text-fig. 140.

1968 P. (Metaclimacograptus) hughesi (Nicholson); Bulman and Rickards, pp. 3-6, fig. 1.

1970 Pseudoclimacograptus ( Metaclimacograptus ) hughesi (Nicholson); Rickards, p. 33, text-fig. 14,
fig. 6.

.1970 P. (Metaclimacograptus) hughesi (Nicholson); Hutt, Rickards and Skevington, p. 4, pi. 1, figs
1-4.

1974 Pseudoclimacograptus (Metaclimacograptus) hughesi (Nicholson); Hutt, p. 22, pi. 2, figs 6, 7, 13,
14.

1975 Pseudoclimacograptus undulatus (Kurck, 1882); Bjerreskov, p. 26, pi. 4, fig e.

Neotype. Designated Pribyl 1948, p. 18, as the specimen figured by Elies and Wood 1906, pi. 27, fig. 1 \ a: BM
(NH) P1890.

Material. Several hundred specimens.

Description. The rhabdosome appears essentially parallel-sided, its width increasing very gradually from
0-5-0-6 mm at th 1 1 to a distal maximum of 0-75 mm. 2TRD is 10-L25 mm at th2L Distally it lies between 1-2
and L5 mm. The thecae have sharp, slightly overhanging and thickened genicula. Their apertures alternate.
They are introverted, with excavations comprising c. one-third of the rhabdosome’s dorso-ventral width. The
medium septum (seen on cleared specimens) undulates.

One specimen has a circular cyst-like growth on the obverse side of the rhabdosome close to the aperture
of th2' (PI. 1, figs 3 and 6).

Remarks. Bulman and Rickards (1968) describe this species in great detail. The thickening of the
geniculum is presumably the same structure as the genicular hood described in this species by
Bulman and Rickards (op. cit.). Hutt, Rickards and Skevington (1970) noted faint traces of
genicular hoods on some of their isolated, but flattened, material.

The cyst-like growth may be the result of parasitism. Similar growths have been recognised
before, for example by Jackson (1971). Conway Morris (1981) gives a useful review of parasitism
in the fossil record, which includes examples on graptolites.

Subfamily retiolitinae Lapworth, 1973 (emend. Mitchell, 1987)

Genus agetograptus Obut and Sobolevskaya, 1968

Type species. Original designation; Agetograptus secundus Obut and Sobolevskaya (in Obut, Sobolevskaya and
Merkureva 1968), pp. 79-80, pi. 8, figs 9-12 ; pi. 9, figs 1-13 ; pi. 10, figs 1-5 ; from the Llandovery of a borehole
in the Norilsk region, USSR.

Diagnosis. Species with th 1 2 longer than th 1 1 and 21 so that its aperture is higher than that of th2L
Throughout the rhabdosome the aperture of the thecae of the second series open above the
succeeding thecae of the first series.

Remarks. Bulman (1970) placed this genus into synonymy with Dimorphograptus Lapworth, 1876.
However, it differs from this genus in that in Agetograptus th 1 2 is elongated rather than being
suppressed. The superficial appearance of the proximal end is similar in both genera.

676

PALAEONTOLOGY, VOLUME 34

Rickards, Hutt and Berry (1977, p. 25) note that some specimens of Ag. secundus have a normal
proximal end (i.e. thecal apertures alternating). Neither of the species described herein exhibits this
feature, however, which Rickards, Hutt and Berry (op. cit.) consider as being indicative of a
dithyrial population (Jaanusson 1973).

Agetograptus primus Obut and Sobolevskaya, 1968

Plate 1, figs 11, 16, 18, 19; Plate 2, fig. 2

*.1968 Agetograptus primus Obut et Sobolevskaya; Obut, Sobolevskaya and Merkureva, pp. 80-1,
pi. 10, figs 6-12.

1970 Orthograptus bellulus Tornquist; Churkin and Carter, p. 28, pi. 3, fig. 1; text-fig. 12g.

71970 Orthograptus cf. insectiformis (Nicholson, 1869); Rickards, pp. 46-47, text-fig. 14, fig. 18.

p 1 978 Orthograptus insectiformis (Nicholson); Chen and Lin, p. 39, pi. 7, fig. 3 ( non 1 and 2).

Holotype. By original designation, the specimen illustrated by Obut, Sobolevskaya and Merkureva 1968, pi.
10, fig. 6, from the Middle Llandovery of borehole N-l, Norilsk district, USSR.

Material. Approximately 100 specimens.

Description. The rhabdosome increases moderately rapidly in dorso-ventral width from 0-4-0-65 mm at th 1 1 to
0-9-1 -2 mm at th5x, the maximum distal value recorded being I -9 mm. 2TRDs are characteristically low,
0-65-0-9 mm at th2L increasing to 1 - 1 5—1-4 mm distally. The thecae are straight tubes inclined at c. 30° to the
rhabdosome axis. Each aperture bears a pair of short, gently curving spines directed antero- and posterio-
laterally. These are situated on the dorsal margin of the aperture. The virgella is long and robust.

Remarks. The position of the spines results in them being visible only when an oblique or lateral
view is presented. They were not observed by Obut and Sobolevskaya (1968) in their material.

Poorly preserved specimens of Ag. primus might easily be mistaken for ‘ Orthograptus ’ bellulus
(Tornquist, 1890).

EXPLANATION OF PLATE 1

Fig. I . Glyptograptus aff. incertus Rickards, 1970, UCWG961C, obverse view; note lateral compression, x 10.

Fig. 2. Glyptograptus sinuatus (Nicholson, 1869), UCWG943B, obverse view, x 10.

Figs 3, 4, 6, 9, 12. Metaclimacograptus hughesi (Nicholson, 1869), 3, UCWG956B, obverse view; note cyst-like
structure adjacent to aperture of th21, x 10. 4, UCWG919A, obverse view; x 20. 6, UCWG956B, cyst-like
structure, x 100. 9, UCWG930D, early growth stage, x 50. 12, UCWG930C, thecal aperture, x75.

Figs 5 and 14. Clinoclimacograptus retroversus Bulman and Rickards, 1968. 5, UCWG943A, obverse view;
x 10. 14, UCWG961B, distal fragment showing slight narrowing of rhabdosome and decrease in angularity
of geniculum, x 10.

Figs 7, 8, 13. ‘ Orthograptus ' insectiformis (Nicholson, 1869). 7, UCWG956D, viewed from distal end of
rhabdosome, x 20. 8, UCWG956A, mesial fragment, x 10. 13, UCWG956A, showing form of thecal
apertures and spines, x 15.

Figs 10 and 15. ‘ Orthograptus ’ cyperoides (Tornquist, 1897). 10, UCWG1027, reverse view, x 20. 15,
UCWG1027, ancora, x 150.

Figs 1 1, 16, 18, 19. Agetograptus primus Obut and Sobolevskaya, 1968. 11, UCWG911B, reverse view, x 10.
16, UCWG960D, early growth stage, x40. 18, UCWG958C, early growth stage; note long, robust virgella,
x40. 19, UCWG911B, view from proximal end; note thecal spines, x 20.

Fig. 17. Agetograptus sp., UCWG919D, reverse view, x 15.

All scanning electron micrographs.

PLATE 1

LOYDELL, Swedish Llandovery graptolites

678

PALAEONTOLOGY, VOLUME 34

Agetograptus sp.

Plate 1, fig 17; Plate 2, figs 1 and 3

Material. Two specimens.

Description. The rhabdosome is gently tapering. Its dorso-ventral width increases from 05-055 mm at thl 1 to
0-9—10 mm at th5h 2TRD at th2x is 0-75 mm. At th5* it is IT mm. The thecae are simple tubes, inclined at
c. 25° to the rhabdosome axis, with slightly everted thecal apertures. The nema is partially enclosed in a
sheath-like structure (PI. 2, fig. 1).

Remarks. This species is similar to Ag. tenuilongissimus Obut and Sobolevskaya, 1968 but has more
closely spaced thecae.

Genus ‘orthograptus’ Lapworth, 1873
‘ Orthograptus' cyperoides (Tornquist, 1897)

Plate 1, figs 10 and 15

*.1897 Diplograptus cyperoides Tornquist, p. 16, pi. 2, figs 30-32.

1907 Diplograptus ( Orthograptus ) cyperoides Tornquist; Elies and Wood, pp. 238-239, pi. 29, fig. 8;
text-fig. 158.

1970 Orthograptus cyperoides (Tornquist); Rickards, pp. 45-46, text-fig. 14, figs 12 and 17.

71974 Orthograptus cyperoides (Tornquist); Hutt, p. 35, pi. 6, figs 2-5; text-fig. 9, figs 6 and 7.

71975 Orthograptus cyperoides (Tornquist); Bjerreskov. pp. 28-29, text-fig. 10D.

.1985 Orthograptus cyperoides (Tornquist); Storch, p. 90, pi. 1, figs 1-4, (7 5); text-fig. 2F, G.

Type specimen. Not yet designated. Tornquist’s material was from the cometa Biozone of Tomarp, Sweden.
Material. One specimen, a proximal end, bearing thecae up to th3 1 .

Description. The specimen has a width of 0-8 mm at thl1 and 0-9 mm at th2k The simple thecae are inclined
at 30° to the rhabdosome axis and have apertures almost perpendicular to the thecal axis. A median septum
is not present. The virgella divides a short distance from the sicula aperture into four branches, all of which
are broken.

Remarks. The dimensions of the specimen agree well with previous descriptions. Both Hutt (1975)
and Bjerreskov (1975) noted thecal apertural spines, prolonged into a fine network in the case of
the latter’s material, on some of their specimens of this species. Storch (1985), however, observed
no such spines, but did note a virgellar meshwork, often covered with a membrane, on about half
of his flattened specimens.

Hutt (1975) suggested that Orthograptus cyperoides and O. insectiformis might be conspecific. The
present material of the two species illustrated herein would indicate that this is probably not the
case.

‘ Orthograptus ' insectiformis (Nicholson, 1869)

Plate 1, figs 7, 8, 13

*.1869 Diplograpsus insectiformis Nicholson, p. 237, pi. 11, fig. 13.

.1907 Diplograptus ( Orthograptus ) insectiformis Nicholson; Elies and Wood, pp. 228-229, pi. 28,
fig. 7 ; text-fig. 1 50.

1970 Orthograptus 7 sp. ; Hutt, Rickards and Skevington, p. 5, pi. 1, figs 8-10.

.1974 Orthograptus insectiformis (Nicholson); Hutt, pp. 34—35, text-fig. 9, figs 1-3, 13.

.1974 Orthograptus insectiformis (Nicholson); Rickards and Koren', pp. 200-201, figs 1-4.

.1975 Orthograptus insectiformis (Nicholson); Bjerreskov, p. 29, text-fig. 10c.

LOYDELL: GRAPTOLITES FROM SWEDEN

679

Holotype. By monotypy; specimen figured by Nicholson 1869, pi. 11, fig. 13; BM (NH) Q3113; from the
Llandovery of Dob’s Linn, Moffat, Scotland.

Material. Approximately 20, mostly fragmentary, specimens.

Description. The rhabdosome increases in dorso-ventral width (excluding spines) from 0-7-0-8 mm at th 1 1 to
a maximum of 1-2 mm. 2TRD distally is L4 mm. The thecae are simple straight tubes with horizontal apertures
each of which is furnished with two pairs of gently curved spines (up to 0-75 mm long) directed antero- and
posterio-laterally. The sicula bears an ancora of type 3 of Bates and Kirk (1984).

Remarks. The ancora of this species has been described and its possible function discussed by Bates
and Kirk (1984).

‘ Orthograptus' inopinatus Boucek, 1943 is very similar in its dimensions to 'O.' insectiformis. It
differs in its possession of spines projecting from the nema (see illustrations in Storch 1985).

Subfamily monograptinae Lapworth, 1873
Genus monoclimacis Freeh, 1897

Type species. Original designation; Graptolithus vomerinus Nicholson, 1872, emend. Lapworth; from the
Comston Flags of northern England.

Diagnosis. Thecae geniculate, with straight supragenicular walls approximately parallel to the rhabdosome
axis. Proximal apertures may be hooked or bear lateral lappets; distal apertures often simpler, of
climacograptid appearance, sometimes with genicular hoods.

Monoclimacis sp.

Plate 3, figs 1-3, 7

Material. Several fragments, the majority of which are distal.

Description. Proximal fragments are gently dorsally curved whilst distal fragments are almost straight. The
minimum dorso-ventral width observed is 0 45 mm, the maximum is 0-65 mm. 2TRD values lie between 1 -4 mm
and 2 0 mm, with the lowest 2TRDs being recorded distally. Distal thecae are of fairly typical monoclimacid
appearance, with sharp genicula, and genicular hoods growing out from above simple, slightly everted
thecal apertures. Supragenicular walls are, however, gently inclined to the rhabdosome axis. One specimen
(UCWG935A, PI. 3, fig. 2) illustrates the change in thecal morphology from hooked at the first theca preserved,
such that the aperture faces proximally, to monoclimacid as seen in the distal thecae. This change takes place
rapidly, over four thecae, and involves the retreat of both the dorsal and ventral margins of the hook, so that
the former disappears and the latter is modified into the genicular hood.

Remarks. The shape of the rhabdosome and the form of the thecae are very similar to those of Mcl.
crenularis (Lapworth, 1880), which has only been recorded from the M. convolutus Biozone. The
latter species reaches a greater width and has more widely spaced thecae, particularly distally, than
the present material, from which it may have been derived.

Genus pristiograptus Jaekel, 1889

Type species. Original designation; P. frequens Jaekel, 1889, p. 669, pi. 28, figs 1 and 2; from the Silurian of
Germany.

Diagnosis. Thecae simple and straight, or almost straight, throughout length of rhabdosome.

680

PALAEONTOLOGY, VOLUME 34

Pristiograptus concinnus (Lapworth, 1876)

Plate 2, fig. 4

v*1876 Monograptus concinnus Lapworth, pp. 320-321, pi. 11, fig. 1.

v.191 1 Monograptus concinnus Lapworth; Elies and Wood, pp. 368-369, pi. 36, fig. 5 a-e (? f)\ text-fig.

240.

1970 Pristiograptus concinnus (Lapworth); Rickards, pp. 60-61, pi. 5, fig. 5.

.1975 Pristiograptus concinnus (Lapworth); Hutt, pp. 57-58, pi. 12, figs 1, 2, 7, 8.

1988 Pristiograptus concinnus (Lapworth); Storch, pp. 14-15, pi. 6, figs 2-4; text-fig. 2b.

Lectotype. Designated Pribyl 1948, p. 68; specimen figured Lapworth 1876, pi. 11, fig. la, from the Llandovery
of Dob's Linn, Moffat, Scotland.

Material. Three fragments.

Description. The fragments are very gently ventrally or dorsally curved, up to L2 mm wide with 2TRDs of

21-2-2 mm. The thecae are simple tubes, inclined at c. 25° to the rhabdosome axis, with horizontal to sub-

horizontal apertural margins, and a slight, often rounded, geniculum at the base of the free ventral wall.

Remarks. The slight geniculum, or convexity of the free ventral wall, of the thecae in this species
has been widely illustrated. Hutt (1975) and Rickards, Hutt and Berry (1977) considered this feature
to suggest that Atavograptus atavus (Jones, 1909) was ancestral to P. concinnus.

Genus pribylograptus Obut and Sobolevskaya, 1966

Type species. Original designation, Obut and Sobolevskaya 1966, p. 33; Monograptus incommodus Tornquist,
1899, p. 1 1, pi. 2. figs 1-5; from the Llandovery of Sweden.

Diagnosis, (emended herein). Rhabdosome usually long and slender with flexuous curvature, but in
one species more robust and straight, and in one with stiff dorsal curvature proximally. Sicula
known in only one species, where it is small, reaching to about the aperture of th. 1 . Thecae are long,
slender, and usually inclined at less than 20° to the axis of the rhabdosome. Thecal apertures usually
furnished with lateral lappets (producing ‘horns' in pyrite internal moulds), although in one species
the distal thecae are of simple pristiograptid form, but with slightly introverted apertures. Thecal
apertures usually overhung by a geniculum and, in two species at least, also by a genicular hood.

Remarks. The author recognizes that it is not correct taxonomic procedure to emend a generic
diagnosis on a species other than the type species, but feels that emendation is justified in this
instance as the internal moulds of the thecae of Pr. incommodus and the proximal thecae of Pr.
leptotheca are so similar.

EXPLANATION OF PLATE 2

Figs 1 and 3. Agetograptus sp., 1, UCWG919D, distal end of rhabdosome; note the sheath-like structure
surrounding the nema, x 50. 3, UCWG919D, proximal end; note that the aperture of th 1 2 opens distally to
that of th2\ x 40.

Fig. 2. Agetograptus primus Obut and Sobolevskaya, 1968. UCWG959D, proximal end, x 40.

Fig. 4. Pristiograptus concinnus (Lapworth, 1876), UCWG958A, rhabdosome fragment; note slight genicula,
x 10.

Figs 5-10. Pribylograptus leptotheca (Fapworth, 1876). 5, UCWG909C. proximal fragment, x 15. 6,
UCWG925B, mesioproximal fragment, x 15. 7, UCWG913A, mesiodistal fragment, x 15. 8, UCWG910A,
distal fragment, x 15. 9, UCWG913A, rounded genicular hood of mesiodistal theca, x 50. 10, UCWG909C,
proximal theca, x 100.

All scanning electron micrographs.

PLATE 2

LOYDELL, Swedish Llandovery graptolites

682

PALAEONTOLOGY, VOLUME 34

Rickards (1976) discussed this genus in detail and included within it Monograptus incommodus
Tornquist, 1899, M. argutus (Lapworth, 1876), M. leptotheca Lapworth, 1876, M. cf. incommodus
sensu Hutt and Rickards 1970, and M. sandersoni Lapworth, 1876. The present author is in favour
of the retention of those forms with proven or inferred lateral apertural lappets with the genus, i.e.
M. incommodus , M. argutus argutus and M. leptotheca , and, in addition, probably M. cf.
incommodus sensu Hutt and Rickards 1970 and M. sandersoni. Until more detailed studies of the
thecal apertures of M. jonesi Rickards, 1970, M. argutus sequens Rickards, 1970, and M. angustus
Rickards, 1970 are carried out, their assignation of this genus must be, at most, questionable.

Pribylograptus leptotheca (Lapworth, 1876)

Plate 2, figs 5-10; Plate 3, figs 4 and 5

*.1876 Monograptus leptotheca Lapworth, p. 352, pi. 12, fig. 4.

1911 Monograptus leptotheca Lapworth; Elies and Wood, p. 371, pi. 37, fig. 2; text-fig. 242.

v.1968 Monograptus leptotheca Lapworth; Rickards and Rushton, p. 268, text-figs 2 and 3.

.1970 Monograptus leptotheca Lapworth; Rickards, pp. 68-69, pi. 6, figs 3 and 4; text-fig. 14, fig 37;

text-fig. 16, fig. 2.

?p 1 970 Monograptus sp. 2; Hutt, Rickards and Skevington, p. 13, pi. 3, figs 58-62 ( non figs 56 and 57).

.1975 Monograptus leptotheca Lapworth; Bjerreskov, p. 51, pi. 7, figs F and G; text-fig. 16c.

.1975 Pribylograptus leptotheca (Lapworth); Hutt, p. 73. pi. 16, figs 1-3, 7.

.1988 Pribylograptus leptotheca (Lapworth); Storch, pp. 28-29, pi. 6, fig. 1 ; text-fig. 2c.

Lectotvpe. Strachan (1971, p. 57) has noted that the lectotype designated by Pribyl (1948, p. 73), the specimen
figured Lapworth 1876, pi. 12, fig. 4 a. has not been recognized amongst Lapworth’s collection and is probably
unrecognizable as a figured specimen.

Material. Several hundred proximal, mesial and distal fragments, but possibly belonging to only a few
individual rhabdosomes. Most are slightly diagenetically flattened, but some are almost perfectly preserved in
three dimensions.

Diagnosis. Rhabdosome more or less straight, of considerable length (probably up to several
hundred mm), and with a dorso-ventral width proximally of 0 4 mm increasing gradually to a distal
maximum of over 2 mm. Thecae biform; proximally the aperture bears a ventral median saddle
together with lateral lappets and is overhung by an angular geniculum furnished with a single-spined
genicular hood. Distally the lappets, geniculum and genicular hood retreat until the thecae become
simple tubes, although with slight lateral apertural expansion. Overlap of thecae is considerable,
increasing from five-eighths proximally to over three-quarters distally. 2TRD is approximately
2-0 mm throughout the rhabdosome. The sicula has not been seen.

Description. The fragments are straight or show slight dorsal curvature. Dorso-ventral width, excluding spines,
increases from 0-4 mm proximally to 2 mm distally. Proximal thecae (PL 2, figs 5 and 10; PI. 3, fig. 5) have
complex apertures, with a horizontal, median ventral saddle flanked by two distally-directed lateral lappets
around which the aperture runs. A broadly triangular genicular hood extending from the angular geniculum
of the succeeding theca is furnished with a proximo-ventrally directed spine, 30-40 /an in diameter and with
a length of 0-4 mm. Thecal overlap (determined by embedding a rhabdosome fragment in resin and grinding
down with carborundum powder) is five-eighths of the thecal length. Distal thecae (PI. 2, fig. 8; PI. 3, fig. 4)
are simple tubes, inclined at a low angle to the rhabdosome axis, with slightly introverted and slightly laterally
expanded thecal apertures. Distal 2TRD is approximately 2 0 mm, with overlap being greater than three-
quarters of the thecal length. Both proximal and distal thecae have a ridge running axially along the middle
of the ventral wall of the interthecal septum. This may be a preservational feature. The transition from
proximal to distal thecal morphologies takes place gradually in the mesial portion of the rhabdosome (PI. 2,
figs 6, 7, 9) and involves the retreat firstly of the genicular spine and then of the genicular hood, accompanied
by the retreat also of the lateral lappets, together with a ‘smoothing out' of the geniculum. The sicula has not
been seen.

LOYDELL: GRAPTOLITES FROM SWEDEN

683

Remarks. Rickards and Rushton (1968) and Bjerreskov (1975) have described well-preserved pyrite
internal moulds of this species, but misinterpreted the proximal thecal structure. Their apertural
transverse ‘horns’ and ‘hood’ are, respectively, a pyrite infilling of the lateral lappets, and of the
area between the median saddle and the genicular hood. Text-figure 1 compares the appearance of
an internal mould with an isolated specimen. That the genicular spine has not been recognised
previously, despite some described specimens (e.g. those of Rickards and Rushton 1968) having
some preserved periderm, is surprising. However, as Hutt (1974, p. 35) notes, ‘spines may easily be
overlooked or inadvertently destroyed when examining specimens preserved in full relief.' The fact
that, before examination of the Kallholen fauna, spines had not been recorded on the thecae of
Agetograptus primus , Monograptus millepeda and M. communis , the latter two of which are well-
known and widely recorded taxa, lends support to Hutt’s (1974) observation.

The author agrees with Rickards, Hutt and Berry’s (1977) suggested derivation of Pr. leptotheca
from Pr. argutus at the D. magnus-M . argenteus Biozone boundary.

spinose

genicular

hood

lappet

text-fig. 1. Pribylograptus leptotheca (Lapworth, 1876). a, internal mould of proximal theca (after Rickards
and Rushton 1968, fig. 3). b, isolated proximal theca, UCWG926B. Both x25.

Genus monograptus Geinitz, 1852

Type species. By subsequent designation (Bassler 1915, p. 822): Lomatoceras priodon Bronn, 1835, p. 56, pi.
1, fig, 13; from the Silurian of Germany.

Remarks. Detailed discussion of the use and scope of the genus Monograptus is to be found in
Rickards (1970), Bulman and Rickards (in Bulman 1970) and Bjerreskov (1975).

Monograptus communis communis Lapworth, 1876
Plate 3, figs 6 and 10

pi 876 Monograptus convolutus Hisinger, sp. Var. (a.) communis Lapworth, p. 358, pi. 13, fig. 4 a ( non
b).

p 1 9 1 3 Monograptus communis (Lapworth); Elies and Wood, pp. 480^181, pi. 49, fig. la, c (non b, d , e);
text-fig. 336a ( non b ).

.1958 Monograptus communis communis Lapworth; Sudbury, pp. 520-522, pi. 23, figs 97-101; text-
figs 18 and 20.

1970 Monograptus communis communis Lapworth; Rickards, pp. 84—85, pi. 6, fig. 7; text-fig. 17, figs
1, 19, ?9.

?p 1 982 Monograptus communis Lapworth, 1876; Lenz, pp. 67-69, figs 5f, g, t and 21g, h, j, l.

684

PALAEONTOLOGY, VOLUME 34

Lectotype. Designated by Elies and Wood 1913, explanation to pi. 49, fig. la; specimen figured by Lapworth
1876, pi. 13, fig. 4a, and by Elies and Wood 1913, pi. 49, fig. la; BU1684a; from the Lower Birkhill Shales of
Dob’s Linn, Moffat, Scotland.

Material. Rare fragments, up to c. 5 mm in length.

Description. Proximal fragments are gently dorsally curved, whilst distal ones are almost straight. The thecae
are triangular and hooked, with the free ventral wall inclined at c. 20° to the rhabdosome axis. The thecal
apertures face proximo-dorsally. They are furnished with a pair of proximo-laterally directed spines up to
0-25 mm long. The maximum dorso-ventral width recorded is 115 mm. 2TRD is 1-8-2T5 mm.

Remarks. The thecal apertural spines have not previously been observed, although Sudbury (1958)
noted a pair of rudimentary lappets on the lateral parts of the dorsal margins of the thecal apertures
on her internal moulds of this species. These lappets represent the infilling of the slight outgrowth
of the dorsal thecal apertural margin at the base of each spine.

Lenz ( 1982) describes (but does not illustrate) uncompressed specimens, which he assigned to M.
communis , from the Cape Phillips Formation, Cornwallis Island, the metathecal portions of which
show strong torsion. These must be specimens of a different species, with M. denticulatus Tornquist,
1899, a likely candidate. Both Tornquist (1899) and Bjerreskov ( 1975) have noted that the apertural
regions of the distal thecae of the latter species are twisted laterally, always to the reverse side of
the rhabdosome.

Rickards, Hutt and Berry (1977) tentatively suggest that M. denticulatus was derived from M.
communis. The morphological information above would suggest that this is unlikely. However, the
discovery of spines on the proximal thecal apertures of M. millepeda , described for the first time
herein, must strengthen their case for this latter species having been derived from M. communis , an
evolutionary relationship supported by stratigraphical evidence: M. communis appears in the
triangulatus Biozone whilst M. millepeda is first seen in the argenteus Biozone to which it is confined.

Storch (1988) has recently identified spines on the thecae of M. clingani (Carruthers, 1867)
( = M. communis obtusus Rickards, 1970) another species which Rickards, Hutt and Berry (op. cit.)
consider to be derived either from M. millepeda or M. communis. I would favour derivation from
the latter species as overall thecal morphology is more similar. Text-figure 2 illustrates evolution
from M. communis s.l.

Monograptus denticulatus sensit Sudbury 1958
Plate 3, fig. 8

1899 Monograptus denticulatus Tornquist, p. 18, pi. 3, figs 19-23.

1913 Monograptus denticulatus Tornquist; Elies and Wood, pp. 474-475, pi. 48, fig. 2a, b (?c-/); text-

fig. 330.

.1958 Monograptus denticulatus Tornquist ; Sudbury, pp. 509-510, pi. 21 , figs 72 and 73 ; text-figs 4 and
12.

1970 Monograptus denticulatus Tornquist; Rickards, pp. 83-84, pi. 7, fig. 3; text-fig. 17, figs 5 and 6.
non 1975 Monograptus denticulatus Tornquist; Bjerreskov, pp. 79-80, text-fig. 23b, c.

.1975 Monograptus denticulatus Tornquist; Hutt, p. 88, pi. 22, fig. 2; pi. 23, figs 1 and 4.

Material. Approximately 50 fragments, though, due to the tenuity of the prothecae, these bear only one or two
thecae.

Description. Proximal thecae are rastritiforin, whilst distal thecae are more triangular. The protheca is very
narrow. Thecal apertures are slightly expanded laterally. They face the dorsal side of the rhabdosome. The
maximum dorso-ventral width recorded is 1-25 mm with 1TRD values being 0-95—1 -3 mm.

Remarks. The thecal morphology is precisely that interpreted by Sudbury (1958) Irom slightly
compressed material.

LOYDELL: GRAPTOLITES FROM SWEDEN

685

Biozone

"~1

convolutus

M clingoni A

Ei «f.

M. lobiferus

argenteus

M millepeda

magnus

^ c /

M. communis s. 1 /

triangulatus

B

text-fig. 2. Evolution from Monograptus communis Lapworth, 1876 s.l. during the Aeronian (= Middle
Llandovery), a, M. clingcini (Carruthers, 1867), after Storch 1988. b. c, M. c. communism b, after Sudbury 1958;
c, UCWG908B. d, M. millepeda (M'Coy, 1850), UCWG957A. e-g, M. lobiferus (M'Coy, 1850); e, proximal
end, after Bjerreskov 1975; F, G, distal and proximal fragments respectively, after Storch 1988. All x 5.

686

PALAEONTOLOGY, VOLUME 34

The thecae of M. denticulatus s.s. are less rastritiform proximally, and distally the apertural
regions are twisted so as to face the reverse side of the rhabdosome (see Bjerreskov 1975 and Text-

text-fig. 3. A, B, Monograptus denticulatus Tornquist, 1899. a, proximal end, showing twisting of thecal
apertures x 10; b, complete rhabdosome, though with damaged proximal metathecae, x 5 (a and b after
Bjerreskov 1975). c, Monograptus denticulatus sensu Sudbury, 1958, almost complete specimen; note
rastritiform proximal thecae, x5 (after Sudbury 1958).

Clearly, the comments regarding the evolutionary position of M. denticulatus given by Sudbury
{op. cit .) and Rickards, Hutt and Berry (1977) are to the species sensu Sudbury, rather than sensu
Tornquist.

Monograptus millepeda (M'Coy, 1850)

Plate 3, figs 9, 11, 14, 15
*.1850 Graptolites millepeda M'C’oy, p. 270.

.1913 Monograptus millepeda (M'Coy); Elies and Wood, pp. 465^466, pi. 46, fig. 10; text-fig. 323.

.1975 Monograptus millepeda (McCoy); Hutt, pp. 96-97, pi. 22, figs 1 and 5; text-fig. 19, fig. 4.

v.1978 Oktavites spinatus , Ni, pp. 411^412, pi. 3, fig. 20; text-fig. 5, figs 1 and 2.

.1988 Campograptus millepeda (McCoy); Storch, pp. 41^42, pi. 9, fig. 2; pi. 10, fig. 4.

Type specimen. Not traced (see Hutt 1975).

Material. Approximately 1 50 specimens.

Description. The fish-hook shaped rhabdosome bears laterally expanded, triangular, hooked thecae which do
not overlap. The hook is retroverted so that the apertures face proximo-dorsally. The proximal thecae each
bear a pair of laterally-directed spines, up to 0-3 mm long. These retreat distally, such that by thlO the thecal
apertural margin is smooth and non-spinose. Dorso-ventral width at thl is 0-45-0-5 mm. The maximum width
recorded is L2 mm. Distal 2TRD is 1 -8—2-2 mm. The sicula is 0-9 mm long. Its apex reaches to just above the
top of thl.

LOYDELL: GRAPTOLITES FROM SWEDEN

687

Remarks. The proximal thecae of M. millepeda have not previously been appreciated as being
spinose, although Ni (1978) illustrated thecal spines on the conspecihc Oktavites spinatus.

It is interesting to note that both Pedersen (1922) and Bjerreskov (1975) illustrate paired thecal
apertural spines in M. lobiferus (M'Coy, 1850). Bjerreskov (op. cit.) noted that they were most
obvious in the proximal portion of the rhabdosome. It would thus seem likely that M. lobiferus
evolved from M. millepeda by increased retroversion of the thecal aperture and by reduction in
rhabdosome curvature (Text-fig. 2). This is at variance with Rickards, Hutt and Berry’s (1977)
conclusions. They state that there can be little doubt that the lineage M. sp. A (M. sp. 1 of Hutt
1975) ( D . magnus Biozone) to M. undulatus Elies and Wood, 1913 (M. convolutus Biozone) led to
M. lobiferus. The derivation suggested herein, from M. millepeda , has a sounder biostratigraphical
basis, in that M. lobiferus first appears in either the M. argenteus or the M. convolutus Biozone,
whilst, to the author’s knowledge, M. undulatus has only ever been recorded with certainty from the
upper part of the M. convolutus Biozone. It should be noted here that Storch (1988) suggested that
the M. lobiferus group was derived from Campograptus Obut, 1949, a genus into which he placed
M. millepeda.

Genus rastrites Barrande, 1850

Type species. By subsequent designation of Hopkinson 1869, p. 158: R. peregrinus Barrande, 1850, p. 67,
pi. 4, fig. 6; from the Llandovery of Bohemia.

Diagnosis. Rhabdosome with narrow prothecae and straight, completely isolated metathecae,
inclined at a high-angle to the rhabdosome axis. The apertures are hooked and laterally expanded
in many species.

Rastrites peregrinus Barrande, 1850
Plate 3, figs 12 and 13

*.1850 Rastrites peregrinus Barrande, p. 67, pi. 4, fig. 6.

p 1 9 1 4 Monograptus (Rastrites) peregrinus (Barrande); Elies and Wood, pp. 488^489, pi. 50, fig. 1 a-d
(non e) \ text-fig. 343.

.1941 Rastrites peregrinus peregrinus Barrande; Pfibyl, pp. 4-6, pi. I, figs 8 and 9; pi. 2, fig. 8; pi 3,
fig. 13; text-figs 5 and 6.

1967 Rastrites peregrinus peregrinus Barrande; Schauer, p. 176, pi. 2, figs 5-7.

.1975 Rastrites peregrinus peregrinus Barrande; Bjerreskov, p. 83, pi. 13, fig. a.

Lectotype. Designated Pfibyl 1941, p. 3, as the specimen figured by Barrande 1850, pi. 4, fig. 6; from the
Llandovery of Bykos, Bohemia.

Material. Approximately 50 fragmentary specimens.

Description. Parallel-sided metathecae, with a height of up to 2-35 mm, arise from narrow prothecae. They are
inclined at 80-90° to the rhabdosome axis. The thecal apertures are slightly hooked so that the crescentic
apertures face proximally. 2TRD lies between 2 0 and 2-5 mm. The proximal end has not been seen.

Remarks. This material agrees well with previous descriptions. The thecal apertural region is
considerably less hooked than in many other rastritids and this would support Sudbury’s (1958)
suggested evolution from M. triangulatus praedecipiens Sudbury, 1958, although, as Schauer ( 1967)
and Rickards, Hutt and Berry (1977) suggest, this may have been via R. socialis Tornquist, 1907.

688

PALAEONTOLOGY, VOLUME 34

PALAEOBIOGEOGRAPHICAL NOTE

Melchin (1989) has recently noted close biogeographical affinities between Arctic Canada and
Siberia and, to a lesser extent, China from the middle Rhuddanian (= Lower Llandovery) onwards
to the top of the Llandovery. Certain genera and species occurring in the above areas are considered
to be absent from Europe. These are the genera Agetograptus , Comograptus Obut and Sobolevskaya,
1968, and ‘ Paramonoclimacis' Wang and Ma, 1977 ( Monograptus sidiachenkoi (Obut and
Sobolevskaya, 1965) and M. falcata (Chen and Lin, 1978)) and the species and subspecies
Climacograptus janischewskyi (Obut, 1949), Metaclimacograptus orientalis (Obut and Sobolevskaya,
1966), ‘ Diplograptus' tcherskyi Obut and Sobolevskaya, 1967 spp., Lagarograptus inexpeditus Obut
and Sobolevskaya, 1968, Petalograptus ankyratus NIGP, 1974, Monograptus cf. M. arciformis Chen
and Lin, 1978, and M. millepeda curtus (Obut and Sobolevskaya, 1968). This faunal difference is
related by Melchin to the ambient current systems.

It is undoubtedly true that a number of the taxa Melchin (1989) quotes appear to exhibit a
distribution which is restricted to Arctic Canada, Siberia and China, including, notably, the highly
distinctive M. sidiachenkoi and La. inexpeditus.

However, the records of Agetograptus primus and Agetograptus sp. herein (being the first definite
records from Europe) necessitate the removal of this genus from Melchin’s list of geographically
restricted taxa. It should also be noted that Rickards, Hutt and Berry (1977) record Me. orientalis
from Sweden, and that M. falcata may be a junior synonym of M. argenteus (Nicholson, 1869), a
species widely recorded from Europe.

This is not to say that Melchin's (1989) ‘faunal provinces' do not exist, only that the number of
geographically restricted taxa involved may be less that he envisages.

Acknowledgements. Special thanks are due to Dr M. G. Bassett who collected the nodules from which the
graptolites for this study were isolated and who also provided locality information and to Dr D. E. B. Bates,
who initiated the study and assisted with electron microscopy. The research for this work was carried out whilst
in tenure of a University of Wales Postgraduate Studentship.

EXPLANATION OF PLATE 3

Figs 1-3, 7. Monoclimacis sp. 1, UCWG935A, distalmost thecae of fragment illustrated in fig. 2, x25. 2,
UCWG935A, fragment showing transition from hooked to monoclimacid thecae, x 10. 3, UCWG935A,
fourth theca of fragment illustrated in fig. 2, x 50. 7, UCWG935A, first theca of fragment illustrated in fig.
2, x 50.

Figs 4 and 5. Prihylograptus leptotlieca (Lapworth, 1876). 4, UCWG926A, distal thecae, ventral view, x 15.
5, UCWG909C, proximal theca, x 50.

Figs 6 and 10. Monograptus communis communis Lapworth, 1876. 7, UCWG908B, mesial fragment; note
thecal spines, x 10. 10, UCWG908B, oblique view of thecal apertures, x 35.

Fig. 8. Monograptus denticulatus sensu Sudbury 1958, UCWG923D, fragment of two thecae, x20.

Figs 9, 1 1, 14, 15. Monograptus millepeda (M‘Coy, 1850). 9, UCWG957A, distal thecae of specimen illustrated
in fig. 15; note absence of thecal spines, x 20. 11. UCWG941A, showing lateral expansion of metatheca,
x 50. 14, UCWG941D, view from proximal end; note thecal spines, x 30. 15, UCWG957A, long fragment,
but with damaged proximal end, x 10.

Figs 12 and 13. Rastrites peregrinus Barrande, 1850. 12, UCWG924B, rhabdosome fragment with damaged
metathecae, x 10. 13, UCWG924A, thecal aperture, x 50.

All scanning electron micrographs.

PLATE 3

LOYDELL, Swedish Llandovery graptolites

690

PALAEONTOLOGY, VOLUME 34

REFERENCES

baily, w. H. 1871. Palaeontological remarks. 22-23. In traill, w. a. and egan, f. w. Explanatory memoir to
accompany Sheets 49, 50 and part of 61 of the Maps of the Geological Survey of Ireland including the country
around Downpatrick, and the shores of Dundrum Bay and Strangford Lough, County of Down. Alexander
Thorn, Dublin and London, 71 pp.

barrande, j. 1850. Graptolites de Boheme. Theophile Haase Fils, Prague, vi + 74 pp., pis 1-4.
bassler, r. 1915. Bibliographic index of American Ordovician and Silurian fossils. Bulletin of the United States
National Museum , 92, 1-718.

bates, D. E. B. and kirk, N. H. 1984. Autecology of Silurian graptoloids. Special Papers in Palaeontology , 32,
121-139.

— zhao yu-hong and loydell, d. k. 1988. Ultrastructural studies on graptolites using scanning
electron microscope. Acta Palaeontologica Sinica, 27, 527-543, pis 1-6.
berry, w. b. n. 1965. Graptolites, 103-9. In kummel, b. and raup, d. (eds). Handbook of paleontological
techniques. Freeman, San Francisco, 852 pp.

bjerreskov, m. 1975. Llandoverian and Wenlockian graptolites from Bornholm. Fossils and Strata, 8. 1-94,
pis 1-13.

1981. Silurian graptolites from Washington Land, western N. Greenland. Bulletin Gronlands Geologiske
Undersogelse , 142, 1-71.

boucek, b. 1943. Uber einige gedornte Diplograptiden des bohmischen und sachsischen Silurs. Mitteilungen der
Tschechischen Akademie der Wissenschaften, 53, (25), 1-6, pi. 1.
bronn, H. G. 1835. Lethaea geognostica, 1. E. Schweizerbart, Stuttgart, 768 pp.

bulman, o. m. b. 1932. On the graptolites prepared by Holm. 3. The structure of some Monograph from the
Retiolites-Shales of Stygforsen, Dalarne. Arkiv for Zoologi , 24A, (9), 10-15, pi. 6.

1970. Graptolithma, with sections on Enteropneusta and Pterobranchia. V1-V163. In teichert, c. (ed.).
Treatise on invertebrate paleontology. Part V. Geological Society of America and University of Kansas Press,
Boulder, Colorado and Lawrence, Kansas, xxxii+163 pp.

— and rickards, r. b. 1968. Some new diplograptids from the Llandovery of Britain and Scandinavia.
Palaeontology, 11, 1-15.

1970. Classification of the graptolite family Monograptidae Lapworth, 1873. V149-V157. In
bulman, o. m. b. 1970. Graptolithina, with sections on Enteropneusta and Pterobranchia. V1-V163. In
teichert, c. (ed.). Treatise on invertebrate paleontology. Part V. Geological Society of America and
University of Kansas Press, Boulder, Colorado and Lawrence, Kansas, xxxii + 163 pp.
carruthers, w. 1867. Graptolites: their structure and systematic position. Intellectual Observer, 11. (4), No.
64, 283-292, pi. 1 ; (5), No. 65, 365-374, pi. 2.

CHEN, xu 1986a. Early Silurian isolated graptolites from Yichang in W. Hubei, China. Acta Palaeontologica
Sinica , 25, 229-238, pis 1-3.

— 19866. On Streptograptus and its paleoautecology. Selected papers from the 13th and 14th Annual
Conventions of Palaeontological Society of China, 115-137, pis 1-3.

and lin, yao-kun 1978. Lower Silurian graptolites from Tongzi, northern Guizhou. Memoirs of Nanjing
Institute of Geology and Palaeontology, Academia Sinica, 12, 1-76, pis 1-19.

Churkin, m. and carter, c. 1970. Early Silurian graptolites from southeastern Alaska and their correlation
with graptolite sequences in North America and the Arctic. United States Geological Survey Professional
Paper , 653, 1-51, pis 1-4.

Conway morris, s. 1981. Parasites and the fossil record. Parasitology , 82, 489-509, pis 1 and 2.
crowther, p. 1981. The fine structure of graptolite periderm. Special Papers in Palaeontology, 26, 1-119.
dumican, l. w and rickards, R. b. 1985. Optimum preparation, preservation, and processing techniques for
graptolite electron microscopy. Palaeontology, 28, 757-766.
elles, G. l. and wood, e. m. r. 1901-18. A monograph of British graptolites. Monograph of the
Palaeontographical Society, 55 (260), 56 (265), 57 (271), 58 (277), 60 (288), 61 (297), 62 (305), 64 (316), 66
(323), 67 (327), 70 (339), i-clxxi, a-m, 1-539, 52 pis.

fortey, R. a, 1989. Review of Early Ordovician (Arenig) graptolites of the Cow Head Group, western
Newfoundland, Canada by S. Henry Williams and Robert K. Stevens. Palaeontology Newsletter, No. 2, 44-46.

and cooper, r. a. 1986. A phylogenetic classification of the graptoloids. Palaeontology, 29, 631-654.
frech, F. 1897. Lethaea geognostica, 1. Lethaea Paleozoica, 1, 11, Graptolithen. Schweizerbach, Stuttgart, pp.
544-684.

LOYDELL: GRAPTOLITES FROM SWEDEN

691

geinitz, H. b. 1842. Uber die Graptolithen. Neues Jahrbuchfur Mineralogie , Geografie , Geologie und Petrefak.-
Kunde , 697-701, pi. 10.

— 1 852. Die Versteinerungen der Grauwackenformation in Sachsen und den angrenzenden Lander- Abtheilungen
( Die Graptolithen ), Wilhelm Engelmann, Leipzig, 58 pp., 6 pis.

hall, J. 1865. Graptolites of the Quebec Group. Geological Survey of Canada; Canadian Organic Remains,
Dec. 2, 151 pp., text-figs 1-31.

hopkinson, J. 1869. On British graptolites. Journal of the Quekett Microscopical Club , 1, 151-166, pi. 8.
howe, m. p. a. 1983. Measurement of thecal spacing in graptolites. Geological Magazine , 120, 635-638.
hutt, J. E. 1974. The Llandovery graptolites of the English Lake District. Part 1 Monograph of the
Palaeontographical Society , 128 (540), 1-56, pis 1-10.

— 1975. The Llandovery graptolites of the English Lake District. Part 2. Monograph of the
Palaeontographical Society , 129 (542), 57-137, pis 11-26.

— and rickards, R. b. 1970. The evolution of the earliest Llandovery monograptids. Geological Magazine ,
107, 67-77.

— and skevington, d. 1970. Isolated Silurian graptolites from the Bollerup and Klubbudden stages of
Dalarna, Sweden. Geologica et Palaeontologica , 4, 1-23.
jaanusson, v. 1973. Morphological discontinuities in the evolution of graptolite colonies. 515-521. In
boardman, R. s., cheetham, a. H. and Oliver, w. a. jr (eds). Animal colonies: their development and function
through time. Dowden, Hutchinson and Ross, Stroudsberg, USA, 603 pp.
jackson, d. E. 1971. Development of Glyptograptus hudsoni sp. nov. from Southampton Island, North-West
Territories, Canada. Palaeontology , 14, 478-86.

— and etherington, j. r. 1969. New Silurian cyrtograptid graptolites from northwestern Canada and
Northern Greenland. Journal of Paleontology, 43, 1114-1121.

jaekel, o. 1889. Uber das Alter des sogen. Graptolithengesteins mit besonderer Beriicksichtigung der in
demselben erhalten Graptolithen. Zeitschrift der Deutschen Geologisclren Gesellschaft , 41, 653-690, pis 28
and 29.

Jones, 0. t. 1909. The Hartfell-Valentian succession in the district around Plynhmon and Pont Erwyd (North
Cardiganshire). Quarterly Journal of the Geological Society of London , 65, 463-537, pis 24 and 25.
koren’, t. n. 1968. Novye Rannesiluriiskie graptolity yuzhnogo Uralu. Paleontologicheskii Zhurnal , 4, 101-3.
[In Russian].

kurck, c. 1882. Nagra nya Graptolitarter fran Skane. Geologiska Foreningens i Stockholm Forhandlingar , 6,
294-304, pi. 14.

lapworth, c. 1873. Notes on the British graptolites and their allies. 1. On an improved classification of the
Rhabdophora. Geological Magazine, (1). 10, 500-504, 555-560, Table I.

— 1876. On Scottish Monograptidae. Geological Magazine, (2), 3, 308-321, 350-360, 499-507, 544-552, pis
10-13, 20.

— 1880. On new British graptolites. Annals and Magazine of Natural History, (5), 5, 149-177, pis 4 and 5.
lenz, A. c. 1974. Evolution in Monograptus priodon. Lethaia, 7, 265-272.

— 1982. Llandoverian graptolites of the Northern Canadian Cordillera: Petalograptus, Cephalograptus,
Rhaphidograptus, Dimorphograptus , Retiolitidae, and Monograptidae. Life Sciences Contributions, Royal
Ontario Museum, 130, 1-154.

— and melchin, m. j. 1989. Monograptus spiralis and its phylogenetic relationship to early cyrtograptids.
Journal of Paleontology , 63, 341-348.

Matthews, s. c. 1973. Notes on open nomenclature and on synonymy lists. Palaeontology, 16, 713-719.
m‘coy, f. 1850. On some new genera and species of Silurian Radiata in the Collection ol the University of
Cambridge. Annals and Magazine of Natural History, (2), 6, 270-290.
melchin, m. j. 1989. Llandovery graptolite biostratigraphy and paleobiogeography. Cape Phillips Formation,
Canadian Arctic Islands. Canadian Journal of Earth Sciences, 26, 1726-1746.

— and lenz, a. c. 1986. Uncompressed specimens of Monograptus turriculatus (Barrande, 1850) from
Cornwallis Island, Arctic Canada. Canadian Journal of Earth Sciences, 23, 579-582.

mitchell. c. e. 1987. Evolution and phylogenetic classification of the Diplograptacea. Palaeontology , 30,
353-405.

NANJING INSTITUTE OF GEOLOGY AND PALAEONTOLOGY, ACADEMIA SINICA (NIGP) 1974. Handbook of Stratigraphy
and palaeontology of Southwest China. Science press, Beijing, 454 pp., 202 pis. [In Chinese].
ni, yu-nan 1978. Lower Silurian graptolites from Yichang, Western Hubei. Acta Palaeontologica Sinica, 17,
387-416, pis 1-4.

692

PALAEONTOLOGY, VOLUME 34

nicholson, h. a. 1868. On the graptolites of the Coniston Flags; with notes on British species of the genus
Graptolites. Quarterly Journal of the Geological Society of London, 24, 521-545, pis 19 and 20.

— 1869. On some new species of graptolites. Annals and Magazine of Natural History, (4), 4, 231-242, pi.

11.

1872. A monograph of the British Graptolitidae. Part 1. General introduction. Blackwood and Sons,
Edinburgh and London, xii + 132 pp.

obut, a. M. 1949. Polevoi Atlas rukovodyashchich graptolitov verchnego silura Kirgizskoy SSR. Publishing
House of the Kirgiz branch of the Academy of Sciences of the USSR, Frunze, 57 pp. [In Russian],

and sennikov, n. v. 1980. Graptolitit silura Sibirskoy platformit (llandoveri basseyna rek Gorbiachin i
Kylumbe). 3-51, 77-78, pis 1-5. In sokolov, b. s., tesakov, u. i. and predmechenskie, n. n. (eds). Silur
Sibirskoy Platformit, Graptolitit i Chitinozoi. Izdatelvstvo ‘Nauka’ Sibirskoe Otdelenie, Novosibirsk, USSR,
88 pp. [In Russian].

— and sobolevskaya, r. f. 1966. Graptolitit rannego silura i Kazachstane. Akademii Nauk SSSR, Sibirskoe
Otdelenie, Institut Geologii Geofiziki. Ministerstvo Geologii SSSR, Nauchno-issledovatelskie Institut Geologii
Arktiki, 1-56, pis 1-8. [In Russian],

-and bondarev, r. f. 1965. Graptolitit silura Taymira. Akademii Nauk SSSR, Sibirskoe Otdelenie,
Institut Geologii Geofiziki. Ministerstvo Geologii SSSR, Nauchno-issledovatelskie Institut Geologii Arktiki,

1 120, pis 1-19. [In Russian].

— and merkureva, a. p. 1968. Graptolitit llandoveri v kernach burovitch skvazhin Norilskogo raiona.
Akademii Nauk SSSR, Sibirskoe Otdelenie , Institut Geologii Geofiziki. Ministerstvo Geologii SSSR, Nauchno-
issledovatelskie Institut Geologii Arktiki, 1-137, pis 1-35. [In Russian],

— and nikolaev, a. a. 1967. Graptolitit i strayigrafiya nizhnego silura okrainnykh podnyatii
Kolymskogo Massiva (Severo-Vostok SSSR). Akademii Nauk SSSR, Sibirskoe Otdelenie, Institut Geologii
Geofiziki. Ministerstvo Geologii SSSR, Nauchno-issledovatelskie Institut Geologii Arktiki, 1-162, pis 1-20. [In
Russian],

packham, G. H. 1962. Some diplograptids from the British Lower Silurian. Palaeontology, 5, 498-526, pis 71
and 72.

paskevicius, j. 1976. On some new Llandoverian diplograptids of the eastern Baltic. 140-151, pis 1 and 2. In
kaljo, d. and koren’, t. n. (eds). Graptolites and stratigraphy. Academy of Sciences of Estonian SSR
Institute of Geology, Tallinn, 256 pp.

pedersen, t. b. 1922. Rastritesskiferen pa Bornholm. Meddelelser fra Dansk geologisk Forening, 6, no. 1 1, 1-29.
portlock, J. E. 1 843. Report on the geology of the county of Londonderry, and of parts of Tyrone and Fermanagh.
Andrew Milliken; Hodges and Smith; Longman; Brown, Green, and Longmans; Dublin and London,
xxi + 784 pp., 40 pis.

pribyl, a. 1941. O ceskych a cizich zastupcich rodu Rastrites Barrande 1850. Rozpravy II. Tridy Ceske
Akademie, 51, no. 6, 1-22, pis 1-3.

1948. Bibliographic index of Bohemian Silurian graptolites. Knihovna Statniho Geologickeho ustavu
Republiky Ceskoslovenske, 22, 1-97.

rabien, a. 1954. Zur Taxionomie und Chronologie der oberdevonischer Ostracoden. Abh. hess. Landesamt.
Bodenforsch., 9, 1-268 [fide Matthews 1973].

richter, R. 1948. Einfiihrung in die Zoologische Nomenklatur. Kramer, Frankfurt a. M., 252 pp.
rickards, R. b. 1970. The Llandovery (Silurian) graptolites of the Howgill Fells (Northern England).
Monograph of the Palaeontographical Society, 123 (524), 1-108, pis 1-8.

1976. Classification of Monograptus : a redefinition of some Llandovery graptolite genera. 155-163. In
kaljo, d. and koren’, t. n. (eds). Graptolites and stratigraphy. Academy of Sciences of Estonian SSR
Institute of Geology, Tallinn, 256 pp.

— hutt, j. e. and berry, w. b. n. 1977. Evolution of the Silurian and Devonian Graptoloids. Bulletin of the
British Museum (Natural History), Geology, 28. 1-120, pis 1-6.

— and koren’, t. n. 1974. Virgellar meshworks and sicular spinosity in Llandovery graptoloids. Geological
Magazine, 111, 193-204, pis 1 and 2.

and rushton, A. w. A. 1968. The thecal form of some slender Llandovery Monograptus. Geological
Magazine, 105, 264-274.

schauer, m. 1967. Biostratigraphie und Taxionomie von Rastrites (Pterobranchiata, Graptolithina) aus dem
anstehenden Silur Ostthuringens und des Vogtlandes. Freiberger Forsclmngshefte, (O, 213, 171-199.
storch, p. 1985. Orthograptus s.l. and Cystograptus (Graptolithina) from the Bohemian lower Silurian. Vestnik
ustredm'ho ustavu Geologickeho, 60, 87-99, pis 1-4.

LOYDELL: GRAPTOLITES FROM SWEDEN

693

- 1988. Earliest Monograptidae (Graptolithina) in the lower Llandovery sequence of the Prague Basin
(Bohemia). Sbomik Geologickych Ved Paleontologies 29, 9-48, pis 1 12.

strachan, i. 1971. A synoptic supplement to ‘A monograph of British graptolites by Miss G. L. Elies and
Miss E. M. R. Wood’. Monograph of the Palaeontographical Society , 125 (529), 1-130.
sudbury, m 1958. Triangulate monograptids from the Monograptus gregarius Zone (Lower Llandovery) of the
Rheidol Gorge (Cardiganshire). Philosophical Transactions of the Royal Society of London , Series B , 241,
485-554, pis'! 9-23.

tornquist, s. L. 1881. Om nagra graplolitarter Iran Dalarne. Geologiska Foreningens i Stockholm
Forhandlingar , 5, 434-445, pi. 17.

— 1890. Undersokningar ofver Siljansomradets graptoliter. L Lunds Universitets Arsskrifter , 26, 1-33, pis
I and 2.

1892. Undersokningar ofver Siljansomradets graptoliter. II. Lunds Universitets Arsskrifter , 28, I 47, pis
1-3.

— 1893. Observations on the structure of some Dipriomdae. Lunds Universitets Arsskrifter , 29, 1-12, 1 pi.
1897. On the Diplograptidae and Eleteroprionidae of the Scanian Rastrites-Beds. Lunds Universitets

Arsskrifter , 33, I -24, pis 1 and 2.

- 1899. Researches into the Monograptidae of the Scanian Rastrites-Beds. Lunds Universitets Arsskrifter ,
35, 1-25, pis 1-4.

1907. Observations on the genus Rastrites and some allied species of Monograptus. Lunds Universitets

Arsskrifter , ( N.S. ), Afd. 2, 3, No. 5, 1-22, pis 1-3.

WANG XIAO-FENG, JIN YU-QIN, WU ZHAO-TONG, FU HAN-YIN, LI ZHO-CHONG and MA GUO-GAN 1977. A
palaeontological atlas of central-south China. Geological Publishing Elouse, Beijing, 470 pp., 116 pis. [In
Chinese].

DAVID K. LOYDELL

Institute of Earth Studies

Typescript received 26 February 1990 University College of Wales, Aberystwyth

Revised typescript received 31 May 1990 Dyfed, SY23 3DB, UK

BELEMNITES FROM THE CONIACIAN TO LOWER
CAMPANIAN CHALKS OF NORFOLK AND
SOUTHERN ENGLAND

by WALTER KEGEL CHRISTENSEN

Abstract. Coniacian to Lower Campanian belemnite faunas from the chalks of Norfolk and southern
England are described, using univariate and bivariate biometric analyses. The faunas include 1 3 species,
subspecies and groups, referred to the following genera: Actinocamax Miller, Gonioteuthis Bayle,
Belemnellocamax Naidin, and Belemnitella d’Orbigny. Two taxa, ‘ Actinocamax' lundgreni Stolley and
Belemnitella p. propinqua (Moberg) of the Belemnitella lineage, are new for England. The nominate subspecies
of Actinocamax verus Miller and two other subspecies are discussed. Samples of Gonioteuthis Bayle are
compared with samples of Gonioteuthis from NW Germany, Belgium and France, the biostratigraphy of which
are well-known. The British specimens of Belemnellocamax ex gr. grossouvrei (Janet) are listed and reviewed.
The concept of B. praecursor Stolley is discussed, as well as Us relationship to the coeval B. alpha Naidin and
to B. mucronata (Schlotheim) from the uppermost Lower Campanian and Upper Campanian. The localities
are placed in the international stratigraphic framework on the basis of their belemnite faunas as well as other
evidence. The traditional zonation of the Coniacian to Lower Campanian of the British Isles is tentatively
correlated with the zonation of NW Germany. The German Gonioteuthis zonation of the upper Coniacian to
Lower Campanian is critically assessed. The palaeobiogeography of the Upper Cretaceous belemnites of the
North European Province is outlined, and the English Coniacian to Lower Campanian belemnite faunas are
commented upon with respect to this framework.

Sharpe (1853) monographed the Upper Cretaceous cephalopods, including the belemnites, of the
Chalk of England. In his work he described five belemnite species representing the genera
Neohibolites Stolley, Actinocamax Miller, Gonioteuthis Bayle, and Belemnitella d'Orbigny. Almost
a century later, Wright and Wright (1951) revised and commented briefly on the species described
by Sharpe, as well as on species described by later authors. They listed 1 3 species which were referred
to the following genera: Neohibolites , Actinocamax , Gonioteuthis , Belemnocamax Crick,
Belemnitella , and Belemnella Nowak.

Since 1853 very little systematic palaeontology has been done on the Upper Cretaceous
belemnites of Great Britain. This is surprising in view of the large amount of work done on English
Lower Cretaceous belemnites and Cretaceous ammonites. As long ago as the 1930s, Brydone ( 1933,
p. 287) stated that: ‘Indeed there is hardly any more crying need than that for a thorough study of
the belemnites of the Upper Chalk’.

Belemnites are uncommon in the Coniacian to Lower Campanian chalks of England. In the
Upper Campanian and Maastrichtian they become more common. The belemnite faunas of the
Coniacian to Lower Campanian include species of the Gonioteuthis lineage and A. verus , while
species of the Belemnitella lineage occur sporadically and Belemnellocamax is very rare. The Upper
Campanian belemnite faunas include only species of Belemnitella , and the Maastrichtian faunas
comprise species of Belemnella with subordinate Belemnitella.

The scope of the present paper is to describe the belemnite faunas from the Coniacian-Lower
Campanian chalks of Norfolk and southern England, utilizing biometric analysis. Southern
England here includes Kent, Essex, Hampshire, and Wiltshire. The uppermost Lower Campanian
belemnite fauna from southern England, consisting of G. quadrata gracilis (Stolley) and B.
mucronata (Schlotheim), is not included in the present paper.

IPalaeontology, Vol. 34, Part 3, 1991, pp. 695-749, 6 pls.|

© The Palaeontological Association

696

PALAEONTOLOGY. VOLUME 34

STAGE

ZONATION
NW GERMANY

TRADITIONAL ZONES

STANDARD
BELEMNITE ZONES
NW EUROPE

SOUTHERN ENGLAND

NORFOLK

>WER CAMPANIAN

upper part

gracilis /
mucronata Zone

Gonioteuthis

quadrata

Zone

post-

Applinocrinus

beds

Gonioteuthis Zone
(=’Zone of granulated
Actinocamax’ sensu
Brydone)

G. quadrata gracilis /
B mucronata

conica /
gracilis Zone

G. quadrata gracilis

papillosa Zone

Applinocrinus

cretaceus

Subzone

Hagenowia

blackmorei

Horizon

U

G. q. quadrata

L

senonensis Zone

pilula /

senonensis Zone

LC

lower part

pilula Zone

Offaster

pilula

Zone

Abundant
0. pilula
Subzone

Echinocorys

depressula

Subzone

lingua /

quad rata Zone

granulata-
qu ad rat a Zone

G. granulataquadrata

Z)

Marsupites /
granulata Zone

Uintacrinus anglicus Zone
Marsupites Zone

Marsupites Zone

U

Uintacrinus /
granulata Zone

Uintacrinus socialis Zone

Uintacrinus Zone

G. granulata

L

z

<

z

O 2

i-

Z

<

co

rogalae /westfalica-
granulata Zone

M. coranguinum Zone

M coranguinum Zone

G. westfalica-
granulata

rogalae /
westfalica Zone

U

_l

corangumum /
westfalica Zone

G. w. westfalica

L

pachti lundulato-
p Heat us Zone

z>

bucailli fprae-
westfalica Zone

G. westfalica
praewestfalica

CONIACIAN
M |

involutus /
bucailli Zone

koeneni Zone

deformis Zone

M. decipiens Zone
M normanniae Zone

M. cortestudinarium
Zone

text-fig. I Zonation and correlation of the Coniacian-Lower Campanian of NW Germany and England,
standard belemnite zones in NW Europe, distribution of belemnites in NW Europe and England, and the age
of localities according to the present study. The stratigraphical age of the localities is based on belemnites and

other evidence.

CHRISTENSEN: CHALK BELEMNITES

697

CHRISTENSEN: CHALK BELEMNITES

698

PALAEONTOLOGY, VOLUME 34

EARLIER WORK

In addition to Sharpe's monograph (1853), belemnites were described and figured by Miller
(1823(7, b), J. de C. Sowerby (1829), Blackmore (1896), Crick (1906, 1907, 1910), Sherborn (1906),
and Brighton (1930). These papers, however, generally are characterized by descriptions of rare
species. Jeletzky (1955a), in his paper on the evolution of Santonian and Campanian species of
Belemnitella, also discussed and figured species of Belemnitella from the Campanian of England.
Christensen (1974) made a morphometric analysis of Actinocamax plenus (Blainville) from the
Plenus Marls of England, described A. bohemicus Stolley from the Middle Coniacian of Yorkshire
(Christensen 1982), and discussed A. primus Arkhangelsky from England (Christensen 1990).

In addition to the papers mentioned above, English Upper Cretaceous belemnites were discussed,
commented upon, and/or used biostratigraphically by Bailey et al. (1983a, 1984), Ernst (1966),
Fletcher and Wood (1978, 1982) Jefferies (1961), Jeletzky (1951a, b), Peake and Hancock (1961,
1970), Reid (1971), Rowe (1901, 1904), Schulz (1979), and Wood (1967), to mention the most
important papers.

MATERIAL

The belemnite material studied in the present paper was placed at my disposal by the Natural
History Museum, London (BMNH), British Geological Survey, Keyworth, Nottingham (BGS),
and Sedgwick Museum, Cambridge (SM). It consists of collections made during the last part of the
nineteenth century and the beginning of the twentieth century, and includes the collections of
H. P. Blackmore, A. W. Rowe, G. E. Dibley, and R. M. Brydone. In addition to these old
collections, belemnites collected recently by C. J. Wood, P. Hammond, A. S. Gale and others are
also included. The belemnites of the old collections are generally labelled only with locality name
and zone. In contrast, the belemnites collected recently are very accurately horizoned with reference
to lithological marker beds.

STRATIGRAPHY

The traditional zonation of the Coniacian-Maastrichtian of the British Isles was summarized by Hancock
(1972) and Rawson et al. (1978) (cf. Text-fig. 1), and later discussed by, among others, Bailey et al. (1983a, 6,
1984), Gale et al. (1987), Gale and Woodroof (1981 ), Mortimore (1986, 1987), Mortimore and Pomerol ( 1987),
Pomerol et al. (1987), Robinson (1986, 1987), Stokes (1975, 1977), Wood (1981), and Wood and Mortimore
(1988).

The Lower Campanian of southern England has been subdivided into several units (fossil horizons) (see
Peake in Hancock 1972, table I, p. 114; see also Wood and Mortimore 1988); five of these units are shown in
Text-figure 1. The Coniacian to Maastrichtian of Norfolk was studied by R. M. Brydone and A. W. Rowe at
the beginning of the century ; this work was summarized by Peake and Hancock (1961, 1970) and Wood (1988).

The Coniacian-Maastrichtian of NW Germany has been carefully studied during the last twenty-five years
by a number of workers (see reviews by Ernst et al. 1979; Schulz et al. 1984). The zonation applied in Germany
is used as a stratigraphic framework in the present paper.

The definition of the Coniacian-Lower Campanian chosen in the present paper is shown in Text-figure 1,
and the belemnite zonations of NW Europe, Balto-Scandia, and the Russian Platform are shown in Text-figure
2. The correlation of the zonations in NW Germany and England is tentative and based mainly upon
information presented by Bailey et al. (1983a, b , 1984). The Uintacrinus socialis Zone in England is defined by
the first occurrence of its index fossil (Bailey et al. 1983a), whereas in Germany U. socialis occurs in both the
Uintacrinus / granulata Zone and the upper part of the subjacent westfalicagranulata Zone (Ulbrich 1971 ; Ernst
and Schulz 1974). The boundary between the Lower and Upper Campanian is placed at the extinction level
of the genus Gonioteuthis. At the beginning of this century this boundary was placed at a lower level by British
workers (e.g. Brydone 1912), namely at the first occurrence of B. mucronata. Fletcher and Wood (1978)
suggested that the boundary as marked by the extinction level of Gonioteuthis was diachronous, being lower
in Northern Ireland and England than in Germany.

CHRISTENSEN: CHALK BELEMNITES

699

Belemnite zones,
NW Europe

Zonal belemnites,
Balto-Scandia

Zonal belemnites,
Russian Platform

w D

03

B casimirovensis

U Maastr,
L 1 U

U Maastr.
L 1 U

6 casimirovensis

D

B. junior

B junior

L Maastrichtian
L 1 U

B fastigata

C

03

£ =3

O

09

03

I-

_i

L.Maastrichtian

5 B sumensis
0)

| B lanceolata
B licharewi

00

B cimbrica

B. sumensis

B obtusa

B pseudobtusa

B lanceolata

B lanceolata

ampanian
| upper part

B. langei
B. minor'

Upper Campanian

Upper Campanian
L 1 U

B. 1. naidini

<b

c S /. langei

trj

T S. /, minor

CD

Upper C;
lower part

B. mucronata

B mucronata

B mucronata

B balsvikensis/B. mucronata

Lower Campanian
lower part | upper part

G. q gracilis/ B mucronata

Lower Campanian

B. mammillatus/G.q.scaniensis
B mucronata

Lower Campanian

L | M | U

B- mucronata/G.q gracilis/
B mammillatus

G. q. gracilis

B alpha/ B praecursor/
G. q. quadrata

z>

G. q. quadrata

_i

G. granulataquadrata

G. granulataquadrata /
B alpha

B. praecursor/A laevigatus/
G granulataquadrata
('Pteria-beds')

Santoman

L | M | U

D

G. granulata —

_i

D

c

03 _

c

o

§ ^
C 0

_J

G. granulata

D

c

03

B praecursor/ G, granulata

G. westtalicagranulata

G. westtalicagranulata/
B propinqua

C

o

c

03

if) _j

B propinqua /

'A. lundgreni uilicus

G. w. westfalica

_i

G w. westtahca/B. propinqua/
A’, lundgreni

Coniacian

L | M | U

G. westfalica praewestfalica

Coniacian

L | M | U

A[ lundgreni

Coniacian

L | U

A. lundgreni

Turonian
L |M|U

Turonian
L |M|U

Turonian
L 1 U

A. plenus triangulus

Cenomanian

L | M | U

Cenomanian

L | M | U

Cenomanian

L | U

A. plenus

A. plenus

A. plenus

A primus

A. primus

A. primus/ N. ultimus

text-fig. 2. Upper Cretaceous belemnite zones in western Europe, Balto-Scandia, and the Russian Platform

(after Christensen 1988).

700

PALAEONTOLOGY, VOLUME 34

The more important localities (Text-figs 3 and 4) mentioned in the text are listed below, with appropriate
bibliographical information and National Grid References.

Norfolk (Text-fig. 3)

The chalk pits in the Gonioteuthis Zone of Norfolk were discussed by Peake and Hancock (1961, 1970) and
Wood (1988).

Wells (TF 928429). According to Peake and Hancock (1970, p. 339A) this pit exposed about 40 m of chalk.
The lowest beds were tentatively referred to the upper third of the depressula Subzone, the chalk in the middle
of the quarry to the cincta belt of the Subzone of abundant O. pilula , and the top beds to the lowest third of
the H. blackmorei Horizon.

The belemnites examined during the course of the present study were collected by P. Hammond. The entire
section has yielded Gonioteuthis , but there is only sufficient material of Gonioteuthis for biometrical analysis
from the lower third of the section. The specimens from this part of the section are referred to G. q. quadrata

CHRISTENSEN: CHALK BELEMNITES

701

and are from the lower part of the lingua / quadrat a Zone of the lower Lower Campanian. This part of the
section also yielded A. veins. Moreover, I have seen a single specimen of Belemnitella sp., from the pilula
Zone.

Stiffkey (TF 975428). The chalk pit was placed in the higher part of the Gonioteuthis Zone by Peake and
Hancock (1970, p. 339A). The belemnites examined by me were collected by P. Hammond and include G. q.
quadrata and B. cf. praecursor. The sample of G. q. quadrata is regarded to be from the upper Lower
Campanian papillosa Zone.

Banham (TM 065878?). This is locality no. 129 of Rowe (MS and field notebooks preserved in the
Palaeontology Library, BMNH, see Wood 1988). Peake and Hancock (1961, p. 313) recorded ‘...a typical
Echinocorys tectiformis Brydone, such as would be expected from the lowest part of the O. pilula Zone;’. I have
seen five specimens of G. granulataquadrata and three specimens of A. verus from Banham; the chalk of this
pit is thus basal Campanian, G. granulataquadrata Zone.

Attlebridge (TG 134165). This is locality no. 168 of Rowe (MS and field notebooks, see Wood 1988). Peake and
Hancock (1961) recorded several pits on Alderford Common, 1 mile north of Attlebridge. On the zonal map
of Peake and Hancock (plate 1), Attlebridge is placed in the top of the Gonioteuthis Zone. 1 have seen a single
specimen which may be referred to G. quadrata gracilis.

Ring land (TG 13951253). This is possibly locality no. 171 of Rowe (MS and field notebooks, see Wood 1988).
According to Wood (1988) the pit is now overgrown and no chalk is visible. The pit was placed very high in
the Gonioteuthis Zone by Wood (1988). I have seen a single specimen, which may be referred to G. quadrata
gracilis.

Kent and Essex (Text-fig. 4)

Gravesend and Grays. The chalk pits of the Gravesend area, Kent, and the chalk pit at Grays, Essex, were
mentioned briefly by Dibley (1900, 1918). He placed them in the top part of the coranguinum Zone. Dibley
(1900) listed nine pits in the Gravesend area, including Fletcher and Co’s Pit. Pits presently available expose
chalk of latest Coniacian to Middle Santonian age (A. S. Gale pers. comm. 1985; C. J. Wood pers. comm.
1988).

The chalk pit at Grays has yielded the following belemnites: G. w. westfalica, B. p. propinqua , and 'A', ex
gr. lundgreni. This belemnite fauna is Lower Santonian.

Belemnites from the Gravesend area include G. vv. westfalica , G. westfalicagranulata , B. p. propinqua , ‘Ad
lundgreni, A. verus and "Ad ex gr. lundgreni', these belemnites are from the upper Lower-Middle Santonian,
Zones of G. westfalica and G. westfalicagranulata.

East Kent. The east Kent succession has yielded G. westfalica praewestfalica, G. vr. westfalica, G.
westfalicagranulata, G. granulata, and A. verus; G. granulata is common in the Marsupites Zone and A. verus
in the Uintacrinus Zone (Bailey et al. 1983a).

I have studied three samples of G. granulata from east Kent (A. W. Rowe Colin, BMNH) (see below). They
are said to come from the Marsupites Zone. On the basis of the mean Riedel Quotient, the sample from the
Northdown brickworks pit near Margate can be assigned to the Upper Upper Santonian Marsupites / granulata
Zone. The two samples from Rifle Butts and Margate, respectively, can be referred to the Lower Upper
Santonian Uintacrinus / granulata Zone. This is at variance with the supposed age given on the labels and the
discrepancy may be due to the small number of specimens in the two samples.

Hampshire (Text-fig. 4)

The chalk pits in Hampshire were listed by Griffith and Brydone (1911) and Brydone (1912).

Micheldever. This is locality no. 295 of Brydone (1912), which is a railway cutting north of the station (SU
518434). The cutting exposed chalk of coranguinum Zone age. The material in the Blackmore Colin, BMNH
probably came from a working pit in the area, exposing chalk of the same age (A. S. Gale pers. comm. 1985).
Micheldever has yielded G. w. westfalica, "Ad lundgreni, B. ex gr. grossouvrei, and A. verus. Moreover, several
specimens of Gonioteuthis, which are destroyed at the alveolar end and therefore not determinable, are also
known from this locality. These belemnites are Lower Santonian.

702

PALAEONTOLOGY, VOLUME 34

text-fig. 4. Map of eastern England showing localities and distribution of the Upper Cretaceous.

CHRISTENSEN: CHALK BELEMNITES

703

Shawford (SU 338274). This is locality no. 1086 of Brydone (1912), also known as Southampton Waterworks
New Pit. This pit formerly exposed chalk of the lower quadrata Zone, possibly the H. blackmorei Horizon
according to C. J. Wood (pers. comm. 1983) and A. S. Gale (pers. comm. 1985).

A small sample of Gonioteuthis from the 'lower course' sensu Brydone was analysed biometrically and
compared with German samples of Gonioteuthis. The Shawford sample is referable to G. q. quadrata and is
regarded to be from the middle Lower Campanian (see below).

I have also studied a small fauna of Belemnitella from this pit, mostly specimens without horizon details. The
Belemnitella fauna is heterogeneous, in contrast to the Belemnitella fauna from East Harnham (see below), and
includes B. cf. praecursor and B. mucronata. The Belemnitella specimens from the ‘lower’ or ‘bottom course’
are all B. cf. praecursor. B. mucronata appears in the uppermost Lower Campanian mucronata /gracilis Zone,
and it thus seems that more zones were present at Shawford, spanning the middle to upper Lower Campanian.
It is, however, possible that the specimens of B. mucronata came from another pit and were purchased from
quarry workers.

Mottisfont (SU 337274). Brydone (1912) listed three pits situated north of the Mottisfont Station, locality nos
1065-1067, placed in the quadrata Zone. Brydone (1914) discussed locality no. 1067, also referred to as
Mottisfont Whiting Pit, and he assigned the lower quarter to the Subzone of abundant O. pilula , whereas the
upper three-quarters was assigned to the quadrata Zone.

Bailey et al. (1983a, p. 35) mentioned that the pit exposed chalk of basal quadrata Zone age, by and large
equivalent to the H. blackmorei Horizon. According to Mortimore (1986, fig. 19), however, the pit showed
chalks from the Old Nore Marl to above the Pepper Box Marls; the old Nore Marl is low in the Subzone of
abundant O. pilula and the Pepper Box Marls are high in the H. blackmorei Horizon (Mortimore 1986, figs 20
and 22).

I have examined B. ex gr. grossouvrei and A. vents from Mottisfont. A. verus was also recorded by Brydone
(1912) in addition to G. quadrata. The presence of A. verus is enigmatic, because this species is not recorded
from above the depressula Subzone in southern England (see Bailey et al. 1983a, fig. 3; Mortimore 1986, fig.
20). In Germany, A. verus ranges up to the top of the pilula Zone sensu germanico at Misburg/Hover near
Hannover (Ernst 19636). The presence of A. verus at Mottisfont may therefore be explained in one of the
following ways: (1) A. verus has a greater stratigraphic range than assumed by English workers, or (2) older
beds may have been exposed formerly at Mottisfont.

Wiltshire (Text-fig. 4)

East Harnham (SU 140288). This pit on the south side of Salisbury is now backfilled, and the collections of
Blackmore, now in BMNH, were probably mostly purchased from quarry workers (A. S. Gale pers. comm.
1985). The nearby section at West Harnham currently exposes high pilula Zone to basal quadrata Zone chalk,
and it appears likely that East Harnham exposed approximately the same levels (A. S. Gale pers. comm. 1985).

The belemnites from East Harnham include G. q. quadrata , B. praecursor , and B. ex gr. grossouvrei. The
sample of G. q. quadrata is regarded to be from the middle Lower Campanian (see discussion below).
According to Blackmore (1896), B. lanceolata ( = B. praecursor) occurred in a thin band in the lower third of
the quadrata Zone. The sample of B. praecursor is homogeneous in contrast to the Belemnitella fauna from
Shawford (see above).

West Harnham. The only belemnite studied by me from this pit is the holotype of Actinocamax blackmorei
Crick, which is referred to the B. grossouvrei group in the present paper. The chalk is presumed to be from the
high pilula Zone-basal quadrata Zone (see above).

PALAEOBIOGEOGRAPHY OF THE UPPER CRETACEOUS BELEMNITES OF

NW EUROPE

Christensen (1975c/, 1976, 1982) surveyed the palaeobiogeography of the Upper Cretaceous
belemnites of Europe (see also Combemorel et al. 1981), and consequently only a brief outline is
given here. Belemnitellidae characterize the North Temperate Realm, but occur also in the northern
part of the Tethyan Realm. They are unknown from the southern part of the Tethyan Realm and
the South Temperate Realm.

The North Temperate Realm includes the North American and North European Provinces, and

704

PALAEONTOLOGY, VOLUME 34

the latter comprises the Central European and Central Russian Subprovinces (Text-fig. 5). The
subprovinces are well-defined in the Coniacian-Lower Campanian and characterized by
independently evolving belemnite lineages: the Gonioteuthis lineage inhabited the Central European
Subprovince and the Belemnitella lineage (including ‘ Actinocamax' lundgreni) inhabited the Central
Russian Subprovince. A. veins and B. ex gr. grossouvrei are widely distributed in both subprovinces;
A. verus occurs commonly, while B. ex gr. grossouvrei is very rare (Christensen 1986).

text-fig. 5. Distribution of Upper Cretaceous biogeographic units in Europe based on belemnites. Upper
Cretaceous land and sea areas represent maximum inundation for all stages. The boundaries are not reliable
in detail, and the biogeographic units are typically gradational in character. After Christensen (1976).

After the extinction of the genera Actinocamax (in the middle Lower Campanian), Gonioteuthis
(at the Lower-Upper Campanian boundary), and Beiemnellocamax (in the lower Upper
Campanian), the genus Belemnitella spread all over the North European Province in the Upper
Campanian. The genus Belemnella appeared in the basal Maastrichtian and continued to the top
of the Maastrichtian. In the Upper Campanian-Maastrichtian, however, no well-defined
subprovinces can be recognized.

Norfolk and southern England are situated in the Central European Subprovince as defined by
belemnites, but the genus Gonioteuthis is generally rarer here than elsewhere.

BELEMNITE FAUNAS OF THE UPPER CONIACIAN-LOWER CAMPANIAN

The late Coniacian chalk of east Kent has yielded a single specimen of Gonioteuthis westfalica
praewestfalica.

The Lower-Middle Santonian belemnite faunas from Grays, Gravesend, and Micheldever consist
of G. w. westfalica , G. westfalicagranulata , 'A.' lundgreni , 'A.' ex gr. lundgreni , B. propinqua , A.
verus , and B. ex gr. grossouvrei. This diverse fauna includes taxa from both the Central European
Subprovince (species of the Gonioteuthis lineage) and the Central Russian Subprovince (species of
the Belemnitella lineage), in addition to the widespread A. verus and B. ex gr. grossouvrei. The
species are very rare with one exception ; A. verus occurs commonly at Gravesend. ’’A.' lundgreni has
its main occurrence on the Russian Platform and in Balto-Scandia. Two specimens of "A.' lundgreni
have been recorded earlier as Gonioteuthis lundgreni /aflf. westfalica sensu Birkelund (see Ernst

CHRISTENSEN: CHALK BELEMNITES

705

1964u, pi. 3, figs 5 and 6) from the Munster Basin in NW Germany (Christensen 1973). B. propinqua
has also its main occurrence on the Russian Platform and in Balto-Scania. Outside this area it has
only been recorded from southern England (this paper). The specimens of "A.' lundgreni, ‘ A .’ ex
gr. lundgreni, and B. propinqua from southern England may be regarded as stray specimens.

The Upper Santonian belemnite fauna consists of G. granulata and A. veins. According to earlier
records, G. granulata is common in the Marsupites Zone and A. verus in the Uintacrinus socialis
Zone.

The lower to middle Lower Campanian fauna includes G. granulataquadrata and G. quadrata,
with subordinate B. praecursor, B. sp., A. verus, and B. ex gr. grossouvrei. This fauna includes taxa
from both of the European subprovinces, as does the Lower to Middle Santonian fauna (see above).
The specimens of B. cf. praecursor from Shawford, Stiflfkey, and Porchester, as well as the specimens
of B. sp. from Wells and the Sussex coast, may be regarded as stray specimens. B. praecursor
occurred fairly commonly at East Harnham, and the population consists of all growth stages. This
may indicate that B. praecursor lived and reproduced in that area during the middle Lower
Campanian.

Fletcher and Wood (1978, pp. 93-94) suggested a B. praecursor event of considerable
biostratigraphic significance in the lower part of the English quadrata Zone, by and large equivalent
to the Hagenowia Horizon (= middle Lower Campanian). They believed that B. praecursor
occurred at this stratigraphic level in Northern Ireland, Shawford in Hampshire, Bramford and
Claydon near Ipswich in Suffolk, Stiffkey in Norfolk, and Hallembaye in eastern Belgium.

In Scania, NW Germany, and Belgium, the genera Gonioteuthis and Belemnitella occur together
in the basal Lower Campanian: G. granulataquadrata and B. alpha in Scania (Christensen 1975u,
1986), G. granulataquadrata and B. praecursor at Braunschweig (Ernst 19646, 1968), G. quadrata
and B. alpha, B. aff. mucronata/ praecursor or B. aff. senior / praecursor in the Munster Basin in NW
Germany (Ernst 19646, Christensen 1986), and G. quadrata and B. praecursor at Hallembaye in
Belgium (Christensen and Schmid 1987). In Scania and NW Germany, the basal Lower Campanian
Gonioteuthis / Belemnitella assemblage also includes B. ex gr. grossouvrei and A. verus. As shown
above, B. praecursor from East Harnham and B. cf. praecursor from Shawford are from the middle
Lower Campanian, whereas B. cf. praecursor from Stiffkey is from the papillosa Zone of the upper
Lower Campanian. I have briefly studied a sample of Belemnitella sp., consisting of more than 100
specimens, from Bramford. This sample was collected by R. M. Brydone and is housed in the
Ipswich Museum. On the basis of the external characters only, this sample seems to be more
advanced than the samples of B. praecursor from Hallembaye and East Harnham, and it may have
come from a higher part of the Lower Campanian. In the Corbieres area of the French Pyrenees,
B. praecursor has recently been recorded from the uppermost Santonian (Christensen et al. 1991).

It can thus be concluded that there is no evidence for a Belemnitella praecursor event in the middle
Lower Campanian as suggested by Fletcher and Wood (1978), even if B. praecursor is recorded from
the middle Lower Campanian of Northern Ireland and southern England.

SYSTEMATIC PALAEONTOLOGY

Morphology of the guard and terminology . The guard is usually the only part of the skeleton preserved, and both
external and internal characters are used for taxonomic classification.

The following characters are generally considered to be of taxonomic value in describing Upper Cretaceous
belemnites belonging to the Belemnitellidae Pavlow: (1) size of guard; (2) shape of guard; (3) structure of the
anterior end; (4) surface markings; (5) internal characters; and (6) ontogeny. The characters were discussed
recently by Christensen (1986).

The Riedel Quotient (Ernst 1964n) is the ratio of length of guard divided by depth of pseudoalveolus, and
the Schlankheits Quotient ( = Slenderness Quotient) of Ernst (1964«) is the ratio of length of guard divided by
dorso-ventral diameter at the anterior end.

Measurements and abbreviations. A list of measured characters and abbreviations is given below (cf. Text-fig.
6): total length of guard (L), length from apex to protoconch (LAP), dorso-ventral diameter at protoconch

706

PALAEONTOLOGY, VOLUME 34

' l

\

\

\ /
X'

\

DVDAE

ventral
fissure

bottom of
ventral
fissure

text-fig. 6. Diagram showing the morphological elements of the guard in (A) Belemnitella praecursor and (B)
Gonioteuthis quadrata. L = length of the guard; LAP = length from the apex to the protoconch; D = depth
of the pseudoalveolus; SD = Schatzky distance; DVDP = dorso-ventral diameter at the protoconch;
DVDAE = dorso-ventral diameter at the alveolar end; FA = fissure angle; and AA = alveolar angle. After

Christensen and Schmid (1987).

(DVDP), lateral diameter at protoconch (LP), dorso-ventral diameter at alveolar end (DVDAE), lateral
diameter at alveolar end (LDAE), maximum lateral diameter (MLD), Schatzky distance (SD), alveolar angle
(AA), fissure angle (FA), Riedel Quotient (RQ), and Slenderness Quotient (SQ).

Linear measurements were made with a vernier caliper to an accuracy of 01 mm, and angles were measured
with a goniometer ocular fitted on a WILD stereomicroscope to an accuracy of 0-5°.

Biometric methods. Species and subspecies variability is analysed using univariate and bivariate statistical
methods and is summarized by descriptive statistics, histograms, and scatter diagrams. The statistics were
calculated according to standard formulae presented by Simpson et al. (1960) and Sokal and Rohlf (1969). The
statistical methods and tests used in the present paper were discussed at length by Christensen (1975a, pp.
31-33).

In the univariate analyses estimates of the following statistics were calculated : arithmetical mean value (X),
standard deviation (SD), and coefficient of variation (CV). In addition, observed range (OR) is reported, and
N is the number of specimens.

The regression line is written: y = a + bx, and the original measurements were used in the calculations,
because of the linear trend on ordinary graph paper and the homoscedastic variance around the regression line.
Estimates of the following statistical parameters were calculated : the slope ( b ) and standard deviation of the
slope (sdQ; the intercept on the y-axis (a) and the standard deviation of the intercept (sdJ; the variance (sd^)
and the standard deviation (SD,/r) of the regression line; and the correlation coefficient (r). N is the number of

CHRISTENSEN: CHALK BELEMNITES

707

specimens. The correlation coefficients were tested for significance by using table Y in Rohlf and Sokal (1969),
and /-tests on the y-intercepts were performed in order to see if the intercept differed significantly from zero.
The regression lines of two samples were compared in the way described by Hald (1957, pp. 571-579).

I have earlier discussed the disadvantages of using ratios in palaeontological studies (Christensen 1973. 1974,
1975a, 1988) especially in cases where growth is allometric. In the present paper, various ratios were calculated
only in order to facilitate comparison with samples of belemnites described by earlier authors.

Order belemnitida Zittel, 1895
Suborder belemnopseina Jeletzky, 1965
Family belemnitellidae Pavlow, 1914

Type genus. Belemnitella d'Orbigny, 1840.

Diagnosis. See Christensen (1975a).

Distribution. Belemnitellidae are restricted to the Upper Cretaceous and are reported from the Lower
Cenomanian to the Upper Maastrichtian. They are mainly distributed in the North Temperate Realm. A few
representatives are also recorded from the northern part of the Tethyan Realm.

Genus actinocamax Miller, 1823
Type species. Actinocamax verus Miller, 1823 by original designation.

Remarks. Naidin (1964) recognized three subgenera within Actinocamax : A. ( Actinocamax ), type
species A. verus Miller, 1823; A. ( Praeactinocamax ), type species A. plentts (Blainville, 1825); and
A. ( Par actinocamax ), type species A. grossouvrei Janet, 1891. This classification was discussed by
Christensen (1982, 1986) and is not followed here. Species belonging to A. ( Actinocamax ) and A.
( Praeactinocamax ) differ only in size, and species referred to A. ( Par actinocamax ) by Naidin are
placed in the genus Belemnellocamax Naidin (see discussion below).

Distribution. Actinocamax ranges from the Lower Cenomanian to the middle Lower Campanian. It is
distributed in the North European and North American Provinces.

Actinocamax verus Miller, 1823
Plate 1, figs 1-9

18236 Actinocamax verus Miller, p. 64, pi. 9, figs 17 and 18 (non pi. 3, figs 16-20).

1906 Actinocamax verus Miller; Sherborn, p. 152, pi. 15, figs 4 and 5.

1952 Actinocamax verus var. dnestrensis Naidin, p. 66, pi. 1, fig. 9; pi. 2, figs 1 and 2.

1962 Actinocamax verus dnestrensis Naidin; Kongiel, p. 113, pi. 20, figs 14-17.

1964 Actinocamax (Actinocamax) verus verus Miller; Naidin, p. 28, text-figs 9 and 10, 18.

1964 Actinocamax ( Actinocamax ) verus dnestrensis Naidin; Naidin, p. 29.

Type. Lectotype, here designated, the original of Miller (18236, pi. 9, figs 17 and 18).

Material. More than two hundred specimens in the BMNH, BGS, and SM collections: about one hundred
specimens from the coranguinum Zone of Gravesend, Kent (e.g. Red Lion Pit and Fletcher’s Pit) and
Micheldever, Hampshire (including BMNH C42708-99 and SM B836-9 from Gravesend, BMNH C43367-77,
C44167 and SM B100082-3 from Micheldever); about one hundred specimens from the Uintacrinus socialis
Zone of southern England (including BMNH C43393-486, SM B65905-19, BGS Zt848); and about one dozen
specimens from the pilula and quadrata Zones of southern England and the Gonioteuthis Zone of Norfolk
(including BMNH C43357-8, BMNH C44876, BMNH C43359-61, BGS GSM 101426).

Description. An Actinocamax with a small, stout guard which is lanceolate in lateral and ventral views. The

708

PALAEONTOLOGY, VOLUME 34

guard is flattened ventrally and the anterior end is compressed. The anterior end may have a low or high cone-
shaped alveolar fracture, be flat, or even have a shallow pseudoalveolus. There is a small pit in the anterior end
for housing the posterior part of the phragmocone. The alveolar fracture is symmetric or asymmetric, and in
the latter case the dorsal side is more incised than the ventral side. The alveolar fracture may be sharply
demarcated from the surface of the guard or may continue gradually into the surface of the guard. The anterior
end exhibits concentric growth layers of the guard and radiating ribs. Dorso-lateral longitudinal depressions
are faintly developed, and the vascular imprints are weakly developed or not present. The ventral fissure and
ventral furrow are not preserved. The surface of the guard may be covered by granules which may form
corrugated transverse lines.

Remarks. A. vents can be distinguished easily from other species of Actinocamax by its small size
and stout guard. Specimens of A. verus having a flat anterior end or a shallow pseudoalveolus differ
from juvenile specimens of G. westfalica in being stouter and more lanceolate in ventral view. Naidin
(1964) recognized three subspecies of A. vents : A. v. verus , A. vents fragilis Arkhangelsky, and A.
vents dnestrensis Naidin, mainly on the basis of the structure of the anterior end. The subspecies
were discussed by Christensen (1986), who placed A. verus dnestrensis in synonymy with A. v. verus ,
and suggested that A. vents fragilis may be considered as a geographic subspecies prevailing on the
Russian Platform.

Two samples of A. verus , one from the Red Lion Pit. Gravesend (basal Santonian) (BMNH
C42708-99), and another from Margate, Kent (Lower Upper Santonian) (BMNH C43393-486),
were analysed with respect to the structure of the anterior end (Table 1). In addition, the analysis
of two samples of A. verus from Scania, Sweden is also reported (Christensen 1986).

All samples showed a series of forms ranging from specimens with a high cone-shaped alveolar
fracture (fragilis- like specimens), through specimens with a low cone-shaped alveolar fracture
(veras-like specimens), to specimens with a shallow pseudoalveolus (dnestrensis- like specimens),
with all intermediate forms. The specimens of the four samples, nevertheless, were subdivided into
three groups, although it was difficult in some cases to decide to which group an individual specimen
should be assigned. Most specimens in all four samples belong to group 1 (Table 1). A few
specimens belong to groups 2 and 3. In the opinion of the author the samples exhibit a normal
variation with respect to the structure of the anterior end; similar variations are also seen in other

EXPLANATION OF PLATE 1

Figs 1-9. Actinocamax vents Miller. 1-3, BGS GSM 101427, great cutting north of Micheldever, locality 295
of Brydone (1912), coranguinum Zone; 1, dorsal view; 2, lateral view; 3, view of the anterior end, x 2. 4-6,
BGS GSM 101426, Mottisfont, locality 1067 of Brydone (1912), quadrata Zone; 4, dorsal view; 5, lateral
view; 6, view of the anterior end, x 2. 7-9, BGS Zt848, Grenham Bay, Birchington, Kent; uppermost 2 m
of U. socialis Zone. 7, dorsal view; 8, lateral view; 9, view of anterior end, x 2.

Figs 10-13. Gonioteuthis westfalica praewestfalica Ernst and Schulz, BGS GSM 1 18339, West Cliff, Ramsgate,
Kent; coranguinum Zone, 10-0—10-5 m beneath Bedwell's Columnar Band; 10, dorsal view; 1 1, lateral view;
12, ventral view; 13, view of the anterior end, x2.

Figs 14-29. Gonioteuthis westfalica westfalica (Schliiter). 14—17, BMNH C7341, Grays, Essex, coranguinum
Zone; 14, dorsal view; 15, lateral view; 16, ventral view; 17, view of the anterior end, x 2. 18-21, BMNH
C44381, Micheldever, coranguinum Zone; 18, dorsal view; 19, lateral view; 20, ventral view; 21, view of the
anterior end, x 2. 22-25, BMNH C43497, Gravesend, coranguinum Zone; a short and stout specimen; 22,
dorsal view; 23, lateral view; 24, ventral view; 25, view of the anterior end, x2. 26-29, BMNH C43496,
Gravesend, coranguinum Zone; 26, dorsal view; 27, lateral view; 28, ventral view; 29, view of the anterior
end, x 2.

Figs 30-37. Gonioteuthis westfalicagranulata (Stolley). 30-33, BMNH C43521, Gravesend, coranguinum Zone;
30, dorsal view; 31, lateral view; 32, ventral view; 33, view of the anterior end, x 2. 34-37, BGS Ztl960,
White Ness near Kingsgate, Kent, coranguinum Zone, 1 m above Barrois Sponge Bed; 34, dorsal view; 35,
lateral view; 36, ventral view; 37, view of the anterior end, x 2.

All specimens are coated with ammonium chloride and the figures are natural size unless otherwise stated.

PLATE 1

CHRISTENSEN, Actinocamax , Gonioteuthis

710

PALAEONTOLOGY, VOLUME 34

samples of Actinocamax, i.a. A. plemis (see Christensen 1974), and in G. westfalica (see Ernst 1964a;
Christensen 1975a). It should, however, be stressed that there is apparently an increase in the
number of fragilis-like specimens from the basal Santonian to the basal Campanian. Ernst (19646,
p. 181 ) has also reported that fragilis-Yike specimens are more common in the basal Campanian than
in the basal Santonian. This apparent trend in the development of the structure of the anterior end
may be real, but should be tested by further studies.

Locality

Age

Group 1

Group 2

Group 3

E

Kullemolla, Scania

basal Campanian

122 (79%)

26 (17%)

6 (4%)

154

Margate, Kent

Lower Upper Santonian

61 (77%)

16 (20%)

2 (3%)

79

Eriksdal, Scania

upper Middle Santonian

100 (86%)

9 (8%)

7 (6%)

116

Red Lion Pit, Gravesend

basal Santonian

80 (94%)

1 (1%)

4 (5%)

85

table I . Estimates of relative abundance of three groups of Actinocamax vents. Group 1 contains specimens
with a low cone-shaped alveolar fracture (vm«-like specimens); group 2 includes specimens with a high cone-
shaped alveolar fracture (fragilis- like specimens); and group 3 contains specimens with a shallow
pseudoalveolus (dnestrensis- like specimens).

Distribution. A. verus is widespread in the North European Province. In NW Europe it is recorded mainly from
the uppermost part of the Lower Santonian I. undulatoplicatus Zone to the middle Lower Campanian Offaster
pilula Zone sensu germanico. In off-shore chalks A. verus is most common in the Upper Santonian. A. verus
is also known from the Lower to Middle Coniacian of Bornholm, Denmark (personal observation). On the
Russian Platform it occurs in the Turonian to lower Lower Campanian (Naidin 1964).

Genus gonioteuthis Bayle, 1878

Type species. Belemnites quadratus Blainville, 1827 by original designation.

Remarks. The evolutionary lineage of Gonioteuthis (in ascending order): G. westfalica praewestfalica
Ernst and Schulz, G. tv. westfalica (Schliiter), G. westfalicagranulata (Stolley), G. granulata
(Blainville), G. granulataquadrata (Stolley), G. quadrata quadrata (Blainville), and G. quadrata
gracilis (Stolley), has been studied especially by Stolley (1897, 1916, 1930), Ernst (1963a, 6, 1964a, 6,
1966, 1968), Ernst and Schulz (1974), Christensen (1975a, 6, 1986, 1988), Christensen and Schmid
(1987), and Jarvis (1980).

Naidin (1964) recognized two subgenera of Gonioteuthis'. G. ( Gonioteuthis ), type species
Gonioteuthis quadrata (Blainville); and G. ( Goniocamax ) Naidin, type species Actinocamax lundgreni
Stolley. The subgenera were treated later as genera by, among others, Naidin and Kopaevich (1977)
and Naidin (1981). Naidin (1964) placed the earliest members of the Gonioteuthis lineage in the
subgenus Goniocamax, in addition to ‘ Actinocamax ’ lundgreni Stolley, and other species from the
Russian Platform, Greenland, and North America. The classification of Naidin was discussed by
Ernst and Schulz (1974) and Christensen (1982, 1986) and is not followed here. '’Ad lundgreni and
G. ( Goniocamax ) are discussed further below.

The Gonioteuthis lineage provides a good tool for biostratigraphy. It is, however, necessary to
analyse homogeneous samples of a certain size in order to make a reliable specific determination,
and limited material has little stratigraphic value (Ernst 1964a; Christensen 1975a) (see also below).
Only the earliest member of the lineage, G. westfalica , is easily recognizable owing to its structure
of the anterior end.

Ernst (1963a, h, 1964a, 6, 1966, 1968) and Ernst and Schulz (1974) characterized samples of
Gonioteuthis on the basis of the mean values of various ratios, including the Riedel Quotient and
the Slenderness Quotient (Schlankheits Quotient of Ernst).

CHRISTENSEN: CHALK BELEMNITES

711

In order to analyse the growth relationship of various characters, I have calculated regression
lines by the least squares method of samples of Gonioteuthis from NW Germany on the basis of
Ernst’s original measurements, which he kindly placed at my disposal. In addition, samples from
Sweden, Bornholm (Denmark), Belgium, France, and England have also been analysed (Christensen
1971, 1973, 1975/7, b, 1986, herein; Christensen and Schmid 1987).

1. Length of guard vs depth of pseudoalveolus. 24 samples of Gonioteuthis were analysed, including
3 samples of G. westfalica from NW Germany and Bornholm, 3 samples of G. westfalicagranulata
from NW Germany and Sweden, 1 sample of G. granulata from NW Germany, 1 sample of G.
granulataquadrata from NW Germany, 13 samples of G. q. quadrata from different levels within the
Lower Campanian of NW Germany, Belgium, France, and England, 2 samples of G. quadrata
gracilis from the top Lower Campanian of NW Germany, and 1 sample of G. quadrata scaniensis
Christensen from the top Lower Campanian of Sweden.

An isometric relationship of the two variates was found in 21 samples. The three samples showing
an allometric relationship are G. quadrata gracilis from the conica/ gracilis and gracilis / mucronata
Zones of Misburg/Hover in NW Germany, respectively, in addition to G. q. quadrata from the
lower linqua/ quadrata Zone of the C.P.L. Quarry at Hallembaye in Belgium. In G. quadrata gracilis
from the gracilis /mucronata Zone the growth relationship is strongly allometric, while in G.
quadrata gracilis from the conica /gracilis Zone and G. q. quadrata from Hallembaye it is slightly
allometric.

2. Length of guard vs dorso-ventral diameter at the alveolar end. 19 samples were analysed : 3 samples
of G. westfalica , 1 sample of G. westfalicagranulata , 1 sample of G. granulataquadrata, 1 1 samples
of G. q. quadrata, 2 samples of G. quadrata gracilis, and 1 sample of G. quadrata scaniensis.

An allometric relationship was found in 12 samples. The seven samples showing an isometric
relationship are G. granulataquadrata from Weinberg, G. q. quadrata from the lower lingua/ quadrata
Zone of Hover and Hallembaye, G. q. quadrata from the lower senonensis Zone of Hover, in
addition to three samples from England and France: Shawford, Mower course’ (middle Lower
Campanian); Stiffkey (papillosa Zone); and Hardivillers (middle Lower Campanian).

In my opinion, samples of Gonioteuthis are better characterized by statistical parameters of the
regression analyses than by ratios. Ernst's Gonioteuthis zonation based on the mean Riedel Quotient
is valid, however, because the relationship of length of guard vs depth of pseudoalveolus is isometric
in almost all analysed samples of Gonioteuthis. The Gonioteuthis zonation is discussed further below.

In the analyses of length of guard v.v dorso-ventral diameter at the alveolar end, the relationship
is allometric to strongly allometric in most samples. It is therefore not valid to calculate the mean
Slenderness Quotient.

According to Ernst (1964c/, fig. 12), samples of G. q. quadrata and G. quadrata gracilis from the
upper senonensis Zone and above are characterized by having a mean Slenderness Quotient of 6-0
or more, with the mean value increasing stratigraphically upwards. The increasing mean value of
the Slenderness Quotient, however, is a function of the allometric relationship of length of guard
vs dorso-ventral diameter in connection with the diminishing mean length of the guard in samples
of Gonioteuthis from the upper Lower Campanian. Due to the allometric relationship, small guards
are generally more slender than large guards, and consequently the mean Slenderness Quotient will
increase when the mean length of the guard decreases.

Phylogeny of Gonioteuthis. The Gonioteuthis lineage is an outstanding example of phyletic
gradualism, namely slow gradual transformation of a suite of characters within populations
through time. The Gonioteuthis stock existed for about 10 Ma and the general trend in evolution is
the gradual calcification of the anterior part of the guard. In G. westfalica the anterior end may be
convexly conical, flat, or developed as a shallow pseudoalveolus, while in stratigraphically younger
representatives the depth of the pseudoalveolus increases and may be up to one-third of the entire
length of the guard. Simultaneously with the development of a deeper pseudoalveolus, the guard

712

PALAEONTOLOGY, VOLUME 34

becomes increasingly stout and large, reaching a maximum in G. quadrata. Another characteristic
feature is the gradual development of granulation. The oldest member, G. westfalica praewestfalica,
does not possess granulation (Ernst and Schulz 1974). The succeeding taxon, G. w. westfalica , shows
a wide variation with respect to this character; some specimens carry scattered granules on the
dorsal and/or ventral side of the guard, while in others the granules appear to be arranged in
longitudinal rows. The same pattern is also valid for G. westfalicagranulata. In stratigraphically
younger species, however, the granulation becomes a very prominent character. In samples of
G. westfalica the shape of the guard is highly variable, and a great proportion of the guards are
lanceolate in ventral view. In younger populations the variation in the shape of the guard is smaller,
and guards which are lanceolate in ventral view are relatively rare. The cross-section of the anterior
end is oval to pointed oval in G. westfalica and subrectangular to subquadrate in stratigraphically
younger representatives, with all intermediate stages. In the uppermost Lower Campanian, the
genus Gonioteuthis is characterized by a return of earlier features; e.g. reduced length of the guard,
diminished depth of the pseudoalveolus, and increased slenderness of the guard.

text-fig. 7. Zonation of the Coniacian-Lower Campanian of NW Germany, Gonioteuthis zones, and the mean
value and observed range of the Riedel Quotient. Column 2 shows the mean value of the Riedel Quotient of
typical samples, column 3 the mean value of small, closely-spaced, samples from the chalk of Lagerdorf,
column 4 the mean value of samples from broad stratigraphic horizons of the marls of Misburg/Hover, and
column 5 the mean value of samples from Braunschweig and Essen. After Christensen (1988).

CHRISTENSEN: CHALK BELEMNITES

713

Gonioteuthis zonation. Ernst (1964a) established a Gonioteuthis zonation of the upper Coniacian-
Lower Campanian, mainly on the basis of the mean Riedel Quotient (Text-fig. 7). The following
points should be stressed :

1. In the earliest members variation of the Riedel Quotient is large and this variation gradually
diminishes upwards (see column 6).

2. There is a rapid change in the mean Riedel Quotient from the upper Coniacian to the basal
Lower Campanian, but virtually no change in stratigraphically younger species. The mean Riedel
Quotient is about 4 during most of the Lower Campanian except in the uppermost Lower
Campanian. In G. quadrata gracilis the mean Riedel Quotient is slightly larger than in G. q. quadrata
from the middle Lower Campanian. This is due to the allometric relationship of the length of the
guard vs the depth of the pseudoalveolus in G. quadrata gracilis in connection with the diminishing
mean length of the guard. Due to allometry, small specimens generally have a shallower
pseudoalveolus than large specimens, and consequently the mean Riedel Quotient increases when
the mean length decreases. Gonioteuthis zonation of the Lower Campanian is therefore hardly
workable using only the mean Riedel Quotient. By using other characters, however, it is possible
to determine samples of Gonioteuthis with some confidence. G. quadrata gracilis is smaller than
G. q. quadrata. Moreover, G. quadrata gracilis has notches in the pseudoalveolus and it is more
strongly granulated than G. q. quadrata from the lower Lower Campanian.

3. Specimens with a very deep pseudoalveolus (Riedel Quotient about 2-5) occur only in the
middle Lower Campanian (see column 6).

4. There is a discrepancy between the samples from Lagerdorf and Misburg/Hover in the lower
and middle Lower Campanian. The samples from Misburg/Hover generally have a deeper
pseudoalveolus (compare columns 3 and 4).

5. A single specimen or some few specimens cannot be assigned safely to a species. Lor instance,
a specimen with a Riedel Quotient of 7 might belong to either G. w. westfalica , G.
westfalicagranulata , G. granulata , or G. granulataquadrata.

The Gonioteuthis zonation of Ernst (1964a), was shown to be workable by Ernst (1966, 1968),
Ernst and Schulz (1974), Ulbrich (1971), Christensen (1975a, b , 1986, 1988), and Jarvis (1980).

Distribution. Gonioteuthis is known from the Middle Coniacian /. involutus Zone to the boundary between the
Lower and Upper Campanian, and the extinction level of the genus has been proposed by several authors,
including Jeletzky (1958) and Schulz et ai (1984), as the boundary between the Lower and Upper Campanian.

The genus had its evolutionary centre in northwestern Europe and is found almost exclusively in the Central
European Palaeogeographic Subprovince. A few specimens are reported from the northernmost part of the
Tethyan Realm.

Gonioteuthis westfalica praewestfalica Ernst and Schulz, 1974
Plate 1, figs 10-13

1974 Gonioteuthis westfalica praewestfalica Ernst and Schulz, p. 49, pi. 5, figs 2-9.

1983a Gonioteuthis praewestfalica Ernst and Schulz; Bailey et al ., p. 35.

Holotvpe. The specimen figured by Ernst and Schulz (1974, pi. 5, fig. 4) is the holotype.

Material. BGS GSM 118339 from West Cliff, Ramsgate, Kent, coranguinum Zone, 1 0 0—10-5 m beneath
Bedwell’s Columnar Band (i.e. top Coniacian sensu Bailey et al. 1984).

Remarks. The subspecies was described in detail by Ernst and Schulz (1974) on the basis of German
material. They regarded it as the earliest member of the Gonioteuthis lineage, appearing in the
Middle and Upper Coniacian chalks of Lagerdorf. They stated that praewestfalica can be
distinguished from the nominate subspecies only on the basis of biometric analysis of a sample. The
main characters separating praewestfalica from westfalica are the ventrally flattened and club-
shaped guard. In addition, praewestfalica is not granulated, whereas westfalica may be granulated.

714

PALAEONTOLOGY, VOLUME 34

The important characters of the English specimen are as follows: length of guard, 49T mm;
Riedel Quotient, 17-5; Slenderness Quotient, 7-2; maximum lateral diameter divided by the lateral
diameter at the alveolar end, T2. Moreover, the specimen is not granulated. The specimen falls
within the variation of G. westfalica praewestfalica and is from the Upper Coniacian. It is therefore
referred to as G. westfalica praewestfalica.

Distribution. G. westfalica praewestfalica is recorded from NW Germany and England. It occurs in NW
Germany in the Middle and Upper Coniacian (Ernst and Schulz 1974). The English specimen came from the
Upper Coniacian.

Gomoteuthis westfalica westfalica (Schltiter, 1876)

Plate 1, figs 14-29

Synonymy. See Christensen (1975a).

Lectotvpe. The specimen figured by Schltiter (1876, pi. 53, fig. 10) was designated as lectotype by Ernst and
Schulz (1974, p. 50).

Material. BMNH C44381, Micheldever, coranguinum Zone; BMNH C7341 and C59277, Grays, coranguinum
Zone; BMNH C43522-3, Red Lion Pit, Gravesend, coranguinum Zone; BMNH C43496-9, C43502, C43504,
C43510, C43518 and C43520, Gravesend, coranguinum Zone; BMNH C43524, Northfleet, Kent, coranguinum
Zone; BMNH C43525, North Foreland, Kent, coranguinum Zone.

Dimensions.

L

D

DVDAE

LDAE

MLD

RQ

SQ

BMNH C4438I

46-3

3-2

6-8

5-7

7-4

14 5

6-8

BMNH C7341

56-5

1-8

8-2

7-7

9-2

314

6-9

BMNH C43522

501

3-8

7-8

7-3

90

13-2

6-4

BMNH C43496

65-4

4-5

9-9

9-0

10-7

14-5

6-6

BMNH C43497

54-8

6-8

110

9-6

1 15

81

50

BMNH C43498

60*

3-2

9-5

8-5

10 5

18-8

6-3

BMNH C43499

62-2

3-8

9-4

8-6

9-8

16-4

6-6

BMNH C43502

65-4

4-5

10 1

9-3

10 8

14-5

6-5

BMNH C43504

55-3

4-5

9-6

8-2

9-7

12-3

5-8

BMNH C43510

51-2

2-0

—

—

—

25-6

—

BMNH C43518

65-6

4-0

90

8-2

10 0

16 4

7-3

BMNH C43520

58*

8-2

1 16

10-3

110

7-1

5-0

* = estimated.

Univariate analysis. The mean value, standard deviation, coefficient of variation, and observed range of length
of guard (in mm), Riedel Quotient, and Slenderness Quotient of the nine complete specimens from Gravesend
are reported below.

Gravesend, Kent:

Character

N

X

SD

CV

OR

L

9

59-8

5-3

8-9

51-2-65-6

RQ

9

14-8

5-6

37-7

7-1-25-6

SQ

8

61

0-8

131

5-0-7-3

Remarks. Ernst (1964a, 1968) recognized two zones of G. westfalica in the Lower and Middle
Santonian. The lower zone is characterized by samples of G. westfalica having a mean Riedel
Quotient above 1 1-5, and the upper zone by samples with a mean Riedel Quotient between 9-5-1 1-5
(Text-fig. 7).

CHRISTENSEN: CHALK BELEMNITES

715

The small sample from Gravesend has a mean Riedel Quotient of 14-8, suggesting the lower G.
westfalica Zone. The large observed range of the Riedel Quotient, however, in connection with the
fact that G. westfalicagranulata occurs at Gravesend too (see below), may indicate the presence of
the upper G. westfalica Zone at Gravesend. The specimens from Micheldever, Grays, and the Red
Lion Pit are regarded to be from the lower G. westfalica Zone on the basis of their Riedel Quotient.

Distribution. G. w. westfalica is very common in NW Germany and Scania. Outside these areas it has been
recorded from most parts of the Central European Subprovince except east of Ukraine. It occurs in the Lower
and Lower Middle Santoman.

Gonioteuthis westfalicagranulata (Stolley, 1897)

Plate 1, figs 30-37

Synonymy. See Christensen (1975a, b).

Lectotype. The specimen figured by Stolley (1897, pi. 2, fig. 16; pi. 3, fig. 6) was designated as lectotype and
refigured by Christensen (1975b, pi. 10, fig. 1; text-fig. 2a).

Material. BGS Ztl960, White Ness, near Kingsgate, Kent, uppermost coranguinum Zone immediately below
entry of Uintacrinus. BMNH C43521, Gravesend, coranguinum Zone.

Dimensions.

L

D

DVDAE

LDAE

MLD

RQ

SQ

BGS Ztl960

50-6

5-6

7-9

7-2

8-1

90

6-4

BMNH C43521

60*

90

9-9

9-2

9-2

6-7

61

* = estimated.

Remarks. BMNH C43521 carries granules and the Riedel Quotient is outside the range of G.
westfalica. It is therefore tentatively referred to G. westfalicagranulata. Ztl960 came from beds
correlative with the German G. westfalicagranulata Zone and falls within the variation of G.
westfalicagranulata .

Distribution. The species occurs in the Upper Middle Santonian G. westfalicagranulata Zone.

Gonioteuthis granulata (Blainville, 1827)

Synonymy. See Christensen (1975a).

Type. Lectotype, here designated, the original of Blainville (1827, pi. 1, fig. 10).

Material. This species occurs fairly commonly at the level of the Marsupites acme in east Kent and Sussex
( Bailey et al. 1 983a). I have analysed biometrically three small samples : ( 1 ) a sample labelled 1 Marsupites band,
Northdown brickworks pit, near Margate’ (BMNH C42650-5, C42666, C42668-9); (2) a sample labelled
‘ Marsupites Zone, Rifle Butts, Margate’ (BMNH C43250-1, C43254-7); and (3) a sample labelled 'Marsupites
band, Margate’ (BMNH C43242-5, C43248, C43236); all A. W. Rowe Colin.

Univariate analysis.

Northdown brickworks pit:

Character

N

X

SD

cv

OR

L

9

613

6-8

111

50-7-71-6

RQ

9

6-5

0-8

1 1-8

5-5— 7-5

SQ

9

5-7

0-5

10-6

48-6-4

716

PALAEONTOLOGY, VOLUME 34

Rifle Butts:

Character

N

X

SD

CV

OR

L

6

67-3

7-2

10 7

55-0-81 0

RQ

6

74

11

15-3

66-9-6

SQ

6

6-2

II

17-7

4-9-7-9

Margate :

Character

N

X

SD

CV

OR

L

6

60-8

54

8-9

56-3-71-2

RQ

6

7-7

1-3

16-3

6 0-9-1

SQ

6

6-2

0-7

1 1-0

51-6-9

Remarks. Ernst (1964a) recognized two zones

of G.

granulata in

the Upper Santonian: a lower

Uintacrinus / granulata Zone characterized by samples of Gonioteuthis having a mean Riedel
Quotient of 7-0-8 0, and an upper Marsupites / granulata Zone defined by samples of Gonioteuthis
with a mean Riedel Quotient of 6-0-7 0 (Text-fig. 7).

The three samples can be referred to G. granulata on the basis of their mean Riedel Quotients.
The sample from Northdown brickworks pit has a mean Riedel Quotient of 6-5 and is thus from
the upper Santonian Marsupites /granulata Zone. The samples from Rifle Butts and Margate have
mean Riedel Quotients of 74 and 7-7, respectively, suggesting the lower Upper Santonian
Uintacrinus / granulata Zone. This is at variance with the supposed age given on the labels, a
discrepancy possibly due to the small number of specimens in the two samples.

Distribution. G. granulata occurs in the Upper Santonian.

Gonioteuthis granulataquadrata (Stolley, 1897)

Synonymy. See Christensen (1975a, b )

Lectotype. The specimen figured by Stolley (1897, pi. 2, fig. 23; pi. 3, fig. 13) was designated as lectotype and
refigured by Christensen (19756, pi. 10, fig. 2; text-fig. 2a).

Material. BMNH C43298, C43000, C43302-4, Banham (pit no. 129 of Rowe), Norfolk.

EXPLANATION OF PLATE 2

Figs 1-27. Gonioteuthis quadrata quadrata (Blainville), Wells, Norfolk (1-12); East Harnham, Wiltshire
(13-18); and Stiffkey, Norfolk (19—27). 1-4, SM B953241, a large average-shaped specimen, dorsal view; 2,
lateral view; 3, ventral view; 4, view of the anterior end. 5-8, SM B95375, a medium-sized specimen with
an average shape; 5, dorsal view; 6, lateral view; 7, ventral view; 8, view of the anterior end. 9-12, SM
B95323, a small specimen with an average shape; 9, dorsal view; 10, lateral view; 1 1, ventral view; 12, view
of the anterior end. 13, BMNH C44308, a medium-sized guard in ventral view. 14-15, BMNH C44326, a
medium-sized guard; 14, lateral view; 15, ventral view. 16-18, BMNH C44281, a large stout specimen; 16,
dorsal view; 17, ventral view; 18, view of the anterior end. 19-21, SM B97232, a large specimen; 19, dorsal
view; 20, dorsal view; 21, ventral view. 22-24, SM B97237, a slender medium-sized specimen; 22, dorsal
view; 23, lateral view; 24, ventral view. 25-27, SM B97247, a small slender specimen; 25, dorsal view; 26,
ventral view; 27, view of the anterior end, x 2.

All specimens are coated with ammonium chloride. All figures are natural size unless otherwise stated.

PLATE 2

sag

CHRISTENSEN, Gonioteuthis

718

PALAEONTOLOGY, VOLUME 34

Univariate analysis.

Character

N

X

SD

C V

OR

L

5

680

9-3

13-7

59-4-80-7

RQ

5

5-5

0-8

14 1

4-2-6- 1

SQ

5

60

0-8

13-4

51-6-8

Remarks. According to Ernst ( 1964c/) G. granulataquadrata has a mean Riedel Quotient of 50-60
(Text-fig. 7). The small sample from Banham has a mean Riedel Quotient of 5-5 and may be referred
to this species.

Distribution. G. granulataquadrata occurs in the basal Lower Campanian ( G , granulataquadrata Zone).

Gonioteuthis quadrata quadrat a (Blainville, 1827)

Plate 2, figs 1-27

1827 Belemnites quadratus Blainville, p. 62, pi. 1, fig. 9.

1878 Gonioteuthis quadrata (Blainville); Bayle, pi. 23, figs 6-8.

1892 Actinocamax quadratus (Blainville) var. ampullacea Stolley, p. 232, pi. 8, fig. 1.

1892 Actinocamax quadratus (Blainville) var. oblonga Stolley, p. 233, pi. 7, fig. 5.

1897 Actinocamax quadratus (Blainville); Stolley, p. 284, pi. 2, fig. 24; pi. 3, fig. 14.

1952 Gonioteuthis quadrata (Blainville) var. cylindrica Naidin, p. 81, pi. 5, figs 4 and 6; pi. 7, fig. 5.

1964a Gonioteuthis quadrata quadrata (Blainville); Ernst, pi. 1, figs 6 and 7; pi. 2, figs 5 and 6; pi. 4,

fig. 1.

1971 Gonioteuthis quadrata quadrata (Blainville); Ulbrich, pi. 1, Reihe 1, figs 1- 16, Reihe 2, figs 1-14,
Reihe 3, figs 1-8.

1980 Gonioteuthis quadrata quadrata (Blainville); Jarvis, pi. 116, figs 1-15.

1987 Gonioteuthis quadrata quadrata (Blainville); Christensen and Schmid, p. 16, pi. 3, figs 8-1 1.
Type. Lectotype, here designated, the original of Blainville (1827, pi. 1, fig. 9).

Material. Wells: 30 near-complete specimens, E. depressula Subzone of the O. pilula Zone, Hammond Colin,
SM B95314-1 5, B95318-24, B95328, B95330-35, B95338-39, B95341-42, B95345-47, B95351, B95353,
B95367-70, B95375, B95382; Stiffkey: 27 near-complete specimens, higher part of the Gonioteuthis Zone,
Hammond Colin, SM B97232^J8, B97250, B97252-56, B97263, B97268-69, B97273; East Harnham: 38 near-
complete specimens, top pilula Zone or base of quadrata Zone, Blackmore Colin, BMNH C44260, C44266,
C44272-73, C44275-76, C44281-87, C44290, C44293, C44299-305, C44307-09, C44314, C44322, C44326,
C44330, C44332, C44336-37, C59164, C59 167-68, C59 170-71 ; Shawford, Tower course’, quadrata Zone: 15
near-complete specimens, Brydone Colin, BGS GSM 101087, 101090-91, 101093, 101096, 101098-103,
101106, 101108-1 1. In addition, I have seen many apical fragments from these localities.

Univariate analyses. The mean value, standard deviation, coefficient of variation, and observed range of the
length of the guard, the Riedel Quotient, and the Slenderness Quotient are shown in Table 2. Histograms for
the samples from Stiffkey, Wells, and East Harnham are shown in Text-figure 8. Histograms were not made
for Shawford owing to the small number of specimens. The size-frequency distributions were tested for
normality using the Kolmogorov-Smirnov test for goodness of fit, and none of the distributions differed
significantly from normal (see Table 3). It should be emphasized that juvenile specimens are absent in the
samples. On the basis of the univariate analyses the samples can be regarded as homogeneous.

Bivariate analyses. Scatter plots of the length of the guard vs the depth of the pseudoalveolus, and the length
of the guard vs the dorso-ventral diameter at the alveolar end of the samples are shown in Text-figures 9-12,
as are the regression lines. The equation of the regression lines are shown in Table 4. Values of the correlation
coefficients are very highly significant (P < 0 001 with N-2 degrees of freedom). The ^-intercepts were tested
by the /-test to see if they differed significantly from zero. In two cases. Wells and East Harnham, the /a-values
are significant (002 > P> 001, with N-2 degrees of freedom), implying an allometric relationship of the

CHRISTENSEN: CHALK BELEMNITES

719

text-fig. 8. Histograms of the length of the guard (L), Riedel Quotient (RQ), and Slenderness Quotient (SQ)
of three samples of Gonioteuthis quadrata quadrata from Stitfikey, East Harnham, and Wells. The figures above

the bars are the actual number of specimens.

length of the guard with the dorso-ventral diameter at the alveolar end. In the other cases, the ^-intercepts do
not differ significantly from zero at the 5% level, implying an isometric relationship of the variates.

On the basis of these bivariate analyses, the samples are regarded as homogeneous, which is consistent with
the univariate analyses.

Remarks. The mean values of the length of guard are of little taxonomic value, because juvenile
specimens are absent in all samples. On the basis of the histograms and observed ranges of the

720

PALAEONTOLOGY, VOLUME 34

WELLS

Character

N

A'

SD

CV

OR

L

30

63.8

8.2

12.8

45.1-80.0

RQ

29

4.4

0.6

14.0

3.4- 5.9

SQ

30

5.6

0.5

9.2

4.7- 6.8

EAST HARNHAM

Character

N

X

SD

CV

OR

L

38

66.3

7.3

11.0

54.0-83.3

RQ

38

4.5

0.6

13.9

3.4- 5.5

SQ

32

6.2

0.6

9.0

5.0- 7.5

SHAWFORD, ’lower course’

Character

N

X

SD

CV

OR

L

15

66.9

7.2

10.8

53.0-78.8

RQ

15

4.4

0.5

12.4

3.5- 5.4

SQ

13

6.0

0.5

7.7

5.3- 7.1

STIFFKEY

Character

N

X

SD

CV

OR

L

27

62.7

6.5

10.3

51.8-75.0

RQ

27

4.1

0.4

10.1

3.2- 4.8

SQ

25

6.2

0.6

9.4

4.9- 7.6

table 2. Univariate analyses of four samples of Gonioteuthis quadrata quadrata from Wells, East Harnham,
Shawford, Mower course’, and Stiffkey. L = length of the guard; RQ = Riedel Quotient; SQ = Slenderness
Quotient; N = number of specimens, X = mean value; SD = standard deviation; CV = coefficient of
variation ; and OR = observed range.

length of the guard, it is obvious, nevertheless, that the specimens from StilTkey generally are smaller
than the specimens from the other samples.

The regression lines of the length of the guard vs the depth of the pseudoalveolus of the four
samples are shown in Text-figure 13. The samples from Shawford and Wells were compared and t-
tests showed that the two samples do not differ significantly at the 5% level. As regards Wells vs
Stiffkey and East Harnham vx Stiffkey, Mests showed that the positions of the regression lines are
significantly different (0 05 > P > 0-01, with N-2 degrees of freedom). The specimens from
Stiffkey and Shawford generally have a deeper pseudoalveolus than the specimens from Wells and
East Harnham. The regression lines of the length of the guard vs the dorso-ventral diameter at the
alveolar end of the four samples are shown in Text-figure 14 from which it is obvious that the
samples from East Harnham, Shawford, and Stiffkey are virtually identical. On the other hand, the
specimens from Wells are stouter. The samples from Wells and Stiffkey were compared and the t-

CHRISTENSEN: CHALK BELEMNITES

721

WELLS

Character

N

D

Probability

L

30

0.1200

P > 0.50

RQ

29

0.0583

P > 0.50

SQ

30

0.1067

P > 0.50

EAST HARNHAM

Character

N

D

Probability

L

38

0.1105

P > 0.50

RQ

38

0.1111

P > 0.50

SQ

32

0.1125

P > 0.50

STIFFKEY

Character

N

D

Probability

L

27

0.1104

P > 0.50

RQ

27

0.0815

P > 0.50

SQ

25

0.0984

P > 0.50

table 3. Results of the Kolmogorov-Smirnov tests for goodness of fit for three samples of Gonioteuthis
quadrata quadrata from Wells, East Harnham, and Stiffkey.

test showed that the difference of the positions is very highly significant (P < 0 001 with 51 degrees
of freedom).

In conclusion, the specimens from Stiffkey are generally smaller, more slender, and have a deeper
pseudoalveolus than the specimens from Wells. The specimens from East Harnham and Shawford
are virtually identical with respect to the analysed characters. As regards the shape of the guard, the
samples from East Harnham and Shawford cannot be differentiated from the sample from Stiffkey.
With respect to the depth of the pseudoalveolus in relation to the length of the guard there is a slight
difference between East Harnham and Stiffkey; the specimens from Stiffkey have a deeper
pseudoalveolus.

Comparisons. The four samples from England were compared to seven German samples described by Ernst
(1964a, b) and two samples from Belgium and France. The sample from the C.P.L. Quarry at Hallembaye,
Belgium was described by Christensen and Schmid (1987), and the sample from Hardivillers, France was
described by Jarvis (1980), and was remeasured for the present study. The results of the univariate analyses of
the nine samples are shown in Table 5, and the equations of the regression lines are reported in Table 6.

1. Wells. The sample from Wells was compared to samples of G. q. quadrata from Bremer and Hallembaye
from the lower part of the lingua/ quadrata Zone.

The maximum length of the guard is about 80 mm in the three samples, as in samples of Gonioteuthis from
the lower and upper parts of the Lower Campanian of NW Germany (Ernst 1964a, fig. 8). As regards the length
of the guard vs the depth of the pseudoalveolus, the sample from Wells does not differ significantly at the 5%
level from the Bremer and Hallembaye samples. With regard to the length of the guard vs the dorso-ventral
diameter at the alveolar end, the positions of the regression lines of the Wells and Hallembaye samples differ
significantly (0 05 > P > 0 025, with 84 degrees of freedom). The specimens from Wells are generally stouter
than the specimens from Hallembaye.

722

PALAEONTOLOGY, VOLUME 34

4-

'To 1 20 1 30 1 40 1 50 1 6(0

L in mm
70 1 80~

. D in
mm

20-

16-

Wells

12-

8-

4

L in mm

10

20

30

40

60

70

80

text-fig. 9. Scatter plots and regression lines for Gonioteuthis quadrata quadrata from Wells, a, length of the
guard (L) v.y dorso-ventral diameter at the alveolar end (DVDAE). b, length of the guard (L) vs depth of the

pseudoalveolus (D). + = mean value.

CHRISTENSEN: CHALK BELEMNITES

723

DVDAE
in mm

12-

8-

4-

To ' 20 ' 30 ' 40 ' 50 ' (dO

L in mm

“i 1 1 —

70 80

20'

D in
mm

16

Shawford

12-

8-

L in mm

' 10 ' 20 ' 30 ' 40 ' 50 ' 60 ' 70 ' 8o"

text-fig. 10. Scatter plots and regression lines for Gonioteuthis quadrata quadrata from Shawford. a, length
of guard vs dorso-ventral diameter at the alveolar end (DVDAE). b, length of the guard (L) vs depth of the

pseudoalveolus (D). + = mean value.

On the basis of the comparisons made above, the sample from Wells is referred to G. q. quadrata and is
considered to be from the lower part of the lingua/ quadrata Zone, as are the samples from Bremer and
Hallembaye.

2. Stiffkey. The sample from Stiffkey was compared to samples of G. q. quadrata and G. quadrata gracilis from
the upper senonensis, papillosa, and conica/ gracilis Zones of Hover/Misburg.

The maximum length of guard in the Stiffkey sample is 75 mm, as in German samples of Gonioteuthis from
the upper part of the Lower Campanian (Ernst 1964a, fig. 8). As regards the length of the guard vs the depth

724

PALAEONTOLOGY, VOLUME 34

WELLS

y = a + bx

N

r

Probability

SDa

sd6

SDyr

t«

Probability

D = -3.9641 + 0.2903L

29

0.7387

P < 0.001

3.1438

0.0488

2.1277

0.9659

0.40 > P > 0.30

DVDAE = -3.6437 + 0.2394L

30

0.9025

P < 0.001

1.4270

0.0222

0.9745

2.5534

0.02 > P > 0.01

EAST HARNHAM

y = a + bx

N

r

Probability

SDa

sd6

SD„r

t a

Probability

D = -2.7180 + 0.2697L

38

0.6762

P < 0.001

3.2645

0.0491

2.1736

0.8326

0.50 > P > 0.40

DVDAE = -3.7598 + 0.2203L

32

0.8609

P < 0.001

1.4334

0.0219

0.8244

2.6230

0.02 > P > 0.01

SHAWFORD, ’lower course’

y = a + bx

N

r

Probability

SDa

sd6

SDyr

t«

Probability

D = -5.8376 + 0.3 193L

15

0.7693

P < 0.001

4.9853

0.0741

2.0056

1.1710

0.30 > P > 0.25

DVDAE = -3.3254 + 0.2191L

13

0.9356

P < 0.001

1.6986

0.0255

0.6676

1.9577

0.10 > P > 0.05

STIFFKEY

y = a + bx

N

r

Probability

SDa

sd6

SDyx

ta

Probability

D = -4.0218 + 0.3088L

27

0.7671

P < 0.001

3.2297

0.0513

1.6867

1.2453

0.25 > P > 0.20

DVDAE = -2.4907 + 0.2042L

25

0.8110

P < 0.001

1.9224

0.0307

0.9492

1.2956

0.25 > P > 0.20

table 4. Estimates of the statistical parameters of four regression analyses of samples of Gonioteuthis quadrata
quadrat a from Wells, East Elarnham, Shawford, ‘lower course’, and Stiflfkey. L = length of the guard;
D = depth of the pseudoalveolus; DVDAE = dorso-ventral diameter at the anterior end.

of the pseudoalveolus, the Stiflfkey sample cannot be differentiated statistically at the 5 % level from the samples
from the papillosa and conica/ gracilis Zones. The position of the regression lines of the samples from Stiflfkey
and the upper senonensis Zone differ significantly (P < 0 001, with 77 degrees of freedom). The specimens from
the upper senonensis Zone have a deeper pseudoalveolus than the specimens from Stiflfkey. With respect to the
length of the guard vs the dorso-ventral diameter at the alveolar end, the Stiflfkey sample cannot be
differentiated statistically at the 5% level from the samples from the upper senonensis, papillosa , or
conica/ gracilis Zones.

In conclusion, the sample from Stiffkey cannot be differentiated statistically from G. q. quadrata from the
papillosa Zone or G. quadrata gracilis from the conica /gracilis Zone. However, the Stiflfkey belemnites do not
possess notches in the pseudoalveolus, as in G. quadrata gracilis, and are therefore referred to G. q. quadrata
and regarded to be from the papillosa Zone.

3. East Elarnham and Shawford. The two samples are virtually identical with respect to the shape of the guard
and the depth of the pseudoalveolus (see above), and therefore only the sample from East Harnham was
compared with other samples of Gonioteuthis.

On the basis of the maximum length of the guard it is suggested that the two samples are from the middle
Lower Campanian (cf. Ernst 1964a, fig. 8). As regards the length of the guard vs the depth of the
pseudoalveolus, the sample from East Harnham was compared to samples of G. q. quadrata from Hallembaye
and Bremer (lower part of the lingua/ quadrata Zone), Hover (upper lingua /quadrata Zone to pilula Zone), and
Hardivillers (pilula Zone sensu anglico), in addition to G. quadrata gracilis from the conica/ gracilis Zone of
Hover. The East Harnham sample does not differ significantly at the 5 % level from the samples of G. q.
quadrata from Hallembaye, Bremer, and Hardivillers, or G. quadrata gracilis from the conica /gracilis Zone of
Hover. In the case of the comparison of East Harnham and Hallembaye, the variances were found to differ
significantly (F = 2-0100 with 36 and 56 degrees of freedom; 001 > T> 0-005); the test for non-equal

CHRISTENSEN: CHALK BELEMNITES

725

To

L in mm

20 ' 30 ' 40 ' 50 ' 60 ' 70

text-fig. 11. Scatter plots and regression lines for Gonioteuthis quadrata quadrata from East Harnham. a,
length of the guard (L) vs dorso-ventral diameter at the alveolar end (DVDAE). b, length of the guard (L) vs

depth of the pseudoalveolus (D). + = mean value.

variances was therefore used. The difference of the positions of the regression lines of the samples from East
Harnham and G. q. quadrata from the upper lingua / quadrata Zone to pilula Zone is very highly significant
(P < 0001 with 101 degrees of freedom); specimens from the upper lingua/ quadrata Zone to pilula Zone
generally have a deeper pseudoalveolus than the specimens from East Harnham.

With respect to the length of the guard vs the dorso-ventral diameter at the alveolar end, the East Harnham

726

PALAEONTOLOGY, VOLUME 34

L in mm

K> ' 20 ' 30 ' 40 ' 50 ' 60 ' 70

D in
mm

L in mm

' 10 ' 20 * 30 ' 40 T 50 T 60 ' 70

text-fig. 12. Scatter plots and regression lines for Gonioteuthis quadrata quadrata from Stiffkey. a, length of
the guard vs (L) dorso-ventral diameter at the alveolar end (DVDAE). b, length of the guard (L) vs depth of

the pseudoalveolus (D). + = mean value.

sample was compared to samples of G. q. quadrata from Hallembaye, Hardivillers, and the lower senonensis
Zone of Hover, in addition to G. quadrata gracilis from the conica/ gracilis Zone of Hover. The East Harnham
sample does not differ significantly at the 5% level from the samples from Hardivillers and the conica /gracilis
Zone. On the other hand. East Harnham vs Hallembaye and East Harnham vs the lower senonensis Zone of

CHRISTENSEN: CHALK BELEMNITES

727

DVDAE

- ®

10

L m mm

20 ' 30 ' 40 ' 50 ' 60 ' 70 ' 80

D in
mm

20-

16-

12-

L in mm

i 1 1 1 1 1 1 i i i i i i r i i

10 20 30 40 50 60 70 80

text-fig. 13. Regression lines of four samples of Gonioteuthis quadrata quadrata. a, length of the guard (L)
vs dorso-ventral diameter at the alveolar end (DVDAE). b, length of the guard (L) vs depth of the
pseudoalveolus (D). + = mean values. 1, East Harnham; 2, Shawford; 3, Stiffkey; 4, Wells.

728

PALAEONTOLOGY, VOLUME 34

Hover comparisons showed that the differences in positions of the regression lines are very highly significant
(P < 0 001). The specimens from Hallembaye and the lower senonensis Zone are generally stouter than the East
Harnham specimens.

In conclusion, the sample from East Harnham cannot be differentiated with respect to slenderness of the
guard and depth of the pseudoalveolus from samples of G. q. quadrata from Hardivillers, and G. quadrata
gracilis from the conica/ gracilis Zone of Hover. The East Harnham specimens are generally more slender than
the specimens from Hallembaye and have a shallower pseudoalveolus than the specimens from the upper
lingua/ quadrata Zone to pilula Zone of Hover.

The samples from East Harnham, Shawford, and Hardivillers, however, are not to be referred to G. quadrata
gracilis, because the maximum length of the guard is larger than in G. quadrata gracilis, and they do not possess
notches in the pseudoalveolus as in G. quadrata gracilis. Moreover, the English top pilula Zone and basal
quadrata Zone are not to be correlated with the German conica/ gracilis Zone (see Text-fig. 1).

On the basis of the maximum length of the guard and the slenderness of the guard, in addition to the mean
Riedel Quotient (4-5 in the East Harnham sample, 4-4 in the Shawford sample, and 4-3 in the Hardivillers
sample) (see Table 5), the three samples are regarded as middle Lower Campanian. In samples of G. q. quadrata
from the middle Lower Campanian of Lagerdorf the mean value of the Riedel Quotient varies from 4-2-4- 5
(see Text-fig. 7); these samples have thus a more shallow pseudoalveolus than contemporaneous samples from
Misburg/Hover.

Gonioteuthis quadrata gracilis ? (Stolley, 1892)

1892 Actinocamax quadratus var. gracilis Stolley, p. 234, pi. 7, fig. 6.

1952 Gonioteuthis quadrata var. gracilis (Stolley); Naidin, p. 79, pi. 4, fig. 3; pi. 5, fig. 5; pi. 6, fig. 5.

1964a Gonioteuthis quadrata gracilis (Stolley); Ernst, p. 166, pi. 1, fig. 8; pi. 2, figs 7-10.

Type. Lectotype, here designated, the original of Stolley (1892, pi. 7, fig. 6).

Material. BMNH C43330 from Attlebridge, and BMNH C43351 from Ringland, Norfolk.

L

D

DVDAE

LDAE

MLD

RQ

SQ

BMNH C43330

56-6

9-9

8-7

8-0

8-0

5-7

6-5

BMNH C4335I

73-9

19 9

13-1

11-8

12-4

3-7

5-6

Remarks. G. quadrata gracilis is distinguished from the nominate subspecies by being smaller and
slenderer, with a shallower pseudoalveolus. Moreover, the edge of the pseudoalveolus often has
dorsal, lateral, and ventral notches.

The two specimens from Norfolk are tentatively referred to G. quadrata gracilis on the basis of
their slender guards. Moreover, Peake and Elancock (1961) placed Attlebridge at the top of the
Gonioteuthis Zone, and Wood (1988) placed Ringland very high in the same zone.

Distribution. G. quadrata gracilis occurs in the upper Lower Campanian conica/ gracilis and gracilis / mucronat a
Zones.

Genus belemnellocamax Naidin, 1964
[= Actinocamax {Par actinocamax) Naidin, 1964, p. 62]

Type species. Belemnites mammillatus Nilsson, 1826 by original designation.

Diagnosis. See Christensen (1975a).

Remarks. The evolutionary lineage of Belemnellocamax , in ascending order, B. ex gr. grossouvrei
(Janet, 1891), B. mammillatus (Nilsson, 1826), and B. balsvikensis (Brotzen, 1960), was studied by
Christensen (1975a, 1976, 1986). The general trend in evolution is the gradual calcification of the
anterior end of the guard. Moreover, the guard becomes more slender and less lanceolate through
time. Normally, the genus is not granulated, but a few granulated specimens of B. ex gr. grossouvrei
and B. mammillatus have been recorded (cf. Christensen 1986).

CHRISTENSEN: CHALK BELEMNITES

729

BREMER, lower part of lingua/quadrata Zone

Character

N

A'

SD

CV

OR

L

38

67.3

6.4

9.5

58.5-82.5

RQ

38

4.6

0.6

13.1

3.7- 6.3

HALLEMBAYE, lower part of lingua/quadrata Zone

Character

N

X

SD

CV

OR

L

60

67.3

7.2

10.7

52.0-80.3

RQ

58

4.5

0.5

11.1

3.5- 5.9

HOVER, upper part of lingua/quadrata Zone

Character

N

X

SD

CV

OR

L

24

60.3

10.5

17.4

41.0-78.5

RQ

24

4.0

0.5

12.2

2.8- 4.9

HOVER, upper lingua/quadrata Zone to pilula Zone

Character

N

X

SD

CV

OR

L

67

60.9

9.8

16.1

40.0-81.5

RQ

67

4.1

0.5

11.2

2.8- 5.4

HOVER, lower senonensis Zone

Character

N

X

SD

CV

OR

L

47

64.1

9.2

14.3

47.3-85.0

RQ

47

3.8

0.5

12.6

2.9- 4.9

HOVER, upper senonensis Zone

Character

N

X

SD

CV

OR

L

54

64.3

9.5

14.7

39.0-80.5

RQ

54

3.7

0.4

12.0

2.6- 4.6

HOVER, papillosa Zone

Character

N

X

SD

CV

OR

L

28

61.8

9.1

14.7

44.7-78.0

RQ

27

4.0

0.5

12.5

3.3- 5.0

HOVER, conica/gracilis Zone

Character

N

X

SD

CV

OR

L

65

57.8

6.9

12.0

41.0-73.0

RQ

65

4.5

0.6

14.1

3.4- 6.4

HARDIVILLERS, middle Lower Campanian

Character

N

X

SD

CV

OR

L

61

65.8

6.7

10.2

50.0-86.6

RQ

61

4.3

0.5

12.6

3.2- 5.5

table 5. Univariate analyses of nine samples of Gonioteuthis quadrata quadrata and G. quadrata gracilis.
L = length of the guard; RQ = Riedel Quotient; N = number of specimens; X = mean value; SD = standard
deviation; CV = coefficient of variation; and OR = observed range.

730

PALAEONTOLOGY, VOLUME 34

BREMER, lower pari of lingua/quadrata Zone

y = a + bx

N

r

Probability

SD„

SD„

SDyr

Probability

D = -3.1563 + 0.2687L

38

0.6924

P < 0.001

3.1534

0.0467

1.8027

1.0009

0.30 > P > 0.25

HALLEMBAYE, lower part of lingua/quadrata Zone

y = a + bx

N

r

Probability

SD„

SD„

SDyj-

t<I

Probability

D = -4.6948 + 0.2970L

58

0.8192

P < 0.001

1.8903

0.0280

1.5332

2.4836

0.02 > P > 0.01

DVDAE = -2.4293 + 0.2136L

58

0.8326

P < 0.001

1.2636

0.0187

1.0293

1.9225

0.10 > P > 0.05

HOVER, upper part of lingua/quadrata Zone

y = a + bx

N

r

Probability

SD„

SD„

SDyJr

td

Probability

D = -0.8883 + 0.2685L

24

0.7982

P < 0.001

2.6404

0.0432

2.1657

0.3364

0.80 > P > 0.70

DVDAE = -1.8504 + 0.2018L

24

0.9193

P < 0.001

1.1258

0.0184

0.9234

1.6437

0.20 > P > 0.10

HOVER, upper lingua/quadrata Zone to pilula Zone

y = a + bx

N

r

Probability

SD„

SD„

SDyj-

ia

Probability

D = -1.7270 + 0.2770L

67

0.8393

P < 0.001

1.3722

0.0223

1.7695

1.2586

0.25 > P > 0.20

DVDAE = -2.6490 + 0.2136L

67

0.9517

P < 0.001

0.5273

0.0086

0.6800

5.0236

P < 0.001

HOVER, lower senonensi s Zone

y = a + bx

N

r

Probability

SD„

SD„

SDyjp

Probability

D = -4.1994 + 0.3336L

47

0.8033

P < 0.001

2.4022

0.0371

13022

1.7482

0.10 > P > 0.05

DVDAE = -1.7965 + 0.201 1L

46

0.9031

P < 0.001

0.9273

0.0143

0.8886

1.9374

0.10 > P > 0.05

HOVER, upper senonensis Zone

y = a + bx

N

r

Probability

SDU

SD,

SDyj

U

Probability

D = -3.3223 + 0.3286L

54

0.8248

P < 0.001

2.0134

0.0310

2.1343

1 .6501

0.20 > P > 0.10

DVDAE = -3.4106 + 0.2224L

53

0.8960

P < 0.001

0.9749

0.0149

1.0066

3.4984

P < 0.001

HOVER, papillosa Zone

y = a + bx

N

r

Probability

SD„

SD„

SDyl

t a

Probability

D = -4.9266 + 0.3353L

27

0.8582

P < 0.001

2.5120

0.0403

1.9013

1.9612

0.10 > P > 0.05

DVDAE = -3.1714 + 0.2127L

26

0.9014

P < 0.001

1.2719

0.0205

0.8210

2.4933

0.02 > P > 0.01

HOVER, conica/gracilis Zone

y = a + bx

N

r

Probability

SD,

SDt

SDyr

Probability

D = -3.7756 + 0.2917L

65

0.7649

P < 0.001

1.8000

0.0309

1.7110

2.0977

0.05 > P > 0.025

DVDAE = -2.5297 + 0.2028L

65

0.9013

P < 0.001

0.7138

0.0123

0.6785

3.5442

P < 0.001

HARDI VILLERS, middle Lower Campanian

y = a + bx

N

r

Probability

SD,

SD„

SDyr

U

Probability

D = -3.7036 + 0.2944L

61

0.7271

P < 0.001

2.2900

0.0340

1.7676

1.6173

0.20 > P > 0.10

DVDAE = -0.6492 + 0.1778L

59

0.8146

P < 0.001

1.0922

0.0165

0.8569

0.5944

0.60 > P > 0.50

table 6. Estimates of the statistical parameters of nine regression analyses of samples of Gonioteuthis quadrata
quadrat a and G. quadrata gracilis. L = length of the guard; D = depth of the pseudoalveolus; DVDAE =
dorso-ventral diameter of the anterior end.

Naidin (1964) placed the grossouvrei group in Actinocamax ( Paractinocamax ), type species
Actinocamax grossouvrei Janet, while mammillatus was placed in Belemnellocamax. The
classification of Naidin was fully discussed by Christensen (1986) and is not followed here. The
subgenus Paractinocamax was considered a junior synonym of Belemnellocamax.

Distribution. B. ex gr. grossouvrei is widely distributed but rare in the North European Province. B.
mammillatus is extremely common in Scania, but rare outside this area; it has been recorded Irom northern
Germany, Poland, and the eastern part of the Russian Platform. B. balsvikensis also occurs commonly in
Scania, but outside this area it is unknown except for a find of two specimens from NW Germany (Christensen

CHRISTENSEN: CHALK BELEMNITES

731

and Schulz 1976). It can thus be concluded that the area of distribution of the genus Belemnellocamax gradually
diminished through its stratigraphic range.

Belemnellocamax is recorded from the lower Santoman (possibly highest Coniacian) to the lower Upper
Campanian (Christensen 1986).

Belemnellocamax ex gr. grossouvrei (Janet, 1891)

Plate 3, figs 1-21

Material. BMNH C42818, West Harnham, top pilula Zone-basal quadrata Zone, (holotype of Actinocamax
blackmorei Crick, 1907); BMNH C44331 and C44254, East Harnham, top pilula- basal quadrata Zone; BGS
GSM 101425, Mottisfont, top pilula- basal quadrata Zone; BMNH C44382, Micheldever, coranguinum Zone.

In addition to these specimens from southern England, three specimens from northern England and France
are also included in the present study : ( 1 ) BMNH C46392, Ruston Parva, Yorkshire, pilula Zone (see Brighton
1930), (2) BMNH C10896, from flinty Chalk at Fimber, Yorkshire, coranguinum Zone (see Crick 1906); (3)
BMNH C32498, Breteuil, France, probably top pilula Zone or base of quadrata Zone. According to I. Jarvis
(in litt. 1986) this specimen most likely came from the large complex of abandoned phosphatic chalk quarries
referred to as Hardivillers by Jarvis (1980).

Remarks. Specimens of the B. grossouvrei group are characterized by their large, ventrally flattened
guards, which are lanceolate to strongly lanceolate in ventral view. Moreover, they generally have
a shallow pseudoalveolus, the cross-section of which is triangular, and juvenile guards are long and
elongated (Christensen 1986). About 100 specimens have been recorded from England, France,
West Germany, Scania in southern Sweden, and the Russian Platform, and these specimens have
been assigned to eleven species and subspecies (see Christensen 1986). The group has been the
subject of excessive subdivision by earlier authors, and hopefully, the revision of this group by F.
Schmid (Kiinzelsau) and M.-G. Schulz (Kiel) will help in solving these taxonomic problems. It
should be stressed that only adult specimens have been figured previously; this may be due to the
fact that juvenile and adolescent specimens are extremely rare.

Those English and French specimens of the B. grossouvrei group that have not been described or figured
earlier, and the holotype of A. blackmorei are briefly commented upon below.

1. BMNH C42818. The holotype of A. blackmorei (PI. 3, figs 1-4). The most anterior end of the guard is
missing. It seems that the anterior end was ground, because it is completely flat and there is no trace of a
shallow central pit for housing the phragmocone, or of a ventral furrow.

2. BGS GSM 101425 (PI. 3, figs 5 and 6). An adult specimen which is strongly lanceolate in ventral view.
The anterior part of the guard is badly preserved due to weathering, and the pseudoalveolus is not present.

3. BMNH C44331 (PI. 3, figs 18-21). An adolescent specimen, which is only very slightly lanceolate in
ventral view. The most anterior part of the guard is not preserved due to weathering, and only the most
posterior part of the pseudoalveolus is present.

4. BMNH C44254 (PI. 3, figs 10-13). A juvenile guard, which is strongly lanceolate in ventral and lateral
views. The anterior end of the guard is flat with a pit in its centre and shows the concentric growth layers of
the guard and radiating ribs. Moreover, the cross-section of the anterior end is subtriangular and a deep, short,
ventral furrow is present. This specimen differs from other specimens of the B. grossouvrei group by having a
flat anterior end. In the structure of the anterior end it resembles Turonian and early Coniacian species of
Actinocamax (cf. Christensen 1982). The specimen is tentatively placed in the B. grossouvrei group due to its
elongated guard which is strongly lanceolate in ventral view.

5. BMNH C44382 (PI. 3, figs 14-17). A juvenile guard which is lanceolate in ventral view. The guard has
anteriorly a shallow pseudoalveolus the cross-section of which is subtriangular. This specimen is somewhat
similar to specimen no. A4331-1 (Ernst 1964r/, PI. 3, fig. 7) from the Lower Santonian of Kellermanshof near
Essen, GFR. It was figured as Gonioteuthis lundgreni / westfalica .

6. BMNH C32498 (PI. 3, figs 7-9). An adult specimen which is strongly lanceolate in ventral and lateral
views. The guard has a shallow pseudoalveolus, the cross-section of which is subtriangular. A short ventral
fissure is present.

732

PALAEONTOLOGY, VOLUME 34

Distribution. The stratigraphical range of the B. grossouvrei group was surveyed by Christensen (1986). B. ex
gr. grossouvrei occurs in NW Germany and Scania in the basal Lower Campanian G. granulataquadrata Zone
and the uppermost Lower Campanian G. quadrata gracilis/ B. mucronata Zone or equivalent zones. In France
and on the Russian Platform the group occurs in the Upper Santonian and Lower Campanian. The specimen
from France commented upon above is probably from the top pilula Zone and/or base of the quadrata Zone.
The two specimens from Fimber and Micheldever are from the Lower Santonian (possibly the latest
Coniacian), whereas the remaining specimens are from the middle Lower Campanian (top pilula Zone and/or
base of quadrata Zone). To sum up, the grossouvrei group occurs from the Lower Santonian (possibly the
highest Coniacian) to the boundary between the Lower and Upper Campanian.

Genus belemnitella d’Orbigny, 1840

Type species. Belemnites mucronatus Schlotheim, 1813; ICZN Opinion 1328 (1985); name no. 2979.
Diagnosis. See Christensen (1975u).

Remarks. The International Commission on Zoological Nomenclature has designated under the
plenary powers, specimen no. kca 5/2 in the collections of the Niedersachsisches Landesamt fur
Bodenforschung, Hannover, Germany, as neotype for Belemnites mucronatus Schlotheim (see
ICZN 1985). The neotype was described and figured by Christensen et al. (1975, pi. 1, fig. 1).

Dozens of species and subspecies of Belemnitella from the Upper Campanian and Lower
Maastrichtian have been established, and the majority of these taxa were erected on the basis of
limited material by eastern European workers. The systematics of many of these taxa is in a state
of disorder and they are poorly understood. Some of these taxa were discussed by Christensen
(1986) and Christensen in Robaszynski and Christensen (1989).

Distribution. Belemnitella is widely distributed in the North European Province and has also been recorded
from the northern part of the Tethyan Realm and the North American Province. The genus is known from
the Lower Santonian to the uppermost Maastrichtian.

Belemnitella propinqtta group

Remarks. Christensen (1986) included ‘ Actinocamax' lundgreni Stolley from the Coniacian to
Middle Santonian and Belemnitella propinqtta (Moberg) from the Lower and Middle Santonian in

EXPLANATION OF PLATE 3

Figs 1-21. Belemnellocamax ex gr. grossouvrei (Janet) from England (1-6, 10-21) and France (7-9). I 4. cast
of holotype of Actinocamax blackmorei Crick, BMNH C10895 (plaster cast) and BMNH C42818, W.
Harnham, top pilula Zone-basal quadrata Zone; 1, dorsal view; 2, lateral view; 3, ventral view; 4, view of
the anterior end, x L5. 5-6, BGS GSM 101425, specimen from Mottisfont, locality no. 1067 of Brydone
(1914), top pilula Zone-basal quadrata Zone; 5, ventral view; 6, lateral view. 7-9, BMNH C32498, specimen
from Breteuil, France, probably top pilula Zone-base quadrata Zone; the specimen probably came from the
large complex of abandoned phosphatic chalk quarries, referred to as Hardivillers by Jarvis (1980) ; 7, dorsal
view; 8, ventral view; 9, view of the anterior end, x F5. 10-13, BMNH C44254, a juvenile specimen from
East Harnham, top pilula Zone-basal quadrata Zone; note the flat anterior end with a pit in its centre, and
the short, deep ventral furrow; 10, dorsal view; 1 1, lateral view; 12, ventral view; 13, view of the anterior
end, x 3. 14—17, BMNH C44382, a juvenile specimen from Micheldever, coranguinum Zone; 14, dorsal view;
15, lateral view; 16, ventral view; 17, view of the anterior end, x 2. 18-21, BMNH C44331, a specimen from
East Harnham, top pilula Zone-basal quadrata Zone; 18, dorsal view; 19, lateral view; 20, ventral view; 21,
view of the anterior end, x 2.

All specimens are coated with ammonium chloride. All figures are natural size unless otherwise stated.

PLATE 3

CHRISTENSEN, Belemnellocamax

734

PALAEONTOLOGY, VOLUME 34

the Belemnitella propinqua group. B. propinqua is generally considered to be the earliest
representative of the genus Belemnitella. It is a well-defined species which was redescribed by
Christensen (1971, 1973). It evolved from "A.' lundgreni.

'A.' lundgreni was placed in Gonioteuthis (Goniocamax) by Naidin (1964) together with several
other species, including the earliest members of the Gonioteuthis lineage. This suggestion was
criticized by Ernst and Schulz (1974) and Christensen (1982). Ernst and Schulz suggested that the
subgenus Goniocamax , type species 1 A.' lundgreni , should be elevated to a genus or considered a
subgenus of Belemnitella. They also suggested that only ‘A.’ lundgreni and its ancestors should be
assigned to Goniocamax. This suggestion must await further studies and is outside the scope of the
present paper. ‘A.' lundgreni , however, is here placed in the B. propinqua group because it is closely
allied to B. propinqua.

‘ A ctinocamax ’ lundgreni Stolley, 1897
Plate 4, figs 1-6

1897 Actinocamax lundgreni Stolley, p. 285, pi. 3, figs 16-20 (non pi. 3, fig. 15).

1897 Actinocamax mammillatus mut. (ant.) bornholmensis Stolley, p. 288, pi. 4, fig. 1.

1897 Actinocamax propinquus Moberg mut. (var.) nov. Stolley, p. 295, pi. 3. fig. 23.

1912 Actinocamax propinquus Moberg; Arkhangelsky, p. 585, pi. 10, figs 14 and 15, ? non 23-27,

34—36.

1912 Actinocamax intermedius Arkhangelsky, p. 582, pi. 9, figs 30 and 31 ; pi. 10, fig. 6, 16-18, ? non
27.

1915 Actinocamax plenus Miller var. excavata Sinzow, p. 144, pi. 8, figs 14-17, ? non 18.

1918 Actinocamax bornholmensis Stolley; Ravn, p. 33, pi. 2, fig. 7.

1918 Actinocamax sp. (cfr. Act. strehlenensis Fritsch); Ravn, p. 34, pi. 2, fig. 8.

1946 Actinocamax lundgreni Stolley; Ravn, p. 30.

1949a Belemnitella propinqua (Moberg); Jeletzky, p. 416, text-figs 1 and 2 (non text-figs 3 and 4).
1957 Actinocamax lundgreni lundgreni Stolley; Birkelund, p. 13, pi. 1, figs 5 and 6.

1957 Actinocamax lundgreni excavata (Sinzow); Birkelund, p. 18, pi. 1, figs 7 and 8.

1957 Actinocamax aff. westfalicus (Schliiter); Birkelund. pp. 27-28, pi. 2, fig. 3.

1958 Actinocamax intermedius Arkhangelsky; Nikitin, p. 5, pi. 1, figs 4-8.

1958 Actinocamax propinquus Moberg; Nikitin, p. 12, pi. 1, figs 9-15; pi. 3, fig. 7.

1964a Gonioteuthis lundgreni /aff. westfalica sensu Birkelund; Ernst, p. 161, pi. 3, figs 5 and 6.

1964 Gonioteuthis (Goniocamax) lundgreni lundgreni (Stolley); Naidin, p. 127, pi. 7, figs 5-7.

1964 Gonioteuthis ( Goniocamax ) lundgreni excavata (Sinzow); Naidin, p. 133, pi. 7, fig. 8.

1972 Actinocamax (Actinocamax) propinquus propinquus Moberg; Glazunova, p. 106, pi. 45, figs 1-5.

1972 Actinocamax (Actinocamax) aff. propinquus propinquus Moberg; Glazunova, p. 107, pi. 46,

fig. 1.

1972 Gonioteuthis (Goniocamax) cf. lundgreni lundgreni (Stolley); Glazunova, p. 113, pi. 46, figs 3
and 4.

1973 Gonioteuthis lundgreni (Stolley); Christensen, p. 131. pi. 10, figs 6-9.

1974 Gonioteuthis (Goniocamax) lundgreni lundgreni (Stolley); Naidin, p. 211, pi. 73, fig. 8.

1974 Gonioteuthis (Goniocamax) lundgreni excavata (Sinzow); Naidin, p. 211, pi. 73, fig. 9.

1975a Actinocamax lundgreni Stolley; Christensen, p. 28.

1982 Actinocamax lundgreni Stolley; Christensen, p. 76.

1986 ‘ Actinocamax ’ lundgreni Stolley; Christensen, p. 30.

Lectotvpe. The specimen figured by Stolley (1897, pi. 3, fig. 18) was designated as lectotype by Birkelund (1957,
p. 4). '

Material. BMNH C44380, C44383, Micheldever, Hants, coranguinum Zone; BMNH C43506, Gravesend,
Kent, coranguinum Zone.

CHRISTENSEN: CHALK BELEMNITES

735

Dimensions.

L

D

DVDAE

LDAE

MLD

RQ

SQ

BMNH C44380

49*

4-7

5-6

4-7

7-2

10 4

8-8

BMNH C44383

53*

4-2

6-4

5-6

6-8

12 6

8-3

BMNH C43506

54*

5-2

7-8

—

8-9

10 4

6-9

(* estimated)

Short description. The length of the guard is up to 70 mm. The guard is stout and lanceolate in ventral view
and slightly lanceolate, subcylindrical or high conical in lateral view. It is markedly flattened ventrally and the
apex is acute. The Riedel Quotient varies from about 6 to about 12, and the cross-section of the anterior end
is subtriangular. The walls of the pseudoalveolus are straight or convex and often have conellae. The surface
of well-preserved guards has longitudinal striae and rather prominent vascular markings.

Remarks. Stolley (1897) distinguished three taxa from the Coniacian of Bornholm: Actinocamax
lundgreni from the ‘Glass marl’ at Muleby, A. mammillatus mut. hornholmensis from the Arnager
Limestone west of Arnager, and A. propinquus mut. (var.) nov. from the marl at Stampe A
(brooklet). Ravn (1946) placed A. mammillatus mut. hornholmensis in the synonymy of A. lundgreni ,
and this view was followed by later authors, including Birkelund (1957), Jeletzky (1958), Naidin
(1964), and Christensen (1973). A propinquus mut. (var.) nov. is also considered a synonym of A.
lundgreni (Christensen unpublished).

Affinity. The first member of the Gonioteuthis lineage, G. westfalica praewestfalica from the Middle
and Upper Coniacian, and the first member of the Belemnitella lineage , 'AG lundgreni from the
Lower Coniacian to Lower Santonian, are both characterized by having ventrally flattened guards
which are lanceolate in ventral view. Moreover, both taxa have vascular markings and longitudinal
striae. ‘AG lundgreni differs, however, from G. westfalica praewestfalica by having a deeper
pseudoalveolus and in being larger. In addition, ‘ A G lundgreni often has conellae on the walls of the
pseudoalveolus.

Distribution. ' A .' lundgreni occurs commonly on the Russian Platform, on Bornholm, Denmark, and in
southern Sweden. Outside this area it is rare and has been recorded from NW Germany (Ernst 1964u;
Christensen 1973, p. 133) and southern England (herein).

Christensen (1973) reviewed the biostratigraphical age of the species and concluded that it had its first
occurrence in the late Coniacian and continued into the early Santonian. This was based on the assumption
that the Arnager Limestone and ‘Glass marl', stratum typica of 'AG lundgreni and A. mammillatus mut.
hornholmensis, are of late Coniacian age. Recent studies of the inoceramid bivalve faunas of the Arnager
Limestone Formation by K.-A. Troger (Freiberg) have shown, however, that the Arnager Limestone west of
Arnager is Lower Coniacian, the ‘Glass marl' at Muleby is upper Lower Coniacian, and the marl at Stampe
A is lower Middle Coniacian in inoceramid terms. On ammonite evidence, however, the Arnager Limestone
west of Arnager is at least Middle Coniacian (Kennedy and Christensen 1991). 'AG lundgreni also occurs
together with G. westfalica and Belemnitella propinqua in parts of the Bavnodde Greensand Formation which
is Lower Santonian (Christensen unpublished). It can thus be concluded that the stratigraphic range of 'AG
lundgreni is Lower Coniacian to Lower Santonian.

‘ Actinocamax ’ ex gr. lundgreni Stolley, 1897
Plate 4, figs 7-12

1907 Actinocamax sp. Crick, p. 392, text-fig. 2.

Material. BMNH C10576, Fletcher and Co’s Pit, Gravesend, Kent, coranguinum Zone; BMNH C59278,
Grays, Essex, coranguinum Zone.

Short description. BMNH Cl 0576 is a 70-6 mm long fragment, consisting of approximately the anterior two-
thirds of the guard. It is estimated that the total length of the guard may have been 105-1 10 mm. It is lanceolate

736

PALAEONTOLOGY, VOLUME 34

in ventral view and subcylindrical in lateral view. The guard has anteriorly a shallow pseudoalveolus 15 mm
deep. The cross-section of the anterior end is subtriangular and the walls of the pseudoalveolus are covered by
large, closely spaced conellae. Ventrally the guard is flattened and the surface of the guard is smooth.

BMNH C59278 is 54 mm long, but the most anterior part of the guard is not preserved. It is lanceolate in
ventral view and slightly lanceolate in lateral view. The guard has anteriorly a shallow pseudoalveolus with
walls covered by conellae. The depth of the preserved part of the pseudoalveolus is 6-8 mm and the cross-
section of the alveolar end seems to have been subtriangular. Ventrally the guard is markedly flattened and the
surface is smooth.

Remarks. BMNH C10576 was fully described by Crick (1907) who assigned it to Actinocamax sp.,
after having discussed its affinity to the grossouvrei group. Later on, it was assigned to the
Mammillata group by Jeletzky (1949 b), the grossouvrei group by Christensen (1975a), and possibly
the 'A.' lundgreni group by Christensen (1986). Cl 0576 shares many characters with the grossouvrei
group, including the size, shape and ventral flattening of the guard, the cross-section of the alveolar
end, and the surface sculpture. It has, however, a deeper pseudoalveolus than the grossouvrei group,
and it is estimated that the Riedel Quotient may have been about 7. In addition, it has many closely
spaced conellae. Cl 0576 resembles 'Ad lundgreni in the shape and ventral flattening of the guard,
and the depth of the pseudoalveolus. It is, however, larger than "Ad lundgreni. On the basis of the
depth of the pseudoalveolus, it is tentatively referred to as ‘Ad ex gr. lundgreni.

BMNH C59278 also has characters in common with the grossouvrei group and 'Ad lundgreni. On
the basis of the depth of the pseudoalveolus (Riedel Quotient estimated to be about 5) it is
tentatively referred to as 'Ad ex gr. lundgreni.

Distribution. The specimens from Grays and Gravesend are considered to be from the Lower Santonian (see
above).

Belemnitella propinqua propinqua (Moberg, 1885)

Plate 4, figs 13-18; Plate 5, figs 11-14

Synonymy. See Christensen (1971).

Holotype. By monotypy the specimen figured by Moberg (1885, pi. 5, fig. 25). It was refigured by Christensen
( 1971, pi. 1, fig. 1).

Material. BMNH C5502, Grays, Essex, coranguinum Zone; BMNH C43508 and C43519, Gravesend, Kent,
coranguinum Zone.

EXPLANATION OF PLATE 4

Figs 1-6, 'Actinocamax' lundgreni Stolley. 1-3, BMNH C44380, Micheldever, coranguinum Zone; 1, dorsal
view; 2, ventral view; 3, view of the anterior end, x 2. 4-6, BMNH C44383, Micheldever, coranguinum Zone;
4, dorsal view; 5, ventral view; 6, view of the anterior end, x 2.

Figs 7-12. ‘ Actinocamax ' ex gr. lundgreni Stolley. 7-8, BMNH C59278, Grays, coranguinum Zone; 7, dorsal
view; 8, lateral view. 9-12, BMNH C10576, Fletcher and Co’s Pit, Gravesend, coranguinum Zone; specimen
figured as Actinocamax sp. by Crick (1907, text-fig. 2); 9, dorsal view; 10, lateral view; 1 1, ventral view; 12,
view of the anterior end showing conellae, x 2.

Figs 13-18. Belemnitella propinqua (Moberg). 13-15, BMNH C43519, Gravesend, coranguinum Zone; 13,
dorsal view; 14, ventral view; 15, view of the anterior end, x 2. 16-18, BMNH C5502, Grays, coranguinum
Zone; 16, dorsal view; 17, lateral view; 18, view of the anterior end, x 2.

Figs 19-22. Belemnitella cf. praecursor Stolley, BMNH C43288, Porchester Pit, Hampshire; specimen labelled
‘ mucronata ’ Zone, but according to Griffith and Brydone (1911) this pit is placed in the Subzone of
G. quadrata; 19, dorsal view; 20, lateral view; 21, ventral view; 22, view of the split anterior end showing
conellae, c. x 2-5.

All specimens are coated with ammonian chloride, except figs 9-12, and are natural size unless otherwise stated.

PLATE 4

CHRISTENSEN, ‘ Actinocamax', Belemnitella

738

PALAEONTOLOGY, VOLUME 34

Short description. A Belemnitella with a rather sturdy guard, lanceolate in ventral view and slightly lanceolate
or subcylindrical in lateral view. The cross-section of the pseudoalveolus at the alveolar end is subtriangular
to pointed oval. The walls of the pseudoalveolus are covered by closely spaced concellae. The Riedel Quotient
varies from a little less than three to about four.

The most anterior part of the guard in C43508 and C43519 is not preserved. It is estimated that the Riedel
Quotient is about 4 in C43508 and about 3 in C43519.

Dimensions.

L D DVDAE LDAE MLD LVF RQ SQ

BMNH C5502 605 19 6 10 0 8-9 102 6 0 31 6 1

Remarks. Christensen (1971, 1973) described the species in detail on the basis of Swedish and
Danish material, including the holotype. Christensen (1971) listed synonyms and showed that the
species had been misinterpreted by Russian palaeontologists. The following taxa were considered
to be synonyms of B. p. propinqua : Actinocamax propinquusl Moberg, 1885, Belemnitella mucronata
mut. anterior Stolley, 1897, B. ex gr. mirabilis Jeletzky, 1948, A. propinquus ravni Birkelund, 1957,
Gonioteuthis jeletzkyi Kongiel, 1962, and B. rylskiana Nikitin, 1958.

Naidin (1974) employed a different concept of B. p. propinqua from that of Christensen (1971,
1973, 1986, and herein). He recognized B. p. propinqua from the Lower Santonian and B. propinqua
rylskiana from the upper Lower to Upper Santonian, in addition to the dubious B. propinqua
mirabilis Arkhangelsky, 1912 from the Santonian of northern Kazakhstan.

Distribution. B. p. propinqua occurs on the Russian Platform, on Bornholm, Denmark, and in southern
Sweden. Outside this area it has only been recorded from southern England (herein). It occurs in the Lower
and Middle Santonian.

Belemnitella praecursor group

Remarks. Christensen (1986) tentatively placed B. alpha Naidin and B. praecursor Stolley from the
lower and middle Lower Campanian in this group. The two taxa were fully discussed, including
their mutual relationship, and their relationship to B. mucronata (Schlotheim) from the uppermost
Lower Campanian and Upper Campanian (see also Christensen 1975«; Christensen and Schmid
1987).

B. praecursor differs from B. alpha in several characters (see below), but it should be stressed that
it is not possible on the basis of only a few specimens to assign them safely to either B. praecursor
or B. alpha because the range of variation overlaps.

EXPLANATION OF PLATE 5

Figs 1-10. Belemnitella praecursor Stolley. 1-2, BMNH C43960, East Harnham, top pilula Zone-basal quadrata
Zone; 1, lateral view; 2, ventral view. 3—4, BMNH C43964, East Harnham, top pilula Zone- basal quadrata
Zone; a smooth and slender specimen; 3, lateral view; 4, ventral view. 5-7, BMNH C44149, East Harnham,
top pilula Zone-basal quadrata Zone; specimen figured as B. lanceolata by Blackmore (1896. pi. 1, fig- 1);
5, dorsal view; 6, lateral view; 7, ventral view. 8-10, BMNH C43954, East Harnham, top pilula Zone-basal
quadrata Zone; 8, dorsal view; 9, lateral view; 10, ventral view.

Figs 1 1-14. Belemnitella propinqua (Moberg), BMNH C43508, Gravesend, coranguinum Zone; 1 1, dorsal view;

12, lateral view; 13, ventral view; 14, view of the anterior end, x 2.

All specimens are coated with ammonium chloride, and are natural size unless otherwise stated.

PLATE 5

CHRISTENSEN, Belemnitella

740

PALAEONTOLOGY, VOLUME 34

text-fig. 14. Histogram of the length from apex to the
protoconch (LAP) of Belemnitella praecursor from East
Harnham. The figures above the bars are the actual number
of specimens.

Belemnitella praecursor Stolley, 1897
Plate 5, figs 1-10; Plate 6, figs 1-4, 8-10
Synonymy. See Christensen and Schmid (1987).

Holotype. By monotypy the original of Stolley (1897, pi. 3, fig. 24). A cast of the holotype was figured by
Christensen (1986, pi. 3, fig. 4) and Christensen and Schmid (1987, pi. 3, figs 4 and 5).

Material. 82 specimens (BMNH C5776, C43953-67, C43972-80, C43986-44007, C44009-31, C44134,
C44160-2, C44245, and C59152-8) from East Harnham, top pilula Zone-basal quadrata Zone.

Description. The guard is long, well-preserved guards ranging up to 150 mm. In ventral view the guard is
generally lanceolate with a constriction at the base of the ventral fissure, and in lateral view it is high conical.
The guard is flattened ventrally over its entire length. The apical angle is acute in both juvenile and adult
specimens, and the mucro is only slightly delimited.

The guard is slender; the ratio of the length from apex to protoconch and the dorso-ventral diameter at the
protoconch varies from about 3 to about 6, being 3-4-4-6 in most specimens, and the mean value is about 4.

The depth of the alveolus is about half the length of the guard in well-preserved specimens. The shape of the
bottom of the ventral fissure is generally straight or almost straight, but it may also be straight with an outward
bend, S-shaped, curve, or undulating. The walls of the alveolus may be covered by conellae. The fissure angle
and Schatzky distance are small, and the alveolar angle varies from 18-22° (see below).

EXPLANATION OF PLATE 6

Figs 1-4, 8-10. Belemnitella praecursor Stolley. 1-4, BMNH C43962, East Harnham, top pilula Zone-basal
quadrata Zone; 1, dorsal view; 2, lateral view; 3, ventral view; 4, view of the anterior end showing internal
characters. 8-10, BMNH C59154, East Harnham, top pilula Zone-basal quadrata Zone; 8, lateral view; 9,
ventral view; 10, view of the split anterior end showing internal characters.

Figs 5-7, 1 1-13. Belemnitella cf. praecursor Stolley. 5-7, SM B97228, Stiffkey, quadrata Zone; 5, dorsal view;
6, lateral view; 7, ventral view. 11-13, BGS GSM 101391, Shawford, quadrata Zone; 11, lateral view; 12,
ventral view; 13, view of the split anterior end showing internal characters.

All specimens are coated with ammonium chloride, and are natural size.

PLATE 6

CHRISTENSEN, Belemnitella

742

PALAEONTOLOGY, VOLUME 34

Juvenile specimens are, by and large, smooth. Adolescent and adult specimens have weakly developed
vascular markings, in addition to dorso-lateral depressions, dorso-lateral double furrows and longitudinal
striae. The vascular markings are most prominent around the ventral fissure, and the longitudinal striae are
typically more distinct than the vascular markings.

Biometry. A sample of B. praecursor from East Harnham was analysed by univariate and bivariate methods.
Only very few specimens were split in the median plane, and thus the internal characters are known in fewer
than ten specimens. The length from the apex to the protoconch was measured in unsplit specimens.

B. praecursor from East Harnham :

Character

N

X

SD

cv

OR

LAP

51

55-6

8-6

15-5

27-8-69-9

DVDP

51

14-2

2-7

19 0

8-1-18-6

SD

8

8-2

1-9

22-8

6-4-11-3

FA

9

15-2

5-5

36-4

7-0-23-0

AA

8

20-2

0-6

2-9

19-5-21-5

LAP/DVDP

51

4-0

0-4

10-2

3-1-4-8

A histogram of the length from apex to protoconch is shown in Text-figure 14. It was tested for normality using
the Kolmogorov-Smirnov-test for goodness of fit, and the test showed that the size-frequency distribution does
not differ significantly from normality at the 5% level (D = 01 361 ; 0-5 > P > 0-2 with 51 degrees of freedom).

Bivariate analysis. The scatter plot of the length from the apex to the protoconch vs the dorso-ventral diameter
at the protoconch is shown in Text-figure 15, as is the regression line. The value of the correlation coefficient
is very highly significant (P < 0 001 with 49 degrees of freedom). A /-test on the ^-intercept gave a value of
01 289 with 49 degrees of freedom which is not significant (0 9 > P > 0-8), implying an isometric relationship
of the variates.

On the basis of the univariate and bivariate analyses the sample from East Harnham can be regarded as
homogeneous.

text-fig. 15. Scatter plot and regression line for Belemnitella praecursor from East Harnham. LAP = length
from apex to the protoconch; DVDP = dorso-ventral diameter at the protoconch; + = mean value.
DVDP = -0 1 777 + 0-2567 LAP; N = 51 ; r = 0-8347; SDa = 1-3782; SD, = 00239; SDyj. = 1-4542.

Remarks. B. praecursor was established on the basis of only one specimen and characterized as being
smooth (Stolley 1897). Later Jeletzky (1955nr) discussed the concept of B. praecursor and established
two varieties, var. media and var. mucronatiformis , in addition to var. praecursor. He failed,
unfortunately, to present appropriate biometrical data and/or a differential diagnosis for the three
varieties.

CHRISTENSEN: CHALK BELEMNITES

743

Christensen and Schmid (1987) analysed the variation of the critical characters in a large sample
of B. praecursor from the Vaals Formation (lower Lower Campanian, lower part of the I. lingua/G.
quadrata Zone sensu germanico) of the C.P.L. Quarry at Hallembaye, Belgium. The results of the
univariate and bivariate biometric analyses of the Hallembaye sample are reported below for
comparison.

B. praecursor from Hallembaye (Christensen and Schmid 1987): Univariate analysis.

Character

N

X

SD

cv

OR

LAP

46

60-9

11-8

19-4

20- 1-79-2

DVDP

60

15-3

3-6

23-3

4-9-19-0

SD

57

6-4

1-3

20-9

4-2-10-2

FA

52

17-6

5-0

28-2

10-0-28-0

AA

59

20-1

0-9

4-3

18-0-22-0

LAP/DVDP

46

4 1

0-5

12-7

31-5-9

Bivariate analysis.

DVDP = -2-0831+0-2854 LAP; N = 46; r = 0-9208; SD„ = 0-6684; SD6 = 00108; SD,JX = 0-8534.

The relationship of the length from the apex to the protoconch vs the dorso-ventral diameter at the protoconch
of this sample is very strongly allometric, and the juvenile and adolescent specimens are more slender than the
adult specimens (Christensen and Schmid 1987).

The sample from East Harnham is closely similar to the sample from Hallenbaye with respect to slenderness
and shape of the guard, surface sculpture, Schatzky distance and fissure angle. The East Harnham sample
differs, however, from the Hallambaye sample in two ways: (1 ) the guard is generally slightly smaller, and (2)
the relationship of the length from the apex to the protoconch vs- the dorso-ventral diameter at the protoconch
is isometric. Christensen (1986) and Christensen and Schmid (1987) considered the allometry to be diagnostic
for the species. This assumption is now known to be incorrect.

The three varieties of B. praecursor , which ‘...are only morphologically varieties of the same specific type
(parts of the same populations . . . ’ ( Jeletzky 1 955u, p. 482), must be treated as subspecies, following Article 45g
of the International Code of Zoological Nomenclature (1985), because they have been used as subspecies in
papers published before 1985. Christensen and Schmid (1987). however, did not recognize subspecies of B.
praecursor , and subspecies are not recognized in the present paper. Extreme variants with a smooth guard,
referred to as var. praecursor by Jeletzky, do occur in the East Harnham sample (see PI. 5, figs 3 and 4).

Affinity. B. praecursor is closely allied to B. alpha Naidin from the lower and middle Lower
Campanian, and B. mucronata (Schlotheim) from the uppermost Lower Campanian and Upper
Campanian. The affinity was fully discussed by Christensen (1986) and Christensen and Schmid
(1987). B. praecursor differs from B. alpha by being more slender, having a smaller Schatzky
distance, and a larger fissure angle. Moreover, in B. praecursor the ratio of the length from the apex
to the protoconch and the dorso-ventral diameter at the protoconch varies from about 3 to about
6, generally being 3-6^E6, and the mean value is 4. In B. alpha , the ratio varies from about 3 to about
4, being 3-2— 3-8 in most specimens, and the mean value is 3-5.

B. praecursor differs from B. mucronata by having weakly developed vascular markings, lacking
a well-defined mucro, and being longer and more slender. The fissure angles of B. mucronata from
the uppermost Lower Campanian and basal Upper Campanian and B. praecursor are very similar,
whereas samples of B. mucronata from the middle Upper Campanian generally have larger fissure
angles.

Distribution. B. praecursor has been recorded from Northern Ireland, through England, France, Belgium,
northern Germany, Poland to Russia (Christensen 1986; Christensen and Schmid 1987).

The specimens from East Harnham are regarded to be from the middle Lower Campanian (see above). B.

744

PALAEONTOLOGY, VOLUME 34

praecursor occurs in West Germany in the lower Lower Campanian G. granulataquadrata Zone and the
overlying I. lingua/ G. quadrata Zone. A single specimen has also been collected in the middle Lower
Campanian Galeola senonensis Zone. In Belgium, it occurs in the lower part of the I. lingua/ G. quadrata Zone
of the basal Lower Campanian. In Northern Ireland and southern England it occurs in the middle Lower
Campanian. In Russia, B. praecursor appears in the uppermost Sanlonian and continues into the lower and
middle Lower Campanian (Naidin and Kopaevich 1977; Naidin 1979, 1983). In the Corbieres area of the
French Pyrenees B. praecursor has recently been recorded from the uppermost Santonian (Christensen et al.
1991). Jeletzky (1955b) recorded a single specimen from the U.S.A. which probably came from the Niobrara
Formation in Kansas.

Belemnitella cf. praecursor
Plate 4, figs 19-22; Plate 6, figs 5-7, 11-13

Material. BMNH C43288, Porchester Pit, Portsdown Hills, Hampshire; SM B97228-31, Stiffkey; BGS GSM
101370, 101376, 101379-81, Shawford.

Dimensions.

LAP

DVDP

LDP

SD

FA

AA

LAP/DVDP

BMNH C43288

6L8

15-3

15-3

6-4

2L5

23-0

4-0

SM B97231

59-0

14-8

14-2

—

—

—

4-0

SM B97228

50-6

1 3-6

13-7

—

—

—

3-7

SM B97230

48-4

12-5

—

10 6

—

—

3-9

SM B97229

36-6

8-8

8-3

—

—

—

4-2

BGS GSM 101370

64-3

15-3

15-6

—

—

—

4-2

BGS GSM 101376

38-7

8-7

8-2

—

—

—

4-5

BGS GSM 101379

64-6

15-4

15-3

—

—

—

4-2

BGS GSM 101380

53-4

13-1

131

—

—

—

41

BGS GSM 101381

57-8

15-5

15-2

—

—

—

3-7

Remarks. The specimen from Porchester Pit is relatively slender (LAP/DVDP is 4-0), has weakly
developed vascular imprints, and no well-defined mucro. The alveolus carries conellae (PI. 4, figs
19-22) as in B. praecursor and B. alpha. The specimens from Stiffkey have weakly developed
vascular imprints and the mucro is not well-defined. The mean value of the LAP/DVDP ratio of
four specimens is 3-9, with an observed range from 3-7-4-2. The specimens from Porchester Pit
and Stiffkey may be B. praecursor on the basis of their LAP/DVDP ratios, but they are referred to
B. cf. praecursor owing to the small number of specimens.

I have also studied a small Belemnitella sample from Shawford, locality 1086 of Brydone (1912).
Most of the specimens are unhorizoned, and the remaining specimens are from ‘bottom course’,
'lower course’, ‘2nd course’, and ‘below upper marl layers’ sensu Brydone. The Belemnitella fauna
is heterogeneous in contrast to the Belemnitella sample from East Harnham (see above). It seems
that specimens of Belemnitella from several horizons within the middle and upper Lower
Campanian are present. Five specimens from the ‘bottom course’ and ‘lower course’ are referred
to B. cf. praecursor on the shape and slenderness of the guard (the mean value of the LAP/DVDP
ratios is 41 with an observed range of 3-7-4-5), and the weakly developed vascular markings. Some
of the unhorizoned specimens may also be B. cf. praecursor. B. mucronata , however, is also present
and this species appears in the uppermost Lower Campanian. The specimens of B. mucronata ,
however, may have come from another pit.

Distribution. The specimens from Stiffkey are regarded as coming from the upper Lower Campanian papillosa
Zone (see above), and the specimens from Shawford from the middle Lower Campanian.

CHRISTENSEN: CHALK BELEMNITES

745

Belemnitella sp.

Material. SM B95274, Wells, pilula Zone; BMNH C44865, Sussex coast, base of G. quadrata Zone.

Remarks. SM B95274 is a fragment of a juvenile specimen consisting of the middle part of the guard.
BMNH C44865 is a stout specimen (LAP: DVDP is 3-5) with vascular imprints. It was recorded as
Actinocamax mercyi by Rowe (1900, p. 343). A specific determination of the two specimens is not
possible.

Acknowledgements. This paper is dedicated to the memory of Tove Birkelund (1928-1986) in token of her
contribution to Mesozoic stratigraphy and palaeontology. Tove’s first field of activity was Upper Cretaceous
belemnites, and in the 1950s she monographed the belemnites from Denmark and Greenland. Later she mainly
focussed on Jurassic and Cretaceous ammonites, particularly from Greenland.

I thank the curators and staff of the following museums and institutes who allowed me to study specimens
in their care: Mr D. Phillips (formerly Natural History Museum, London), Mr C. J. Wood (formerly British
Geological Survey, Keyworth), Dr D. Price (Sedgwick Museum, Cambridge), Mr R. A. D. Markham (Ipswich
Museum), and Dr W. J. Kennedy (University Museum, Oxford). I also thank Mr C. W. Wright (Seaborough,
Dorset), who on behalf of Prof. J. M. Hancock (London), Dr W. J. Kennedy, and Mr C. J. Wood, invited me
to monograph the Upper Cretaceous belemnites of the United Kingdom. I am grateful to Mr C. J. Wood and
Dr A. S. Gale (London) for providing me with information on specimens and outcrops. The paper was read
critically by C. J. Wood who offered helpful suggestions and improved the English text. I wish to express my
sincere thanks for this help. I gratefully acknowledge the technical assistance of Mr C. Rasmussen, Mr S. L.
Jakobsen, Mrs Nina Topp, and Mrs Annemarie Brantsen, all from the Geological Museum, Copenhagen. This
study is supported by the Carlsberg Foundation.

REFERENCES

arkhangelsky, a. d. 1912. The Upper Cretaceous deposits in the eastern part of the European Russia.

Material)1 dlya Geologii Rossii , 25, 631 pp. [In Russian],
bailey, h. w., gale, a. s.. mortimore, r. N., swiECiCKi, a. and wood, c. J. I983u. The Coniacian-Maastrichtian
stages of the United Kingdom, with particular reference to southern England. Newsletter in Stratigraphy , 12,
29-A2.

19836. Criteria for defining the Coniacian to Maastrichtian stage boundaries in
England. 9-12. In birkelund, t., bromley, r., Christensen, w. k., hakansson, e. and surlyk, f. (eds).
Cretaceous stage boundaries. Geologisk Centralinstitut, University of Copenhagen, Copenhagen, 210 pp.

1984. Biostratigraphical criteria for the recognition of the Coniacian to

Maastrichtian stage boundaries in the Chalk of north-west Europe, with particular reference to southern
England. Bulletin of the Geological Society of Denmark , 33, 31-39.
bayle, E. 1878. Fossils principaux der terrains de la France. Explication de la Carte Geologique de la France ,
4(1), Atlas, 79 plates.

birkelund, t. 1957. Upper Cretaceous belemnites from Denmark. Biologiske Skrifter. Det Kongelige Danske
Videnskabernes Selskab , 9, 1-69.

blackmore, H. p. 1896. Some notes on the aptychi from the Upper Chalk. Geological Magazine , (4), 3,
529-533.

blainville, H. M. D. de 1827. Memoire sur les Belemnites , considerees zoologiquement et geologiquement .
Levrault, Paris, 136 pp.

Brighton, a. g. 1930. Actinocamax from the Upper Chalk of Yorkshire. The Naturalist , 1930, 117-120.
brydone, r. m. 1912. The stratigraphy of the Chalk of Hants. Dulau and Co., London, 116 pp.

— 1914. The Zone of Off aster pilula in the South English Chalk. I-IV. Geological Magazine , (6), I, 359-369,
405-411, 449^157, 509-513.

— 1933. The Zone of granulated Actinocamax in East Anglia. Transactions of the Norfolk and Norwich
Naturalists' Society , 13, 285-293.

Christensen, w. K. 1971. Belemnitella propinqua propinqua (Moberg, 1885) from Scandinavia. Bulletin of the
Geological Society of Denmark , 20, 369-384.

— 1973. The belemnites and their stratigraphical significance. 113-140. In bergstrom, j., Christensen,
w. k., johansson, c. and norling, e. An extension of Upper Cretaceous rocks to the Swedish west coast at
Sardal. Bulletin of the Geological Society of Denmark. 22, 83-154.

746

PALAEONTOLOGY, VOLUME 34

1974. Morphometric analysis of Actinocamax plenus from England. Bulletin of the Geological Society of
Denmark , 23, 1-26.

1975n. Upper Cretaceous belemnites from the Kristianstad area in Scania. Fossils and Strata , 7, 1-69.

— 1975 b. Designation of lectotypes for Gonioteuthis westfalicagranulata and G. granulataquadrata.
Palaontologisches Zeitschrift, 49, 126-134.

1976. Palaeobiogeography of Late Cretaceous belemnites of Europe. Palaontologisches Zeitschrift , 50.
113-129.

- 1982. Late Turonian-early Coniacian belemnites from western and central Europe. Bulletin of the
Geological Society of Denmark , 31, 63-79.

- 1986. Upper Cretaceous belemnites from the Vomb Trough in Scania, Sweden. Sveriges geologiska
Undersokning, Ca57, 1-57.

1988. Upper Cretaceous belemnites of Europe: State of the art. 5-16. In streel, m. and bless, m. j. m.
(eds). The Chalk District of the Euregio Meuse- Rhine . Natuurhistorisch Museum and Laboratoires de
Paleontologie de l'Universite d’Etat, Maastricht and Liege, 116 pp.

1990. Actinocamax primus Arkhangelsky (Belemnitellidae; Upper Cretaceous), Biometry, comparison
and biostratigraphy. Palaontologisches Zeitschrift , 64, 75-90.

bilotte, m. and melchior, p. 1991. Upper Cretaceous belemnitellids from the Corbieres area in the
French Pyrenees. Cretaceous Research , 11, 359-369.

— ernst, g., schmid, f., schulz, m.-g. and wood, c. j. 1975. Belemnitellci mucronata mucronata (Schlotheim,
1813) from the Upper Campanian: Neotype, biometry, comparison and biostratigraphy. Geologisches
Jahrbuch , A28. 27-57.

— and schmid, f. 1987. The belemnites of the Vaals Formation from the C.P.L. quarry at Hallembaye in
Belgium - Taxonomy, biometry and biostratigraphy. Geologisches Jahrbuch , A94, 3-37.

- and schulz, m.-g. 1976. First record of Belemnellocamax balsvikensis (Brotzen, 1960) from NW
Germany. Neues Jahrbuch fur Geologie und Paldontologie , Monatshefte , 9, 522-531.

combemorel, r., Christensen, w. K., naidin, D. p. and Spaeth, c. 1981. Les Belemnites. Cretaceous Research ,
2, 283-286.

CRICK, G. C. 1906. Note on a rare form of Actinocamax (A.) grossouvrei from the Chalk of Yorkshire. The
Naturalist , 1906, 155-158.

1907. Note on two rare forms of Actinocamax from the English Upper Chalk. Geological Magazine , 4,

389-395.

- 1910. On Belemnocamax boweri, n.g. et sp. A new cephaloped from the Lower Chalk of Lincolnshire.
Proceedings of the Geologists' Association , 21, 360-365.

dibley, g. e. 1900. Zonal features of the chalk pits in the Rochester, Gravesend, and Croydon areas.
Proceedings of the Geologists' Association , 16, 484-499.

- 1918. Additional notes on the chalk of the Medway Valley, Gravesend, west Kent, north-east Surrey, and
Grays (Essex). Proceedings of the Geologists' Association , 29, 68-96.

ernst, G. 1963a. Stratigraphische und gesteinchemische Untersuchungen im Santon und Campan von
Lagerdorf (SW-EIolstein). Mitteihmgen aus clem Geologischen Staatsinstitut in Hamburg , 32, 71-127.

1963b. Zur Feinstratigraphie und Biostratonomie des Obersanton und Campan von Misburg und Hover
bei Hannover. Mitteihmgen aus dem Geologischen Staatsinstitut in Hamburg , 32, 128-147.

1964a. Ontogenie, Phylogenie und Stratigraphie der Belemnitengattung Gonioteuthis Bayle aus dem
nordwestdeutschen Santon/Campan. Fortschritte in der Geologie von Rheinland und Westfalen , 7, 113-174.

1964b. Neue Belemnitellenfunde in der Bottroper Mulde und die stratigraphische Stellung der ‘Bottroper
Mergel’. Fortschritte in der Geologie von Rheinland und Westfalen , 7, 175-198.

1966. Fauna, Okologie und Stratigraphie der Mittelsantonen Schreibkreide von Lagerdorf (SW-
Holstein). Mitteilungen aus dem Geologischen Staatsinstitut in Hamburg , 35, 115-145.

1968. Die Oberkreide-Aufschliisse im Raume Braunschweig-Hannover und ihre stratigraphische
Gliederung mit Echinodernren und Belemniten. Beihefte zu den Berichten der Naturhistorischen Gese/lschaft
zu Hannover , 5, 235-284.

— schmid, f. and klischies, g. 1979. Multistratigraphische Untersuchungen in der Oberkreide des Raumes
Braunschweig-Hannover. 1 1^16. In wiedmann, j. (ed.). Aspekte der Kreide Europas. IUGS, A6, 1-680.

— and schulz, m.-g. 1974. Stratigraphie und Fauna des Coniac und Santon im Schreibkreide-Richtprofil
von Lagerdorf (Holstein). Mitteihmgen aus dem Geologisch-Paldontologischen Institut der Universitdt
Hamburg , 43, 5-60.

Fletcher, t. p. and wood, c. J. 1978. Cretaceous rocks. 85-115. In wilson, h. e. and manning, p. i. (eds).
Geology of the Causeway Coast. Memoir of the Geological Survey of Northern Ireland, Sheet 7, 2, 73-172.

CHRISTENSEN: CHALK BELEMNITES

747

- 1982. Cretaceous. 44-54. In Griffith, a. e. and wilson, h. e. (eds). Geology of the country around
Carrickfergus and Bangor. Memoir Geological Survey of Northern Ireland, Sheet 29, viii+ 1 18 pp.
gale, a. s., wood, c. J. and bromley, r. g. 1987. The lithostratigraphy and marker bed correlation of the White
Chalk (Late Cenomanian-Campanian) in southern England. Mesozoic Research , 1, 107-118.

— and woodroof, p. b. 1981. A Coniacian ammonite from the ‘Top Rock’ in the Chalk of Kent.
Geological Magazine , 118, 557-560.

Glazunova, a. e. 1972. Paleontological evidence for the stratigraphical classification of the Cretaceous in
Povolzh’e. Ministry of the Geology of the USSR. Geological Institute VSEGEI , 114 pp. [In Russian].
Griffith, c. and brydone, r. m. 1911. The zones of the chalk in Hants. Dulau and Co., London, 35 pp.
hald, a. 1957. Statistical theory with engineering applications. John Wiley and Sons Inc., New York, 783 pp.
Hancock, j. m. (ed.) 1972. Cretace de l’Angleterre, Pays de Galles. Ecosse. Lexique Stratigraphique
International , 1 (3a XI), 1-162.

international commission on zoological nomenclature. 1985. Opinion 1328. Belemnites mucronatus
Schlotheim, 1813 (Coleoidea): Conserved and neotype designated. Bulletin of Zoological Nomenclature , 42,
222-225.

jarvis, i. 1980. Palaeobiology of Upper Cretaceous belemnites from phosphatic chalk of the Anglo-Paris
Basin. Palaeontology , 23, 889-914.

jefferies, r. p. s. 1961 . The palaeoecology of the Actinocamax plenus Subzone (Lowest Turonian) in the Anglo-
Paris Basin. Palaeontology , 4, 609-647.

jeletzky, j. a. 1949a. Some notes on ‘ Actinocamax ' propinquus Moberg 1885, its taxonomic position and
phylogenetic relations within the Family Belemnitellidae Pavlow 1913, morphological characters, and
synonymy. Geologiska Foreningens i Stockholm Forhandlingar , 71, 415^124.

— 1949a. Uber den taxonomischen Wert einiger morphologische Elemente des Rostrums der belemnitel-
lenartigen Formen (Familie Belemnitellidae Pavlow, 1913), sowie liber die Gattung Belemnella (Nowak,
1913, subg.) Jeletzky, 1941, ihre Phylogenie und einige Vertreter. Neues Jahrbuch fur Mineralogie, Geologie
und Paldontologie, Monatshefte , 9, 257-287.

— 1951a. The place of the Trimingham and Norwich chalk in the Campanian-Maestrichtian succession.
Geological Magazine , 88, 197-208.

19516. Die Stratigraphie und Belemnitenfauna des Obercampan und Maastricht Westfalens, Nordwest-
deutschlands und Danemarks, sowie einige algemeine Gliederungs - Probleme der jtingeren borealen
Oberkreide Eurasiens. Beihefte Geo/ogisches Jahrbuch , 1, 1-142.

— 1955a. Evolution of Santonian and Campanian Belemnitella and paleontological systematics: exemplified
by Belemnitella praecursor Stolley. Journal of Palaeontology, 29, 478-509.

19556. Belemnitella praecursor , probably from the Niobrara of Kansas and some stratigraphic
implications. Journal of Palaeontology , 29, 876-885.

— 1958. Die jiingere Oberkreide (Oberconiac bis Maastricht) Siidwestrusslands und ihr Vergleich mit der
Nordwest- und Westeuropas. Beihefte Geo/ogisches Jahrbuch , 33, 1-157.

— 1965. Taxonomy and phylogeny of fossil Coleoidea (= Dibranchiata). Geological Survey of Canada ,
Paper 65-2, No. 42, 12-16.

Kennedy, w. j. and Christensen, w. k. 1991. Coniacian and Santonian ammonites from Bornholm, Denmark.

Bulletin of the Geological Society of Denmark , 38, 203-226.
kongiel, r. 1962. On belemnites from Maastrichtian, Campanian and Santonian sediments in the Middle
Vistula Valley (Central Poland). Prace Muzeum Ziemi , 5, 1-148.
lake, r. d., young, b., wood, c. J. and mortimore, r. n. 1987. Geology of the country around Lewes. Memoir
British Geological Survey, Sheet 319, viii+ 1 17 pp.
miller, j. s. 1823a. Observations on belemnites. Transactions of the Geological Society, (2), 2, 45-62.

— 18236. Observations on the genus Actinocamax. Transactions of the Geological Society , (2), 2, 63-67.
moberg, J. c. 1885. Cephalopoderne i Sveriges Kritsystem. 2. Artsbeskrifning. Sveriges Geologiska

Undersokning, C73, 1-45.

mortimore, r. n. 1986. Stratigraphy of the Upper Cretaceous White Chalk of Sussex. Proceedings of the
Geologists' Association , 97, 97-139.

— 1987. Upper Cretaceous Chalk in the North and South Downs, England: a correlation. Proceedings of
the Geologists' Association , 98, 77-86.

— and pomerol, b. 1987. Correlation of the Upper Cretaceous White Chalk (Turonian to Campanian) in
the Anglo-Paris Basin. Proceedings of the Geologists' Association, 98, 97-143.

naidin, d. p. 1952. The Upper Cretaceous belemnites of western Ukraine. Trudy Moskovskogo Geologo-
Razvedoclmogo Instituta imeni S. Ordzhinikidze , 27, 1-170. [In Russian],

748

PALAEONTOLOGY, VOLUME 34

1964. The Upper Cretaceous belemnites of the Russian Platform and contiguous regions , Actinocamax,
Gonioteuthis, Belemnellocamax. Moscow University Press, Moscow, 190 pp. [In Russian],

1974. Subclass Endocochlia - Order Belemnitidae. 197-240. In krumgol’ts, g. ya. (ed.). Atlas of the
Upper Cretaceous fauna of Donbass. ‘Nedra’, Moscow, 639 pp. [In Russian].

1979. Vergelichende Stratigraphie der Oberen Kreide der Russischen Tafel und West-Europas. 497-510.
In wiedmann, j. (ed.). Aspekte der Kreide Europas. IUGS, A6, 1-680.

1981 . The Russian Platform and Crimea. 29-68. In reyment, r. a. and bengtson, p. (eds). Aspects of Mid-
Cretaceous Regional Geology. Academic Press, London, 327 pp.

- 1983. Upper Cretaceous zonation of the European palaeobiogeographical region. 136-139. In birkelund,
t., bromley, r. G., Christensen, w. K., hakansson, E. and surlyk, F. (eds). Symposium on Cretaceous stage
boundaries. Geologisk Centralinstitut, University of Copenhagen, Copenhagen, 210 pp.

and kopaevich, l. f. 1977. On the zonal division of the Upper Cretaceous in the European
palaeobiogeographical province. Byulleten Moskovskogo Obschestra Ispytatelei Prirody, Otdel Geologiche-
skii , 52, 92-U2. [In Russian],

nikitin, i. i. 1958. Upper Cretaceous belemnites of the northeastern slope of the Dnjepr-Donez-basin. Trudy
Instituta Geologicheskikh Akademia Nauk Ukrainskoi SSR , Seriya Stratigrafii i Paleontologii, 20, 1-92. [In
Ukrainian],

pavlow, a. p. 1914. Jurassic and Lower Cretaceous Cephalopoda of northern Siberia. Imperatovskoi Akademii
Nauk , St Petersburg , Zapiski Seriia <5, Fizino-mathematicheskii Otdel , 21, 68 pp. [In Russian].
peake, N. b. and HANCOCK, J. M. 1961. The Upper Cretaceous of Norfolk. Transactions of the Norfolk and
Norwich Naturalists' Society , 19, 269-339.

1970. The Upper Cretaceous of Norfolk. Revised edition. Transactions of the Norfolk and Norwich
Naturalists' Society , 19, 269-339J.

ravn, j. p. j. 1918. Kridtaflejringerne paa Bornholms sydvestkyst og deres fauna. II. Turonet. Danmarks
geologisk Undersogelse , II. Raekke, 31, 1-39.

- 1946. Om Nyker-omraadets kridtaflejringer. Biologiske Skrifter. Det Kongelige danske Videnskabernes
Selskab , 4, 1-36.

RAWSON, P. F., CURRY, D., DILLEY, F. C., HANCOCK, J. M., KENNEDY, W. J., NEALE, J. W., WOOD, C. J. and WORSSAM,
b. c. 1978. A Correlation of Cretaceous rocks in the British Isles. Geological Society London , Special Report ,
9, 1-70.

reid, r. e. H. 1971. The Cretaceous rocks of north-eastern Ireland. Irish Naturalists' Journal , 17, 105-129.
robaszynski, f. and Christensen, w. k 1989. The Upper Campanian-Lower Maastrichtian chalks of the
Mons Basin, Belgium : a preliminary study of belemnites and foraminifera in the Harmignies and Ciply
areas. Geologie en Mijnbouw , 68. 391-408.

robinson, n. d. 1986. Lithostratigraphy of the Chalk Group of the North Downs, southeast England.
Proceedings of the Geologists' Association. 97, 141-170.

1987. Upper Cretaceous Chalk in the North and South Downs, England: a reply. Proceedings of the
Geologists' Association, 98, 87-93.

rohlf, F. j. and sokal, r. r. 1969. Statistical Tables. W. H. Freeman and Co., San Francisco, 253 pp.
rowe, A. w. 1900. The zones of the white chalk of the English coast. I. Kent and Sussex. Proceedings of the
Geologists' Association. 16, 289-368.

- 1901. The zones of the white chalk of the English coast. II. Dorset. Proceedings of the Geologists'
Association , 17, 1-76.

1904. The zones of the white chalk of the English coast. IV. Yorkshire. Proceedings of the Geologists'
Association , 18. 193-296.

schluter, c. 1876. Cephalopoden der oberen deutschen Kreide. Teil 2. Palaeontographica , 24, 123-263.
schulz, m.-g. 1979. Morphometrisch-variationsstatische Untersuchungen zur Phylogenie der Belemniten-
Gattung Belemnella im Untermaastricht NW-Europas. Geologisches Jahrbuch , A47, 3-157.

- ernst, G., ernst, h. and schmid, f. 1984. Coniacian to Maastrichtian stage boundaries in the standard
section for the Upper Cretaceous white chalk of NW Germany (Lagerdorf-Kronsmoor-Hemmoor) :
Definitions and proposals. Bulletin of the Geological Society of Denmark , 33, 203-215.

sharpe, d. 1853. Description of the fossil remains of Mollusca found in the Chalk of England. Cephalopoda.

Monograph of the Palaeontographical Society , 68 pp.
sherborn, c. d. 1906. On the belemnites of the chalk of Yorkshire. The Naturalist , 1906, 152-154.
simpson, g. G., roe, a. and lewontin, r. c. I960. Quantitative Zoology. Revised Edition. Harcourt, Brace and
World Inc., New York, 440 pp.

CHRISTENSEN: CHALK BELEMNITES

749

sinzow, i. 1915. Upper Cretaceous sediments of the Saratow province. Zapiski Imperatorskago Mineralogi-
cheskogo Obshchestra , (2), 50, 133-162. [In Russian],

SOKAL. r. r. and rohlf, f. j. 1969. Biometry - The principles ami practise of statistics in biological research.
W. H. Freeman and Co., San Francisco, 776 pp.

sowerby. 3. de C. 1829. The Mineral Conchology of Great Britain , 6. The author, London, 230 pp.
stokes, r. b. 1975. Royaumes et provinces faumque de Cretace etablis sur le base d’une etude systematique de
genre Micraster. Memoires du Museum National d'Histoire naturelle, C31. 1-94.

1977. The echinoids Micraster and Epiaster from the Turonian and Senonian chalk of England.
Palaeontology , 20, 805-821

stolley, e. 1892. Die Kreide Schleswig-Holsteins. Mitteilungen aus dem Mineralogischen Institut der Universitdt
Kiel , 1, 191-309.

1897. Ueber die Gliederung des norddeutschen und baltischen Senon sowie die dasselbe character-
iserenden Belemniten. Archiv fur Anthropologie und Geologie Schleswig-Holsteins , 2, 216-302.

- 1916. Neue Beitrage zur Kenntnis der norddeutschen oberen Kreide. III. Die Bedeutung der Actinocamax-
Arten als Leitfossilien der oberen kreide. Jahresbericht des Niedersachsischen geologischen Vereins zu
Hannover. 9, 95-104.

- 1930. Einige Bemerkungen liber die Kreide Siidskandinaviens. Geologiske Foreningens i Stockholm
Forhandlinger, 52, 157-190.

ulbrich, h. 1971. Mitteilungen zur Biostratigraphie des Santon und Campan des mittleren Teils der
Subhercynen Kreidemulde. Freiberg Forschungshefte , C267, 47-71.
wood, c. J. 1967. Some new observations on the Maestrichtian stage in the British Isles. Bulletin of the
Geological Survey of Great Britain , 27, 271-288.

1981. Upper Cretaceous. 92 1 15. In rent, p. (ed. ). Eastern England from the Tees to the Wash. Institute
of Geological Sciences, London, vii + 155 pp.

- 1988. The stratigraphy of the chalk of Norwich. Bulletin of the Geological Society of Norfolk , 38, 3-120.
— and mortimore, r. n. 1988. Biostratigraphy of the Newhaven and Culver Members. 58-64. In young, b.
and lake, R. d. (eds). Geology of the country around Brighton and Worthing. Memoir British Geological
Survey, Sheets 318 and 333, viii+ 1 14 pp.

wright, c. w. and wright, e. v. 1951. A survey of the fossil Cephalopoda of the Chalk of Great Britain.
Monograph of the Palaeontographical Society , 40 pp.

zittel. K. a. von 1895. Grundzuge der Pa/aeontologie (Paldozoologie). R. Oldenburg. Munich and Leipzig,
971 pp.

Typescript received 11 April 1990
Revised typescript received 24 August 1990

W. K. CHRISTENSEN

Geological Museum
University of Copenhagen
Oster Voldgade 5-7

DK-1350 Copenhagen, Denmark

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform
in style to those already published in this journal, and should be sent to Dr Dianne Edwards, Department of Geology,
University of Wales College of Cardiff CF1 3YE, who will supply detailed instructions for authors on request (these are
published in Palaeontology 1990, 33, pp. 993-1000).

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 1991 is £45 00 (U.S. $80) for Institutional
Members, and £20 00 (U.S. $36) for Ordinary and Student Members. A single copy of each Special Paper is available on a
non-subscription basis to Ordinary and Student Members only, for their personal use, at a discount of 25% below the listed
prices: contact the Marketing Manager. Non-members may obtain Nos 35-43 (at cover price) from Basil Blackwell Ltd.,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, England, and older issues from
the Marketing Manager. For all orders of Special Papers through the Marketing Manager, please add £1-50 (U.S. $3) per
item for postage and packing.

PALAEONTOLOGICAL ASSOCIATION PUBLICATIONS

Special Papers in Palaeontology

For full catalogue and price list, send a self-addressed, stamped A4 envelope to the Marketing Manager. Numbers 2-32 are
still in print and are available together with those listed below:

33. (for 1985): Evolutionary Case Histories from the Fossil Record. Edited by J. c. w. cope and p. w. skelton. 202 pp., 80
text-figs., 4 plates. Price £30 (U.S. $60).

34. (for 1985): Review of the upper Silurian and lower Devonian articulate brachiopods of Podolia, by o. l. Nikiforova,
t. L. modzalevskaya and m. g. bassett. 66 pp., 6 text-figs., 16 plates. Price £10 (U.S. $20).

35. (for 1986): Studies in palaeobotany and palynology in honour of N. F. Hughes. Edited by d. j. batten and d. e. g.
briggs. 178 pp., 29 plates. Price £30 (U.S. $60).

36. (for 1986): Campanian and Maastrichtian ammonites from northern Aquitaine, France, by w. J. Kennedy. 145 pp., 43
text-figs., 23 plates. Price £20 (U.S. $40).

37. (for 1987): Biology and revised systematics of some late Mesozoic stromatoporoids, by rachel wood. 89 pp., 31 text-
figs., 7 plates. Price £20 (U.S. $40).

38. (for 1987): Taxonomy, evolution, and biostratigraphy of late Triassic-early Jurassic calcareous nannofossils, by p. r.
bown. 118 pp., 19 text-figs., 15 plates. Price £30 (U.S. $60).

39. (for 1988): Late Cenomanian and Turonian ammonite faunas from north-east and central Texas, by w. j. Kennedy,
131 pp., 39 text-figs., 24 plates. Price £30 (U.S. $60).

40. (for 1988): The use and conservation of palaeontological sites. Edited by p. r. crowther and w. a. Wimbledon. 200 pp.,
31 text-figs. Price £30 (U.S. $60).

41. (for 1989): Late Jurassic-early Cretaceous cephalopods of eastern Alexander Island, Antarctica, by p. j. howlett.
72 pp., 9 text-figs., 10 plates. Price £20 (U.S. $40).

42. (for 1989): The Palaeocene flora of the Isle of Mull, by m. c. boulter and z. kvacek. 149 pp., 23 text-figs., 23 plates.
Price £40 (U.S. $80).

43. (for 1990): Benthic palaeoecology of the Late Jurassic Kimmeridge Clay of England, by p. r. wignall, 74 pp., 50 text-
figs. Price £30 (U.S. $60).

Field Guides to Fossils and Other Publications

These are available only from the Marketing Manager. Please add £100 (U.S. $2) per book for postage and packing plus
£1-50 (U.S. $3) for airmail. Payments should be in Sterling or in U.S. dollars, with all exchange charges prepaid. Cheques
should be made payable to the Palaeontological Association.

1. (1983): Fossil Plants of the London Clay, by M. E. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $16) (Members £6

or U.S. $12).

2. (1987): Fossils of the Chalk, compiled by e. owen; edited by A. b. smith. 306 pp., 59 plates. Price £11 -50 (U.S. $23)

(Members £9-90 or U.S. $20).

3. (1988): Zechstein Reef fossils and their palaeoecology, by N. hollingworth and T. Pettigrew, iv + 75 pp. Price £4-95

(U.S. $10) (Members £3.75 or U.S. $7-50).

4. (1991): Fossils of the Oxford Clay, edited by d. m. martill and j. d. Hudson. 286 pp., 44 plates. Price £15 (U.S. $30)

(Members £12 or U.S. $24).

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $40).

1985. Atlas of Invertebrate Macrofossils. Edited by s. w. Murray. Published by Longman in collaboration with the
Palaeontological Association, xiii 4-241 pp. Price £ 1 3-95. Available in the USA from Halsted Press at U.S. $24-95.

© The Palaeontological Association, 1991

Palaeontology

VOLUME 34 • PART 3

CONTENTS

The Upper Jurassic diapsid Lisboasaurus estesi - a maniraptoran
theropod

A. R. MILNER and S. E. EVANS 503

The conchostracan fauna of the Great Estuarine, Middle Jurassic
Group, Scotland

CHEN PEI-JI and J. D. HUDSON 515

A new, primitive, dinocephalian mammal-like reptile from the
Permian of southern Africa

B. S. RUBIDGE 547

Fishes and amphibians from the late Permian Pedra de Fogo
Formation of northern Brazil

c. b. cox and p. hutchinson 561

Silurian cryptospores and miospores from the type Llandovery
area, south-west Wales

N. D. BURGESS 575

Silurian cryptospores and miospores from the type Wenlock area,
Shropshire, England

N. D. BURGESS and J. B. RICHARDSON 601

A new marsupiate cidaroid echinoid from the Maastrichtian of
Antarctica

D. B. BLAKE and W. J. ZINSMEISTER - 629

A new species of machaeridian from the Silurian of Podolia,

USSR, with a review of the Turrilepadidae

J. M. ADRAIN, B. D. E. CHATTERTON and L. R. M. COCKS 637

Mosasaurs from the Upper Cretaceous of Niger

T. LINGHAM-SOLIAR 653

Isolated graptolites from the Llandovery of Kallholen, Sweden

D. K. LOYDELL 671

Belemnites from the Coniacian to Lower Campanian chalks of
Norfolk and southern England

W. K. CHRISTENSEN 695

Printed in Great Britain at the University Press , Cambridge

ISSN 0031-0239

Palaeontology

VOLUME 34 ■ PART 4 NOVEMBER 1991

101

Pi55

THE PALAEONTOLOGICAL ASSOCIATION

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

COUNCIL 1991-1992

President -. Professor J. W. Murray, Department of Geology, The University, Southampton S09 5NH
Vice-Presidents : Dr P. R. Crowther, City of Bristol Museum and Art Gallery, Queen’s Road, Bristol BS8 1RL
Dr P. A. Selden, Department of Extra-Mural Studies, University of Manchester, Manchester M13 9PL
Treasurer : Dr M. E. Collinson, Department of Biology, King’s College, London W8 7AH
Membership Treasurer : Dr H. A. Armstrong, Department of Geological Sciences, The University, South Road,

Durham DH1 3LE

Institutional Membership Treasurer : Dr A. R. I. Cruickshank, Leicestershire Museums Arts and Records Service,

96 New Walk, Leicester LEI 6TD

Secretary : Dr J. A. Crame, British Antarctic Survey, High Cross, Madingley Road, Cambridge CB3 OET
Circular Reporter : Dr D. Palmer, Department of Geology, National Museum of Wales, Cardiff CF1 3NP
Marketing Manager (Sales): Dr C. R. Hill, Department of Botany, The Natural History Museum,

London SW7 5BD

Marketing Manager ( Publicity ) : Dr P. B. Wignall, Department of Earth Sciences, The University, Leeds LS2 9JJ
Public Relations Officer : Dr D. M. Martill, Department of Earth Sciences, The Open University,

Milton Keynes MK7 6AA

Editors

Dr J. E. Dalingwater, Department of Environmental Biology, University of Manchester, Manchester M13 9PL
Dr P. Doyle, Department of Earth Sciences, Thames Polytechnic, London El 2NG
Dr D. Edwards, Department of Geology, University of Wales College of Cardiff, Cardiff CF1 3YE
Dr P. D. Lane, Department of Geology, University of Keele, Keele, Staffordshire ST5 5BG
Dr A. R. Milner, Department of Biology, Birkbeck College, Malet Street, London WC1E 7HX
Dr P. D. Taylor, Department of Palaeontology, The Natural History Museum, London SW7 5BD

Other Members

Dr E. A. Jarzembowski, Brighton Dr W. J. Kennedy, Oxford Dr R. B. Rickards, Cambridge

Overseas Representatives

Argentina : Dr M. O. Mancenido, Division Paleozoologia invertebrados, Facultad de Ciencias Naturales y Museo, Paseo del
Bosque, 1900 La Plata. Australia: Dr K. J. McNamara, Western Australian Museum, Francis Street, Perth, Western
Australia 6000. Canada: Professor S. H. Williams, Department of Earth Sciences, Memorial University, St John’s,
Newfoundland A1B 3X5. China: Dr Chang Mee-mann, Institute of Vertebrate Palaeontology and Paleoanthropology,
Academia Sinica, P.O. Box 643, Beijing. Dr Rong Jia-yu, Nanjing Institute of Geology and Palaeontology, Chi-Ming-Ssu,
Nanjing. France: Dr J.-L. Henry, Institut de Geologie, Universite de Rennes, Campus de Beaulieu, Avenue du General
Leclerc, 35042 Rennes Cedex. Iberia: Prof. F. Alvarez, Departamento de Geologia, Universidad de Oviedo, C/. Jesus Arias
de Velasco, s/n. 33005 Oviedo, Spain. Japan : Dr I. Hayami, University Museum, University of Tokyo, Hongo 7-3-1, Tokyo.
New Zealand: Dr R. A. Cooper, New Zealand Geological Survey, P.O. Box 30368, Lower Hutt. Scandinavia: Dr R.
Bromley, Fredskovvej 4, 2840 Holte, Denmark. U.S.A. : Prof. A. J. Rowell, Department of Geology, University of Kansas,
Lawrence, Kansas 66044. Prof. N. M. Savage, Department of Geology, University of Oregon, Eugene, Oregon 97403. Prof.
M. A. Wilson, Department of Geology, College of Wooster, Wooster, Ohio 44961. Germany : Prof. F. T. Fursich,
Institut fur Palaontologie, Universitat, D8700 Wurzburg, Pliecherwall 1

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1991 are:
Institutional membership . . £60-00 (U.S. $108) Student membership . . £1 1-50 (U.S. $20)

Ordinary membership £28 00 (U.S. $50) Retired membership .... £14 00 (U.S. $25)

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr A. R. I.
Cruikshank, Leicestershire Museums Arts and Records Service, 96 New Walk, Leicester LEI 6TD. 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 H. A. Armstrong, Department of
Geological Sciences, The University, South Road, Durham DH1 3LE. Subscriptions cover one calendar year and are due each
January; they should be sent to the Membership Treasurer. All members who join for 1991 will receive Palaeontology ,
Volume 34, Parts 1 4. Enquiries concerning back numbers should be directed to the Marketing Manager.

Non-members may subscribe, and also obtain back issues up to 3 years old, at cover price through Basil Blackwell Ltd,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, England. For older issues
contact the Marketing Manager.

Cover: BolboJ'orma intermedia Daniels and Spiegler (Incertae Sedis, possibly a calcified algal cyst) from Site 552A, southwest
margin of Rockall Plateau, late Miocene NN9-10. x 800. Bolboforma was planktonic and cysts are found in epicontinental
shelf sea deposits, thus providing a useful biostratigraphic link with oceanic sequences.

XX^rarses^

NOVEL ULTRASTRUCTURE IN WATER-
CONDUCTING CELLS OF THE LOWER DEVONIAN
PLANT SENNICAULIS H I P PO C RE P IFO RM I S

by P. KENRICK, D. EDWARDS and R. C. DALES

Abstract. A description of the ultrastructure of the water-conducting cells in Sennicaulis hippocrepiformis
Edwards (1981), an early land plant of uncertain affinity, is based on pyrite and limonite permineralizations
from two Lower Devonian localities in Dyfed and Powys, Wales. The ultrastructure of the two-layered cell wall
is unique among land plants, although a simple large helical thickening suggests affinity with the Tracheophyta.
The lumen of each cell is lined with a thin microporate layer that overlies the bulk of the wall, including the
simple helical thickening, which has a spongy texture. Tapering end walls like those seen in tracheids have not
been observed. The reconstruction of the cell wall is based on an analysis of the mineral and coalified material
using polished thick sections and scanning electron microscopy; the interpretation relies on comparative
morphology and recent advances in knowledge of the process of sedimentary pyrite formation. This novel cell
type is shown to be very different from that recently described in a similarly preserved plant, Goss/ingia
breconensis Heard, and comparisons are also made with presumed water-conducting cells of other early land
plants. The microporate layer resembles that found in some extant hepatics, although a convincing argument
for a close phylogenetic relationship requires more information on the chemical structure of the wall layers and
the morphology of the whole plant.

The thirty-second annual address to the Palaeontological Association entitled 'Pioneering plants’,
while celebrating the variety in organization exhibited by early land plants, cautioned that
preconceptions based on extant vegetation can impose restrictions on interpretation. The latter may
be particularly important at a time of anatomical, biochemical and morphological innovation
associated with the colonization of a new and highly stressful environment. Anticipation
undoubtedly colours perception. An excellent example relates to the water-conducting tissues of
land plants and, more specifically, to the tracheid or vessel which characterizes the vascular plant
and which is essential to the homoiohydric condition.

Water-conducting cells of early land plant fossils have usually been interpreted as tracheids
similar to those in extant plants. The presence of tracheids of apparently similar structure in
different higher taxa (e.g. Rhyniophytina, Zosterophyllophytina, Trimerophytina and Lycophytina)
has reinforced the concept of a monophyletic origin for these diverse groups. Recently, the
identification and redescription of major cell wall features (e.g. thickening or pitting type) using
scanning electron microscopy has shown that previous descriptions based on light microscope
observations were often incorrect (Kenrick and Edwards 1988). Further technical refinement and
a better understanding of how certain minerals form have also facilitated analysis and interpretation
of cell wall ultrastructure. Thus detailed studies of the water-conducting cells ('tracheids’) of the
oldest well-preserved vascular plants have identified two distinctly different types of wall
ultrastructure in cells that are superficially similar. The previously described Goss/ingia- type
(Kenrick and Edwards 1988) is comparable to protoxylem elements in some extant pteridophytes,
whereas the Sennicaulis- type, described here, combines helical thickenings - a tracheid feature -
with a thin microporate wall - a feature of the water-conducting cells of some hepatics.

IPalaeontoIogy, Vol. 34, Part 4, 1991, pp. 751-766, 2 pls.|

© The Palaeontological Association

752

PALAEONTOLOGY, VOLUME 34

MATERIAL AND LOCALITIES

Three sets of sterile, permineralized axes showing similar wall structure were obtained from two
localities in South Wales: Mill Bay West, a new locality in Dyfed, and Brecon Beacons Quarry in
Powys. Most of the material illustrated here was collected from Mill Bay (PI. 1, figs 1, 2, 6, 7; PI.
2, figs 1-3, 6; Text-fig. 1a, b) and contains a high proportion of limonite (see p. 758). Two
illustrations (PI. 1, figs 3 and 4) are of pyritized material from the Edwards collection of Sennicaulis
hippocrepiformis from Brecon Beacons Quarry and now housed in the British Museum (Natural
History) - BM(NH). This material contains little or no limonite and is a good example of the type
of wall structure described here as seen in polished transverse section. Further limonitic material,
collected from Brecon Beacons Quarry, is used to demonstrate the helical nature of the wall
thickenings because it etched well (PI. 1, figs 6 and 7). These two localities are in separate local units
of the Lower Old Red Sandstone, namely the Cosheston Group and the Senni Beds (Allen 1974;
Thomas 1978).

Mill Bay West ( SN 002 049). This locality on the southern shore of the Daugleddau estuary, c.
3-5 km north-east of Pembroke Dock, is in the Mill Bay Formation (Thomas 1978) of the Cosheston
Group. The sequence consists of dark, grey-green, highly indurated, c. 3 m thick, interlaminated
very fine-grained sandstones and coarse siltstones which have been ‘pillowed’. The pillowing is
thought to be of tectonic origin and has severely deformed the bedding planes. The locality,
frequently submerged at high tide, is unit 269 of Thomas (1978) and is 126 m above the base of his
MBW log D. The lithology is a difficult one from which to extract plants, but has yielded abundant,
sterile, coalified axes with a central strand preserved in pyrite or limonite. These new fossils have
been assigned to the form genus Sennicaulis based on anatomical observations of the central strand.

A comparative analysis of the lithostratigraphy and miospore assemblages of the Cosheston
group and the Senni Beds led Thomas (1978) to equate the latter to the lower two-thirds of his Mill
Bay Formation. This places the locality in the uppermost Gedinnian to lower Siegenian on
palynological correlations (Richardson et al. 1982).

Brecon Beacons Quarry (SN 9715 2084). This is a disused roadside quarry c. 12 km south of Brecon
on the main road (A470) to Merthyr Tydfil. About 14 m of the predominantly blue-grey fluviatile
sediments of the Senni Beds are exposed. These consist of a series of massive sandstones (channel-
fill deposits) which cut into and alternate with silty sandstones and siltstones. Some specimens
assignable to Sennicaulis , and collected from the Tarella trowenii plant bed (Edwards and Kenrick
1986), contain significant amounts of limonite. One non-oxidized, pyritic specimen from the
Edwards Sennicaulis hippocrepiformis collection at the BM(NH) was examined. It was originally
collected from plant bed 3, some 2 m above the main Gosslingia horizon (Friend and Williams
1978). The Gosslingia breconensis material illustrated here and first figured in Kenrick and Edwards
(1988) was collected from horizon 4, the main Gosslingia plant bed (Friend and Williams 1978).
Spore assemblages suggest a lowermost Siegenian age for both the Tarella and Gosslingia beds
(lower part of Zone III, Richardson et al. 1982).

Figured specimens are housed in the National Museum of Wales, Cardiff (prefix NMW) or the
British Museum (Natural History), London (prefix V).

TECHNIQUES

Specimens were examined using incident light and scanning electron microscopy. Employing these
two techniques in combination enables accurate reconstructions and interpretations of the cell wall
of pyritic and limonitic fossils to be made. Incident light microscopy illustrates the distribution of
mineral and coalified material within the wall which is essential for interpreting the electron

KENRICK ET AL.\ ULTRASTRUCTURE OF LOWER DEVONIAN PLANT

753

text-fig. 1. a, b, Sennicaulis hippocrepiformis; NMW 90.42G.1 ; Lower Old Red Sandstone; Mill Bay West,
South Wales; transverse section of xylem strand at different levels along an axis, x 70; a. More or less terete
strand composed of two zones of cells: inner zone of small cells surrounded by outer zone of large cells
(specimen is cracked in upper left quarter). Pyrite in cell lumens (high reflectance), oxide in cell walls (low
reflectance); b, elliptical shape to strand caused by collapse of outer zone cells into non-preserved middle

region. Mineral is limonite.

micrographs. Scanning electron microscopy clearly illustrates the three-dimensional form of the cell
and its ultrastructure.

For light microscopy, highly polished thick sections (PI. 1; Text-fig. 1a) were made using a
standard mineralogical polishing technique outlined in Kenrick and Edwards (1988). Bright field
incident light microscopy produced the best results for pyritized axes and also for specimens with
oxidized cell walls and pyritic lumens. Brief etching of the polished surface heightened contrast in
pyritized specimens. The most effective etchant proved to be a saturated solution of ammonium
ceric (IV) nitrate in dilute sulphuric acid (Biggs and Rocken 1983). The entirely limonitic specimen
(Text-fig. Ib) was not polished, but saw marks were removed with 600 grit carborundum and the
specimen mounted under a coverslip in Histomount and photographed using dark field illumination
on an Olympus Vanox. A monochromatic green filter improved the photographic results obtained
with Ilford PAN F or FP4 35 mm black-and-white film.

For scanning electron microscopy, a deep etch or a fractured surface was preferred. Concentrated
solutions of nitric acid and hydrochloric acid were used to etch pyrite and limonite respectively.
Etching time varied depending on the depth of etch required and the mineral present, but was
typically between 5 and 15 minutes. In many instances, better results were obtained from fractured
surfaces, which also served as a control for the effects of acids on wall structure in etched sections.
All specimens were coated with gold and examined using a JEOL 35CS or Cambridge 360 scanning
electron microscope.

SYSTEMATIC PALAEONTOLOGY

Edwards (1981) erected the genus Sennicaulis as a: ‘Form genus for sterile axes, each with a
prominent terete xylem strand lacking parenchyma, but for which surface characteristics and
branching pattern are uncertain. Centrarch xylem composed of tracheids with annular and helical
secondary thickenings surrounded by a zone of compressed thick-walled cells. Outer cortex
consisting of several layers of elongate thick-walled cells. Inner cortex and epidermis not preserved. ’

754

PALAEONTOLOGY, VOLUME 34

The specimen illustrated here (Text-fig. 1a, b) shows that the central region of these strands can
be quite variable. The absence of a well-defined inner zone in Text-figure 1b, taken from the same
axis as the section illustrated in Text-figure 1a, is probably due to poor preservation.

In her original description, Edwards interpreted the cell walls as being almost entirely replaced
by pyrite, so no organic structure or ultrastructural features were described. Kenrick (1988)
collected more material from Brecon Beacons Quarry and Mill Bay and described the wall
structures illustrated here. Recently, Dales was able to confirm the presence of similar wall
ultrastructure in material from the type collection. We propose, therefore, to emend the generic
diagnosis, but our ignorance of the gross morphology of the plant precludes the use of any taxon
above the generic level.

INCERTAE SED1S
Genus sennicaulis Edwards (1981)

Type species: Sennicaulis hippocrepiformis Edwards (1981).

Original diagnosis. See Edwards (1981, p. 225).

Emended diagnosis. Form genus for sterile axes ± circular in cross section containing a prominent
terete xylem strand but for which surface characteristics and branching pattern are unknown.
Xylem composed of elongate elements with predominantly helical thickenings surrounding a terete
central region of smaller cells. Xylem element wall two layered: a thin, continuous, microporate
layer (next to cell lumen) covers a layer with spongy texture. Outer cortex consisting of several layers
of elongate thick-walled cells. Inner cortex and epidermis not preserved.

Sennicaulis hippocrepiformis Edwards (1981)

Plate 1, figs 1^4, 6, 7; Plate 2, figs 1-6; Text-fig. 1a, b

1981 Sennicaulis hippocrepiformis Edwards, p. 225, plates 1-6, figs 1-46.

Holotype. Sections V60352-V60386 ; figured in Edwards (1981) pi. 1, figs 1, 3, 6; pi. 2, figs 8, 12, 13; pi. 3, figs
15 and 16; pi. 4, fig. 24; pi. 5, figs 31, 33-36; pi. 6, figs 37—44.

Paratypes. Sections V60387-V60404 and V60405-V60429 ; figured in Edwards (1981) pi. 2, fig. 9; pi. 3, figs
17-22; pi. 4, figs 23, 25-28; pi. 5, figs 29, 30, 32; pi. 6, fig. 45.

Type locality. Brecon Beacons Quarry, Wales (Friend and Williams 1978), Lower Old Red Sandstone
( = Siegenian), Lower Devonian.

EXPLANATION OF PLATE 1

Figs 1-4. Sennicaulis hippocrepiformis Edwards; transverse sections (T.S.) of S-type conducting cells. 1 and 2,
NMW 90.42G. 1 ; Lower Old Red Sandstone; Mill Bay West, South Wales; outer and inner (respectively)
zone cells, detail of Text-fig. 1a, pyrite cell lumina show high reflectance and limonitic cell walls low
reflectance, thickenings arrowed, x315. 3 and 4, V 60388; Siegenian; Brecon Beacons Quarry, South
Wales, x 675.

Fig. 5. Gosslingia breconensis (Kenrick and Edwards); NMW 87. 19G. 1 ; Siegenian; Brecon Beacons Quarry,
South Wales; T.S. of G-type xylem cell, x 675.

Figs 6 and 7. S. hippocrepiformis ; NMW 90.42G.2; Siegenian; Brecon Beacons Quarry, South Wales;
longitudinal sections (L.S.) of S-type cells. 6, limonite. 7, pyrite in lumen, limonite in wall. Both x 510.

Figs 8 and 9. G. breconensis ; NMW 87. I9G.2; Siegenian; Brecon Beacons Quarry, South Wales; L.S. of G-
type cells preserved in pyrite, x 510.

PLATE 1

7 f

yji j
J*

2 5,1 a £

Kt

btc

5‘«S

-> 4r

v i*

It

<*.J */

i ^

»c >

£Ufc

KENRICK (?/ a/., Sennicaulis , Gosslingia

756

PALAEONTOLOGY, VOLUME 34

Emended diagnosis. Axes up to 6 mm diameter containing a terete xylem strand, 0-9-0-2 mm in
cross-section. Xylem composed of elongate elements, with predominantly helical elements
surrounding a clearly defined central region of narrower elements also helically thickened mixed
with non-ornamented cells. Xylem element wall two layered: a thin continuous microporate layer
(next to cell lumen) overlies a layer with a spongy texture. Micropores are 100 nm (40-200 nm) in
diameter with a density of c. 16 /mrr2. Means and actual xylem measurements are as follows: width
of thickening = 1 1-2 jum (6-3-27-5 //m); inter-thickening distance = 17-4 /mi (5-0-32-5 /mi) ; distance
thickening protrudes into cell lumen = 9-8 /mi (5 0-13-8 //m); lumen diameter at widest point =
34-5 //m ( — 45-0 //m); inter-thickening wall thickness c. 2-5 /tm. Outer cortex composed of up to five
layers of thick-walled (c. 5 //m) cells, more or less isodiametric in cross-section but decreasing in
diameter and length towards the outside; mean cell width = 50 /nn (25-100 /mi) and cell length up
to 450 //m.

DESCRIPTION OF ANATOMY

Major features of the xylem. The xylem is a more or less terete strand of elongate cells and in well
preserved sections consists of two distinct zones. The outer zone is composed of large cells, up to
six cells thick, and completely surrounds an inner zone of smaller cells. The boundary between the
two zones is quite distinct (PI. 1, figs 1 and 2; Text-fig. 1a). Preservation varies along a single axis
and, in some regions, the central zone cells are not preserved. This appears to have resulted in
collapse of the outer zone into the non-preserved central region and consequently a more elliptical
outline to the strand (Text-fig. 1a).

Outer zone xylem cells. Each cell has a single large helical thickening with frequent reversals in the
direction of the helix (see Table 1 for all measurements). The helical nature of the thickening is most
clearly seen in etched transverse sections (PI. 2, fig. 4); the reversal of the helix is most obvious in
fractured lumen casts (PI. 2, fig. 1). Cells are elongate, but no end walls were observed and so their
total lengths are unknown. Variation in cell diameter within the outer zone suggests some tapering,
but may merely indicate the presence of cells of different diameter (PI. 2, fig. 1).

Scanning electron microscopy of fractured or etched sections shows the most ultrastructural
detail (PI. 2). At low magnification fractured sections commonly reveal lumen casts of coarse
textured mineral with deep grooves marking the path of the helical thickening (PI. 2, fig. 1 : two cells
on the left). Partially coalified cell walls are also present, have a smoother texture and usually retain
the helical thickening itself (PI. 2, fig. 1 : two cells on the right). At higher magnification the wall is
seen to consist of at least two layers: a very thin microporate layer overlies the bulk of the wall
which has a spongy texture (PI. 2, figs 2 and 3). The microporate layer is continuous over the entire
inner surface of the cell and covers the thickenings as well as the wall between. Micropores, about

EXPLANATION OF PLATE 2

Scanning electron micrographs of Sennicaulis- type (S-type) xylem cell wall structure.

Figs 1-3, 6. NMW 90.42G.3, stub 183; Lower Old Red Sandstone; Mill Bay West, South Wales. 1, fractured
and partially etched longitudinal section showing lumen casts of coarse textured mineral with deep grooves
marking the path of the helical thickening (two cells on left) and coalified walls where lumen casts have fallen
away with thickenings arrowed (two cells on right), x 600. 2, higher magnification of area in figure 1 showing
structure of wall thickening: microporate wall continuous over thickening and between thickenings, x 3500.
3, details of spongy interior to thickenings, x 12000. 6, details of microporate wall as seen from cell lumen,
x 14000.

Figs 4 and 5. NMW 85. 18G.27/i stub 169; Siegenian; Brecon Beacons Quarry, South Wales; details of
partially demineralized thickenings in outer and inner zone cells at the same magnification, x450. 4, outer
zone cells, simple helical thickenings. 5, inner zone cells, some with helical thickenings.

PLATE 2

KENRICK et al S-type xylem cell wall structure

758

PALAEONTOLOGY, VOLUME 34

100 nm in diameter (ranging between 200 nm to less than 40 nm diameter), penetrate this layer
completely (PI. 2, figs 2, 3, 6). The density of micropores is approximately 16 /mr2. The microporate
layer is very thin; its actual thickness was difficult to measure but it is estimated as approximately
equal to the micropore diameter of 100 nm or less. In all cells this layer is wrinkled and this is
especially noticeable on and around the thickenings themselves (PI. 2, figs 1 and 2). It should be
noted that the small size of the micropores make them difficult to resolve even at high magnification.
At low magnification the inner surface of the coalified wall appears to be smooth (compare PI. 2,
figs 1 and 2).

The remainder and bulk of the wall underlying the microporate layer has a distinctive appearance
that is observed most easily within the helical thickening. It is a coalified layer with a spongy texture
(PI. 2, figs 2 and 3). A thinner layer of similar composition underlies the wall layer between
thickenings. There appear to be no direct connections or channels through this spongy layer
connecting the micropores of adjacent cells, nor any indication of a layer separating adjacent cells.

Inner zone xylem cells. This zone consists of helically thickened cells of similar structure to those in
the outer zone but also contains cells without thickenings (PI. 2, fig. 5). Thickened cells predominate
and the nature of the unthickened cells is uncertain because of the paucity of well-preserved and
uncompressed material within the inner region. All cells are distinctly smaller in diameter than those
in the outer zone (compare PI. 1, figs 1 and 2).

Cellular preservation outside the xylem. There is no cellular preservation outside the xylem in
specimens from Mill Bay. However, specimens from the original Edwards collection of Sennicaulis
hippocrepiformis show a narrow zone of crushed cells immediately adjacent to the xylem and a
further layer of cells close to or at the edge of the axis and separated from the central tissues by a
region where there is no cellular preservation. A detailed analysis of cell wall structure in these
regions has not yet been undertaken.

MINERALOGICAL OBSERVATIONS

All specimens considered here are preserved in pyrite, limonite, or a combination of both minerals.
Limonite is used as a broad term that can encompass virtually any combination of the various iron
oxides and hydroxides that exist naturally. In fossil plants it forms by the oxidation of pyrite and
pseudomorphs pyrite textural morphologies. The timing of this oxidation cannot be defined more
specifically than post pyrite formation.

Observation of polished thick sections in reflected light clearly shows the distribution of mineral
and coalified material in the wall of pyritized specimens (PI. 1, figs 3 and 4), in that pyrite has high
reflectance and the coalified material has low reflectance, although precise identification of coaly
material in such sections remains a problem (Kenrick and Edwards 1988). While other non-opaque
minerals (e.g. silicates) have a similar low reflectance, we interpret the spongy structure in the
composite wall as organically derived, because it is acid resistant, has no discernible crystalline
structure and is similar in composition and structure in all modes of permineralization in all
occurrences. At very high magnifications both microporate and spongy material show a similar
texture, which is not observed elsewhere in the permineralized cells.

Pyrite textures

Considering those in the lumen of the tracheid, pyrite textures can be divided into three
morphologies.

Equant euhedra of pyrite. These have an average grain size of 4-10 pm and are surrounded by a later
overgrowth of pyrite. In transverse section 2-6 euhedra are visible in each cell lumen, while in
longitudinal section many tens of euhedra were seen (Text-fig. 2a, b). Equant euhedra were also
observed within the coalified helical thickenings (Text-fig. 2a).

KENRICK ET AL.\ ULTRASTRUCTURE OF LOWER DEVONIAN PLANT

759

text-fig. 2. Pyrite textures observed within the xylem of longitudinally fractured axes of Sennicaulis
hippocrepiformis from the Brecon Beacons Quarry, South Wales, a, V 60402; sections through pyrite euhedra
in the helical thickenings following light etch, x 1800. b, V 60402; euhedra within lumen: in such a deeply
etched longitudinal section individual euhedra show their distinctive octohedral form, x 2000. c, V 60415;
strongly etched axis with helical thickenings (cf. PI. 2, fig. 3) in which voids left by the removal of pyrite show
little resemblance to the octohedral pyrite morphology, and are thought to be related to the original structure
of the wall, x 8280. d, V 60415; framboid within lumen of tracheid: individual crystallites ^ 1 pm cluster to

form framboids 10-20 pm diameter, x 3450.

Framboidal pyrite. Framboids (Text-fig. 2d) are often surrounded by a later overgrowth or
overgrowths of pyrite and are less common than in analogous areas of lumen in Gosslingia
breconensis (Kenrick and Edwards 1988).

Bladed pyrite. Some pyrite crystals extend from points on or adjacent to the cell wall out into the
cell lumen (PI. 1, figs 1-4). Sometimes pyrite infilling of the cell lumen is incomplete, leaving a space
which may be filled later with another diagenetic mineral, typically calcite (PI. 1, figs 1-4).

The helical thickenings are filled with euhedra of pyrite ( c . 2-4 //m) within what is interpreted as a
coalified ‘spongy’ matrix. The high reflectance of the pyrite and the low reflectance of the coalified
material give the ‘C ’-shaped thickenings a speckled appearance in polished section (PI. 1, fig. 4,
Text-fig. 2b).

In specimens that are partially pyritic and partially limonitic (PI. 1, figs 1, 2, 7; Text-fig. 1a)
particular care must be taken when interpreting the extent of the mineralization of the cell wall.
Limonite formation in fossil plants generally seems to start within the cell wall and may or may not
proceed to the pyrite of the lumen (Kenrick 1988). We suggest that the microcrystalline pyrite and

760

PALAEONTOLOGY, VOLUME 34

organic material within the wall provide a relatively easy route into the specimen for oxidizing
chemicals. In contrast, the large pyrite crystals and paucity of coalified material within the cell
lumen prove a more effective barrier to oxidation. Oxidation produces a mineral with reflectance
similar to that of the coalified material but much lower than that of pyrite. This makes it difficult
to assess the relative abundance and distribution of mineral within the cell wall of partially oxidized
specimens using reflectance alone. Superficial examination may erroneously suggest (PI. 1, figs 1, 2,
7) that the walls are heavily, or almost entirely, coalified.

Specimens in which most or all of the pyrite has been oxidized are better viewed in dark field
illumination (PI. 1, fig. 6; Text-fig. 1b). Although major wall features such as thickenings are easy
to see, lack of contrast makes it difficult to assess the distribution of mineral and coalified material.

TAPHONOMY

The description and interpretation of wall features in Sennicaulis in terms of original cell wall
structures requires an understanding of taphonomic processes and an analysis of the effects of pyrite
formation on cell wall structure. The process of sedimentary pyrite formation in recent sediments
is relatively well understood (Berner 1984), and the way this process affects the appearance of the
plant cell wall has been discussed at length by Kenrick and Edwards (1988) in relation to another
Lower Devonian plant, Gosslingia breconensis .

The pyritization process in fossil plants is seen as one of selective decay by a consortium of
bacteria ending with sulphate reducers and replacement by iron monosulphides which are
eventually converted to pyrite. Decay of the fossil plant is selective because some tissue systems are
less resistant to bacterial decomposition or more attractive in terms of energy yield than others. For
example, woody tissue is persistent in the fossil record because of the decay-resistant lignin
component within the cell wall, whereas thin-walled, non-lignified parenchyma is comparatively
rarely preserved. Easily decomposed tissues may be entirely replaced by pyrite, whereas woody
tissues often retain their original form although the organic material within the cell wall has been
diagenetically altered. This argument has been extended to explain the distribution of pyrite and
coalified material within a cell wall (Kenrick and Edwards 1988). For example, within the xylem cell
walls of Gosslingia breconensis mineral and coalified material are distributed in a characteristic and
consistent pattern. The most plausible explanation for this is that the distribution of coalified
material faithfully reflects the pattern of decay resistant chemicals within the wall (usually identified
as lignin) and the distribution of pyrite reflects the pattern of decay within the wall (i.e. the middle
lamella and areas of cellulose with little or no aromatic components).

Mineral and coalified material are also distributed in a characteristic and consistent pattern in the
cell wall of Sennicaulis , although in a different manner to that found in Gosslingia. Since there is
little or no cellular preservation outside the xylem strand, the perforate coalified layer and the
underlying spongy coalified material must be relatively decay resistant and therefore are interpreted
as having an aromatic non-polysaccharide component. The microporate layer is of uniform
composition and, by comparison to walls of similar outward appearance in the water-conducting
cells of the Hepaticae, may be plasmodesmata derived, although their presence in the region of the
thickenings is difficult to explain (see p. 764). The micropores are not preservation effects caused by
microcrystalline pyrite growth because: (1) pyrite crystals of this size have not been observed in the
cell wall, (2) the pores are not angular in outline as is the cubic crystal structure of the mineral, (3)
the abundant larger pyrite crystals do not damage this layer, and (4) the feature is regular and
uniform in plants from different localities. We also do not consider that the pores were produced
during etching procedures as they are also present in fractured specimens. The wrinkling and
thinness of the microporate layer suggests that, although resilient, its primary role was not
structural. The underlying spongy layer that forms the bulk of the wall both within and between
thickenings is more difficult to interpret. That it is decay resistant is evident; whether there was an
entirely polysaccharide component occupying the cavities in the spongy texture or whether those
cavities were water or air-filled features of the wall is difficult to assess. It seems unlikely that the

KENRICK ET AL. \ ULTRASTRUCTURE OF LOWER DEVONIAN PLANT

761

spongy texture is an artefact caused by pyrite crystal growth within the cell wall because the shape
of the cavities does not mirror the cubic crystal structure of the mineral.

TERMINOLOGY AND COMPARISONS OF CONDUCTING ELEMENTS

We propose the terms ‘ S-type ’ and ‘ G-type ’ for the types of wall thickening and composition in two
distinctive early land plant water-conducting cells (Text-fig. 3).

The ‘S-type’ as exemplified by Sennicaulis is an elongate cell with a large, simple helical
thickening ( sensu Bierhorst 1960). The wall is two layered : a thin, continuous, coalified, microporate
inner layer (next to the cell lumen) overlies a thick coalified spongy layer forming most of the

text-fig 3. Diagrammatic representation of G-type ( Gosslingia ) (a, d) and S-type ( Sennicaulis ) (b, c) cells and
cell walls, a, b, longitudinal sections through G-type and S-type cells at the same magnification, showing in face
view the shape, size and density of the thickenings: scale bar = 10 //m; a, G-type note the discontinuous nature
of the inner wall layer (i.e. that part of the wall facing the lumen); b, S-type note the continuous wall over and
between thickenings, c, d, S-type and G-type cell walls in section at the same magnification : scale bar = 2 /im ;
c, S-type showing two layered wall the bulk of which is made from a spongy material underlain by the very
thin microporate wall (stippled). Stippling indicates size and density of micropores on surface and is not for
purposes of shading; d, G-type showing two layered wall : an inner decay-resistant layer (black in section, white
on the surface) and an outer layer not resistant to decay often also found inside thickenings (shaded with

diagonal lines).

762

PALAEONTOLOGY, VOLUME 34

interior of the thickening and the wall between thickenings. The coalified remains represent
relatively decay resistant material within the cell wall, but their distinctive form and distribution are
unlike that seen in the lignified walls of tracheids. Since we can find no comparable structures in
tracheids of extant plants, except for the helical thickening, we consider it unwise to go further and
interpret the original chemical composition of the decay resistant component as ‘lignin’, although
presumably it is of an aromatic nature.

The ‘G-type’ is typified by Gosslingia breconensis (Kenrick and Edwards 1988) and has indirectly
attached annular or tilted annular thickenings ( sensu Bierhorst 1960). In fossils the indirect
attachments are in the form of perforated sheets of coalified material between adjacent annular bars.
Some elements also have direct attachment between adjacent annular bars ( sensu Bierhorst 1960).
In pyrite permineralizations the wall is clearly two layered: a relatively thick discontinuous coalified
layer overlies one replaced by pyrite. The distribution of pyrite reflects the distribution of decay
resistant chemicals within the cell wall (Kenrick and Edwards 1988) and since this pattern in the
G-type is similar to that in the tracheids of some extant plants, we tentatively interpret the chemical
composition of the decay resistant component as ‘lignin’, and the pyritized one as predominantly
cellulose.

In polished longitudinal section both types of cell appear superficially similar as it is impossible
to distinguish unequivocally between helical and tilted annular thickenings. However, direct
comparison at the same magnification shows several obvious differences in the size, shape and
distribution of thickenings (compare S-type element in Plate 1, figs 6 and 7, with G-type element in
Plate 1, figs 8 and 9; Table 1). The S-type elements have larger thickenings that are more rounded
and widely spaced, even though cell lumen diameters in both types of cell are very similar. Taking
mean values, thickenings in the S-type element are three times as large and six times further apart
than those in G-type elements of comparable lumen diameter. An interesting result is that in cells
of comparable maximum lumen diameter, the effective diameter for fluid conductance (Jeje and
Zimmermann 1979) in the S-type cell is just over half what it is in the G-type (Table 1). This is most
noticeable in transverse sections of similar magnification viewed side by side (compare S-type
element in Plate 1, fig. 4 with G-type element in Plate 1, fig. 5; Table 1).

Table 1 . Comparative measurements of cell wall dimensions for S-type and G-type cells. For fuller explanation
of 3, £, t, W, D and TW see Kenrick and Edwards (1988). 3 = width of thickening; I = interthickening
distance; t = distance thickening protrudes into cell lumen; W = maximum lumen diameter; D = effective
diameter for fluid conductance; TW = thickness of wall. MBW = Mill Bay West, BBQ = Brecon Beacon
Quarry, N = number of measurements.

Element

type

Locality

3

I

t

W

D

TW

S-type

MBW

mean /*m

11-2

17-4

9-8

34-5

14-9

2-5

max. //m

27-5

32-5

13 8

450

17-4

—

min. jum

6-3

50

5-0

—

—

—

N

74

77

52

11

—

—

G-type

BBQ

mean //m

3-3

2-8

3-6

29-6

22-4

—

max. //m

4-5

5-3

61

500

37-8

3-5

min. /mi

1-5

10

1-5

—

—

1-5

N

36

33

16

23

—

—

Only helical thickenings were observed in the S-type element (clearly seen in scanning electron
micrographs of etched sections, PI. 2, fig. 4). The frequent reversal in direction of the helix in the
S-type cell (PI. 2, fig. 1) has also not been seen in the G-type. Both annular and helical thickenings
clearly occur in the G-type element along with frequent direct connections between adjacent annular
bars.

KENRICK ET A L. ULTRASTRUCTURE OF LOWER DEVONIAN PLANT

763

Ultrastructural differences are still more marked (Text-fig. 3). Both types of cell have a wall of
at least two layers although whether or not homologous is debatable. In the S-type cell, the inner
part of the wall is preserved as a very thin, coalified, microporate layer that is continuous over the
entire inner surface of the cell. In the G-type cell, the inner part of the wall is relatively thick and
continuous over the inner surface of the cell except for relatively large holes of variable size between
thickenings (see Text-fig. 3). Unlike the micropores in the S-type element, these holes never occur
on the thickenings themselves. The micropores of the S-type element are about 100 nni diameter
(ranging from 200 nm to less than 40 nm) with a density of about 16 pm~2, whereas the holes in the
G-type measure about 2-3 /mi diameter (ranging from 4 /mi to about 0-5 /mi).

The second layer or outer part of the wall in the G-type element which is entirely pyritized in
Gosslingia is approximately the same thickness as the inner. In contrast, the outer layer of the S-
type element forms the bulk of the wall and although partially pyritized contains a significant
proportion of structured coalified material (Text-fig. 2c).

DISTRIBUTION OF S- AND G-TYPES

S-type

There is only one other fossil plant in which ultrastructural features comparable to those in the S-
type element have been described. Hueber (1982) described helically strengthened tubes in axes
assignable to Taeniocrada dubia. He observed three components in the structure of the wall: ‘an
outer, thin, fibrillar layer; sponge-textured helical thickenings; and a microporate layer that lines
the lumen of the tube’. Apart from the outer, thin, fibrillar layer which has not been observed in
S-type elements from Sennicaulis the structure of the wall would appear to be identical although no
photographs of Hueber’s material have yet been published. The conducting strand is described as
centrarch in T. dubia but in Sennicaulis from the Mill Bay locality it is clearly divided into two
distinct zones, at least over part of the axis. The inner zone is often not well preserved and in many
sections is totally missing. When this occurs, the outer zone cells collapse inward and this, combined
with the presence of crushed cells in the centre, may give the impression of a centrarch strand.

Few ultrastructural studies have been performed on early land plant water-conducting cells so it
is impossible to identify unequivocally other plants with similar wall structure. However, the shape
and size of the helical thickenings in the S-type cell are strikingly different from those in the G-type
(see Table 1 and p. 761) and may prove to be characteristic. The recent reinvestigation of
Stockmansella langii (formerly Taeniocrada langii) illustrates cells with certain similarities to the S-
type (Fairon-Demaret 1985, 1986), although organic material in the cell wall could not be
recognized. Fairon-Demaret (1985) noted that the spacing and tilt of the thickenings, seen as
grooves on goethite casts of the cell lumen, indicate a simple helical thickening with frequent
inversions in the tilt, and that in this feature these cells are similar to those of Rhynia gwvnne-
vaughanii. These features, together with the dimensions of the thickening, point to a structure
similar to that in the S-type cell.

Edwards (1981) and Edwards, D. S. (1986) have both commented on the similarities between the
helical elements of Sennicaulis hippocrepiformis and Rhynia gwynne-vaughanii, including the size and
shape of thickenings, originally described as annular in R. gwynne-vaughanii (Kidston and Lang
1917), and the frequent reversal of the helix within cells, all typical features of the S-type element.
Edwards D. S. (1986) was unable to find an organic wall in etched tracheids of R. gwynne-vaughanii ,
so the wall ultrastructure is unknown. The xylem strand of R. gwynne-vaughanii is also much smaller
than that illustrated here and is not divided into distinct zones.

There are no known examples of S-type conducting cells among extant vascular plants. Helical
elements are common in the protoxylem, but lignin never forms in a continuous sheet on the inner
surface of the cell, nor are there micropores comparable to those illustrated here.

Although the microporate layer is unknown in tracheids, a wall of similar appearance, but lacking
helical thickenings and, presumably, decay resistant aromatics, is sometimes found in bryophyte

764

PALAEONTOLOGY, VOLUME 34

gametophytes (Hebant 1979). In certain Hepaticae, as well as the enigmatic genus Takakia , the
presumed cellulose walls of water-conducting cells have numerous plasmodesmata-derived
perforations of similar size and density to those found in Sennicaulis. (Compare Plate 2, figures 2
and 6 with Hebant’s (1979) figure 2.) The whole wall thickness of these cells is similar to the
thickness of the microporate layer in the S-type element. Such a microporate wall is not found in
the hydroids of mosses (Hebant 1979).

G-type

A review of the distribution of G-type elements was given by Kenrick and Edwards (1988). Earlier
reports of scalariform or reticulate pitting in zosterophylls have been shown to be incorrect in all
instances where scanning election microscopy has been undertaken. Instead, G-type elements with
annular or helical thickenings characterize the metaxylem of zosterophylls, Barinophytales and a
group of fossil plants ( Asteroxylon , Baragwanathia and Drepanophycus) similar in vegetative
characteristics to modern herbaceous lycopods. With one exception ( Leclercqia , Grierson 1976), the
protoxylem elements of those and other important Devonian taxa have not received the same
detailed consideration as the metaxylem and are described as helical or annular.

GENERAL DISCUSSION

In our studies of Gosslingia tracheids we were able to interpret their wall structure in terms of that
in extant tracheids, although examples are rare. In addition it is possible that the perforate sheet in
the G-type element is homologous with the pits of trimerophytes where there are perforated sheets
of secondary wall material within the pit chamber (Hartman and Banks 1980) and possibly even
with strands of secondary wall material in the pits of Carboniferous lycopods.

In contrast to all these taxa, the gross morphology of the plants that produced S-type cells is
poorly understood and the structure cannot be directly related to other presumed water-conducting
cells. Compression fossils of Sennicaulis hippocrepiformis from Brecon Beacons Quarry are
unknown but permineralizations suggest that the axes were more or less terete in cross section. The
Mill Bay West anatomy was obtained from compressions of long, ribbon-like, sterile axes. The
Gaspe (Quebec) specimens of Taeniocrada dubia , which show a similar cell-wall structure, were
originally described by Dawson as the rhizomata of Psilophyton princeps (Hueber 1967, 1982).
Hueber (1982) described the plant as ribbon-like, with a thin cuticle and stomata. Reproductive
parts were not found on specimens from which he isolated anatomy, although terminal sporangial
trusses have been described from German material (Hoeg 1967; Krausel and Weyland 1930).

Gross morphology is better understood in Stockmansella langii (Fairon-Demaret 1985, 1986) and
Rhynia gwynne-vaughanii (Edwards D. S. 1986), taxa united by sporangial abscission. Ultra-
structural wall features have not been described for either plant and are currently under
investigation.

The novel wall structure in the water-conducting cells of Sennicaulis and Taeniocrada cannot
easily be compared with that in extant tracheids nor indeed with hydroids in mosses. The cell is
superficially similar to an early formed protoxylem tracheid in that it has a helical thickening, but
the wall layers are unlike those in the tracheids of extant plants and the more tracheid-like cells of
some early land plant fossils. Terms such as primary and secondary wall are difficult to apply and
have been avoided. The micropores are probably developmentally similar to the plasmodesmata
derived pores in the water-conducting cells of some hepatics and functionally similar, at least in the
wall layer between thickenings, but their presence in the layer over the latter is more difficult to
interpret developmentally. The cell might be considered plesiomorphic in this respect.

Reinvestigation of the cell walls of the xylem of Rhynia major , now Aglaophyton major , showed
that there are no internal thickenings and that the plant does not have tracheids (D. S. Edwards
1986). This was a surprising find in one of the two species of a genus that has come to represent the
archetypal early vascular plant, and demonstrates that branched sporophytes do not imply vascular
plant status. Our investigation shows that helical thickenings in water-conducting cells do not imply

KENRICK ET AL.\ ULTRASTRUCTURE OF LOWER DEVONIAN PLANT

765

a tracheid-like wall structure. In the case of Sennicaulis, ultrastructural studies were required to
demonstrate the novel organization. Many early land plant fossils of the rhyniophyte grade are
included in Tracheophyta because of helical or annular thickenings in the xylem cells. The
systematic position of these fossils must be re-evaluated because in most cases detailed studies of
water-conducting cells have not been undertaken. It is natural to go further and question whether
‘tracheid’ can be used as a synapomorphy for vascular plants and to speculate on the relationships
of plants with S-type cells. For the moment we do not wish to go beyond the recognition of a level
of organization that would seem to fall between that of the extant bryophytes and vascular plants.
More information is required on the morphology of plants with S-type cells and an exhaustive
phylogenetic analysis is necessary before explicit statements of relationships can be made.

Acknowledgements. Paul Kenrick was financed by NERC research grant No GR3/5069 and a grant from the
Field Museum of Natural History, and Robin Dales by a NERC postgraduate research studentship. All
sources of funding are gratefully acknowledged. P. K. acknowledges the helpful criticism of Peter Crane and
Andrew Drinnan.

REFERENCES

allen, j. r. l. 1974. The Devonian rocks of Wales and the Welsh Borderland. 47-84. In owen, t. r. (ed.).

The Upper Palaeozoic and post- Palaeozoic rocks of Wales. University of Wales Press, Cardiff.
berner, r. a. 1984. Sedimentary pyrite formation: an update. Geochimica et Cosmochimica Acta , 48, 605-615.
bierhorst, D. w. 1960. Observations on tracheary elements. Phytomorphology , 10, 249-305.
biggs, d. l. and rocken, R. J. 1983. Paragenesis of iron sulfide as a function of coal rank. Tenth International
Congress on Carboniferous Stratigraphy and Geology, 3, 145-164.
edwards, d. 1981. Studies on Lower Devonian petrifactions from Britain 2. Sennicaulis, a new form genus for
sterile axes based on pyrite and limonite petrifactions from the Senni Beds. Review of Palaeobotany and
Palynology, 32, 207-226.

— and kenrick, p. 1986. A new zosterophyll from the Lower Devonian of Wales. Botanical Journal of the
Linnean Society of London, 92, 269-283.

Edwards, D. s. 1986. Aglaophyton major, a non-vascular land-plant from the Devonian Rhynie Chert.

Botanical Journal of the Linnean Society of London, 93, 173-204.
fairon-demaret, m. 1985. Les planles fossiles de l’Emsien du Sart Tilman, Belgique. I. Stockmansia langii
(Stockmans) comb. nov. Review of Palaeobotany and Palynology, 44, 243-260.

— 1986. Stockmansella , a new name for Stockmansia Fairon-Demaret (fossil). Taxon, 35, 334.

friend, p. F. and williams, b. p. j. (eds). 1978. A field guide to selected outcrop areas of the Devonian of
Scotland, the Welsh Borderland and South Wales. International Symposium on the Devonian System
(P.A.D.S. 78). The Palaeontological Association, London, 106 pp.

Grierson, j. d. 1976. Leclerccjia complexa (Lycopsida, Middle Devonian): its anatomy, and the interpretation
of pyrite petrifactions. American Journal of Botany, 63, 1184-1202.
hartman, c. m. and banks, h. p. 1980. Pitting in Psilophyton dawsonii , an early Devonian trimerophyte.
American Journal of Botany, 67, 400- 412.

hebant, c. 1979. Conducting tissues in bryophyte systematics. 365-383. In clarke, g. c. s. and duckett, j.
G. (eds). Bryophyte systematics : Systematics Association special volume 14. Academic Press, London and
New York.

hoeg, o. a. 1967. Psilophyta. 191-352. In boureau, e. (ed.). Traite de Paleobotanique. II. Masson et Cie, Paris.
hueber, f. m. 1967. Psilophyton-. the genus and the concept. 815-822. In Oswald, d. h. (ed.). International
symposium on the Devonian System 2. Alberta Society of Petroleum Geologists, Calgary.

— 1982. Taeniocrada dubia Kr. and W. : its conducting strand of helically strengthened tubes. Botanical
Society of America, Miscellaneous Series (Abstract), 162, 58-59.

jeje, a. a. and zimmermann, m. h. 1979. Resistance to water flow in xylem vessels. Journal of Experimental
Botany, 30, 817-827.

kenrick, p. 1988. Studies on Lower Devonian plants from South Wales. Unpublished Ph D. thesis. University
of Wales, Cardiff.

766

PALAEONTOLOGY, VOLUME 34

kenrick, p. and edwards, d. 1988. The anatomy of Lower Devonian Gosslingia breconensis Heard based on
pyritized axes, with some comments on the permineralization process. Botanical Journal of the Linnean
Society , 97, 95-123.

kidston, r. and lang, w. h. 1917. On Old Red Sandstone plants showing structure, from the Rhynie Chert
Bed, Aberdeenshire. Part I. Rhynia gwynne-vaughani Kidston and Lang. Transactions of the Royal Society
of Edinburgh , 51, 761-784.

krausel, r. and weyland, h. 1930. Die Flora des deutschen Unterdevons. Abhandlungen der Preussischen
geologischen Landesanstalt , Berlin , 131, 1-92.

richardson, j. b., streel, M., hassan, a. and steemans, Ph. 1982. A new spore assemblage to correlate between
the Breconian (British Isles) and the Gedinnian (Belgium). Annales de la Societe Geologique de Belgique , 105,
135-143.

thomas, R. G. 1978. The stratigraphy, palynology and sedimentology of the Lower Old Red Sandstone
Cosheston Group, S. W. Dyfed, Wales. Unpublished Ph.D. thesis. University of Bristol.

p. KENRICK

Department of Geology
Field Museum of Natural History
Roosevelt Road at Lake Shore Drive
Chicago, Illinois 60605, USA

D. EDWARDS and R. C. DALES
Department of Geology

Typescript received 20 September 1990 University of Wales College of Cardiff

Revised typescript received 28 November 1990 P.O. Box 914, Cardiff CF1 3YE, UK

THE PANGAEA DICYNODONT RECHNISAURUS
AND THE COMPARATIVE BIOSTRATIGRAPHY OF
TRIASSIC DICYNODONT FAUNAS

by C. BARRY COX

Abstract. A dicynodont from the Triassic Manda Beds of East Africa is found to belong to the kannemeyeriid
genus Rechnisaurus , first described from India, with which the genus Shaanbeikannemeyeria from China and
Mongolia is congeneric. The divergent skull modifications of the Families Lystrosauridae and Kannemeyeriidae
from the primitive Permian dicynodont skull morphology are explained in functional terms, both being
adaptations to increase the length of the jaw adductor muscles. The kannemeyeriid pattern is similar to that
of the ceratopsian dinosaurs, while that of the lystrosaurids seems also to be the result of feeding adaptations
rather than suggesting semi-aquatic life. The relative ages of dicynodont-bearing Triassic faunas are reviewed
in the light of recent changes in taxonomy and biostratigraphy.

The Triassic Manda Beds of Tanzania were discovered by Stockley in 1930 (Stockley 1932), and
his fossils were described by Haughton (1932). Further collections were made by both Parrington
and Nowack in the 1930s. Study of this material showed that the fauna contained a large
amphibian, the rhynchosaur Stenaulorhynchus, thecodontians, cynodonts, and dicynodonts
(Haughton 1932;Huene 1938«, 19386, 1942, 1950; Parrington 1946; Crompton 1955; Cruickshank
1965, 1967, 1986; Cox and Li 1983).

The British Museum (Natural History) and London University Joint Palaeontological Expedition
to Northern Rhodesia and Tanganyika spent six weeks collecting in the Manda Formation in 1963
(Attridge et al. 1964), and 184 specimens of dicynodonts, cynodonts, rhynchosaurs, and
thecodontians were collected from 27 localities. These localities are shown on the accompanying map
(Text-fig. 1 ), which is based on a mosaic of aerial photographs and on survey work by the late Dr
W. W. Bishop (then Curator of the Uganda Museum). Local villagers provided the names of the
streams.

Bishop (1968) described the geology of the area. The Manda Formation is 2000 m thick, being
a series of purple to chocolate-coloured mudstones interbedded with grey to whitish sandstones and
grits of variable thickness and persistence. Though reptile bones were found from 150 m above the
base of the Formation to 60 m below its top, 65% of the specimens collected were found in two
levels, 1 180 m and 1310m above its base. Below the Manda Formation lies the 700 m thick Kingori
Sandstone Formation, from which Cruickshank (1986) reports the dicynodont Kannemeyeria.
Below the Kingori Sandstone lies the 275 m thick Kawinga Formation, and the base of the exposed
sequence in the area is made up of the 535 m thick Ruhuhu Formation, in which the 1963
Expedition found a few specimens of endothiodont dicynodonts, some of which bore canine teeth
lateral to the main maxillary tooth-row.

The dicynodont Angonisaurus from locality 12 in the Manda Formation has already been
described (Cox and Li 1983). The present paper is concerned with a large dicynodont from this
formation, and with its implications for the structure, classification, and comparative stratigraphy
of Triassic dicynodonts.

[Palaeontology, Vol. 34, Part 4, 1991, pp. 767-784.|

© The Palaeontological Association

768

PALAEONTOLOGY, VOLUME 34

text-fig. 1. Map of the main collecting areas of the 1963 Expedition in the Ruhuhu Valley, south-western
Tanzania. Localities (numbered) to the south lie in the Triassic Manda Formation, those to the north-west lie
in the Late Permian Ruhuhu Formation. Curving lines indicate the locations of resistant levels in the Manda
Formation, which caused low ridges in the landscape. Dashed lines indicate footpaths.

SYSTEMATIC PALAEONTOLOGY

Suborder anomodontia Owen, 1859
Infraorder dicynodontia Owen, 1859
Family kannemeyeriidae Huene, 1948
Genus rechnisaurus Chowdhury, 1970

1980 Shaanbeikannemeyeria Cheng, p. 139, text-figs 29-30.

Diagnosis. Large dicynodont : skull 400-650 mm long. Skull tapers anteriorly from wide posterior
squamosal region to rather narrow, blunt snout. Median ridge on premaxillae and nasals.
Maxillary tusk emerges from inner side of caniniform process; no other teeth. Caniniform process
of maxilla antero-ventrally directed, its posterior margin flared laterally, rugose. Intertemporal bar
at sharp angle to interorbital bar, rather narrow but transversely concave, lacking median crest.
Wide lateral wings of squamosal directed sharply antero-ventrally, almost parallel to zygomatic
arch. Quadrate lies far forward. Very short median pterygoid meeting behind short interpterygoid
vacuity. Processus cultriformis of parasphenoid projects strongly antero-dorsally.

Type species. Rechnisaurus cristarhynchus Chowdhury, 1970.

Rechnisaurus cristarhynchus Chowdhury, 1970
Text-figs 2 and 3

Material. The specimen described here, R1 1955 in the Natural History Museum, field no. U17/1, was collected
from locality 17 (Text-fig. 1), nearly 14 km from the Lutumba Mission Station, Songea District, south-western

COX: TRIASSIC DICYNODONT RECHNISAU RUS

769

A

B

text-fig. 2. Rechnisaurus cristarhynchus ; BM(NH) R1 1955, x 0 2. a, dorsal view, at right angles to interorbital
area, b, occipital view. For abbreviations see Text-figure 3.

770

PALAEONTOLOGY, VOLUME 34

Tanzania, from the lower of the two main fossiliferous levels of the Manda Formation. Much of the skull was
found lying on its ventral surface on the top of a low sandstone hillock. Many fragments of the dorsal surface
had already been weathered off and had rolled down the sides of the hillock. They were found scattered over
an area of a few hundred square metres.

Description. The whole of the snout anterior to the caniniform processes of the maxillae is missing, as is the
dorso-median part of the snout as far back as the level of the front of the orbits, and the whole of the dorso-
median part of the skull posterior to the pineal foramen and dorsal to the foramen magnum.

In dorsal view (Text-fig. 2a), the skull is 570 mm broad across the occiput, and tapers anteriorly. The
preserved part of the skull is 590 mm long, but the original length was probably about 650 mm. The interorbital
region is broad. There is a depression in the postero-median region of the frontals, behind which the bone rises
steeply to form the beginning of a high intertemporal bar. The lateral walls of the depression are formed by
ridges which mark the junction of the postorbital and frontal, and which curve antero-laterally, dying out
towards the root of the post-orbital bar. There is no trace of a postfrontal bone. The preparietal forms the
lateral and anterior walls of the pineal foramen, and tapers anteriorly to a sharp point projecting between the
frontals.

The lateral walls of the pineal foramen diverge posteriorly, suggesting that these surfaces extended
posteriorly as the inner surfaces of a pair of ridges along the intertemporal bar, which would then have had
a median groove. The width of the posterior end of the postorbital bone, which extends laterally as a wing from
the more median bones, suggests that the anterior part, at least, of the intertemporal bar, was quite wide. The
ventral surface of the posterior ends of the postorbital bones are at a sharp angle to the interorbital region,
indicating that the intertemporal bar rose abruptly postero-dorsally.

The dorso-medial region of the occiput is missing (Text-fig. 2b). The ventro-lateral corner of the squamosal
extends a considerable distance lateral to the quadrate condyles. As is usually the case in large vertebrates, the
foramen magnum is small in comparison with the size of the occipital condyle.

The nasal region of the snout has been crushed ventrally, and is not shown in lateral view (Text-fig. 3a). The
suspensorium extends at a great angle antero- ventrally and runs smoothly into the zygomatic arch. The lower
end of the post-orbital bar therefore lies not far antero-dorsally from the quadrate condyle. There is a deep
pit between the main body of the quadrate and the quadratojugal, above the lateral quadrate condyle, which
may have been the attachment area for a ligament restricting the anterior movement of the lower jaw. The outer
surface of the posterior end of the zygomatic arch is concave dorso-ventrally. The ventral edge of the zygomatic
arch below the orbit is quite sharp. The squamosal extends anteriorly to overlie the outer surface of the maxilla
as far forwards as the posterior edge of the base of the caniniform process.

The secondary palate (Text-fig. 3b, c) is long and narrow, with almost vertical side walls. It is very deep, its
surface lying 125 mm above the ventral end of the caniniform processes. The moderately-sized tusk emerges
from the inner surface of the caniniform process, not from its apex. A rugose area extends dorsally from the
root of the tusk on the inner side of the maxilla. The posterior edge of the caniniform process extends postero-
laterally as a massive, rugose flange. Further posteriorly, a large foramen between the jugal, the maxilla, and
the pterygoid runs dorsally into the maxillary antrum surrounding the base of the tusk. The mid-region of the
palate, posterior to this point and anterior to the region around the interpterygoid vacuity, is missing.

The quadrate condyles (Text-fig. 3e) are orientated in an antero-medial direction; this area is not distorted,
and this feature raises questions as to how the lower jaw functioned, since it is normally presumed to have had
a simple antero-posterior sliding/rotating movement (e.g. Crompton and Hotton 1967). The interpterygoid
vacuity (or fenestra medio-palatinalis, see Cox 1968, p. 13) and basicranial axis are well-preserved. The whole
basicranial axis is greatly shortened. There is therefore only a very short median contact of the pterygoids
behind the fenestra; this region is raised into a transverse, ventrally-projecting ridge, which lies only just
anterior to the basisphenoid tubera. The internal carotid foramina can be seen piercing the parasphenoid. A

text-fig. 3. Rechnisaurus cristarhynchus ; BM(NH) R1 1955, x 02. A, lateral view, b, anterior view of snout, c,
palatal view of snout. D, ventral view of basis cranii and quadrate region, e, lateral view of braincase.
Abbreviations: BO, basioccipital; BSP, basisphenoid; F, frontal; IP, interparietal J, jugal; L, lacrimal; MX,
maxilla; PAL, palatine; PF, prefrontal; PMX, premaxilla; PO, postorbital; PSP, parasphenoid; PT,
pterygoid; Q, quadrate; QJ, quadrato-jugal ; SQ, squamosal; fen.ov., fenestra ovalis; i.c., internal carotid
foramina; pr.cu., processus cultriformis of parasphenoid; sel.tu., sella turcica; Vid., Vidian canal; VII,

foramen for facial nerve.

COX: TRIASSIC DICYNODONT RECHNISAURUS

771

772

PALAEONTOLOGY, VOLUME 34

little antero-dorsal to these, a pair of foramina on either side are the openings for the anterior branch of the
internal carotid artery and for the palatal branch of the Vllth cranial nerve; they open from within a short
recess, which is therefore a truncated Vidian canal (Text-fig. 3e, f, i.c.). The foramen through which the main
trunk of the Vllth cranial nerve exits from the braincase can be seen anterior to the fenestra ovalis (Text-fig.
3f, VII).

THE TAXONOMIC POSITION OF THE MANDA SPECIMEN

Though the skull is incomplete, it shows a number of features that suggest its taxonomic identity.
The acute angle between the intertemporal bar and the interorbital region, the moderately elongated
snout with a median nasal ridge, and the antero-ventrally directed suspensorium, are all diagnostic
features of the kannemeyeriid dicynodonts, as slightly differently defined by Cox and Li (1983) and
King (1988). Comparisons between the Manda specimen and other kannemeyeriids show that the
most obvious similarities are with Shaanbeikannemeyeria Cheng, 1980, from the Er-Ma-Ying
Formation of China. Both have wedge-shaped skulls, strong rugose maxillae, a narrow
intertemporal bar that nevertheless has a slight median longitudinal groove, and a very short
interpterygoid vacuity and median pterygoid neck between that vacuity and the parasphenoid.
However, further comparison shows that all these features are present also in Rechnisaurus
(Chowdhury 1970; Bandyopadhyay 1989) from the Yerrapalli Formation of India, and detailed
comparison of the drawings and descriptions of Cheng and Chowdhury leave no doubt that the two
genera are identical. Cheng did not compare his Chinese dicynodont with Chowdhury’s 1970 genus,
but that was possibly because Keyser and Cruickshank (1979) had erroneously concluded that
Rechnisaurus was a junior synonym of Kannemeyeria (Bandyopadhyay 1989).

The Chinese species must therefore be known as Rechnisaurus xilougouensis (Cheng), and it can
be distinguished from the Indian species, R. cristarhynchus, by the fact that the premaxillae and
frontals project medially into the frontals. A third species, Shaanbeikannemeyeria buerdongia from
the lower Er-Ma-Ying Formation of Inner Mongolia, was described by Li (1980), and must now be
renamed Rechnisaurus buerdongia (Li); it differs from the other species in having smaller maxillary
processes.

The incompleteness of the African specimen makes it difficult to compare with the other species.
However, it is identical to the Indian specimen in the unusual feature that the squamosal extends
far forwards to reach the base of the caniniform process of the maxilla. The only differences between
the two specimens are that the canine tusk is smaller than in the Indian specimen and that it has
a strong ridge around the posterior rim of the interpterygoid vacuity, while the Indian species shows
no ridge. The Manda specimen is accordingly regarded only as an additional specimen of
Rechnisaurus cristarhynchus Chowdhury. Accordingly, the missing postero-dorsal part of the skull
and the dorsal outline of the snout in Text-figures 2a, b and 3a have been restored as in R.
cristarhynchus.

When it was collected, two features of the Manda specimen recalled advanced kannemeyeriids
such as Placerias : its short interpterygoid vacuity, and the fact that its tusk emerges from the inner
side of the caniniform process rather than from its apex. This is probably the basis for Keyser’s
(1980, p. 62) statement that ‘A species that resembles Ischigualastia from the Santa Maria and
Ischigualasto Formations of South America has been found in the Manda Beds of Tanzania.’ (In
fact, Ischigualastia is not known from the Santa Maria Formation.)

Another dicynodont from the Manda Formation, Sangusaurus (Cox 1969) has been described by
Cruickshank (1986). It shares many features with the holotype of R. cristarhynchus , including a
small boss just posterior to the pineal foramen, a feature otherwise unknown in Triassic
dicynodonts. It is possible that these two genera are congeneric, despite the fact that Sangusaurus
is tuskless and has a more pointed premaxilla (Table 1).

Rechnisaurus was thus widely distributed in Pangaea, being found in two localities in Gondwana,
and two localities in Asia, some 16,000 km away around the western end of the Tethys Sea. Both
these areas lie in the Triassic temperate regions and, as Parrish, Parrish, and Ziegler (1986) point
out, no Lower or Middle Triassic therapsids have been found between palaeolatitudes 35° N and

COX: TRIASSIC DICYNODONT RECHNISAURUS

773

35° S, in the tropical regions. The absence of Rechnisaurus from this intermediate belt of latitudes
is thus merely part of a more general pattern. The presence of species of that genus both north and
south of this tropical band shows that this therapsid, at least, was able to traverse this area. The
absence of fossils therefore suggests that conditions there were unsuitable for fossilization, rather
than that therapsids did not inhabit these regions. A similar pattern of absence of Triassic tetrapod
fossils in general from an equatorial band of latitudes has already been documented (Cox 1973a).

table 1. Cranial features of kannemeyeriid dicynodonts.

IT /IO
at

angle

Quadrate

forwards

Strong

maxilla

Tusks

Midnasal

ridge

Blunt

snout

Wedge-

shaped

skull

Kannemeyeria

+

—

+

+

+

+

+

Rechnisaurus

+

+

+

+

+

+

+

U ralokanneme veria

+

+

+

+

+

+

+

Rahidosaurus

+

+

?

+

+

—

?

Sangusaurus

+

+

+

+

+

7

+

Wadiasaurus

+

+

±

±

+

+

—

Ischigualastia

+

+

-

-

-

—

+

Jachaleria

+

+ *

-

—

—

—

+

Placerias

+

+

+

+

-

—

+

‘ Moghreheria '

+

+

+

+

?

—

+

+ = character present; — = character absent; ± = character present in some specimens, absent in others;
? = character unknown; * = see text; IT = intertemporal bar; IO = interorbital bar.

The relationships of Rechnisaurus. Rechnisaurus shows a number of features that appear to a greater
or lesser extent in several other Triassic dicynodonts (Table 1 ) : pronounced angulation between the
intertemporal bar and the frontal region, pointed out by King (1988); an antero-ventrally directed
suspensorium ; a strong maxilla, usually bearing a tusk and with a rugose antero- ventral flange; a
median nasal ridge; a skull that is wedge-shaped in dorsal view, and a bluntly pointed snout. The
Family Kannemeyeriidae of Cox and Li (1983) begins with the Early Triassic Kannemeyeria , in
which the suspensorium is directed ventrally. It continues with a number of Middle Triassic genera
( Rechnisaurus , Uralokannemeyeria , Rahidosaurus , Sangusaurus , Wadiasaurus), in which the
suspensorium is directed more antero-ventrally. In all these genera the snout is bluntly pointed and
bears a median dorsal nasal ridge. In Wadiasaurus and in the Late Triassic genera ( Ischigualastia ,
Jachaleria , Placerias, ‘ Moghreheria'). the snout has become pointed and lacks this ridge. In the
tuskless members of the group ( Jachaleria , Ischigualastia) the maxilla is, not surprisingly, less well-
developed than in the remaining genera - a correlation between these two features is strongly
supported by the fact that, in Wadiasaurus , some specimens (presumably male) are tusked and have
a strong maxilla, while others (presumably female) are tuskless and have a poorly developed maxilla
(Bandyopadhyay 1988). (The poorly preserved skull of Moghreheria Dutuit, 1980 seems so similar
to that of Placerias Camp and Welles, 1956 that these two genera may well prove to be identical,
while the few fragments of Azarifeneria Dutuit, 1989 are very similar to the corresponding portions
of Ischigualastia Cox 1965.)

King has recently (1988) provided a new classification of Triassic dicynodonts, which differs in
several respects from that of Cox and Li (1983). As she rightly comments (King 1988, p. 69), "Until
a complete revision of the Middle and Late Triassic material, using first-hand information, is
undertaken, no absolute consensus of opinion on their relationships can be expected.’ A difference
between the two classifications is that King separates Ischigualastia and Placerias from the
kannemeyeriids, and links them more closely with the Zambiasaurus-Stahleckeria group. This is

774

PALAEONTOLOGY, VOLUME 34

despite the great difference between the shapes of the occipital plates and snouts of these two groups
(Cox 1965), and despite the fact that it is then necessary to suggest that a long, high intertemporal
bar at an angle to the frontal region evolved independently in the Kannemeyeria group and in the
Placerias group.

The feature that unites the Placerias group to the Stahleckeria group (and also to the
Sinokannemeyeria group) in King’s (1988) cladogram is that the parietals are covered incompletely
by the postorbitals. She here refers to a condition in which the postorbital did not extend as far
posteriorly as the parietal-interparietal suture, so that the posterior parts of the parietals are
exposed in the skull roof. King (1990, p. 225) defines the character as merely ‘Postorbitals do not
extend far posteriorly’. This is indeed true of Ischigualastia and Placerias. But in Stahleckeria the
posterior extent of the postorbital is uncertain: Huene reconstructs it as extending posteriorly to
meet the squamosal in his skull no. 2 (Huene 1935, pi. 5), but Camp’s (1956, fig. 45) drawings of
a cast of the same specimen show a short postorbital. In her diagnosis of the genus Stahleckeria
King (1988, pp. 107-108), makes no statement on this character. In the other stahleckeriid genus,
Zambiasaurus , the postorbital appears to have extended back as far as the junction between the
parietal and the interparietal (Cox 1969). The posterior parts of the parietal are indeed exposed in
all of these genera, but this would be true irrespective of the posterior extent of the postorbital,
because the latter bone is lateral to the wide parietal, rather than dorsal to it.

According to King (1988, 1990), the features of the postorbital and parietal mentioned above are
found in the Sinokannemeyeria group (which includes also Parakannemeyeria , Uralokannemeyeria ,
Dinodontosaurus , and Rhadiodromus ), as well as in the Placerias and Stahleckeria groups. However,
though the postorbital does not extend far posteriorly in Dinodontosaurus, it does so in
Sinokannemeyeria and Parakannemeyeria (Sun 1963) and Uralokannemeyeria (Danilov 1971); the
situation in Rhadiodromus (Kalandadze 1970) is uncertain. The posterior part of the parietal is
widely exposed in all these genera except Parakannemeyeria, but this ‘genus’ may merely consist of
skulls of Sinokannemeyeria that have been compressed laterally, those of the latter genus having
instead been compressed dorso-ventrally. Sinokannemeyeria has a wide snout, with canine teeth set
wide apart and projecting backwards, a wide intertemporal bar between the postorbitals, and a low,
wide occiput. Parakannemeyeria has a narrow snout, with canine teeth set close together, a narrow
intertemporal bar covered by the postorbital bones, and a high, narrow occiput. These are exactly
the differences that might be expected if the two sets of skulls had been affected by the two different
types of plastic deformation that would have resulted from the different orientations of deposition.
The two genera are both from the Er-Ma-Ying Formation of China (Sun 1963, 19896).

The exposure of the parietal, which King does not mention in her (1990) list of taxonomic
characters, thus appears in this case to be more reliable than the degree of posterior elongation of
the postorbitals, which she retains as a character. But such an osteological detail has no obvious
adaptive value in itself, and may arise from quite different skull shapes. For example, the parietal
is likely to be exposed if the intertemporal bar is extended postero-dorsally into a crest, far away
from the centres of ossification of the postorbitals. But it is also likely to be exposed if the
intertemporal bar is very wide, as in Lystrosaurus. It is therefore preferable to try to identify the
underlying adaptive changes that may in turn have led to the appearance of these features. In
the case of the postero-dorsal extension of the intertemporal bar, one of the features that unites
the Placerias group with the kannemeyeriid group, this change seems to involve a fundamental
change in the jaw muscle system, to meet an adaptive need that Lystrosaurus achieved in a different
way, as follows.

Most Permian dicynodont genera have a long, low skull, in which the intertemporal-interorbital
bar, like the zygomatic arch and the palatal surface, are parallel with one another. The posterior
edge of the maxilla and the suspensorium leading down to the palate are also parallel to one
another, but run antero-ventrally. In the Permian dicynodont Emydops (Crompton and Hotton
1967), the adductor muscles run at a shallow angle (c. 35°) down to the lower jaw because of the
long, low skull The horizontal component of their action is therefore much stronger than the
vertical component, and there is little vertical bite force (Text-fig. 4a). Extra attachment surface for

COX: TRIASSIC DICYNODONT RECHNISAURUS

775

text-fig. 4. Lateral views of skulls of a, Emydops ; b, Lystrosaurus Imurrayi ; c, Rechnisaurus cristarhynchus;
D, Lystrosaurus mccaigi ; e, Protoceratops. Axes indicate direction of action of adductor internus jaw muscles
from postero-lateral corner of temporal vacuity, and jaw axis from quadrate condyle along surface of palate.
The distances between the anterior surfaces of the quadrate condyles and the anterior end of the snout are the
same in drawings a-c and e. (a and b, after Crompton and Hotton 1967; c, skull after Cheng 1980, lower jaw
from Kannemeyeria after Camp 1956; d, after Brink 1952; e, after Dodson 1976.)

the inner portion of the adductor muscles is provided by the postorbital bones extending as wings
projecting laterally from the intertemporal bar.

In Lystrosaurus (Text-fig. 4b), a greater vertical component to the action of the adductor muscles
has been achieved by shortening the skull, as noted by Cluver (1971). In order not to reduce the
length of these muscles, which would have reduced the gape, the whole palatal structure has been
moved antero-ventrally, paralleled by a similar antero-ventral extension of the suspensorium. The
whole muscle mass is now at a steeper angle to the palatal surface ( c . 50°), and is therefore
accommodated in a more vertically-orientated column, and the temporal vacuity is shorter. Because
the palate has extended anteriorly, the snout in contrast is longer as well as deeper. Postorbital
wings still project laterally from the intertemporal bar.

The resulting skull morphology, in which the orbit and nostril remain dorsally placed, has been
likened to that of a hippopotamus, and it has therefore been suggested that Lystrosaurus was semi-

776

PALAEONTOLOGY, VOLUME 34

aquatic (Broom 1902; Watson 1912). Though, in the specimen described by Crompton and Hopson
(1967), which appears from their illustrations to be a specimen of Lystrosaurus murrayi, the external
nostril is at the level of the ventral edge of the orbit, that of Lystrosaurus mccaigi is placed below
the orbit (Text-fig. 4d). Brink (1951) states that there was a groove running dorsally from the
postero-dorsal margin of the external naris in this species, and suggests that fleshy nostrils were
directed upwards to open in front of the eyes, but the morphology is certainly not clearly indicative
of semi-aquatic life. The jaw muscles in L. mccaigi would have been the most dorso-ventrally
directed (c. 60°) of those of any dicynodont. They would therefore have been the most powerful and,
as Watson (1912) comments in the context of the massiveness of the jaws of Lystrosaurus, this is
difficult to reconcile with the softness of most aquatic plants. Watson also notes that the powerful
sacrum of Lystrosaurus is surprising in an aquatic animal (though it would not be as surprising
in a semi-aquatic animal that had to support its weight on land for long periods). Earlier, Broom
(1903) had stated that the shoulder-girdle of Lystrosaurus was different from that of other
dicynodonts in having a smaller interclavicle and a larger sternum. He suggested that this was
somehow correlated with an aquatic life, but did not explain this.

Nothing indicates that the earlier or later dicynodonts were semi-aquatic, so Lystrosaurus would
represent a brief, unique aquatic excursion. It would also be surprising to find this unusual
environment suddenly very widespread in the world at the time of Lystrosaurus , for the genus is
known from deposits in Africa, Antarctica, India, European Russia, and China. King (pers. comm.
1989; 1991) has come to the same conclusion, that Lystrosaurus was not aquatic, independently.
None of the arguments is conclusive, but they suggest that the matter requires re-evaluation.

Kannemeyeriids appear to have achieved an orientation of the adductor muscles similar to that
of the lystrosaurids, but partly by an antero-ventral movement of the palate and partly by a postero-
dorsal extension of the intertemporal bar (Text-fig. 4c). The adductor muscle mass now runs at an
angle of c. 50° to the palate, and almost parallel to the intertemporal bar. As a result, the lateral
postorbital wings to the intertemporal bar are lost, because they are no longer at an appropriate
angle to provide attachment for the adductor muscles, and would have obstructed the access of
these muscles to the more postero-dorsal part of the intertemporal bar.

Chrulew (1976) noted the general similarity between the skull architecture of dicynodonts and
that of ceratopsian dinosaurs, comparing only Dicynodon and Lystrosaurus with Triceratops , in
which the parieto-squamosal 'frill' extends more posteriorly than horizontally. The skull
architecture of kannemeyeriids is instead similar to that of the primitive ceratopsian dinosaur
Protoceratops (Text-fig. 4e). In that genus, as in the kannemeyeriids, the skull is postero-dorsally
extended to provide a greater length for the jaw muscles (Haas 1955; Ostrom 1966) which run at
about 50° to the axis of the lower jaw, and there is a narrow, horny beak on the anterior end of both
the upper and lower jaws. In the ceratopsians, however, there is a battery of cutting teeth posterior
to the beak, and later ceratopsians developed a coronoid process and a greatly elongated posterior
'frill' to the skull to improve the mechanics of the use of this posterior dentition. Ostrom (1966)
suggests that this system, in which only antero-posterior jaw movements were possible, was
particularly adapted for chopping up fibrous food, and that the Late Cretaceous ceratopsians may
have fed on the leaves of cycads and palms.

The kannemeyeriids thus appear to have an advanced feeding system, and it may be no
coincidence that they survived (as the Late Carnian Placerias group) longer than any other
dicynodont. These Late Triassic kannemeyeriids are also the only dicynodonts found in the Triassic
tropics (cf. Parrish, Parrish and Ziegler 1986), and had a much more extremely pointed premaxilla
than their earlier Triassic kannemeyeriid relatives. The stahleckeriids, by contrast, appear to have
increased the area available for insertion of the jaw muscles mainly by lateral expansion of the
squamosals, and had a very blunt anterior end to the premaxilla; it may also be no coincidence that
this group became extinct during the Middle Triassic, perhaps because of competition from
advanced rhynchosaurs and gomphodonts, or because of a diminution in their preferred plant food.

COX: TRIASSIC DICYNODONT RECHN1SAURUS

111

RELATIVE AGES OF THE TRIASSIC DICYNODONT- BE A RING FAUNAS

Triassic dicynodonts, being large and common terrestrial herbivores, are useful in documenting
Triassic biogeography and stratigraphy. Anderson and Anderson’s (1970) review of Triassic
stratigraphy was mainly a statement of opinion, it lacked systematic discussion of the data upon
which it was based, and was confined to Gondwana biostratigraphy. Their views, as reflected in a
later paper on Permo-Triassic tetrapod distribution (Anderson and Cruickshank 1978) have been
followed in King’s (1988) review of the Dicynodontia. The following review integrates work since
1970 on new faunas, reassigned genera, and redated faunas and their implications for dicynodont
evolution and taxonomy.

The earliest Triassic dicynodont is Lystrosaurus from the Lystrosaurus Zone of South Africa, and
the Fremouw Formation of Antarctica, the Panchet Formation of India, the Ryabinian horizon of
European Russia, and the Chiu-T’sai-Yuan-Tze Formation of China (Colbert 1982). (An isolated
dicynodont quadrate bone from Australia was identified as that of Lystrosaurus by Thulborn (1983)
but, as pointed out by King (1983), it is impossible to identify it to generic level.)

The Lystrosaurus fauna contains rhytidosteid amphibians, found in several parts of the world, but
only in formations of early Scythian (Griesbachian-Smithian) age (Cosgriff 1984). The most direct
evidence of the age of these faunas is the rhytidosteid Deltasaurus in the marine Kockatea Shales
of Western Australia, a close relative of Rhytidosteus of the Cynognathus Zone of South Africa. The
Kockatea Shale contains ammonites and a microflora suggesting a Lower Scythian age - either
Smithian (Anderson and Cruickshank 1978; Camp and Banks 1978) or, according to Banks (1978),
Dienerian. Lozovskiy (1985) has also independently dated the Lystrosaurus beds of the Moscow
basin as lowest Triassic, based on fresh-water conchostracans.

Above the Lystrosaurus Zone in South Africa lies the Cynognathus Zone, but it has long been
recognized (Cosgriff 1984) that there is a sharp break between these two. The characteristic
dicynodont is Kannemeyeria. An association between Kannemeyeria and other large dicynodonts
has been reported from two faunas, but both these reports have since been shown to be erroneous.
Crozier (1970) identified two dicynodonts from the lower fossiliferous horizon of the N’tawere
Formation of Zambia, recognizing one as a new species of Kannemeyeria , and the other as
Rechnisaurus crist arhynchus . However, Keyser and Cruickshank (1979) and Bandyopadhyay ( 1989)
have shown that the latter specimen is quite different from Rechnisaurus , and appears to belong to
Kannemeyeria , which is thus the only dicynodont from this horizon in the N’tawere Formation. The
presence of the cynodont Diademodon in this horizon as well as in the Cynognathus Zone of South
Africa further supports the view that these two formations are of similar age (Cox 1969).

Kannemeyeria was also described by Cruickshank (1965) from the Manda Formation of
Tanzania, in which Rechnisaurus is also found, but he has since shown (1986) that the specimen
instead came from the underlying Kingori Sandstone Formation.

Both Kannemeyeria and Diademodon are found in the Omingonde Formation of South West
Africa (Keyser 1973), which therefore also appears to be of an age similar to that of the Cynognathus
Zone. However, the presence of Titanogomphodon , which is more advanced than the other
diademodonts in having a secondary basin on the crowns of the molariform teeth (Keyser 1973),
suggests that the Formation may be of slightly later date.

Kannemeyeria is thus known from the Cynognathus Zone of South Africa and from deposits in
other parts of Africa (Zambia, Tanzania, South-West Africa), that are therefore presumably of
similar age. It is also known from the upper part of the Puesto Viejo Formation of Argentina
(Bonaparte 1966), accompanied by the cynodont Cynognathus - the zone fossil of the Cynognathus
Zone. Kannemeyeria is therefore known only from south-western Gondwana.

The Cynognathus Zone is commonly viewed as being of late Early Triassic age (e.g. Anderson and
Anderson 1970), but there is no way in which its age can be determined in terms of the classic
European non-marine German sequence or marine Alpine sequence. All that can be attempted for
the succeeding Triassic faunas is to place them in relative order - and even this is complicated by
ecological biases of some of the faunas, and by recent re-assignments of some synapsid genera.

778

PALAEONTOLOGY, VOLUME 34

The next-youngest diverse, well-documented Triassic fauna is that of the Manda Formation of
Tanzania. This contains traversodontid cynodonts and rauisuchid thecodonts, which are absent
from the Cynognathus Zone. It also contains the rhynchosaur Stenaulorhynchus , while the only
rhynchosaurs in the Cynognathus Zone are Howesia and Mesosuchus , which are more primitive in
lacking the interlocking blade and groove jaw apparatus of the later rhynchosaurs (Benton 1980,
1983). There is therefore little doubt that the Manda Formation is of later age than the Cynognathus
Zone. Its exact age cannot be directly established, but the Manda rhynchosaur Stenaulorhynchus is
closest to Rhynchosaurus (Benton 1983), which is found in England both in the Midlands and in the
Otter Sandstone Formation of Devon (Benton 1990). The age of the Midlands species is difficult to
establish, different elements in the faunal and floral assemblage being interpreted to indicate ages
ranging from Late Scythian to Late Ladinian (Benton 1990). However, the age of the Otter
Sandstone vertebrate fauna, which includes fishes, amphibians and reptiles, may be Anisian (Milner
et al. 1990).

Manda dicynodont genera are found also in three other faunas, Rechnisaurus in the Yerrapalli
Lormation of India and in the Er-Ma-Ying Formation of China, and Sangusaurus in the upper
fossiliferous horizon of the N’tawere Formation of Zambia (Cruickshank 1986). The Yerrapalli
fauna contains the dicynodonts Rechnisaurus and Wadiasaurus , an erythrosuchid and a prestosuchid
thecodontian, a trirachodont cynodont, and a labyrinthodont amphibian (Bandyopadhyay 1988).
The age of the Er-Ma-Ying Formation is somewhat perplexing. It is now divided into upper and
lower portions (Sun 1989). The faunas of each portion contain thecodontian reptiles similar to
genera in the Cynognathus Zone of South Africa: Halazhaisuchus of the lower portion and
Shansisuchus of the upper portion are respectively similar to Euparkeria and to Erythrosuchus of the
Cynognathus Zone (Sun 1989). This seems to suggest that both faunas are similar in age to the South
African fauna, but the lower fauna also contains Rechnisaurus (Shaanheikannemeyeria), which is
found also in the Manda fauna. Rechnisaurus in the Er-Ma-Ying Formation suggests an age similar
to the Manda Formation, while the other elements must be seen as anomalous, perhaps due to the
considerable distance between the Chinese and African faunas. The upper fossiliferous horizon of
the N'tawere Formation of Zambia contains the dicynodonts Sangusaurus (Cruickshank 1986) and
Zambiasaurus , the traversodont cynodont Luangwa , and fragments of thecodontians.

Manda-like dicynodonts in the Donguz Formation of European Russia include Rabidosaurus,
Rhadiodromus (= Rhinocerocephalus), and Rhinodicynodon (Kalandadze 1970), and Uralo-
kannemeyeria (Danilov 1971). Apart from the shansiodont Rhinodicynodon , all of these are very
similar to Rechnisaurus in skull proportions. Three other dicynodont genera from the Donguz
Formation have been described more recently, but these ( Edaxosaurus , Elatosaurus, and
Calleonasus ; Kalandadze and Sennikov 1986) are known only from respectively an ‘upper jaw
bone' and two nasal bones, and are effectively nomina nuda. The rest of the Donguz fauna comprises
temnospondyl amphibians (the capitosaur Eryosuchus , Ochev 1966; plagiosaurs, Shishkin 1986),
the procolophonid Orenburgia (Ochev 1968; Ivakhnenko 1977); thecodontians ( Dongusia ,

‘ Erythrosuchus ' Ochev 1980; Vyushkovisaurus Ochev 1982; Vytshegdosuchus, Dongosuchus
Sennikov 1988), and traversodont cynodonts (Tatarinov 1974, 1988). The thecodontians are
fragmentary, and the other groups are from such long-lived lineages that they provide little
information on the age of the Donguz Formation. Thus, unfortunately, one can only rely on the
slender evidence of the similarities of the Rechnisaurus- like dicynodonts discussed above, and
provisionally view it as if similar age.

As noted earlier, the Manda Formation cannot be dated more precisely than Anisian. The other
four faunas just discussed offer such poor evidence on dating that they are all ascribed provisionally
to the Anisian, this reflecting also the fact that they appear to be later than the Scythian
Cynognathus Zone fauna, and earlier than the South American faunas to be discussed next.

The South American dicynodont faunas include those of the Chanares and Ischigualasto
Formations of Argentina, and of the Santa Maria Formation of Brazil. Cox (1965, 1968, 19736).
and Bonaparte (1982) have discussed the relative ages of these South American Middle- Late
Triassic faunas. An important new element has been the documentation by Barberena (1977) of the

COX: TRIASSIC DICYNODONT RECHNISAURUS

779

consistent differences between two faunal assemblages in the Santa Maria Formation, which now
makes it possible to resolve some apparent anomalies. The fact that the Santa Maria fauna included
the rhynchosaur Scaphonyx (also known from the Ischigualasto fauna) and the early dinosaur
Staurikosaurus gave the impression that the whole Santa Maria fauna must be considerably younger
than the Manda fauna (Cox 1965, 1968, 1973). However, Barberena (1977) has since shown that
these elements are found only in deposits near the town of Santa Maria. He names this the
Rhynchocephalian Assemblage, but now that it has been realized that rhynchosaurs and
sphenodontids are not closely related (Carroll 1977), a more appropriate name would be the
Rhynchosaur Assemblage; its age is discussed below.

Barberena (1977) points out that the rhynchosaurs, abundant in this Rhynchosaur Assemblage,
are wholly lacking in deposits from the Xiniqua and Pinheiros areas, respectively to the west and
to the east of Santa Maria; these localities instead contain an abundance of stahleckeriid
dicynodonts ( Stahleckeria and Dinodontosaurus ), as well as chiniquodont cynodonts. He names this
fauna the Therapsid Assemblage. (Though he shows the kannemeyeriid dicynodont Jcichaleria as
found in this Assemblage, Bonaparte (1982) states that it is instead from the much later Caturrita
Formation of Brazil.)

Dinodontosaurus and the traversodont cynodont M asset ognathus (Hopson and Kitching 1972) are
found also in the Chanares Formation of Argentina. In addition to these similarities at generic level,
the two faunas are similar in containing chiniquodont cynodonts and in lacking rhynchosaurs. They
therefore appear to be of very similar, if not identical, age and character. It is difficult to estimate
that age precisely, since the chiniquodont cynodonts range widely in age, from the Early Triassic
Manda fauna to the Late Triassic Argentinian faunas of Ischigualasto and Las Esquina (Hopson
and Kitching 1972; Bonaparte 1982). However, the Chanares herbivorous cynodonts do appear to
be somewhat more advanced than those of the Manda fauna (Romer 1967). This suggests that the
Chanares/Santa Maria Therapsid Assemblage is later in age than the Manda fauna, and may
therefore be provisionally placed in the Ladinian.

Barberena (1977) rightly points out that the discovery that the Rhynchosaur Assemblage is
younger than the Therapsid Assemblage makes it unnecessary to postulate that merely ecological
differences were the reason why rhynchosaurs and dicynodonts had not been found together.
However, since rhynchosaurs were present in the Manda fauna, which is clearly older than the
Therapsid Assemblage, a problem remains. The absence of rhynchosaurs from the Chanares/Santa
Maria Therapsid fauna must have been caused by either geographical or ecological factors.

A geographical cause, in the sense of a geographical barrier between the South American and
East African localities, seems unlikely. Apart from the general similarities in Triassic faunas
throughout Pangaea (Cox 1973u), the dicynodont Kannemeyeria is found in both Africa and
Argentina, the Brazilian dicynodont Stahleckeria is closely related to the Zambian genus
Zambiasaurus (Cox 1969), and the rhynchosaur Stenaulorhynchus of the Manda Formation is
related to Rhynchosaurus of Britain (Benton 1983).

An ecological explanation therefore remains more plausible, though its precise nature can only
be guessed at. Rhynchosaurs were clearly herbivorous (Benton 1983), and therefore may well have
competed with dicynodonts. Though the two groups co-exist in the Manda fauna, it may be that
the climatic conditions of the South American areas, far closer to the periphery of a Gondwana
continent that still lacked Andean mountains, may have been moister and may have encouraged a
lusher flora, which may have favoured the dicynodonts. The two groups again co-exist in the Late
Triassic Ischigualasto fauna, but the stahleckeriid dicynodonts of the Chanares/Santa Maria
Therapsid Assemblage had by then disappeared, and only the kannemeyeriids co-existed with the
rhynchosaurs.

The next-youngest dicynodont-bearing faunas are those of the Ischigualasto Formation of
Argentina, the Chinle Formation of North America, and the Argana Formation of Morocco. Only
the kannemeyeriid type of dicynodont is found in these. Their ages have recently been re-evaluated
by Olson and Sues (1986), on the basis of palynoflorule and plant megafossil evidence, and they are
now considered to be of Late Carnian age (though this evidence is indirect, rather than direct

780

PALAEONTOLOGY, VOLUME 34

interfingering of terrestrial and marine deposits). The presence of the phytosaur Paleorhinus in the
Moroccan fauna as well as in the Dockum Formation of Texas and in the Popo Agie Member of
the Chugwater Formation of Wyoming also suggests an age no later than the Late Carnian (Ballew
1989), and Litwin (1986) similarly suggests a Late Carnian age for the Lower Unit of the Petrified
Forest Member of the Chinle Formation (which contains Placerias) on the basis of its plant spores.
Hunt and Lucas have also very recently (1991) suggested a Late Carnian age for these faunas on
the basis of the presence in all of them of elements coeval with the phytosaur Paleorhinus of the
German Blasensandstein.

If such a Late Carnian age is provisionally accepted for the Ischigualasto fauna, the question
arises as to what age should be assigned for the Santa Maria Rhynchosaur Assemblage. The
presence of the rhynchosaur Scaphonyx in both faunas suggests that there is no great difference
in their ages. This is supported by Hopson’s (1985) restudy of the traversodont cynodont
Gomphodontosuchus braziliensis from the Rhynchosaur Assemblage, which shows that it is probably
a juvenile specimen closely related to Exaeretodon of the Ischigualasto fauna. The Rhynchosaur
Assemblage, also, is therefore given a Late Carnian age in Table 2.

However, the Ischigualasto fauna differs from the Chinle/Dockum fauna, as it is still dominated
by mammal-like reptiles such as the dicynodonts and cynodonts, while the North American faunas
are instead dominated by archosaurs such as the thecodontians and early dinosaurs. Though there
may also be an ecological component to these differences, they suggest that the Ischigualasto fauna
is rather more archaic and older than the North American faunas.

The last known dicynodont is Jachaleria from the Los Colorados Formation of Argentina
(Bonaparte 1971) and from the Caturrita Formation of Brazil (Araujo and Gonzaga 1980). This
genus is clearly very closely related to Ischigualastia (Table 1 : Araujo and Gonzaga [1980] state that
the posterior orientation of the suspensorium in Jachaleria and the curved shape of the zygomatic
arches in Ischigualastia are probably the result of distortion), which suggests that these formations
are not very much later in time than those mentioned in the preceding paragraph, and they too are
given a Late Carnian date.

table 2. Relative ages of Triassic dicynodont-bearing strata (figures indicate age in millions of years, from
Cowie and Bassett 1989).

220

Late Carnian
Late Carnian

230

Ladinian

235

Anisian

240

Late Scythian
Early Scythian

250

Los Colorados; Caturrita; Chinle

Ischigualasto; Argana; Santa Maria Rhynchosaur Assemblage
Santa Maria Therapsid Assemblage; Chanares

Manda; Upper N’tawere; Yerrapalli; Donguz; Omingonde; Er-Ma-Ying

Cynognathus Zone; Lower N’tawere; upper Puesto Viejo
Lystrosaurus Zone; Panchet; Ryabinian horizon; Chiu-T'sai-Yuan-Tze;
Fremouw; lower Puesto Viejo

The sequence of Triassic dicynodont-bearing faunas that results from the above review is shown
in Table 2. It cannot be too strongly emphasized that the relative ages of these faunas are considered
to be much more reliable than their absolute ages in terms of the Scythian-Carnian classic sequence.

Acknowledgements. I am grateful to the Principal of King’s College, University of London for permission to
take the sabbatical leave during which this work was carried out. and to the Fulbright Commission for kindly
providing a travel grant. I am also grateful to Dr W. A. Clemens of the Museum of Paleontology, flniversity

COX: TRIASSIC DICYNODONT RECHNISAURUS

781

of California, Berkeley, for providing office and library facilities. Dr Gillian King of the South African
Museum, Capetown, Dr Jose Bonaparte of the Bernardino Rivadavia Museum in Buenos Aires, and
Dr Michael Benton of the Geology Department, University of Bristol, all provided valuable comments on the
manuscript. The drawings of the African skull of Rechnisaurus were made by the late Peter Hutchison.

REFERENCES

anderson, h. m. and anderson, j. m. 1970. A preliminary review of the biostratigraphy of the uppermost
Permian, Triassic and lower Jurassic of Gondwanaland. Palaeontographica Africana , 13 (suppl.), 1-22.

- and cruickshank, a. R. I. 1978. The biostratrigraphy of the Permian and Triassic. Part 5. A review of the
classification and distribution of Permo-Triassic tetrapods. Palaeontographica Africana , 21, 15-44.

araujo, d. c. and gonzaga, t. d. 1980. Uma nova especie de Jachaleria (Therapsida, Dicynodonta) do
Triassico do Brasil. Adas II Congreso Argentino de Paleontologia y Bioestratigrafia y I Congreso
Latinoamericano de Paleontologia , Buenos Aires, 1, 159-174.
attridge, j. et al. 1964. The British Museum (Natural History) - University of London Joint Palaeontological
Expedition to Northern Rhodesia and Tanganyika, 1963. Nature , 201, 445^449.
ballew, k. l. 1989. A phylogenetic analysis of Phytosauria from the Late Triassic of the Western United
States. 309-339. In lucas, s. G. and hunt, a. p. (eds). Dawn of the age of dinosaurs in the American Southwest.
New Mexico Museum of Natural History, Albuquerque, 414 pp.
bandyopadhyay, s. 1988. A kannemeyeriid dicynodont from the Middle Triassic Yerrapalli Formation.
Philosophical Transactions of the Royal Society of London, Series B, 320, 185-233.

1989. The mammal-like reptile Rechnisaurus from the Triassic of India. Palaeontology, 32, 305-312.
banks, m. r. 1978. Correlation chart for the Triassic System of Australia. Bulletin of the Bureau of Mineral
Resources, Geology and Geophysics (Canberra), 156C, 1-39.

BARBFRENA, M. c. 1977. Bioestratigrafia preliminar da Formaqao Santa Maria. Pesquisas do Instituto de
Geociencias da UFRGS, Porto Alegre, 7, 111-129.

benton, m. j. 1983. The Triassic reptile Hyperodapedon from Elgin: functional morphology and relationships.
Philosophical Transactions of the Royal Society of London, Series B, 302, 605-717.

- 1990. The species of Rhynchosaurus , a rhynchosaur (Reptilia, Diapsida) from the Middle Triassic of
England. Philosophical Transactions of the Royal Society of London, Series B, 328, 213-306.

bishop, w. w. 1968. The evolution of fossil environments in East Africa. Transactions of the Leicester Literary
and Philosophical Society, 62, 22-44.

bonaparte, j. f. 1966. Sobre nuevos terapsidos Triasicos hallados en en centra de la Provincia de Mendoza,
Argentina. (Therapsida: Dicynodontia y Cynodontia). Acta Geoldgica Lilloana, 8, 91-100.

- 1971. Annotated list of the South American Triassic tetrapods. Proceedings with Papers, Second
Gondwana Symposium, South Africa, 665-682.

- 1982. Faunal replacement in the Triassic of South America. Journal of Vertebrate Paleontology, 2,
362-371.

brink, a. s. 1951. On the genus Lystrosaurus Cope. Transactions of the Royal Society of South Africa, 33,
107-120.

broom, r. 1902. On the leg and toe bones of Ptychosiagum. Transactions of the South African Philosophical
Society, 1 1, 233-235.

- 1903. On the structure of the shoulder girdle in Lystrosaurus. Annals of the South African Museum, 4,
139-141.

camp, c. l. 1956. Triassic dicynodont reptiles. Part 2. Triassic dicynodonts compared. Memoirs of the
University of California, Geology, 13, 305-348.

- and banks, m. r. 1978. A proterosuchian reptile from the Early Triassic of Tasmania. Alcheringa, 2,
143-158.

- and welles, s. p. 1956. Triassic dicynodont reptiles. Part 1. The North American genus Placerias.
Memoirs of the University of California, Geology, 13, 255-304.

Carroll, r. l. 1977. The origins of lizards. 359-396. In Andrews, s. m., miles, r. s. and walker, a. d. (eds).

Problems of vertebrate evolution. Linnean Society Symposium Series, 4, 41 1 pp.

Cheng, z.-w. 1980. Mesozoic stratigraphy and paleontology of the Shaanxi-Gansu-Ningxia Basin. Vol. 2. part
7, vertebrate fossils. Publishing House of Geology, Beijing, fin Chinese].
chowdhury, t. r. 1970. Two new dicynodonts from the Triassic Yerrapalli Formation of central India.
Palaeontology , 13, 132-144.

782

PALAEONTOLOGY, VOLUME 34

chrulew, s. 1976. Biomechanik und Funktion des Systems Schadel/Kiefer bei Dicynodonten und Ceratopsien.

Zentralblatt fur Geologie und Palciontologie, 1976, 291-292.
cluver, m. a. 1971. The cranial morphology of the dicynodont genus Lystrosaurus. Annals of the South African
Museum , 56, 155-274.

colbert, E. H. 1982. The distribution of Lystrosaurus in Pangaea and its implications. Geobios , Memoire
Special , 6, 375-383.

cosgriff, j. 1984. The temnospondyl amphibians of the earliest Triassic. Journal of Vertebrate Paleontology,

4, 30-46.

cowie, j. w. and bassett, m. g. 1989. Global stratigraphic chart. Episodes , 12 (2), supplement,
cox, c. b. 1965. New Triassic dicynodonts from South Africa, their origins and relationships. Philosophical
Transactions of the Royal Society of London , Series B , 248, 457-516.

1968. The Chanares (Argentina) Triassic reptile fauna. IV. The dicynodont fauna. Breviora , Museum of
Comparative Zoology , 205, 1-27.

1969. Two new dicynodonts from the Triassic N’tawere Formation, Zambia. Bulletin of the British
Museum (Natural History ), Geology, 21, 29-71.

- 1973m Triassic Tetrapods. 213-233. In hallam, a. (ed.). Atlas of palaeobiogeography. Elsevier, London
and New York, 531 pp.

1973/). Gondwanaland Triassic stratigraphy. Anales Academia Brasiliense de Ciencias , 45, 115-119.

- and li, j.-l. 1983. A new genus of Triassic dicynodont from East Africa and its classification.
Palaeontology , 26, 389-406.

Crompton, a. w. 1955. On some Triassic cynodonts from Tanganyika. Proceedings of the Zoological Society
of London, 125, 617-669.

— and hotton, n. 1967. Functional morphology of the masticatory apparatus of two dicynodonts (Reptilia,
Therapsida). Postilla Peabody Museum of Natural History, Yale University, 109, 1-51.
crozier, E. a. 1970. Preliminary report on two Triassic dicynodonts from Zambia. Palaeontologia Africana, 13,
39-45.

cruickshank, A. R. I. 1965. On a specimen of the anomodont reptile Kannemeyeria latifrons (Broom) from
the Manda Formation of Tanganyika, Tanzania. Proceedings of the Linnean Society of London , 176,
149-157.

1967. A new dicynodont genus from the Manda Formation of Tanzania (Tanganyika). Journal of
Zoology, London, 153, 163-208.

1986. Biostratigraphy and classification of a new Triassic dicynodont from East Africa. Modern Geology,
10. 121-131.

danilov, a. i. 1971 . A new dicynodont from the Middle Triassic of southern Cisuralia. Paleontological Journal,

5, 265-268.

dodson, p. 1976. Quantitative aspects of relative growth and sexual dimorphism in Protoceratops. Journal of
Paleontology, 50. 929-940.

dutuit, j-m. 1980. Principaux caracteres d’un genre de dicynodonte du Trias marocaine. Comptes Rendus de
T Academie des Sciences, Paris, Serie D, 290, 655-658.

- 1989. Azarifeneria barrati , un deuxieme genre de dicynodonte du Trias superieur marocain. Comptes
Rendus de T Academie des Sciences , Paris, Serie D, 309, Serie II, 303-306.

haas, G. 1955. The jaw musculature in Protoceratops and in other ceratopsians. American Museum Novitates,
1729, 1-24.

haughton, s. H. 1932. On a collection of vertebrates from Tanganyika Territory. Quarterly Journal of the
Geological Society of London, 88, 634—671.

hopson, j. a. 1985. Morphology and relationships of Gomphodontosuchus brasiliensis von Huene (Synapsida,
Cynodontia, Tritylodontoidea) from the Triassic of Brazil. Neues Jahrbuch fur Geologie und Palciontologie ,
Monatshefte, 1985, 285-299.

and kitching, J. w. 1972. A revised classification of cynodonts (Reptilia; Therapsida). Palaeontologia
A fricana, 14, 71-85.

huene, F. von 1935. Die fossilen Reptilien des siidamerikanischen Gondwanalandes an der Zeitenwende. Leif. 1,
Ordnung Anomodontia. 92 pp.

1938m Stenaulorhynchus, ein Rhynchosauride der ostafrikanischen Obertrias. Nova Acta Leopoldina , 6,
83-121.

1938/). Ein grosser Stagonolepide aus der jungeren Trias Ostafrikas. Neues Jahrbuch fur Mineralogie,
Geologie und Palciontologie, Beilagebcinde, 80, 264-278.

1942. Die Anomodontier des Ruhuhu-Gebietes in der Tubinger Sammlung. Palaeontographica, 94,
154-184.

COX: TRIASSIC DICYNODONT RECHNISAURUS

783

1950. Die Theriodontier des ostafrikanischen Ruhuhu-Gebietes in der Tubinger Sammlung. Neues

Jahrbuch fur Geologie und Palaontologie, Abhandlungen , 92, 47-136.
hunt, a. p. and lucas, s. s. 1991. The Paleorhinus biochron and the correlation of the non-marine Upper
Triassic of Pangaea. Palaeontology , 34, 487-501.

ivakhnenko, M. f. 1977. Early Triassic procolophonid genera of Cisuralia. P aleontologicheskii Zhurnal , 9, (1)
86-91 . [In Russian],

kalandadze, N. n. 1970. New Triassic kannemeyeriids from the southern Cis-Ural region. 51-57. In flerov,
k. K. (ed.). Materials on the evolution of terrestrial vertebrates. A collection of articles. Nauka, Moscow. [In
Russian].

- and sennikov, a. g. 1986. New reptiles from the Middle Triassic in the Southern Ural forelands.
Paleontological Journal , 20, 73-80.

keyser, a. w. 1973. A new Triassic vertebrate fauna from South West Africa. Palaeontologia Africana , 16,
1-15.

1980. The stratigraphic distribution of the Dicynodontia of Africa reviewed in a Gondwanaland context.

Proceedings of the Fifth International Gondwana Symposium , 61-63.

— and cruickshank, a. r. i. 1979. The origins and classification of Triassic dicynodonts. Transactions of the
Geological Society of South Africa, 82, 81-108.

king, G. m. 1983. First mammal-like reptile from Australia. Nature, 306, 209.

- 1988. Encyclopedia of paleoherpetology. Part 17C: Anomodontia. Gustav Fischer, Stuttgart and New
York, 174 pp.

- 1990. The dicynodonts. A study in palaeobiology. Chapman and Flail, Fondon and New York,
233 pp.

- 1991. The aquatic Lystrosaurus : a palaeontological myth. Historical Biology, 4, 285-341.

LI, J.-L. 1980. Kannemeyeriid fossil from Inner Mongolia. Vertebrate i Palasiatica, 18, 94—99.

litwin, r. j. 1986. The palynostratigraphy and age of the Chinle and Moenave Formations, southwestern
USA. Ph.D. dissertation, Pennsylvania State University.

Lozovskiy, v. R. 1985. Age of the Lystrosaurus beds in the Moscow syneclise. Doklady Akademii nauk SSSR,
[English translation], 272, 127-131.

milner, a. r., Gardiner, b. G., fraser, N. c. and taylor, m. a. 1990. Vertebrates from the Middle Triassic
Otter Sandstone Formation of Devon. Palaeontology, 33, 873-892.

OCHEV, v. G. 1966. Systematics and morphology of capitosauroid labyrinthodonts. Saratov State University
Press, Saratov, 184 pp. [In Russian],

- 1968. A new representative of Triassic procolophonids from Bashkiria. Ezhegodnik Vsesoyuznogo
Paleontologicheskogo Obshchestva, 18, 298-301.

- 1980. New archosaurs from the Middle Triassic of the southern Ural forelands. Paleontological Journal,
14, (2) 91-96.

1982. Pseudosuchia from the Middle Triassic of the southern Ural forelands. Paleontological Journal, 16,
(2) 95-101.

olsen, p. e. and sues, h.-d. 1986. Correlation of continental Late Triassic and Early Jurassic sediments, and
patterns of the Triassic-Jurassic tetrapod transition. 321-351 In padian, k. (ed.). The beginning of the age
of dinosaurs. Cambridge University Press, Cambridge, 378 pp.
ostrom, j. h. 1966. Functional morphology and evolution of the ceratopsian dinosaurs. Evolution, 20, 290-
308.

parrington, f. r. 1946. On the cranial anatomy of cynodonts. Proceedings of the Zoological Society of London,
116, 181-197.

parrish, j. m„ parrish, j. T. and ziegler, a. m. 1986. Permian-Triassic paleogeography and paleoclimatology
and implications for therapsid distribution. 109-131. In hotton, n., Maclean, p.-d., roth, j. j. and roth,
E.c. (eds). The ecology and biology of mammal-like reptiles. Smithsonian Institution Press, Washington and
London, 326 pp.

romer, a. s. 1967. The Chanares (Argentina) Triassic reptile fauna. III. Two new gomphodonts, Massetognathus
pascuali and M. teruggii. Breviora , Museum of Comparative Zoology , Harvard, 264, 1-25.
sennikov, a. G. 1988. New rauisuchids from the Triassic of the European U.S.S.R. Paleontological Journal, 22,
120-124.

shishkin, m. a. 1986. New data on plagiosaurs from the Triassic of the U.S.S.R. Byulleten Moskovskogo
Obshchestva Ispytatelei Prirody Odtielenie Geologii, 61, (3) 97-102. [In Russian],
stockley, G. M. 1932. Geology of the Ruhuhu coalfields, Tanganyika Territory. Quarterly Journal of the
Geological Society of London, 88, 610-622.

784

PALAEONTOLOGY, VOLUME 34

sun, a.-l. 1963. The Chinese kannemeyeriids. Palaeontologia Sinica , 17, 1-109.

- 1980. Late Permian and Triassic terrestrial tetrapods of North China. Vertebrata Palasiatica , 18, 1 09—

1 11.

1989. Before dinosaurs. China Ocean Press, Beijing, 109 pp.
thulborn, R. a. 1983. A mammal-like reptile from Australia. Nature , 303, 330-331.

tatarinov, L. p. 1974. Theriodonts of the U.S.S.R. Trudiy Paleontologischeskogo Instituta Akademia nauk
SSSR , 143, 1-252. [In Russian],

1988. On the morphology and systematic position of the gomphodont cynodont Antecosuehus ochevi.
Paleontological Journal , 22, 82-90.

watson, d. m. s. 1912. The skeleton of Lystrosaurus. Records of the Albany Museum , 2, 287-295.

C. BARRY COX

Division of Biosphere Sciences

Typescript received 2 October 1989 King’s College, University of London

Revised typescript received 8 March 1991 Campden Hill Road, London W8 7AH

IMMUNOLOGICAL INVESTIGATIONS OF
RELATIONSHIPS WITHIN THE TEREBRATULID

BRACHIOPODS

by M. COLLINS, G. B. CURRY, G. MUYZER, R. QUINN, SHANJIN XU,

P. WESTBROEK and S. EWING

Abstract. Intra-crystalline macromolecules isolated from the skeletons of nine species of Recent articulate
brachiopods were compared by enzyme linked immunosorbent assay (ELISA) to assess the relationships within
the Order Terebratulida (Phylum Brachiopoda, Class Articulata). Immunological distance data indicated that
the sub-division of this order into three suborders (based on the characteristics of the internal skeleton,
particularly the brachial loop) is not valid. Three major clusters were recognized within the Terebratulida
which approximately correspond to recognized superfamilies, with the exception that the family Kraussinidae
(possessing a long brachial loop) is placed within the Terebratulacea (characterized by a short loop). The three
lineages are more closely related to each other than was previously predicted, and yet within most lineages there
is a greater degree of subsequent diversification than has hitherto been recognized. The independent evolution
of long brachial loops in two lineages identified by the immunological data, highlights problems with using the
internal skeleton as a high-level taxonomic character within the terebratellids.

The Terebratulida is the largest extant order of brachiopods, being characterized by the possession
of an internal skeleton (brachidium or brachial loop) which in life supports the lophophore. A
taxonomic treatment incorporating patterns of loop development was first elaborated by Beecher
(1893), extended by Muir-Wood (1955), and utilized in the Treatise (Williams et al. 1965) in which
the Order Terebratulida is divided into three suborders on the basis of the ontogeny and form of
loop development. Two suborders characterized as possessing either short- (Terebratulidina) or
long-loops (Terebratellidina), contain extant members; the third (Centronellida) is extinct. The
prevailing taxonomic situation of the extant families is summarized by Collins et al. (1988);
although inadequacies within the systematics of the Terebratulida have been recognized (e.g.
Richardson 1975; Elliott 1976; Williams and Hurst 1977) the essential sub-division has remained
unchallenged (Text-fig. 1 a).

In the earlier immunological study, we presented data which appeared to challenge the primacy
of loop length as a subordinal character within the terebratulids (Collins et al. 1988). No
immunological distances were illustrated, and the systematic conclusions were restricted, since only
three sera were used to establish the relationships among the various extant genera. The
implications that this earlier study had for the systematics of the brachiopods prompted us to
conduct a detailed investigation using a more robust immunological distancing approach.

Immunological distance measurements have frequently been used in taxonomic investigations,
either using whole organism extracts (Olsen-Stojkovich et al. 1986; Price et al. 1987) or purified
macromolecules (Sarich and Wilson 1967; Lowenstein et al. 1981 ; Lowenstein 1985). The technique
involves the preparation of antisera against each taxon studied and reciprocal determinations of all
sera against all taxa. In interpreting immunological distance, the technique assumes that the rate of
evolution averaged over a large number of antigenic sites is uniform enough to give an accurate
portrayal of the evolutionary branching pattern of the groups examined; for serum proteins it has
been demonstrated that immunological distance closely parallels protein evolution (e.g. Maxon and
Maxon 1979). In addition, within the cancellothyrid genus Terebratulina (which is used in this

| Palaeontology, Vol. 34, Part 4, 1991, pp. 785-796.|

© The Palaeontological Association

786

PALAEONTOLOGY, VOLUME 34

study) there is excellent congruence between genetic (Cohen et al. 1991) and both immunological
(Collins et al. 1988) and morphological (Cohen et al. 1991) data.

The immunological approach is particularly useful for groups such as the brachiopods, where live
sampling is only rarely possible. By isolating protected macromolecules (intra-crystalline skeletal
glycoproteins), sampling of spirit-stored and dried museum specimens was possible, greatly
extending the range of available material. More significantly, fossil organic matter isolated from
Pleistocene and Pliocene skeletons could also be screened immunologically (Collins et al.
in press). If it is possible, using immunological distances, to establish a phenogram for Recent
brachiopods, this could then be used as a benchmark against which to compare the pattern of
reactions of fossil extracts (when tested with the same range of sera), to assess the quality of
preserved molecular information (Collins et al. in press).

a

b

TEREBRATULIDA

rhynchonellida-ORDER-

-SUBORDER-

Terebratellacea Terebratulacea

King, 1850 Gray, 1840

- SUPERFAMILY -

FAMILY —

Dallinacea

Cooper, 1973

Cancellothyrididai

Thompson, 1926

Mt Kr Nl Wi Ds Ln Gv Tr Nn Mt Kr Nl Wi Ds Ln Gv Tr Nn

text-fig. 1. a , systematic sub-division of the Terebratulida based on the Treatise (Williams et al. 1965). b ,
modifications of the Terebratulida by the inclusion of two new post-Treatise superfamilies proposed by Cooper

(1973, 1981).

A total of nine brachiopod antisera were prepared against skeletal glycoproteins. Although the
taxonomic distribution of the genera investigated has been dictated primarily by the availability of
sufficient quantities of material for antiserum preparation, it has been possible to include taxa which
represent all major groups of living articulate brachiopods. Not surprisingly, the number of genera
available from each group closely reflects their relative abundance in Recent brachiopod faunas.
Thus, five genera were available from among the long-looped terebratulids (which dominate
present-day faunas), four of which were assigned to two families of the Superfamily Terebratellacea,
while the remaining genus was classified within the Superfamily Dallinacea (Text-fig. 1 b). Three
short-looped terebratulids were available, representing two discrete families which, depending upon
interpretation, represented either one (Text-fig. 1 a) or two (Text-fig. 1 b) discrete superfamilies. The
rhynchonellid genus Notosaria was a convenient outgroup for the investigation of terebratulid
phylogeny. Roughly 10% of all living brachiopod genera are therefore included in this first
comprehensive investigation of the biochemical systematics of brachiopods.

MATERIALS AND METHODS

Species and sources used in the immunological investigation are listed in Table I. Shells were soaked in
concentrated NaOCl (12% active chlorine) solution overnight prior to further handling to remove surface
contaminants. Additional NaOCl was added for a minimum 48 hours to remove the inter-crystalline organic
matrix, and thereby weaken the shell, liberating the fibres of the secondary shell layer (Collins 1986). It was
necessary to treat the rhynchonellid Notosaria for a week before shell softening was complete. The fibres were
isolated from the remainder of the shell material as previously described (Collins et al. 1988) and exhaustively
rinsed in double-distilled water prior to freeze drying.

COLLINS ET AL. \ BRACHIOPOD SEROTAXONOMY

787

table 1. Species, with localities, used in the immunological distance study.

Taxon

Locality

Prep.

Abbrev.

Serum

Terebratellidae

Waltonia ( Terebratella) inconspicua

Christchurch, N. Zealand

Fibre

Wi

K5040

(Sowerby)

Neothyris lenticularis (Deshayes)

Foveaux Strait, N. Zealand

Protein

NL

427

Dallinidae

Dallina septigera (Loven)

Hebridian Rise, Scotland

Fibre

Ds

K.5007

Kraussinidae
Kraussina rubra (Pallas)

Southern Tip, S. Africa

Fibre

Kr

801

Megerlia truncata (Gemlm)

Corsica, Mediterranean

Fibre

Mt

K5053

Cancellothyridae

Terebratulina retusa (L.)

Firth of Lorn, Scotland

Powder

Tr

K4962

Terebratulidae

Liothyrella neozelandica (Thomson)

Foveaux Strait, N. Zealand

Fibre

Ln

802

Gryphus vitreus (Born)

Corsica, Mediterranean

Powder

Gv

803

Rhynchonellida
Notosaria nigricans (Sowerby)

Christchurch. New Zealand

Fibre

Nn

K5038

Intra-crystalline macromolecules were isolated from the secondary layer fibres by de-mineralization in 20%
wt/vol disodium ethylene diamine tetraacetic acid (EDTA, pH 8), and the EDTA was subsequently removed
by ultra-filtration across an Amicon YM 5 or 10 filter (5 or 10 kD). This treatment yielded approximately 6
mg of water-soluble organic matter per 20 g of fibre preparation.

Special treatments were applied to three samples. For Gryphus vitreus it was necessary to use shell powders
since secondary layer fibres could not be isolated (this genus develops a tertiary shell layer), Megerlia truncata ,
for which little material was available, was dialyzed against EDTA using the technique of Weiner and
Lowenstam (1980), and for Neothyris lenticularis a single protein (45 kD) cut from a sodium dodecyl sulphate
polyacrylamide gel was used for immunization.

Antisera were prepared using 500 pg aliquots of sample immunized subcutaneously into New Zealand white
rabbits according to the following schedule. Primary immunization with Freund's complete adjuvant, followed
by three secondary injections with incomplete adjuvant at two to three week intervals, with bleeds after the
second and third booster immunizations. Antisera were stored with 0 002% NaN3 in aliquots at — 20 °C.

The IgG fraction of each antiserum was prepared using a crude ammonium sulphate precipitation technique.
Saturated (NH4),S04 was added drop by drop to serum on ice to final volume of 50%. The serum was left for
30 minutes at 4 °C and then centrifuged at 14000 rpnr for 10 minutes. The supernatant was carefully pipetted
off and the pellet re-suspended in Tris buffered saline (20 mM Tris, 0-9% wt/vol NaCl, pH 7-5; TBS).

Crude antigen preparations for the immunological assay were produced by dissolving a slight excess of shell
fibre (or powder) preparation in 20% wt/vol EDTA (a ratio of 0 046 g fibre/1 ml EDTA), and then
centrifuging to remove the residual undissolved carbonate.

An enzyme-linked immunosorbent assay (ELISA; Harlow and Lane 1988) was used to measure
immunological distances among taxa. Both antigen (crude EDTA extract) and homologous antiserum were
used for all taxa examined, and all antigen-antibody combinations were produced. 10 p\ of each crude antigen
solution, diluted to 100 /d in 20% wt/vol EDTA, was pipetted onto the microtitre plate which was then
incubated at 37 °C for 90 minutes. Unbound antigen was washed away by rinsing three times with washing
buffer (TBS to which had been added 0-05 % wt/vol of the ionic detergent Tween 20). The wells were blocked
for 30 minutes with 2% wt/vol gelatin diluted in TBS. Antisera diluted in 0-2% wt/vol gelatin/TBS/Tween
20 were added to given wells and incubated overnight at 4 °C for L5 hours at 37 °C, following which any
unbound material was washed away. Alkaline phosphatase labelled goat-anti-rabbit-affinity purified IgG
(GAR-AP, Sigma) diluted I : 5000 in gelatin/TBS/Tween 20 was added to the wells for 1 hour at 37 °C after
which the wells were rinsed five limes with TBS/Tween, once with TBS and then allowed to stand for 2 minutes
with phosphatase substrate buffer (pH 9-2), prior to the addition of disodium p-nitrophenyl phosphate
(Sigma). After 15 minutes the reaction was stopped by the addition of 50 //I of 1 m NaOH, and the absorbence

788

PALAEONTOLOGY, VOLUME 34

at 405 nm of the wells read immediately with a Titretek Multiskan Plus, automated plate reader. The extent
of colour development is proportional to the amount of first antibody bound.

Immunological distance was calculated using the formula ID = lOOxloglO (100/A), where A is the mean
reciprocal % cross-reactivity (homologous antigen reactivity = 100%; see Table 2). These distances were
obtained from the linear regions of semi-logarithmic binding curves plotted using a series of antibody
concentrations for each combination of antigen and antibody (Text-fig. 2). Reciprocal distances were averaged
for each combination and means of duplicates were used for clustering.

Tree diagrams were constructed either using the method of UPGMA (Sneath and Sokal 1973) or Fitch and
Margoliash (1967) , as represented by FITCH in the program package PHYLIP, written by Joseph Felsenstein
(University of Washington, Seattle).

6

c

U->

o

N-

CD

O

c

CO

■8

o

cn

-O

"T

Waltonia

* —

Neothyris

Dallina

-0

Terebratulina

-» —

Gryphus

Liothyrella

-D

Mergerlia

-h

Kraussina

H

Notosaria

1000000

10000000

Antibody dilution

text-fig. 2. Immunological binding curves. A series of dilutions of the antiserum raised against Dallina , tested
against all nine antigens. Immunological distance (ID) between species is 100 times the log of the antiserum
dilution required to give the same binding for two different species. ID can be determined directly from the

linear portions of the binding curves.

RESULTS AND DISCUSSION

Taxonomic framework

In the brachiopod Treatise, the Order Terebratulida was divided into two superfamilies, the long-
looped Terebratellacea and the short-looped Terebratulacea. A simplified graphical representation
of the Treatise classification down to family level is shown in Text-figure \a. The classification is
based entirely on comparative morphology, and the graphical representation should not be
confused with the phenograms used to portray the immunological data. The ‘clustering' unit (i.e.
vertical axis) is taxonomic hierarchy for the representation of the Treatise classifications, while for
the immunological data it is a numerical scale reflecting the degrees of reactivity of the various
antisera. To emphasize this point, the analyses of immunological data are shown with the most
closely related taxa at the top (i.e. the immunological distance between clusters increases
downwards; Text-fig. 3), while the morphological interpretation is shown in the reverse orientation
with the lower levels of the taxonomic hierarchy (i.e. families, genera) at the bottom and orders, etc.
at the top (Text-fig. 1).

COLLINS ET AL.\ BRACHIOPOD SEROTAXONOMY

789

text-fig. 3. UPGMA phenogram of the immunological distances of nine genera of articulate brachiopods

(data from Table 2).

There have been a number of suggested modifications to the Treatise classification some of which
have been incorporated into Text-figure 1 b. The major complication with such modified schemes
is that they only deal with relatively few genera, and may leave many unanswered questions about
the authors' intended taxonomic assignments, if any, for taxa not discussed. The problem can be
circumvented in this paper by including only those groups for which immunological data are
currently available. This involves two new superfamilies which have been proposed by Cooper
(1973, 1981) and which, in effect, subdivides the two terebratulid superfamilies listed in the Treatise.

Immunological distances

Immunological distances among genera of the order Terebratulida are presented in Table 2 and the
UPGMA and Fitch-Margoliash dendrograms based on this data set are given in Text-figures 3 and 4.

At one extreme, the outgroup rhynchonellid Notosaria nigricans was the least reactive with all
eight terebratulid antisera, and hence was well separated from the terebratulids in both phenograms
(Text-figs 3 and 4). At the other extreme, all confamilial genera clustered together in both forms of
analysis (e.g. Megerlia with Kraussina , Gryphus with Liothyrella, and Neothyris with Waltonia;

790

PALAEONTOLOGY, VOLUME 34

table 2. Immunological distances among the articulate brachiopods from ELISA of intra-crystalline
macromolecules. Distances represent means of reciprocal distances obtained from binding curves. See Table
1 for species abbreviations.

Wi

Ds

NL

Tr

Gv

Li

Mt

Kr

Nn

Waltonia inconspicua

0

15

6

100

169

177

204

117

301

Dallina septigera

0

18

77

118

129

154

170

168

Neolhrvris lenticularis

0

88

156

149

130

132

239

Terebratulina retusa

0

134

70

146

206

253

Gryphus vitreus

0

14

45

15

193

Liothyrella neozelandica

0

34

54

272

Megerlia truncata

0

16

284

Kraussina rubra

0

306

Notosaria nigricans

0

o—i

100—

03

O

CD

O

O

c

E 200‘
E

300 —

CO

03

-o

03

Q

03

03

-a

03

05

_Q

03

03

03

"O

03

CO

c

03

O

CD

0)

CC3

cO

X3

"O

c

zs

c o

CC5

CO

_Q

ZJ

CO

CD

5

03

1—

Ds Wi Nl Tr

M t K r Gv Ln

Nn

14

Terebratulida

Order
Rhynchonellida

250

text-fig. 4. Fitch-Margoliash phenogram of the data from Table 2. Note that while the overall topology is
similar to Text-figure 3, the cancellothyrid lineage ( Terebratulina ) arises from the terebratulid, and not the

TLD, lineage.

Text-figs 3 and 4). These results, at the highest and lowest level of the taxonomic hierarchy
investigated, are entirely consistent with established morphology-based brachiopod systematics,
and reinforce our contention that intra-skeletal macromolecules are an important source of
phylogenetic information (Collins et al. 1988).

COLLINS ET AL.\ BRACHIOPOD SEROTAXONOMY

791

Within the terebratulids, the immunological distances distinguish three main clusters (Text-figs
3 and 4). The novel aspect of these three main clusters is that they do not coincide with the long-
and short-looped stocks which, as mentioned above, currently represent the primary subdivisions
of the terebratulids. Instead, two represent respectively the Cancellothyrididacea (raised to
superfamily status by Cooper 1973) and a sub-group of the Terebratellacea (the so-called TLD
group of Collins et al. 1988). The third is more heterogeneous, including both a long-looped family
(the Kraussinidae) and short-looped superfamily Terebratulacea.

The fact that the three-way clustering pattern is confirmed by two different methods of analysing
the data (Text-figs 3 and 4), and is based on fully reciprocal immunological distances, confirms that
this is an accurate reflection of the phylogenetic relationships among living terebratulids.

Plotting the immunological distances of the major nodes against estimated divergence times
(Table 3) reveals a strong correlation (Text-fig. 5). However, given that there is no universal
'molecular evolutionary clock’ (e.g. Britten 1986), it is necessary to put this apparent correlation
into context.

table 3. Estimated divergence time of major branches used in regression of ID against range.

Divergence

mean ID

Time (Ma)

Rationale

(first occurrence of...)

No divergence

0

0

Wi from NL

6

12

genus Waltonia

Mt from Kr

16

3

genus Kraussina

Li from Gv

14

15

genus Liothyrella

D1 from Wi, NL

16

15

subfamily Neothyridinae

Mt, Kr from Li, Gv

37

15

family Kraussinidae

Tr from DI. Wi, NL

100

150

genus Terebratulina

Mt, Kr, Li, Gv from DL, Wi, NL

163

205

subfamily Terebratulinae

Nn from others

295

470

Order Spiriferida
(ancestral to the
Terebratulida)

Divergence time (Ma 12)

text-fig. 5. Plot of immunological distance against range. Divergence times are estimated from the ages given

in Table 3.

792

PALAEONTOLOGY, VOLUME 34

The divergence times (listed in Table 3) used for the analysis are based on the first appearance
of the most recent probable ancestor, but these are not always well defined, and are subject to
significant change as continuing investigation of fossil brachiopods extends or reduces the known
range of taxa. In addition, although the approach of using polyclonal antisera against a large
number of different determinants will generate an averaged clock speed, which is beneficial, the
structure of individual antigenic determinants is not known. The immunological determinants of the
glycoproteins are believed to be mainly sugar moieties (due to their sensitivity to periodate and their
insensitivity to proteinase K treatments; Collins et al. in press), the production of which is
controlled by complex pathways, of whose evolutionary patterns we remain ignorant.

Evolution of the Order Terebratulida

The Order Terebratulida, which is first recorded from the Lower Devonian, is characterized by the
possession of a distinctive internal skeleton, ‘the loop’. The traditional interpretation of evolution
within this order is of considerable ‘experimentation’ with the internal skeleton in the early
Devonian. The period of diversification was relatively short (e.g. Stehli 1965; Rudwick 1970;
Williams and Hurst 1977), with two distinct lineages diverging in the late Devonian, giving rise to
the present-day Terebratulacea (short loop) and Terebratellacea (long loop).

The serotaxonomic data clearly indicate that this interpretation is in considerable need of
revision. Assuming the minimum age for diversification of the rhynchonellid and terebratulid
lineages is approximately 470 Ma at the first appearance of the order Spiriferida (believed to be
ancestral to the Terebratulida), then the great immunological distances between this primary
division and subsequent diversification suggests that surviving lineages diversified considerably later
than the Devonian.

The immunological data would suggest that the three modern-day lineages identified in this study
(TLD terebratellaceans, the cancellothyrids, and the ‘terebratulaceans’, the latter including the
Kraussinidae) diverged from each other at around the same time (Text-figs 3 and 4). An early
Mesozoic diversification is implied by the immunological data, which is consistent with the
appearance of the first terebratellacean and terebratulacean genera in the Upper Triassic, and the
earliest cancellothyrids from the Middle to Upper Jurassic (Muir-Wood et al. 1965).

The three major lineages have subsequently undergone different patterns of evolution. The most
morphologically conservative have been the cancellothyrids, which have changed very little since the
late Jurassic, and which consequently have the largest and older surviving terebratulid genus
Terebratulina , surely a case for revision?

The small immunological distances between the three members of the TLD cluster imply a high
degree of relatedness, which runs contrary to the assumption that members of this lineage diversified
as early as the Upper Cretaceous (Muir-Wood et al. 1965). Conversely, the immunological distances
which separate the Kraussinidae from the Terebratulidae are greater than is anticipated from a
geological record for the former family which extends back only into the Miocene.

The clustering of Macandrevia , previously included in the Dallinidae (Muir-Wood et al. 1965), or
Laqueidae (Richardson 1975), within the ‘kraussinid’ group (Text-figs 3 and 4) illustrates the
limited value of loop morphology for familial assignment. The value of the loop as a taxonomic
character is further reduced when (as is common in fossil samples) its ontogeny is difficult to
decipher. The serological data suggest a much closer relationship between Dallina and the
Terebratellidae than a Cretaceous diversification would imply. It is clear that future studies will
have to determine whether the accepted familial assignments of Cretaceous genera are justified, with
their added implication that both the laqueid and terebratellid lineages ( sensu Richardson 1975)
can be traced back to the Cretaceous.

The phytogeny of the Kraussinudae has recently been discussed by Brunton and Hiller (1990). If
the lineage also includes the Megathyrididae, as suggested by our previous study (Collins et al.
1988), and expected from current classification, the divergence of this lineage would be placed in the
late Mesozoic, which more closely corresponds with the immunological distances observed. The
major revelation arising from the immunological data, although anticipated in our earlier study

COLLINS ET AL.: BRACHIOPOD SEROTAXONOMY

793

(Collins et al. 1988), is not the divergence time of the kraussinid cluster, but its origin. The
immunological data clearly indicate that the lineage is derived, not from the long-looped
terebratellaceans (TLD group), but from the short-looped (terebratulacean) stock. In reviewing the
phylogeny of the kraussinids, Brunton and Hiller (1990) justifiably comment that our preliminary
results (Collins et al. 1988) linking the kraussinids to the terebratulaceans were surprising, given that
there is no obvious kraussinid-like ancestor within this lineage. However, additional antisera raised
for this investigation, against Kraussina , Megerlia , Liothyrella , and Gryphus , all confirm the original
finding. This result has major taxonomic implications, not only for the classification of the
kraussinids, but for the significance of the loop as a subordinal taxonomic character.

Taxonomic Implications

Immunological data cannot on their own determine the taxonomic relationships of terebratulid
brachiopods, because only Recent taxa are involved in the immunological distance experiments.
The data only relate to the phylogenetic relationships of the families to which the included genera
are assigned, and these assignments may be incomplete or wrong (e.g. Macandrevia; Collins et al.
1988). However, the immunological study highlights relationships which contradict current
morphological classifications.

The simplified cardinalia and distinctive loop morphology of the cancellothyrids make this the
most homogeneous of the extant terebratulid lineages.

The TLD terebratellaceans are characterized by supporting hinge plates on the median septum,
which distinguishes these terebratellaceans from the ‘aberrant’ genus Macandrevia , but this latter
organization is also seen in some Palaeozoic terebratulids and in other articulate brachiopod orders.
The loop, which is best developed within this group, has previously been cited as the major
discriminatory feature in the terebratulids (e.g. Stehli 1965). However, ontogenetically, there is a
strong relationship between long- and short-looped forms. The earliest stages in the development
of the descending branches of the calcareous loop up to their fusion with the median septum are the
same for all three extant lineages. Elliott (1953, 1957) notes that the first-formed calcareous support
for all modern long-looped brachiopods (with the exception of Argyrotheca) always includes a
dorsal median septum. Of the Palaeozoic superfamilies of terebratulids believed to be ancestral to
modern forms, all contain genera which possess median septa, but there is no clear evidence for the
involvement of the septum in the ontogeny of the loop in either the Stringocephalacea or
Zeilleriacea. If in these two superfamilies a long loop was derived simply by anterolateral growth
of a short dielasmatacean-like loop, then the role of the median septum in all modern long-looped
lineages is significant. The fixing of one novel character, a link between an acceleration in the time
and rate of growth of the median septum coincident with that of the descending branches was
probably all that was necessary to pave the way for the Mesozoic and Cenozoic diversifications
(Text-fig. 6).

Immunological distances provide an important new perspective on terebratulid evolution, and
pose problems for current morphologically-based taxonomy. Morphological taxonomy of a group
with such a ‘simple’ skeleton is prone to error, yet molecular taxonomy is generally inapplicable to
a predominantly fossil group. Therefore, a major taxonomic revision of this order appears
inevitable, and it should proceed as a rigorous re-analysis of skeletal ontogeny and morphology in
the light of a more widespread molecular investigation.

Acknowledgements. The authors wish to acknowledge the technical assistance of T. Zomerdijk, S. Forbes and
Y. van Zijl. Dr C. H. C. Brunton is thanked for his comments on an earlier version of the manuscript. The
research was supported by grants from the Royal Society (M.C. and G.B.C.), NERC (M.C. and G.B.C.), the
Lounsbery Foundation, New York (G.M. and P.W.), and NSF (P.W.).

j'p sZ'? l

ll m

i if

? a °

§ § *

ML

v?

JS!

I S

s

fi

|[

IS

1

I

I

I I

| Sil | Dev | Carb [ Perm

Dielasmatacea

I Strjingoi

Trias | Jur | Cret

/ Terebratulacea

j Tertiary |

Terebratellacea

TLD group)

i

i Dallinacea

Terebratellacea

Cancellothyrididacea

\ j T

erebratulacea

(Terebratellacea) Kraussinidae

Interpretation Interpretation based

on serotaxonomy

794

PALAEONTOLOGY, VOLUME 34

Traditional Interpretation Interpretation based

on serotaxonomy

CD

03

p

’c
'(/)
C D
13

5

£

03

CD

(J

03

75

-Q

0)

text-fig. 6. Terebratulid phylogeny (after Williams and Rowell 1965; Williams and Hurst 1977) reinterpreted

in the light of the serotaxonomic data.

REFERENCES

beecher, c. e. 1893. Revision of the families of loop-bearing brachiopods. Transactions of the Connecticut
Academy , 9, 376-391.

britten, r. j. 1986. Rates of DNA sequence evolution differ between taxonomic groups. Science , 231,
1393-1398.

brunton, c. h. c. and hiller, n. 1990. Late Cainozoic brachiopods from the coast of Namaqualand, South
Africa. Palaeontology , 33, 313-342.

COLLINS ET AL.\ BRACHIOPOD SEROTAXONOMY

795

cohen, b. l., curry, G. b. and balfe, p. 1986. Genetics of the brachiopod Terebratulina. 55-64. In rachboeuf,
p. r. and emig, c. c. (eds). Les brachiopodes fossites et actuels. Biostratigraphie du Paleozoic , 4, 1 500.

— balfe, p., cohen, m. and curry, g. b. 1991. Genetic divergence within and between populations of
North Atlantic morphospecies of Terebratulina retusa and Terebratulina septentrionalis. 109- 1 14. Proceedings
Volume of the 2nd International Brachiopod Conference. Balkema, Rotterdam.

COLLINS, M. j. 1986. Post mortality strength loss in shells of the Recent articulate brachiopod Terebratulina
retusa (L.) from the west coast of Scotland. 209-218. In rachboeuf, p. r. and emig, c. c. (eds). Les
brachiopodes fossiles et actuels. Biostratigraphie du Paleozoic, 4, 1-500.

— curry, g. b., quinn, r., muyzer, g., zomerdijk, t. and westbroek, p. 1988. Sero-taxonomy of skeletal
macromolecules in living terebratulid brachiopods. Historical Biology , 1, 207-224.

— muyzer, g., curry, g. b., Sandberg, p. and westbroek, p. (in press). Macromolecules in brachiopod
shells: characterization and diagenesis. Lethaia, 24.

— MUYZER, G., WESTBROEK, P., CURRY, G. B., SANDBERG, P. A., XU, S. J., QUINN, R. and MACKINNON, D. 1991.
Preservation of fossil biopolymeric structures: conclusive immunological evidence. Geochimica Cos-
mochimica Acta 55, 2253-2257 .

cooper, G. a. 1973. Fossil and Recent Cancellothyridacea (Brachiopoda). Tohoku University , Scientific
Reports , 2nd Series (Geology) (Special Volume), 6, 371-390.

— 1981. Brachiopoda from the southern Indian Ocean (Recent), Smithsonian Contributions to Paleobiology,
43. 1-193.

elliott, G. 1953. Brachial development and evolution in terebratelloid brachiopods. Biological Reviews, 28,
261-279.

1957. Accessory brachial structures in long-looped brachiopods. Geological Magazine, 94, 334-336.

— 1976. Comments on ‘The loop-development and classification of terebratellacean brachiopods’.
Palaeontology, 19,413^114.

fitch, w. m. and margoliash, e. 1967. Construction of phylogenetic trees. Science, 155, 1279-1284.
harlow, e. and lane, d. 1988 Antibodies : A laboratory manual. Cold Spring Flarbor Laboratory, New York,
420 pp.

lowenstein, s. m. 1985. Molecular approaches to the identification of species. American Scientist, 73, 541-547.

— sarich, v. m. and Richardson, b. j. 1981. Albumin systematics of the extinct mammoth and Tasmanian
wolf. Nature, 291. 1409-1411.

maxon, l. r. and maxon, r. d. 1979. Comparative albumin and biochemical evolution in plethodonlid
salamanders. Evolution, 33, 1057-1062.

muir-wood, H. m. 1955. A history of the classification of the phylum Brachiopoda. British Museum (Natural
History), London, 124 pp.

— STEHLI, f., ELLIOT, G. and hatai, K. 1965. Terebratulida. H728-H730. In moore, r. c. (ed.). Treatise on
invertebrate paleontology. Part H. Brachiopoda. Geological Society of America and University of Kansas
Press, Boulder, Colorado and Lawrence, Kansas, xxxii + 927 pp.

olsen-stojkovich, j., west, J. a. and lowenstein, J. M. 1986. Phylogenetics and biogeography in the
cladophorales complex (Chlorophyta) : some insights from immunological distance data. Botanica Marina ,
29, 239-249.

price, r. a., olsen-stojkovich, j. and lowenstein, j. m. 1987. Relationships among the genera of Pinnacea:
An immunological comparison. Systematic Botany, 12, 91-97.
richardson, j. r. 1975. Loop development and the classification of terebratellacean brachiopods.
Palaeontology, 18, 285-314.

rudwick. M. J. s. 1970. Living and fossil brachiopods. Hutchinson University Library, London, 199 pp.
sarich, v. M. 1973. The giant panda is a bear. Nature, 245, 218-220.

— and wilson, a. c. 1967. Immunological time scale for hominid evolution. Science, 158, 1200-1203.
sneath, p. H. a. and sokal R. p. 1973. Numerical taxonomy. W. H. Freeman, San Francisco, 573 pp.
stehli, f. G. 1965. Paleozoic Terebratulida. H730-H762. In moore, r. c. (ed.). Treatise on invertebrate

paleontology . Part H. Brachiopoda. Geological Society of America and University of Kansas Press, Boulder,
Colorado and Lawrence, Kansas, xxxii+927 pp.

weiner, s. and lowenstam, h. a. 1980. Well-preserved fossil mollusk shells: characterization of mild diagenetic
processes. 95-1 1 3. In hare, p. e., hoering, t. c. and KING, k. jr. (eds). Biogeochemistry of amino acids. Wiley,
New York, 558 pp.

williams, a. and hurst, j. m. 1977. Brachiopod evolution. 79-121. In hallam, a. (ed.). Patterns of evolution
as illustrated by the fossil record. Developments in palaeontology and stratigraphy, 5. Elsevier, Amsterdam,
591 pp.

796

PALAEONTOLOGY, VOLUME 34

— and rowell, a. j. 1965. Evolution and phylogeny. H164-H199. In moore, r. c. (ed.). Treatise on
invertebrate paleontology , Part H. Brachiopoda, Geological Society of America and University of Kansas
Press, Boulder, Colorado and Lawrence. Kansas, xxxii + 927 pp.

— MUIR-WOOD, H. M., PITRAT, C. W., SCHMIDT, H., STEHLI, F. G., AGER, D. V., WRIGHT, A. D., ELLIOTT, G. F.,
AMSDEN, T. W., RUDWICK, M. J. S., HATAI, K.., BIERNAT, G., MCCLAREN, D. J., BOUCOT, A. J., JOHNSON, J. G.,

staton, r. d., grant, r. e. and jope, H. m. 1965. Brachiopods, In moore, r. c. (ed.). Treatise on invertebrate
paleontology. Part H. Brachiopoda. Geological Society of America and University of Kansas Press, Boulder,
Colorado and Lawrence, Kansas, xxxii + 927 pp.

m. j. collins,1 G. muyzer and p. westbroek

Geobiochemistry Unit, Gorlaeus Laboratory
Leiden University
Leiden, Netherlands

G. B. CURRY and R. QUINN
Departments of Geology and Applied Geology
The University
Glasgow G12 8QQ, UK

S. EWING

Department of Immunology
University of Oxford
Parks Road, Oxford, UK

xu s.

School of Pharmaceutical Science
Beijing Medical University
Beijing, China

'Present address

Typescript received 10 August 1990
Revised typescript received 1 January 1991

Department of Geology
University of Bristol
Bristol BS8 1RJ, UK

FEEDING STRATEGIES IN GRAPTOLOIDS

by SUSAN RIGBY

Abstract. The progressive loss of stipes through the Ordovician and the appearance of uniserial forms in the
Silurian mark obvious changes in graptoloid morphology through time. However, several features of
graptoloids are shown here to have remained remarkably constant. These include stipe width and the number
of thecae per cm. The number of zooids in an entire colony (assuming a one-to-one ratio with the thecae)
changes abruptly at the Ordovician-Silurian boundary but the range of zooid numbers remains the same. The
feeding areas of graptoloids, and the intensities with which they fed have been measured, assuming that
graptoloids rotated as they moved through the water. The mean and range of feeding intensity remained
constant through time although the total number of zooids changed. Silurian curved monograptids plot in the
same place as inclined biserial forms and tetragraptids on a feeding intensity-area graph. Multiramous colonies
in both the Ordovician and Silurian plot in the same place. Scandenl biserial forms plot with straight
monograptids. It is postulated that graptoloids with low feeding intensities lived in areas of low food
availability. High intensity feeders would have lived where food was more abundant. Colony size (measured
as total number of zooids) might have related to the dependability of food supply.

There are obvious changes in graptoloid morphology through time. Early Ordovician forms with
many stipes gave rise to forms with progressively fewer stipes, although a multiramous component
to the fauna was almost always present. Monograptids first appeared close to the Ordovician-
Silurian boundary, and came to dominate Silurian faunas. They gave rise to new types of
multiramous graptoloids, produced by cladial generation rather than by dichotomy.

This constant change in colony form allows accurate biostratigraphical correlation of the
Ordovician and Silurian periods. The reasons for the change are unknown. Bates and Kirk (1985)
postulate that the initial radiation of planktonic graptoloids resulted in ‘ ... increased automobility,
giving access to the higher and more food rich layers of the water’. Later reductions in the number
of stipes are suggested to have prevented graptolites from becoming tangled with one another in the
crowded Palaeozoic seas. Rickards (1975) suggests that the initial reduction in stipes reduced the
weight of the colony. Later changes in the orientation of the stipes may have served to remove the
delicate proximal end from turbulent surface waters.

It would be expected that radical changes of form would mirror changes in life habit. However
graptoloids were colonial. Changing the colony form would not necessarily have changed the form
of the zooids. A simple change in 'building instructions’ to a colony would have resulted in radical
changes to the overall form, but the colony would have survived if this form was in any way
advantageous. This contrasts with non-colonial organisms where most mutations to the transcribed
DNA are fatal because they affect so many metabolic processes. Thus changes with such serious
implications for biostratigraphy as the first appearance of monograptids need not be seen as a
response to changes in the environment or in graptoloid life habit.

The graptoloid rhabdosome needed to perform a number of functions for the colony to have
survived. Only some of these are known at the present time. One was the need to present a
hydrodynamically stable form to the water so that the zooids would not interfere with one another
whilst feeding. Scandent colonies needed to remain scandent ; horizontal forms needed to retain that
orientation. Hydrodynamic needs may also have included including rotation to the colony as it
moved through the water (Rigby and Rickards 1989). Another need was to present the zooids to
the water in an efficient feeding array for the available food. The methods by which this was
achieved in multiramous colonies have been investigated by Fortey and Bell (1987).

It is important to note that a range of morphologies could have achieved the same result. This

IPalaeontology, Vol. 34, Part 4, 1991, pp. 797—813.)

© The Palaeontological Association

798

PALAEONTOLOGY, VOLUME 34

is certainly seen in multiramous colonies, where Silurian forms with curved cladia achieved a high
coverage of the available area, just like Ordovician forms which were multiramous due to
dichotomizing branches (Fortey and Bell 1987).

The purpose of this paper is to explore some aspects of graptoloid morphology which are
common to every graptoloid regardless of the number and orientation of the stipes. Changes in
these features through time can be compared to overall changes in morphology, to see whether
these large scale changes necessarily marked a significant change in life habit.

METHODS

The features (Text-fig. 1) chosen were: (1) number of zooids per centimetre of stipe; (2) stipe width;
(3) total number of zooids in the colony (Z); (4) area (A) or diameter (D) of feeding circle; and (5)
feeding efficiency (Z/A or Z/D).

Total zooids in

ameter (D)

colony

text-fig. 1 . Measurements made on graptoloids for this study.

(Z)

Measurements of zooids per cm stipe width, total number of zooids in the colony, feeding area
and feeding intensity were made on the species figured by Elies and Wood (1901-18) in their
Monograph of the British Graptolites , with the exception of the retiolitids and other forms with
skeletal rhabdosomes. Two hundred and forty two species were included. This data set was chosen
because it runs from the Tremadoc to the end of the Silurian. Some species have been synonymized,
and many more defined since this work was done. Some zonal boundaries have been redefined.
However, it remains the only comprehensive view of British graptoloids and so was used in its
entirety, without addition or alteration. This should have avoided any changes in bias derived from
different interpretations of what constitutes a species which would have resulted in different authors’
views had been taken for different zones.

In some cases the drawings and text descriptions were at odds. In these cases the drawings were
used as it was necessary to use these for measurements of feeding area. In the case of some
multiramous forms, e.g. Cyrtograptus murchisoni , graptoloids had to be reconstructed from
incomplete figured specimens. In these cases the branching rules used in the fragments of

RIGBY: GRAPTOLOID FEEDING

799

rhabdosome were followed, and the overall form assumed to have been circular (Bulman 1970).
Forty specimens were measured from the original specimens and these values compared with those
obtained from the drawings in Elies and Wood (1901-18). There were no significant differences in
the readings, with an average discrepancy between the two of 0-5 mm.

The number of thecae per cm is given in species descriptions. It is normally a range, and in these
cases the highest value was taken. A rhabdopleuran model was assumed for the graptoloid zooid
(Rickards 1975) with one zooid filling one theca. It is possible that a Cephalodiscus type strategy
might have been adopted, with several zooids for each thecal cup. This would necessitate
multiplying the number of thecae by a correction factor.

Stipe width was measured as the distance between the tip of a theca and the dorsal wall of the
rhabdosome. It was taken as the maximum width on the drawing of the specimen, or from the width
given in the species description if it was apparent that the colony was flattened. It is possible that
an exaggerated value was given for some specimens where only flattened specimens have been
recovered.

The total number of zooids in the colony was measured from Wood’s drawings or taken from the
maximum colony length given in species descriptions when this gave a greater value. In these cases
the maximum length given was multiplied by the number of thecae recorded per cm. This
measurement would obviously have changed continuously as the colony grew. Some species may
have continued growing indefinitely, for instance Monograptus flemingi, of which there is a specimen
in the Sedgwick Museum (A52567) 75 cm long. Other graptoloids would have reached a maximum
growth size beyond which they could not physically progress. An example is Skiagraptus which
narrows distally to a point. As a first order approximation, the maximum length described by Elies,
or the length of the specimen drawn by Wood was taken, whichever was longer.

Feeding area was calculated assuming that the colony fed during spiralling motion through the
water column (Rigby and Rickards 1989). This is accepted to have been unlikely in all colony forms.
However, almost all of the species so far modelled had the capacity to rotate. These include
Pseudoclimacograptus , Triaenagraptus , Cyrtograptus , dicellograptids, dicranograptids, hooked and
straight monograptids, Orthodichograptus , Trochograptus, Dichograptus, Loganograptus , and
Nemagraptus. Standard orientations were assumed. There is some evidence for these orientations
(Rickards and Crowther 1979; Fortey and Bell 1987) and more is emerging from physical
modelling. It makes no difference which way up the rhabdosome is assumed to have lived, sicula
up (Bates and Kirk 1985 and references therein) or sicula down (Bulman 1964). Each colony would
have spiralled through a column of water of unknown length. The area of a circular section through
this tube can be measured. The diameter of this circle is given by the width of the colony (Text-fig.
1). The assumption of rotation is crucial; if it did not occur then a circular area of water would
have been tapped by few species.

Feeding intensity was measured by dividing the total number of zooids by the area of water from
which they must have fed. In Text-figure 6 this is plotted as Z/D, i.e. as the total number of zooids
divided by the diameter of the feeding circle. This was done to reduce the range needed to plot the
graph which is greatly increased when area is used. The total zooids were divided by the area of
feeding circle for all other graphs.

No allowance was made for possible different rates of rotation, or different speeds of movement
through the water. These factors cannot at present be defined with any degree of accuracy. Some
graptoloids may have achieved larger colonies than those preserved or yet discovered. Post-mortem
effects may have distorted the stipe thickness and colony diameter recorded for some species. So all
of the number given must be regarded as approximations. They are not presented as highly accurate
values but as readings which allow comparison between apparently dissimilar morphologies.

RESULTS

Zooids per an

The mean number of zooids per cm in Ordovician graptoloids was 11 -8. In Silurian forms it was

800

PALAEONTOLOGY, VOLUME 34

11-4 cm-1. In both cases the mode is lOcirr1. Standard deviations are 2-5 cm"1 and 2-9 cm"1
respectively (Text-fig. 2). In effect there is no significant change in these values across the
Ordovician-Silurian boundary. The range of values, however, was greater in the Silurian, with a
maximum of 28 cm"1, and a minimum value of 1 cm"1. This contrasts with an Ordovician range of
14 cm"1, between 5 cm"1 and 19 cm"1. In the Silurian the greater range is largely due to relatively
uncommon or short lived genera such as Rastrites.

These values compare with Fortey’s (1983) 'average graptoloid’ which had 9-10 thecae per cm.

Stipe width

The modal value of stipe width in both Ordovician and Silurian forms is 1-5 mm. The mean rises
from T8 mm in the Ordovician to 2-2 mm in the Silurian. The standard deviation rises from FI mm
to 1-6 mm (Text-fig. 3). These two changes are almost wholly due to the appearance of the genus
Rastrites in the Silurian. These forms have isolated thecae which can be 18 mm long. When these
forms are excluded from the calculation, the mean for the Silurian becomes l-88mm and the
standard deviation IT mm, effectively the same as Ordovician values. Fortey’s (1983) ‘average
graptoloid’ had a stipe 2-2 mm wide.

Total number of zooids in colony

The mean number of zooids in a colony is 171 in the Ordovician, and falls to 95 in the Silurian. The
mode falls from 250 to 25. The standard deviation for the Ordovician forms is 205, showing massive
variability in colony size. In the Silurian it is 94 (Text-fig. 4). This is a significant difference between
graptoloids of different ages.

Feeding diameter

Ordovician feeding diameters range from 1 mm to 640 mm, with a mean of 56 mm and a mode of
2-5 mm. The standard deviation is 10-3 mm. By contrast, the Silurian range is from 0-5 mm to
140 mm, with a mean of 9-7 mm and a mode of 0-25 mm. The standard deviation is 2T mm,
reflecting the smaller, more clustered distribution (Text-fig. 5). The large number of modal
graptoloids in each case make the distributions appear similar. A Spearman’s Rank Correlation
Coefficient test was applied to the data, using the null hypothesis that there was no correlation
between the two frequency distributions. This test was chosen as it is non-parametric and suitable
for grouped data (Hammond and McCullagh 1974). The null hypothesis had to be rejected when
the correlation between the two distributions was found to be significant at the 0-01 level.

Feeding intensity

The range of feeding intensities shown by graptoloids of both Ordovician and Silurian age is from
less than 10 to almost 1500 zooids per diameter cm (informally labelled as .vcm(D)"1). The mean
value is 254 cnr(D)"1 in the Ordovician and 281 cm(D)"1 in the Silurian. In both cases the mode is
750 cm(D)"1. Standard deviation is 291 cm(D)"1 in the Ordovician and 297 cm(D)"1 in the Silurian
(Text-fig. 6). In the light of the enormous range of variation which is present, these differences are
negligible. A Spearman’s Rank Correlation Coefficient was applied, using the null hypothesis that
there was no correlation between the two distributions. This had to be rejected as the data were
found to be significantly correlated only at the 0-5 level.

FEEDING INTENSITY-AREA PLOTS

Feeding intensity can be plotted against the area of the feeding circle for each species of graptoloid.
The resulting graphs are shown in Text-figure 7. The range of variation is great and log scales need
to be used on both axes. This makes it appear that there is a straight line correlation between the
two, but this is misleading. The plots are the result of a range of factors and Z/A is not simply
correlated with A.

The variables on these graphs are linked in such a way that as A increases, Z/A will tend to

RIGBY: GRAPTOLOID FEEDING

801

text-fig. 2. Number of zooids per cm plotted against time. Spot size is proportional to the number of readings
in each position. The bar graph represents the same data, on the same v-axis but summarized for the
Ordovician and Silurian. Note that a species was counted once for every zone in which it was lound. This is

true for all of these graphs.

802

PALAEONTOLOGY, VOLUME 34

text-fig, 3. Stipe width plotted against time (top) and summarized as a bar graph for the Ordovician and

Silurian (bottom).

decrease. This contributes to the negative correlation seen on the graphs. However, the correlation
between the two variables is not the source of interest of the plots. They should instead be regarded
as a crude section through graptoloid morphospace. The variables are plotted separately in Text-
figures 5 and 6, but plotting both on the same graph is convenient shorthand.

The important result from these graphs is the striking similarity of the Ordovician and Silurian
plots. It must be remembered that feeding intensity is a function of several factors, including the
total number of zooids in the colony. The mean of this value changes abruptly at the

No. of zooids in coiony No. of zooids in colony

RIGBY: GRAPTOLOID FEEDING

803

1000 +
600-999
500-599
400-499
300-399
200-299
190-199
180-189
170-179
160-169
150-159
140-149
130-139
120-129
100-119
100-109
90-99
80-89
70-79
60-69
50-59
40-49
30-39
20-29
10-19
0-9

• • •

• • • • •

• •

• • •

• • •

• • •

• • •

• •

• • •

• •
• •

• • • •

• •

• • •

• • •

• •••••• •

• • •• • • •

• • • • •

• • •

• •

1000 +
600-999
500-599
400-499
300-399
200-299
190-199
180-189
170-179
160-169
150-159
140-149
130-139
120-129
110-119
100-109
90-99
80-89
70-79
60-69
50-59
40-49
30-39
20-29
10-19
0-9

Liandovery

Wenlock Ludlow

1

i

t r~

a

?

Ordovician rfl#

Lj Silurian

n. ,

MME. n

fi

v: s

i

40

20

20

40

Frequency

text-fig. 4. Total number of zooids in a colony plotted against time (top) and summarized as a bar graph for

the Ordovician and Silurian (bottom).

Diameter of feeding circle (cm) Diameter of feeding circle (cm)

804

PALAEONTOLOGY, VOLUME 34

Frequency

text-fig. 5. Diameter of feeding circle plotted against time (top) and summarized as a bar graph for the

Ordovician and Silurian (bottom).

RIGBY: GRAPTOLOID FEEDING

805

text-fig. 6. Feeding intensity, measured as total zooids/diameter of feeding circle, plotted against time (top)
and summarized as a bar graph for the Ordovician and Silurian (bottom). Note that the change of scale at high
feeding intensity values makes a false peak appear at the top of the bar graph. This is not a histogram.

806

PALAEONTOLOGY, VOLUME 34

<

N

log

10 000

1000

100

10

0.1

0.001

• Ordovician
o Silurian

«°o*

•o°°

••

log

0.01

0.1

10

100

1000

Area of feeding circle (A) (cm 2)

text-fig. 7. Feeding intensity (measured as total zooids/area of feeding circle) plotted against feeding area.
Note the similarity of plots for the Ordovician and Silurian data. The apparent straight line correlation is a

function of using log values on both axes.

Ordovician-Silurian boundary. But this derived plot remains the same. The area of feeding circle
must have changed at the same time in such a way as to result in the feeding intensity remaining
the same. This is shown to have been the case in Text-figure 5 where the feeding circle of Silurian
graptoloids is seen to be much smaller than that for Ordovician forms.

The shapes of rhabdosome which produce plots in different parts of the graph are shown in Text-
figure 8 for the Ordovician and Text-figure 9 for the Silurian. High intensity feeders with a small
feeding circle are scandent biserial forms in the Ordovician and early Silurian. Later in the Silurian
these are ‘replaced’ by straight monograptids. The word replaced is used in inverted commas
because there is no proof that this was an active or competitive replacement within the environment.
It is simply replacement in a given area of the graph, which could have been the result of passive
processes unrelated to competition. However, it could indicate a refilling, in the Silurian, of feeding
niches which were unoccupied following the end Ordovician extinction. The recovery after this
extinction event follows the classic exponential increase in diversity after a slow initial growth in
diversity. This diversity explosion probably began at the base of the persculptus Zone (Rickards
1988). The lowest diversity of all is found in the extraor dinar ius Zone, with only a few biserial types
present. In the persculptus Zone these become more numerous, and rare uniserial scandent forms
appear. By the top of the acuminatus Zone, a diverse biserial population is supplemented by
dimorphograptids and several genera of scandent monograptids, including Atavograptus ,
Lagarograptus and Coronograptus (Rickards 1988).

RIGBY: GRAPTOLOID FEEDING

807

2

Area of feeding circle (A) (cm )

text-fig. 8. This graph demonstrates where a range of Ordovician graptoloids plot on the feeding
intensity - area of feeding circle diagram. Key: a, Amplexograptus arctus; b, Climacograptus biconus;
c, Petalograptus minor; d, Cryptograptus antennarius; e, Dicranograptus furcatus; f, Didymograptus stabilis;
g, Phyllograptus cf. typus; h, Didymograptus bifidus; i, Tetragraptus reclinatus; j, Didymograptus gibberulus;
k, Didymograptus hirundo; 1, Dichograptus octobrachiatus.

Moderate intensity feeders with moderate feeding areas are inclined biserial forms and
tetragraptids in the Ordovician. In the Silurian this area of the graph is filled by curved or spiral
monograptids. Large feeding area forms always have low feeding intensities. The Ordovician forms
include horizontally disposed didymograptids and multiramous graptoloids like Loganograptus.
These are ‘replaced’ on the Silurian graphs by cyrtograptids and other multiramous forms
generated with cladia.

THEORETICAL AND STRUCTURAL LIMITS TO FEEDING STRATEGIES

Why do graptoloids plot where they do? Some theoretical limits can be applied (Text-tig. 10)
because of the small number of basic morphologies adopted by graptoloids. These limits are not,
in themselves, environmental indicators although the presence of a given type of graptoloid at a
given locality might be.

Vertically disposed forms could not have had a very low feeding intensity unless they had very

808

PALAEONTOLOGY, VOLUME 34

<

N

TO

®

i_

TO

W

“2

©

o

N

tB

e

H

O.OOI 0.01 0.1 1 10 100 1000

Area of feeding circle (A) (cm2 )

text-fig. 9. This graph demonstrates where a range of Silurian forms plot on feeding intensity - area
diagrams. Key: a, Monograptus regularise b, M. barrandei ; c, M. colonus ; d, M. fimbriatus ; e, M. scanicus ;
f, M. turriculatus ; g, M. involutuse h, M. cyphus; i, Cyrtograptus murchisoni bohemicus\ j, M. convolutus.

broad stipes. For instance, a graptoloid with stipes 2 mm wide would need to have had less than one
zooid to have a feeding intensity below 10. This is clearly impossible. A lower limit to the feeding
efficiency of a vertically disposed graptoloid with any width of rhabdosome is given by the need to
have had a minimum of one zooid.

No upper limit to theoretical feeding intensity can be predicted for vertically orientated forms.
Some graptoloids could, perhaps, have carried on growing indefinitely, reaching higher and higher
feeding intensities. The upper limits on this graph may be entirely a function of preservation - long
graptoloids are less likely to be preserved intact. However, some graptoloids can be seen to have
reached a mature stage where growth had ceased (C. M. Mitchell, pers. comm.), so this is not
entirely the case. Long graptoloids may have become structurally weaker as they grew, or their
growth may have been limited by food availability.

A lower possible limit for the feeding efficiency of horizontal forms is the need to cover a given
diameter of feeding circle. Thus, for example, a colony with a feeding diameter of 10 cm needed at

RIGBY: GRAPTOLOID FEEDING

809

graph.

least 10 cm of stipe and would have fed over a circle of 78-5 cm2 (nr, where r— 5). It would
inevitably have contained approaching 100 zooids, as horizontal forms tend to have around 10
zooids per cm (R. B. Rickards pers. comm.). The lowest possible feeding intensity for this form
would have been 100/78-5 cm2. A colony 20 cm in diameter could not have had less than about 200
zooids, and thus a feeding efficiency of 200/3 14-16 cm2. The lowest possible feeding efficiency for
horizontal forms is given by the minimum number of zooids needed to achieve a given feeding
diameter.

An upper possible limit for horizontal forms can also be defined. This is where the total number
of zooids filled the available space completely. In the case of a colony 10 cm in diameter, the area
available for feeding on was 78-5 cm2. If each zooid required, for example, 1 mm2 of space in order
not to compete with its nearest neighbour, then there could have been a maximum 7850 zooids in
the colony. This would have completely filled the available space. It would also have been
impossible to reach because no pattern of branching arms could completely fill a circle. There is a
structural upper limit imposed by the branching patterns which graptoloids evolved. However, there
is good evidence to suggest that many multiramous forms developed patterns of branching which
maximized their coverage of the available area (Fortey and Bell 1987). In order to increase their
feeding intensity further within a given feeding circle it would have been necessary to become dome-
shaped.

Total zooids/area (Z/A) Total zooids/area (Z/A)

810

PALAEONTOLOGY, VOLUME 34

Area of feeding circle (A) (cm2)

text-fig. 1 1 . Inclined biserial and tetragraptid forms modelled mathematically with the region over which they

plot on a feeding intensity - area graph shown.

RIGBY: GRAPTOLOID FEEDING

811

Inclined and curved forms can be regarded as intermediate between vertically and horizontally
disposed forms. For graptoloids with straight inclined stipes, lower and upper limits to feeding
efficiency depended on the inclination of the stipes (Text-fig. 1 1 ). A simplifying assumption is made
that the geometry of curved forms approximated to triangles with the hypotenuse missing (Text-fig.
12). This shape is rare in nature but does occur, e.g. Monograptus limatulus. Upper and lower limits
to feeding efficiency depended on the inclination of this hypotenuse for a given areas of feeding
circle.

Area of feeding circle (A) (cm2)

text-fig. 12. Curved monograptids modelled mathematically with the region over which they plot on a feeding

intensity - area graph shown.

The same area on the graph was covered by curved monograptids and inclined, biserial forms.
Inclined tetragraptids obviously plot with higher feeding efficiencies for a given area of feeding
circle.

CONCLUSIONS AND DISCUSSION

The first conclusion concerns the appearance of monograptids. The decrease in the number of stipes
was matched by greater variability in the number of zooids per cm and in the width of stipe than
that present in earlier forms. The degree of curvature of stipes in some species also increased.

Although the mean total number of zooids in a colony fell sharply when monograptids appeared,
the mean feeding intensity, which is partly a function of total zooids, retained the same mean, mode
and standard deviation across the Ordovician-Silurian boundary. This was achieved by curvature
of some stipes, and by variations in the colony width for straight, scandent forms. As a result, the

PALAEONTOLOGY, VOLUME 34

log log graphs of feeding area versus intensity are strikingly similar for the Ordovician and Silurian
graptoloids. This suggests a uniformity of function beneath the obvious dissimilarity of colony
forms at different times.

Changes in the number of stipes, their orientation, the total number of zooids in the colony and
their packing would all have served to present the zooids of a given colony to the water an effective
feeding array. Competition must have meant that forms as different as Cyrtograptus and
Monograptus priodon , or Petalograptus and Trochograptus were suited to different niches.
Presumably all of these forms presented effective feeding arrays, but in niches with different
availabilities of food.

The feeding area intensity plots are interpreted in the following way. High intensity feeders
must have required an intense concentration of food in the environment. Conversely, a low feeding
intensity is interpreted as meaning that the environment in which the graptolite lived had a low food
abundance. The size of the colony can be interpreted as a measure of dependability of food supply.
In order for a colony to become large, food needed to be continuously available for a relatively long
period of time, unless colony growth could occur discontinuously. Small colonies may have been
the result of highly variable food availability, perhaps of seasonality.

It has been suggested (Bates and Kirk 1985; R. A. Cooper, pers. comm.) that large, multiramous
colonies in the lower Ordovician were confined to deep water positions. If true, it fits well into the
feeding strategy model as multiramous forms were low intensity feeders. Fortey and Cocks (1986)
used observations of graptoloid faunas combined with palaeogeographic reconstructions to suggest
that an isograptid fauna of early Ordovician graptoloids lived in an offshore, oceanic environment.
Typical genera include Pseudisograptus , Oncograptus , Cardiograptus , Goniograptus and
Sigmagraptus. All of these would have had moderate to low feeding intensities.

Two general models have been proposed for graptoloid life habit. The first suggests that neutrally
buoyant graptoloids remained static in the water and fed on particles of food which moved past
them, at least in their early growth stages (Finney 1979). The alternative view is that graptolites were
mobile in some way and moved through a column of water, either passively (Bulman 1964;
Rickards 1975), or by automobility (Kirk 1967; Bates and Kirk 1985). This second view is
supported by modelling experiments conducted by the author (Rigby and Rickards 1989), although
the mechanism of movement remains obscure. If the first view is correct, then availability of food
would have directly controlled the number of zooids per centimetre of stipe. If the second view is
correct, food availability would have controlled the feeding intensity of the colony, i.e. the total
number of zooids buffered by the area of water available for them to feed on.

If the mobile model is accepted, then the results of this paper should be interpreted as follows.
The similarities of feeding intensity in the Ordovician and Silurian indicates the same range of
productivity at both times. A reduction in the mean total zooids in the colony is compensated for
by changes in stipe curvature and thickness.

If the static model of graptoloid life habit is accepted, then the results from the number of zooids
per centimetre are the most important. They suggest the same average food supply across the
Ordovician-Silurian boundary, but with an increased range of abundances, perhaps implying a
greater range of environments. The general decrease in total zooids in a colony suggests that in
general food supply was less predictable in the Silurian.

In either model, a similar level of food availability is predicted for the Ordovician and Silurian.
This was exploited by graptoloids with a range of different rhabdosome shapes. Strategies for
exploiting a given food abundance changed with time but can be reconstructed by means of the
graphs documented here. These bring out similar patterns which have previously been hidden by the
huge range of graptoloid morphology.

FUTURE WORK

It is now necessary to test this model of food availability and its control on graptoloid morphology
in real locations where the palaeogeographic setting is well known. This in itself is still a major
problem. Areas of upwelling should be contrasted with known high latitude sites where productivity

RIGBY: GRAPTOLOID FEEDING

813

would also have been high but possibly more seasonal, especially when permanent ice was present
at the poles. These results in turn should be compared to those from deep water sites where
productivity would have been much lower but more reliable.

It is important to see how different astogenetic stages of various morphologies plot on the graphs.
Horizontal colonies could theoretically have kept the same feeding intensity as they grew, but this
must have changed in other forms. In curved monograptids an abrupt change must have occurred
when curvature first began to develop. This is sometimes several centimetres along the stipe. There
would have been a progressive increase in feeding intensity with growth in straight, scandent forms.

The biggest single problem facing this kind of investigation is that of taphonomy. Graptoloids
are most often found in a broken state, where measurements of feeding areas and intensities are
impossible. What is needed is a method of reconstructing graptolite assemblages from their
preserved remains. Work by Budd (1990) seems to show that the rate of deposition was a primary
control on graptolite break-up, with slow burial resulting in greater breakage. Computer
programs allow these breakage effects to be removed. This opens the way for further study on this
problem. It may eventually be possible not only to understand the ecology of graptolites much
better than at present, but also to produce a useful tool for understanding the oceanography of the
early Palaeozoic seas.

REFERENCES

bates, d. E. b. and kirk, n. h. 1985. Graptolites, a fossil case history of evolution from sessile, colonial animals
to automobile superindividuals. Proceedings of the Royal Society of London, Series B. 228, 207-224.
budd, G. e. 1990. A new approach to post-mortem processes in graptolites. Unpublished undergraduate thesis,
University of Cambridge.

bulman, o. M. B. 1964. Lower Palaeozoic plankton. Quarterly Journal of the Geological Society of London, 120,
455-476.

1970. Graptolithina. V1-V163. In teichert, c. (ed.). Treatise on invertebrate paleontology. Part V, 2nd
edition. Geological Society of America and the University of Kansas Press, Boulder and Lawrence,
xxxii -(- 163 pp.

elles, G. l. and wood, e. m. r. 1901-18. A Monograph of British Graptolites. Monograph of the
Palaeontographical Society , 538 pp. (1901, pp. 1-54; 1902, pp. 55-102; 1903, pp. 103-134; 1904, pp.
135-180; 1906, pp. 181-216; 1907, pp. 217-272; 1908, pp. 213-358; 1911, pp. 359^114; 1913, pp. 415-486;
1914, pp. 487-526; 1918, pp. 527-538).

finney, s. c. 1979. Mode of life of planktonic graptolites: flotation structure in Ordovician Dicellograptus sp.
Paleobiology , 5, 31-39.

fortey, r. a. 1983. Geometrical constraints in the construction of graptolite stipes. Paleobiology , 9, 1 16-125.

— and bell, a. 1987. Branching geometry and function of multiramous graptoloids. Paleobiology , 13, 1-19.

— and cocks, l. r. m. 1986. Marginal faunal belts and their structural implications, with examples from the
Lower Palaeozoic. Journal of the Geological Society of London , 143, 151-60.

hammond, r. and McCULLAGH, p. s. 1974. Quantitative techniques in geography: an introduction. Clarendon
Press, Oxford, 319 pp.

kirk, n. h. 1969. Some thoughts on the ecology, mode of life and evolution of the Graptolithina. Proceedings
of the Geological Society of London , 1659, 273-292.

rickards, r. b. 1975. Palaeoecology of the Graptolithina, an extinct Class of the Phylum Hemichordata.
Biological Reviews of the Philosophical Society of Cambridge, 50, 397-436.

— 1988. Graptolite faunas at the base of the Silurian. Bulletin of the British Museum ( Natural History ),
(Geology), 43, 345-349.

— and crowther, p. r. 1979. New observations on the mode of life, evolution and ultrastructure of
graptolites. 397-410. In larwood, g. and rosen, b. r. (eds). Biology and systematics of colonial organisms.
Systematics Association Special Volume, 11, 1-589.

rigby, s. and rickards, r. B. 1989. New evidence for the life habit of graptoloids from physical modelling.
Paleobiology , 15, 402-413.

SUSAN RIGBY

Typescript received 29 January 1990
Revised typescript received 30 October 1990

Department of Earth Sciences
University of Cambridge
Cambridge CB2 3EQ, UK

MORPHOLOGY AND SHELL MICROSTRUCTURE OF
CRETACEOUS THECI DEI DINE BRACHIOPODS AND
THEIR BEARING ON TH ECI DEI DINE PHYLOGENY

by PETER G. BAKER

Abstract. New morphological and microstructural information from Bifolium faringdonense (Davidson,
1874), Bosquetella campichei (de Loriol. 1872), Thecidiopsis tetragona (Roemer, 1839) and Thecidiopsis
bohemica imperfecta Nekvasilova, 1967 indicates that the specimens assigned to T. bohemica imperfecta do not
belong to Thecidiopsis. The organization and microstructure of the brachial valve, place T. bohemica imperfecta
close to thecidellinids such as the Recent Thecidellina blochmanni Dali, 1920. In comparison with other
Cretaceous monoseptate forms, differences in morphology and shell microstructure clearly separate T.
bohemica imperfecta , B. faringdonense and B. campichei , and are considered to be distinctive enough to allow
T. bohemica imperfecta to be assigned to a new genus, Eothecidellina. The discovery of canopied brachial lobes
in B. faringdonense and rudimentary canopies in the Middle Jurassic Moorellina dubia (d'Orbigny, 1850),
together with the clarification of the systematic position of Eothecidellina imperfecta (Nekvasilova), provides
a much clearer picture of thecidellinid evolution. The correlation of the morphology and shell microstructure
of T. tetragona with basal Middle Jurassic genera and the narrowing of the thecidein plexus of descent, enables
a revised thecideidine phylogeny to be presented.

Discovery of a Middle Jurassic thecideidine with a partially suppressed fibrous secondary shell
mosaic (Baker and Elston 1984), recognition of shell microstructural features traceable over long
periods of time (Baker 1989), and studies on ontogeny (Smirnova 1969, 1984) pointed to weaknesses
in existing taxonomic and phylogenetic frameworks for thecideidines. The evidence indicated that
conclusions about thecideidine evolution based on brachial lobe morphology (Pajaud 1970) and the
time of onset of the suppression of fibrous secondary shell secretion (Williams 1973) had been
premature. Smirnova’s (1984) contention that Thecidellina arose from Upper Aptian Bifolium stock,
prompted the restudy of Upper Cenomanian monoseptate forms previously assigned (Nekvasilova
1967) to Thecidiopsis , and renewal of the search (Baker and Laurie 1978) for better-preserved
specimens of Bi folium. The further suggestion that Thecidellina might be traced back to a Jurassic
moorellinid ancestor (Smirnova 1984) revived interest in the smaller species of Moorellina. When
it was shown (Baker and Elston 1984; Baker 1989) that the Lower Cretaceous (Upper Valanginian)
Thecidiopsis tetragona (Roemer, 1839) could also be traced back to Middle Jurassic (Aalenian)
roots, a restudy of T. tetragona became necessary. Although the general Thecidiopsis shell
succession had been established (Williams 1 973 ; Smirnova 1979, 1984), the requirement for a clearer
understanding of the detailed shell microstructure of T. tetragona was increased following the
discovery by Baker (1989) of a parallel shell development pattern in the early Middle Jurassic
P achy moorellina dundriensis (Rollier, 1915).

MATERIAL AND METHODS

Specimens of Eothecidellina imperfecta (Nekvasilova, 1967) from Zbyslav, Czechoslovakia, were
obtained from Dr O. Nekvasilova, Academy of Sciences Prague. Specimens of Boscpietella
campichei (de Loriol, 1872) and Thecidiopsis tetragona (Roemer, 1839) from Chateau du Marais,
Auberson, were loaned from the Campiche Collection, housed in the Musee Geologique, Lausanne.

! Palaeontology, Vol. 34, Part 4, 1991, pp. 815-836, 5 pls.|

© The Palaeontological Association

816

PALAEONTOLOGY, VOLUME 34

Specimens of Bifolium faringdonense (Davidson, 1874) were selected from material previously
collected from the Faringdon Sponge Gravels (Baker and Laurie 1978) and Moorellina dubia
( d'Orbigny , 1850) from material collected from Crickley Hill, near Cheltenham (Baker 1989).

The holotype of Eothecidellina imperfecta (ON-3) and paratypes ON-1, ON-2, ON-4 to ON-8 and
ON-2661 to ON-2664 are housed in the Geological Institute of the Czechoslovak Academy of
Sciences in Prague. The specimens of E. imperfecta , Bifolium faringdonense and Moorellina dubia
figured in this paper (BD9023-BD9028) are housed in the British Museum (Natural History). The
specimens of Bosquetella campichei (reg. no. 42530) and Thecidiopsis tetragona (reg. nos
42531-42533) are housed in the Musee Geologique, Lausanne.

The techniques used for the recovery and preparation of E. imperfecta were fully documented by
Nekvasilova (1967, p. 117) and require no elaboration here. Similarly, for the recovery and
preparation of B. faringdonense specimens see Baker and Laurie (1978, p. 557), and for M. dubia
see Baker and Elston (1984, p. 777).

SYSTEMATIC PALAEONTOLOGY

Order spiriferida Waagen, 1883
Suborder thecideidina Elliott, 1958
Superfamily thecideacea Gray, 1840
Family thecidellinidae Elliott, 1958
Subfamily thecidellininae Elliott, 1953
Genus eothecidellina gen. nov.

Etymology. From the Greek eos (dawn) after the very early appearance of canopied intrabrachial cavities of
thecidellinid type.

Type species. Thecidiopsis ( Thecidiopsis ) bohemica imperfecta Nekvasilova, 1967.

Age. Upper Cretaceous, Upper Cenomanian.

Diagnosis. Small endopunctate thecidellinin, pedicle valve typically with a flat, well defined ventral
interarea, shallow anterior sulcus and relatively large, rounded-triangular attachment scar, brachial
valve with intrabrachial cavities roofed by perforate canopies and a long, straight, median septum
onto which the characteristic pustulose ornament of the subperipheral rim is extended.

Eothecidellina imperfecta (Nekvasilova, 1967)

Plate 1, figs 1-6; Plate 3, figs 1-3; Plate 4, figs 1-7;

Text-figs 1, 4a-c, 5a-c, 6c

1868 Thecidium sp.; Schloenbach, p. 156, pi. 5, fig. 9

1959 Thecidiopsis ( Thecidiopsis ) bohemica Backhaus; Nekvasilova, p. 147; pi. 11. figs 1-4.

1967 Thecidiopsis ( Thecidiopsis ) bohemica imperfecta Nekvasilova, p. 1 15-136, pis 1-8; text-figs 1-1 1.

1968 Thecidiopsis bohemica imperfecta Nekvasilova; Pajaud, p. 45.

1968 Thecidiopsis bohemica imperfecta Nekvasilova; Smirnova and Pajaud, p 143.

1970 Thecidiopsis bohemica imperfecta Nekv. ; Pajaud p. 208, pi. 4, fig. 2.

1972 Thecidiopsis (Thecidiopsis) bohemica Backhaus; Smirnova, p. 123.

Type specimens. Holotype ON-3, paratypes ON- 1, ON-2, ON-4 to ON-8, ON-2661 to ON-2664, hypotypes
BD9023-BD9025.

Distribution. The locality from which the type material was obtained is given as Zbyslav, near Caslav, Bohemia,
Czechoslovakia. The species has, however, been recorded from a number of other Bohemian localities, namely

BAKER: CRETACEOUS THECIDEIDINE BRACHIOPODS

817

c.s.s. v.c.

brachial valve

pedicle valve

brachial lobe
skeletal elements

text-fig. I Eothecidellina imperfecta (Nekvasilova). a-v, ‘Wild’ stereomicroscope traces of cellulose acetate
peels (8-25, out of series 1-49) of serial sections through the median septum and brachial lobes of specimen
BD 9024. Plane of section horizontal, relative to the surface of the brachial lobes: intersecting the commissural
plane with an angle of about 5° ventral deflection. Abbreviations: b.c., brachial cavity; c., brachial lobe
canopy; c.s.s. canopy skeletal support; i.c., intrabrachial cavity; m.s., median septum; v.c., visceral cavity;
w.e., wing-like extension. Peel interval approximately 20 /<m. Scale bar represents 10 mm.

Kamajka and Starkoc near Caslav, Kank near Kutna Hora, Vitezov near Velim, Skalka near Zehusice and
Predboj near Prague. At all the recorded localities the material was collected from sediments of Upper
Cenomanian age, approximating stratigraphically to a plenus Zone position.

Diagnosis. Eothecidellina up to about 2-5 mm in length, 1-9 mm wide and 2-00 mm thick. Outline
roughly pyriform; dorsiconvex initially, but developing a characteristically triangular lateral profile
after the appearance of the free ventral wall. Pseudodeltidium flat, parallel with the surface of the
ventral interarea, demarcated only by indistinct grooves. Hinge line straight with a ratio of length
to shell width of approximately 0-5. Brachial valve moderately convex with a small hypercline dorsal
interarea.

Description. Small, roughly pyriform thecidellinid, with a relatively large attachment scar and well developed
free ventral wall giving rise to a characteristically triangular lateral profile. The ventral interarea is clearly
developed with a flat indistinct pseudodeltidium. The moderately convex brachial valve has a small but well
defined hypercline dorsal interarea.

PALAEONTOLOGY, VOLUME 34

MORPHOLOGY, GROWTH AND SHELL MICROSTRUCTURE OF
EOT HECIDELLIN A IMPERFECTA

Valve characters

Pedicle valve. Hemispondylium raised, with clearly defined dental ridges, hemispondylial plate and
supporting septum, rising from a weakly developed median ridge. The teeth are clearly
cyrtomatodont and not widely separated. The inner surface of the valve is ornamented by scattered
tubercles. In the anterior sector the tubercles are arranged in more linear series and many coalesce
to form low ridges on the inner surface of the free ventral wall.

Brachial valve. (PI. 1, figs 3-6; Text-figs I, 4b, 5a-c, 6c). The cardinal margin is straight, extending
for slightly more than half the maximum width of the valve. The cardinal process is relatively large,
faintly trilobed, and with a characteristic dorsal deflection. The sides diverge anteriorly to form well
developed inner socket ridges. The lateral adductor muscle scars are erect (orientated almost
perpendicular to the commissural plane) and asymmetrically pyriform as a result of constriction by
the cardinal process. The brachial bridge is well developed. The outer surface of the subperipheral
rim is ornamented by up to four rows of small tubercles, with the tubercles of each row usually offset
by one half phase relative to adjacent rows. The inner margin of the subperipheral rim is
characterized by a row of elongate impressed areas which have been identified by Williams (1973,
p. 48, fig. 10) as lophophore muscle scars. The tuberculate ornament of the subperipheral rim
extends onto the median septum which is typically long and prominent, relatively thick anteriorly
and, posteriorly, arching over the visceral cavity. The brachial cavities are occupied by medium-
sized brachial lobes which are developed as perforate canopies over intrabrachial cavities. The
brachial lobes are invariably broken in separated valves, so that the intrabrachial cavities are
exposed as two shallow, roughly oval depressions, characteristically ringed by the damaged canopy
skeletal supports (PI. 1 fig. 3; PI. 3, fig. 3). Each complete brachial lobe, therefore, represents a sort
of posteriorly open, perforate calcareous sac (Text-fig. 6c), situated on either side of the median
septum and slightly contiguous with it posteriorly. A previously unrecorded feature of the brachial
lobes is the development of a pair of wing-like extensions (Text-fig. 1r) arching backwards into the
visceral cavity. The broken remains of these structures can be identified (PI. 1, figs 4 and 5), even
in brachial valves in which the brachial lobes have been substantially damaged.

Ontogeny

The early ontogeny of E. imperfecta is not known. All the characters of pedicle and brachial valves
are present in the smallest valves available. In the pedicle valve, the beginning of the development
of the free ventral wall is thought to coincide with the onset of sexual maturity (Nekvasilova 1967,

EXPLANATION OF PLATE 1

Figs 1-6. EothecideUina imperfecta (Nekvasilova). Upper Cenomanian of Zbyslav, Czechoslovakia. 1 and 2,
BD 9023 ; brachial and three-quarters profile views, photographic record, of hypotype sectioned complete
shell, x 18. 3-6, BD 9025; brachial anterior, three-quarters profile and posterior views of a hypotype to show
the morphology typical of separated brachial valves in which the brachial canopies have been destroyed,
exposing the intrabrachial cavities, x 20.

Figs 7-11. Bifolium faringdonense (Davidson). Upper Aptian; Faringdon, Oxfordshire. 7-10, BD 9026;
brachial, posterior, lateral and anterior views, photographic record, of hypotype sectioned complete shell,
x 20. 11, BD 9027; interior of hypotype brachial valve showing the remains of brachial canopies and the
intrabrachial cavities filled with quartz grain matrix, x 18.

Figs 12-14. Bosquetella campichei (de Loriol). Upper Valanginian; Auberson; Musee Geologique, Lausanne,
42530; posterior, three-quarters profile and brachial views, photographic record, of hypotype sectioned
brachial valve, x 12.

All scanning electron micrographs of gold-coated material.

PLATE I

BAKER, Eothecidellina, Bifolium , Bosquetella

820

PALAEONTOLOGY, VOLUME 34

text-fig. 2. Bifolium faringdonense (Davidson), a-r, ‘Wild' stereomicroscope traces of cellulose acetate peels
(13-30, out of series 1-36) of serial sections through the median septum and brachial lobes of specimen BD
9026. Plane of section horizontal, approximately parallel with the commissural plane. Abbreviations as in Text-
figure 1, plus: b.b., brachial bridge; c.p., cardinal process; h., hemispondylium of pedicle valve; p.p.s.,
posteriorly projecting spur; t., hinge tooth. Peel interval approximately 25 fim. Scale bar represents TO mm.

BAKER: CRETACEOUS THECIDEIDINE BRACHIOPODS

821

text-fig. 3. Bifolium faringdonense (Davidson). Partial reconstruction of the brachial lobes of serially sectioned
specimen BD 9026, with no attempt to hypothesise the nature of the connection of the skeletal elements
between successive peels, a, brachial view, b, vertical plot of peel data along x-y (Text-figs 2 and 3a) for acetate
peels 19-33 (Text-fig. 2g-r inclusive), to show the brachial lobe skeletal elements in tranverse location, forming
a canopy over the intrabrachial cavities. Abbreviations; b.s., brachial lobe skeletal elements; i.c., intrabrachial
cavity. For ease of correlation of peel traces with original acetate peels, left is plotted as viewed, not

anatomically. Scale bar represents 0-5 mm.

p. 125). In the brachial valve, ontogenetic development is characterized by slight changes in the
relative proportion of the median septum and the brachial lobes, and also by increasing dorsal
deflection of the cardinal process. The median septum, initially quite stubby, becomes increasingly
blade-like. The immature brachial lobe canopies are characterized by relatively large, more irregular
perforations which are reported to be gradually sealed as the brachial lobes mature (Nekvasilova
1967, p. 126).

Microstructure (Text-fig. 4a-c)

Although the shell shows the suppression of fibrous secondary shell typical of Cretaceous
thecideidines, E. imperfecta is unusual among Upper Cretaceous species in that fibrous secondary
shell remains a major component of the pedicle valve.

Pedicle valve. A thick granular primary layer is present, underlain by fibrous secondary shell. The
fibrous secondary shell layer appears to be partially suppressed and is absent from the attachment
scar area. It is, however, present as a continuous thin sheet over the entire area of the free ventral
wall (Text-fig. 4c). In this region the granular calcite layer is approximately 75 pm thick, almost half
the thickness of the shell wall. Sections (PI. 4, figs 1-5) show that, at its inner boundary, the granular
layer is longitudinally invaginated into the fibrous layer and eventually may become pinched oft' to
form flattened tubercle cores surrounded by fibrous secondary shell. The development pattern seems
to restrict the formation of tubercles to a single row in antero-lateral sectors of the valve. However,
the thickened median ridge (PI. 4, fig. 7) can clearly be seen to have originated from a succession
of such structures, sandwiched between sheets of fibrous secondary shell. A normal orthodoxly
stacked fibrous secondary shell succession (Williams 1973) is also seen in the hinge teeth.

Brachial valve. The secretion of fibrous secondary shell is almost completely suppressed, secondary
fibres only occurring on the sides of the cardinal process and the inner socket ridges (PI. 3, fig. 2).
The remainder of the shell (Text-fig. 4b) is composed of granular calcite (PI. 3, fig. 1 ; PI. 4, fig. 6)
in which the mainly equant granules are up to 3-0 pm in diameter.

822

PALAEONTOLOGY, VOLUME 34

MORPHOLOGY AND SHELL MICROSTRUCTURE OF OTHER CRETACEOUS

THECIDEI DINES

Bifolium faringdonense {Davidson), Upper Aptian

Morphology. The morphology of B. faringdonense has been described by Elliott (1948) and Pajaud
(1970) with additional comments by Baker and Laurie (1978). The form of the brachial lobes has
always been interpreted from their morphology as seen in separated brachial valves. The indequacy
of this approach was noted by Nekvasilova (1967), but the weakly coherent matrix of complete
shells frustrated attempts (Baker and Laurie 1978, p. 557) to obtain reliable evidence from
Faringdon specimens.

Recently, however, a brachial valve with partially preserved brachial lobes was discovered (PI. 1,
fig. 11; PI. 2, fig. 6), and also a complete shell (PI. 1, figs 7-10) filled with crystalline calcite. The latter
when serially sectioned (Text-fig. 2) enabled the first full description of the brachial apparatus of
B. faringdonense. The previously described auriform structures (Pajaud 1970, p. 193; Baker and
Laurie 1978, p. 563) are identified as the damaged supports of a network of skeletal strands almost
40 jum thick, anastomosing to form perforate canopies, partially roofing the intrabrachial cavities
(Text-figs 3 and 6b). The posteriorly projecting spur, first observed in juveniles (Baker and Laurie
1978, p. 568) is confirmed by the current study (Text-figs 2f and 3a).

Microstructure. Previous studies of shell microstructure (Williams 1973, p. 465; Baker and Laurie
1978, p. 569) have shown that, with the exception of the teeth and inner socket ridges, fibrous
secondary shell is replaced by granular calcite in B. faringdonense. Splayed acicular crystallites are
associated with the development of tubercles in the subperipheral rim. The current study reveals the
presence of secondary fibres in the sides of the cardinal process. Also, in (presumably aged) adults
the brachial valve is so thickened by the subsequent deposition of granular calcite, that the skeletal
supports of the brachial lobe canopies become substantially buried (Text-figs 2q, r and 3b).

Bosquetella campichei (de Loriol ), Upper Valanginian (PI. 1, figs 12-14)

Microstructure

Pedicle valve. No pedicle valve was available for study from the Campiche Collection, but Smirnova
(1984) described pedicle valves from Solovevka, Crimea, in which tubercle cores of granular calcite
are pinched off from the primary layer and enveloped by secondary fibres in a manner very similar
to that seen in E. imperfecta. In B. campichei , however, unlike E. imperfecta, fibrous shell is
underlain by granular calcite (Smirnova 1984, pi. 23, fig. 1).

Brachial valve. Fibrous secondary shell is strongly suppressed so that, with the exception of the

EXPLANATION OF PLATE 2

Figs 1-5. Thecidiopsis tetragona (Roemer). Upper Valanginian; Auberson; Musee Geologique, Lausanne.
1 and 2, 42531 ; brachial and three-quarters profile view, photographic record, of a sectioned young adult
complete shell, x 12. 3-5, 42532; brachial, tilted lateral and posterior views showing the morphology typical
of separated brachial valves and the interdigitation of brachial lobes (all damaged) and lateral septa, x 15.

Fig. 6. Bifolium faringdonense (Davidson). Angled view (rotation, 45°; tilt angle, 50°) ot the right brachial
cavity of specimen BD 9027, showing the remains of the brachial canopy (arrowed) in more detail, x 30.

Figs 7-9. Moorellina dubia (d'Orbigny). Aalenian; Crickley Hill, near Cheltenham; BD 9028; three-quarters
profile, posterior and brachial views of a hypotype showing the almost intact brachial lobes in the form of
rudimentary canopies; the external surface of this specimen (especially figure 8, lower) is encrusted by a
bryozoan, x 30.

All scanning electron micrographs of gold-coated material.

PLATE 2

BAKER, Thecidiopsis , Bifolium , Moorellina

824

PALAEONTOLOGY, VOLUME 34

cardinal process and inner socket ridges, the brachial valve consists of a single layer of
microcrystalline calcite. This consists of equant grains about 3-5 pm in diameter (PI. 3, fig. 7) but,
in the subperipheral rim and areas close to it where resorption has occurred, tubercle cores are
formed from acicular crystallites which show the same splayed orientation (PI. 4, fig. 8) as that
found in Bifolium faringdonense.

Thecidiopsis tetragona ( Roemer ), Upper V alanginian (PI. 2, figs 1-5)

Microstructure

Smirnova’s (1979, 1984) studies revealed that in T. tetragona fibrous secondary shell is strongly
suppressed. She identified a primary layer of granular or acicular calcite underlain principally by
granular calcite, with or without the presence of fibrous secondary shell. Secondary fibres are
restricted to the hinge teeth and the hemispondylium, and to occurrence as irregular bundles in the
wall of the pedicle valve. In the brachial valve, secondary fibres are only to be found in the inner
socket ridges. Additionally in T. tetragona , the secondary shell mosaic was reported (Smirnova
1979, p. 340) to be complicated by the inclusion of ‘rod-like’ bodies and ‘fir-tree’ structures.

Pedicle valve. (Text-fig. 4f). Detailed investigation of serially sectioned specimens from the
Campiche Collection, confirms the presence of a thin granular primary layer and enables the so-
called, rod-like bodies, to be identified as acicular crystallite tubercle cores (Baker 1989, p. 62) of
the type found in Pachymoore/lina. Sections parallel with the boundary of the primary layer show
that the tubercle cores are initiated through ‘seeding’ acicular crystallites on to its inner surface (PI.
5, fig. 1). Initially, the tubercle cores remain separated by granular calcite (PI. 5, fig. 2) but as they
develop, this is gradually excluded and the tubercle cores come into lateral contact with their
neighbours (PI. 5, fig. 7). Each generation is usually offset by a half-phase so that, in oblique section,
a very characteristic overlapping ‘fish-scale’ effect is produced (PI. 5, fig. 8). In high oblique or
transverse section (PI. 5, fig. 4) the tubercle cores often appear to have a centre column of granular
calcite. This is usually a consequence of the splayed arrangement of acicular crystallites forming the
tubercle core. Crystallites in near axial orientation will be sectioned transversely (PI. 5, fig. 3) and
give the appearance of being granular. Some tubercle cores, however, do appear to have an axial
structure in which scalenohedra-like bodies (PI. 5, fig. 5) are mixed with the axially orientated
acicular crystallites. Away from the hinge teeth and hemispondylium, where it is well represented,
fibrous secondary shell is restricted to the inner margin of the free ventral wall. It occurs as irregular,
attenuated strands (PI. 5, figs 3 and 6), pinched and deflected by the development of the acicular

EXPLANATION OF PLATE 3

Figs I -3. Eothecidellina imperfecta (Nekvasilova). All BD 9025. 1, granular calcite shell and endopuncta (top
left) in the floor of the left brachial cavity, x 2000. 2, right inner socket ridge (socket, lower centre) showing
the development of fibrous secondary shell, x 1 50. 3, left brachial cavity showing detail of the broken canopy
supports, ringing the intrabrachial cavity (upper right), x 100.

Figs 4 and 5. Thecidiopsis tetragona (Roemer). 4, 42531; granular primary layer, x4000. 5, 42532; coarser
secondarily deposited granular calcite (cf. Fig. 4) in the floor of the brachial cavity, associated with the
development of the brachial lobes, x 2000.

Figs 6 and 7. Bosquetella campichei (de Loriol). 6, 42530; enlarged view of the brachial cavities showing the
asymmetrically pyriform brachial lobes and absence of intrabrachial cavities, x 35. 7, fracture surface on the
median septum of the same specimen; showing the granular calcite shell, x 3000.

Figs 8 and 9. Moorellina dubia (D’Orbigny). 8, BD 9028; enlarged view of the right brachial cavity showing
detail of the canopy and the nature of the canopy skeletal supports, x 130. 9, same specimen; showing the
typical partially resorbed moorellinid fibrous secondary shell, x 3000.

All scanning electron micrographs of gold coated material.

PLATE 3

BAKER, Eothecidellina , Thecidiopsis, Bosquetella, Moorellina

826

PALAEONTOLOGY, VOLUME 34

text-fig. 4. Diagrammatic reconstruction of the shell microstructure of a-c, Eothecidellina imperfecta
(Nekvasilova) and d-f, Thecidiopsis tetragona (Roemer). a, d, locational diagrams. B, E, block diagrams
showing the microstructure of brachial valves, c, f, block diagrams showing the microstructure of pedicle
valves. Abbreviations: b. 1 brachial lobe interdigitation ; c., brachial lobe canopy; l.s., lateral septum;
p., endopuncta; s.p.r., subperipheral rim; t.c., tubercle core.

BAKER CRETACEOUS THECIDEIDINE BRACHIOPODS

827

crystallite tubercle cores. In view of their abrupt appearance close the inner surface of the shell wall,
these groups of secondary fibres must, presumably, originate from epithelial cells which had
previously been secreting acicular calcite. Somehow the ‘memory’ of certain cells is triggered and
fibrous secondary shell is secreted.

Brachial valve. (Text-fig 4e). Smirnova (1979, p. 340) described the brachial valve as double layered,
with an outer crystalline layer containing ‘fir-tree’ structures, aligned nearly perpendicular to the
external surface of the valve. The fir-trees were reported to be abruptly cut off from the granular
calcite inner layer. Serially sectioned Campiche specimens show that the fir-trees (subsequently
called ‘herring-bone’ patterns by Smirnova (1984)) are the equivalent of the acicular crystallite
splays (Baker and Laurie 1978, p. 564, text-fig. 5d) found in the tubercle cores of Bifolium. In T.
tetragona the wall of the brachial valve is thickened as a result of the increasing development of the
brachial lobes and their interdigitation with the lateral septa. The inner layer described by Smirnova
is the product of deposition of granular calcite in areas where shell resorptive activity had previously
been operative. The cut-off, therefore, as in Pachymoorellina dundriensis is really the trace of the
transgression across an older resorption surface. As in P. dundriensis the secondarily deposited
granular calcite is coarser (PI. 3, fig. 5) than the granular calcite underlying the transgression surface
(PI. 3, fig. 4). Clearly, the so-called third layer (Smirnova 1979, p. 340) associated with the
development of the hemispondylium in the pedicle valve of T. tetragona has the same origin. Similar
transgression surfaces have been identified in Thecidellina barretti (Davidson, 1864) by Williams
(1973, p. 452, fig. 26).

DISCUSSION

Eothecidellina imperfecta may be distinguished from Bifolium faringdonense and Bosquetella
campichei through differences in morphology and shell microstructure. Brachial, posterior and
lateral views of the dorsal cardinalia (Text-fig. 5) show that the lateral adductor muscle scars differ
in shape, orientation and disposition. In E. imperfecta the cardinal margin is straight and slightly
more than one-half the width of the valve (Text-fig. 5a), whereas in B. faringdonense it is very short,
usually not extending beyond the boundary of the dental sockets (Text-fig. 5d). B. campichei also
has a straight cardinal margin but the cardinal process is relatively, much smaller (Text fig. 5g) than
that of E. imperfecta. The cardinal process differs in attitude (Text-fig. 5b, e, h). E. imperfecta has
a small hypercline dorsal interarea (Text-fig. 5c), a feature not seen in B. faringdonense or B.
campichei (Text-fig. 5f, i). The configuration of the brachial lobes as indicated by their skeletal
elements is oval in E. imperfecta and B. faringdonense (PI. 1, figs 3 and 11) and the canopies
overarching the intrabrachial cavities are more completely developed in E. imperfecta than in B.
faringdonense. In B. campichei the brachial lobes are asymmetrically pyriform (PI. 1, fig. 14). There
are no intrabrachial cavities (PI. 3, fig. 6) and, as far as is known, the skeletal elements of the
brachial lobes are not overarching.

With the exception of the inner socket ridges and sides of the cardinal process, the brachial valve
of E. imperfecta appears to be composed entirely of granular calcite. There is no differentiation into
acicular crystallites as seen in B. faringdonense and B. campichei. In the pedicle valve, the tubercle
cores clearly originate in the same way in E. imperfecta and B. campichei , but in B. campichei , the
tubercle cores are not stacked to form a median ridge. In E. imperfecta fibrous secondary shell forms
a complete inner lining, whereas, in B. campichei the inner shell layer is composed of granular
calcite. No tubercle cores have been seen in the pedicle valve of B. faringdonense and, except in the
hinge teeth, there is no development of fibrous secondary shell.

The brachial lobe canopies of E. imperfecta are very similar to those of thecidellinids such as
Thecidellina blochmanni Dali. T. hlochmanni may be distinguished by its much smaller, incipiently
trilobed cardinal process. In Eothecidellina the posterior projecting spur (Text-fig. 6c) may be the
precursor of the much more prominent posterior extension of the Thecidellina brachial lobe skeleton
(Text-fig. 6d). E. imperfecta has a brachial valve microstructure which is closely comparable with
the strongly suppressed fibrous secondary shell succession of Thecidellina barretti (Davidson),

828

PALAEONTOLOGY, VOLUME 34

text-fig. 5. Drawings to show brachial, lateral and posterior views of dorsal cardinalia. A-c, Eothecidellina
imperfecta (Nekvasilova). d-f. Bifolium faringdonense (Davidson), g-i, Bosquetella campichei (de Loriol).
Abbreviations: c.p., cardinal process; i., interarea; l.a.s., lateral adductor muscle scar; m.s., median septum;
s., dental socket; v.c., visceral cavity. Scale bars represent 0-5 mm.

mapped by Williams (1973). In T. barretti the secretion of fibrous secondary shell is also strongly
suppressed in the pedicle valve, in marked contrast to the well developed fibrous secondary layer
found in E. imperfecta.

It is probable that the specimens alleged to be juveniles of Backhausina schlueteri (Lundgren) and
Backhausina groenwalli (Nielsen) figured by Pajaud (Pajaud 1970, p. 221, figs a a and c b) are
assignable to E. imperfecta. However, because of their different age (B. schlueteri , Campanian; B.
groenwalli , Danian) and geographical occurrence, they are not included in the synonymic list. In
view of what is now known about ontogeny the specimens are, nevertheless, clearly separable from
Backhausina , which destroys the argument for the neotenous derivation of Thecidellina from
Backhausina (Pajaud 1970).

The differences in morphology, microstructure and stratigraphic occurrence are thought to be
indicative of the taxonomic separation of Eothecidellina and Thecidellina, but similarities,
particularly in the development of the brachial lobes (Text-fig. 6c, d), are thought to imply a close
phylogenetic relationship. Similarly, the new evidence indicates that Eothecidellina and Bifolium are
phylogenetically related. Although morphologically distinct, the similarity of the brachial lobes is

EXPLANATION OF PLATE 4

Figs 1-7. Eothecidellina imperfecta (Nekvasilova). 1, BD 9024/4; section through the free ventral wall of the
pedicle valve, showing the general shell fabric and invaginated granular calcite in longitudinal section
(section orientation horizontal, almost parallel with the shell surface; section location - anterior sector),
x 200. 2, BD 9023/33; section through the free ventral wall, showing the fibrous secondary shell and
invaginated granular calcite in transverse section (section orientation - parallel with the plane of the
commissure; section location - anterior sector), x 700. 3, same specimen; showing invaginated granular
calcite almost pinched off to form a tubercle core, x 1000. 4, enlarged portion of figure 1, as indicated; to
show the junction between granular calcite and fibrous secondary shell in detail, x 1200. 5, as figure 4; but
showing detail of the convergence of fibrous secondary shell to pinch off granular calcite in the formation
of a tubercle core, x400. 6, uniform granular calcite in the median septum of the brachial valve, x 3000. 7,
BD 9023/48; transverse section through the free ventral wall showing the succession of tubercle cores,
stacked to form a median ridge (section orientation - parallel with the plane of the commissure; section
location - anterior sector, mid-line of shell), x 250.

Fig. 8. Bosquetella campichei (de Loriol). 42530/23 ; section through the brachial valve showing acicular calcite
tubercle cores in transverse section (section orientation - parallel with the plane of the commissure; section
location - antero-lateral sector), x 700.

All scanning electron micrographs of gold-coated cellulose acetate peels of sectioned specimens.

PLATE 4

BAKER, Eothecidellina, Bosquetella

830

PALAEONTOLOGY, VOLUME 34

text-fig. 6. Diagrams to show the possible sequence (a-d) in the development of canopied brachial lobes, a,
Moorellina dubia (d’Orbigny), Aalenian, drawn from specimen BD 9028. b. Bifolium faringdonense (Davidson),
Upper Aptian, reconstructed from specimen BD 9026. c, Eothecidellina imperfecta (Nekvasilova), Upper
Cenomanian, reconstructed from specimen BD 9024. d, Thecidellina blochmanni Dali, Recent, after
Nekvasilova (1967, pi. 6, fig. 1). Abbreviations: b.b, brachial bridge; b.c., brachial cavity; c., brachial lobe
canopy; c.p., cardinal process; i.c., intrabrachial cavity; m.s., median septum; p.a.s., posteriorly arching spur.

Scale bar represents 0-5 mm.

again striking. Its earlier occurrence (Upper Aptian) and the less complete development of the
canopies of the brachial lobes (Text-fig. 6b) places Bifolium in an attractive position from an
evolutionary point of view. However, the heavily suppressed fibrous secondary mosaic of the pedicle
valve seems to rule out Bifolium from a direct ancestral position. Pajaud (1970) regarded Rioultina
as being ancestral to Bifolium. Smirnova (1984) considered Rioultina to be close to, but, slightly off
the main line of Bifolium descent from Moorellina stock. Smirnova’s interpretation is probably more
accurate, since it can now be demonstrated that the small. Middle Jurassic (Aalenian) Moorellina

EXPLANATION OF PLATE 5

Figs 1-8. Thecidiopsis tetragona (Roemer). 1, 42532/3; inner surface of the primary layer showing acicular
crystallites of future tubercle cores ‘seeded’ on to granular calcite (section orientation - parallel with the
external shell surface; section location antero-lateral sector), x 350. 2, same specimen; showing acicular
crystallite tubercle cores developed but still separated by granular calcite, x 350. 3, 42533/49; transverse
section through the free ventral wall showing the attenuated fibrous strands at the inner boundary (lower
margin) and transversely sectioned tubercle cores (section orientation - parallel with the plane of the
commissure; section location - as fig. 1), x 375. 4, same specimen; showing tubercle core in greater detail,
x 800. 5, enlarged portion of figure 4; showing the development of scalenohedra-like bodies (lower left) in
the axial region, x4000. 6, 42533/49; attenuated strand showing detail of fibrous secondary shell in
transverse section (section orientation and location as in figure 3), x400. 7, 42533/7; oblique section
through the free ventral wall showing the acicular crystallite tubercle cores in lateral contact (section
orientation and location as in figure 3), x 325. 8, same specimen; showing the characteristic ‘fish-scale’
overlap of acicular crystallite tubercle cores in oblique section, x 250.

All scanning electron micrographs of gold-coated cellulose acetate peels of sectioned specimens.

PLATE 5

BAKER, Thecidiopsis

832

PALAEONTOLOGY, VOLUME 34

BAKER: CRETACEOUS THECIDEIDINE BRACHIOPODS

833

text-fig. 7. Phylogenetic reconstruction, showing, where known, the essential ontogeny of the brachial
supports, and, shell microstructure of thecideidine genera. In the diagrammatic illustrations of shell
microstructure, horizontal lines indicate continuous layers, diagonal lines indicate restricted distribution. The

presence of periostracum is assumed in all genera.

834

PALAEONTOLOGY, VOLUME 34

dubia possessed brachial structures (PI. 2, figs 7-9; Text-fig. 6a) which may be regarded as
rudimentary canopied brachial lobes (PI. 3, fig. 8), in the early stages of their evolution. The shell
microstructure of M. dubia is typically moorellinid (PI. 3, fig. 9). Smirnova’s (1969) discovery that
the ontogeny of Thecidiopsis did not pass through a rioultinid phase (undivided median septum and
auriform brachial lobes (Pajaud 1966, p. 630)) clearly separates Eothecidellina from Thecidiopsis.

The distribution of secondary fibrous shell in E. imperfecta and T. tetragona is instructive.
Although fibrous secondary shell forms a continuous thin sheet in the pedicle valve of E. imperfecta ,
its appearance coincides with the beginning of the development of the free ventral wall. In T.
tetragona the residual strands of fibrous secondary shell appear at the inner boundary of the free
ventral wall, after substantial deposition of acicular calcite. The diagram of T. tetragona shell
microstructure in Smirnova’s (1979) paper is misleading because the fibrous secondary layer is
depicted as a well-differentiated, continuous internal layer. The seeding of patches of acicular
crystallites on to the inner surface of the primary layer is very reminiscent (see Williams 1968, p. 7,
text-fig. 4) of the initiation of normal secondary fibres. In both genera the sudden appearance of
fibrous secondary shell, suggests a latent ability of the epithelial cells to switch from the secretion
of acicular crystallites, to the secretion of normal secondary fibres. As already noted, the acicular
crystallite tracts discovered in Pachymoorellina, and well established in T. tetragona , equate with the
rod-like bodies of Smirnova (1979, 1984) not, as previously stated (Baker 1989, p. 67), with the fir-
tree structures.

Thecideidine phytogeny

The various phylogenetic models proposed for the Thecideidina have been thoroughly reviewed in
a recent paper (Baker 1990) and need only brief further mention here. The earlier interpretations
of Elliott (1948, 1953) and Backhaus (1959) were modified by Rudwick (1968), Smirnova (1969) and
Pajaud (1970). Although the rectilinear evolutionary pattern of the Lacazellinae had been well
understood for some time (Pajaud 1970) the evolution of the Thecidellinidae and Thecideinae had
remained problematical. Smirnova’s (1969) demonstration of the distinctiveness of thecidellinid and
thecideid ontogenetic development patterns revived interest in Backhaus’s (1959) idea of two
phyletic groups, and introduced the notion of parallel lineages within the groups. Smirnova’s (1969)
interpretation, with additional refinement (Smirnova 1984), was much more satisfactory, since it
avoided most of the problems (Baker 1990, pp. 177, 180) of heterochronous expression, mutation,
neoteny and undiscovered adults attendant to the other schemes. Williams’s (1973, p. 468, fig. 100)
phylogenetic chart of thecideidine shell structure variation had been modelled on the phylogenetic
framework proposed by Pajaud (1970). Subsequent studies (Baker and Elston 1984; Smirnova
1984; Baker 1989) showed that Pajaud’s phylogeny was flawed. Also, the suppression of fibrous
secondary shell occurred much earlier in the history of the group than Williams envisaged.
Interpretation of the new morphological and microstructural evidence obtained from the current
study of E. imperfecta , B. faringdonense, B. campichei and M. dubia , not only strengthens the
lineages identified by Smirnova (1984), but also enables a clearer understanding of the phylogeny
of thecidelliniform genera to be reached. Similarly, the new interpretation of the shell microstructure
of T. tetragona , allied to the evidence from study of Mimikonstcintia sculpta Baker and Elston, 1984,
and P. dundriensis (Baker 1989), gives a much clearer picture of lineage within the Thecideinae.

My review of existing taxonomic and phylogenetic models (Baker 1990) clearly stated the case for
including thecospiraceans and Bactrynium in the Thecideidina and supported the view of Dagis
(1973) that early thecospiraceans were very close to thecideidine ancestral stock. I concluded that
sufficient information was available to enable a reliable taxonomic and phylogenetic framework to
be established. A revised classification was presented, including the Thecospiracea and the
Bactryniidae in the Thecideidina which, in turn, was included in the Spiriferida. A recent study
(Benigni and Ferliga 1989), assigning Carnian thecospirids from Italy to the Strophomenida, failed
to address the arguments advanced by Williams (1973) and Baker (1984) in favour of removal of
the Thecospiracea to the Spiriferida. In content, if not conclusion, the study confirms the very close
relationship between thecospiracean and thecideacean shell microstructure and strengthens the

BAKER: CRETACEOUS THECIDEIDINE BRACHIOPODS

835

argument for inclusion of the Thecospiracea in the Spiriferida, because the microstructure of the so-
called taleolae of Thecospira semseyi Bittner, 1912, is so close to the microstructure of the denticles
of juvenile spiriferaceans (compare Benigni and Ferliga 1989, p. 527, fig. 9b, with Baker 1984,
pi. 77, fig. 3) as to almost be interchangeable.

Proposals about phylogeny were deferred (Baker 1990, p. 183; research in progress) pending the
interpretation of new morphological and microstructural evidence. The evidence from the current
study, incorporated into the existing framework, now enables a revised phylogeny to be presented
(Text-fig. 7). Shell microstructure must not be regarded as diagnostic in every case. For example,
although uniformity of shell microstructure is characteristic of the various species of Moorellina ,
shell microstructure varies in Thecidiopsis. Williams’s (1973, p. 465) study of T. essenensis indicated
a more heavily suppressed fibrous secondary mosaic than that found in T. tetragona. In
T. essenensis, fibrous shell is found in the hinge teeth and base of the outer wall of the hemi-
spondylium but, unless overlooked, is absent from the free ventral wall. Therefore, although
considerable variation of shell structure may be found among species assigned to a particular
thecideidine genus, the status of observed differences remains uncertain. It is clear, however, that
the range of shell microstructure differences reflects the operation of biologically distinct
physiological processes, which are probably no less diagnostic than differences in morphology.
Nevertheless, in the absence of quantifiable morphological variation, it would be premature to
identify morphogenera solely on the basis of shell microstructure. Therefore, in the current study
shell microstructure of the genera is illustrated in so far as is possible by that of the type species
(Text-fig. 7).

CONCLUSIONS

It is clear that ideas about morphology of the brachial lobes of species such as E. imperfecta and
B. faringdonense , based solely on interpretation of broken skeletal supports as seen in separated
brachial valves, may lead to confusion over the systematic arrangement of taxa. The current study
further emphasises the critical importance of the study of serially sectioned whole shells in the
determination of the form of thecideidine brachial lobes, because the often distinctively sculptured
structures, arising from differential shell accretion and resorption, are almost invariably destroyed
in separated valves.

As for shell microstructure, the essential limitation of morphogeneric taxonomic or phylogenetic
frameworks remains the inability to incorporate distinctions which may be perfectly valid bio-
generically, but remain undifferentiated at morphogeneric level. Flowever, in the Thecideidina,
chronologically arranged changes in shell microstructure and microstructures persistent over long
periods become important indicators of lineage, especially when allied to morphological and
ontogenetic evidence.

Acknowledgements. I thank Dr O. Nekvasilova (Academy of Sciences, Prague) for the generous gift of
numerous specimens of Eothecidellina imperfecta , and Dr M. Weidmann (Musee Geologique, Lausanne) for
the loan of specimens of Bosquetella campichei and Thecidiopsis tetragona. Thanks are due to Mr S. A. Taylor
(Derbyshire College of Higher Education) for help with the preparation of the plates.

REFERENCES

backhaus, e. 1959. Monographic der cretacischen Thecideidae (Brachiopoden). Mitteilungen aus dem
Geologischen Stdatsinstitut in Hamburg, 28, 5-90.

baker, P. G. 1984. New evidence of a spiriferide ancestor for the Thecideidina (Brachiopoda). Palaeontology ,
27, 857-866.

— 1989. Evaluation of a thecideidine brachiopod from the Middle Jurassic of the Cotswolds, England.
Palaeontology, 32, 55-68.

— 1990. The classification, origin and phylogeny of thecideidine brachiopods. Palaeontology, 33, 175-191.

836

PALAEONTOLOGY, VOLUME 34

— and elston, d. g. 1984. A new polyseptate thecideacean brachiopod from the Middle Jurassic of the
Cotswolds, England. Palaeontology, 27, 777-791 .

and laurie, k. 1978. Revision of the Aptian thecideidine brachiopods of the Faringdon Sponge Gravels.
Palaeontology, 21, 555-570.

benigni, c. and ferliga, c. 1989. Carnian Thecospiridae (Brachiopoda) from San Cassiano Formation
(Cortina D'Ampezzo, Italy). Revista Italiana di Paleontologia e Stratigrafia, 94, 515-560.
bittner, a. 1912. Brachiopoden aus der Trias des Bakonyer Waldes. Resultate der Wissenschaftlichen
Erforschung des Balatonsees, Wien, Paldontologie Anhang 1, Pt 1, 1-60.
dagis, a. s. 1973. Ultrastructure of thecospirid shells and their position in brachiopod systematics.
Paleontological Journal , 6, 359-369.

dall, w. h. 1920. Annotated list of the Recent Brachiopoda in the collection of the United States National
Museum, with the description of thirty-three new forms. Proceedings of the United States National Museum,
Washington, 57, 261-317.

davidson, T. 1864. On the Recent and Tertiary species of the genus Thecidium. Geological Magazine, 1, 12-22.

— 1874. A monograph of the British Fossil Brachiopoda (Supplement to the British Cretaceous
Brachiopoda). Monograph of the Palaeontographical Society, 4(1), 1-72.

elliott, G. f. 1948. Palingenesis in Thecidea (Brachiopoda). Annals and Magazine of Natural History, (12), 1,
1-30.

1953. The classification of the thecidean brachiopods. Annals and Magazine of Natural History, (12), 6,
693-701.

1958. Classification of thecidean brachiopods. Journal of Paleontology , 32, 373.

GRAY, j. e. 1840. Brachiopoda. 151 . In Synopsis of the contents of the British Museum, (42nd edn). G. Woodfall,
London, 370 pp.

loriol, p. de 1872. In pictet, f. j. Description des fossiles du terrain Cretace des environs de Sainte-Croix, 5.

Materiaux pour la Paleontologie Suisse, Ser. 6, 1, Geneve, Bale , Lyon, 1-158.
nfk v asilov a , O. 1967. Thecidiopsis ( Thecidiopsis ) bohemica imperfecta n. subsp. (Brachiopoda) from the
Upper Cretaceous of Bohemia. Sbornik geologikych ved Praze, Paleontologie , 9, 115-136.
orbigny, A. d’. 1850. Prodrome de paleontologie stratigraphique universelle des animaux mollusques et rayonnes,
Vol. 2. Masson, Paris, 394 pp.

pajaud, d. 1966. Problemes relatifs a la determination des especes chez les Moorellininae (Thecideidae,
brachiopodes). Bulletin de la Societe Geologique de France, (7), 8, 630-637.

1970. Monographic des thecidees (brachiopodes). Memoires de !a Societe Geologique de France, 112,
1-349.

roemer, f. a. 1839. Die Versteinerungen des norddeutschen Oolith-Gebirges, Nachtrag. Hanovre, 1-22.
rudwick, m. j. s. 1968. The feeding mechanisms and affinities of the Triassic brachiopods Thecospira
Zugmayer and Bactrynium Emmrich. Palaeontology, II, 329-360.

Smirnova, t. n. 1969. Ontogeny and phylogeny of Early Cretaceous brachiopods of the suborder Thecideidina.
Paleontological Journal, 3, 64-78.

— 1979. Shell microstructure in Early Cretaceous thecidean brachiopods. Paleontological Journal, 13,
339-344.

1984. [Early Cretaceous brachiopods ( morphology , systematics, phylogeny and their significance in
biostratigraphy and palaeozoogeography)]. Nauka Press, Moscow, 199 pp. [In Russian],
waagen, w. 1882-1885. Salt Range fossils. Productus-limestone fossils (fasc. 2). Brachiopoda. Memoirs of the
Geological Survey of India. Palaeontologia Indica, (13), 1, 329-770.
williams, a. 1968. Evolution of the shell structure of articulate brachiopods. Special Papers in Palaeontology,
2, 1-55.

1973. The secretion and structural evolution of the shell of thecideidine brachiopods. Philosophical
Transactions of the Royal Society of Fondon, Series B, 264, 439-478.

peter g. baker
Department of Geology

Typescript received 19 September 1990 Derbyshire College of Higher Education

Revised typescript received 15 January 1991 Kedleston Road, Derby DE3 1GB, UK

THE FIRST DICYNODONT FROM THE FATE
PERMIAN OF MALAGASY

by JEAN MICHEL MAZIN ClJld GILLIAN M. KING

Abstract. Until now, no therapsid was known with any certainty from Malagasy. The present specimen,
discovered in 1948, but never described, is the first record of a complete dicynodont skull from this country.
A description of the skull is given and it is referred to a new species of the genus Oudenodon. The rarity of
dicynodonts in Malagasy may be because of environmental differences from the adjoining landmass of Africa.

Resume. La presence de therapsides a Madagascar n’avait jamais ete montree avec certitude. Le present
specimen, collecte en 1948 mais jamais decrit, est la premiere decouverte d'un crane presque complet de
dicynodonte dans cette region. Une description du crane en est donnee et il est attribue a une nouvelle espece
du genre Oudenodon. La rarete des dicynodontes a Madagascar peut etre due aux conditions de milieu,
differentes de celles regnant sur la masse continentale africaine contigue.

The Permo-Triassic of Malagasy, sometimes called the ’Malagasian Karoo’ is known for the
richness of its fossils which have been found in reptile-bearing nodules from the Lower Sakamena
Formation (Late Permian), (Piveteau 1926, 1957; Carroll 1981; Currie 1981). Reptiles, such as
Hovasaurus , Thadeosciurus , Tangasaurus , and C Y audio saurus, are well represented and constitute a
typical Malagasian fauna, unknown elsewhere. Until now, except for a fragmentary and
unpublished specimen noted by Piveteau (1957), from below the Cistecephalus zone and
provisionally referred to as a theropsid, the therapsids did not seem to be represented in this fauna.
The abundance of diapsids and the rarity of therapsids characterize the Malagasian Karoo in
comparison with the South African Karoo, where therapsids are abundant. However, the study of
a new specimen, housed in the collections of the Universite Paris 6, confirms the presence of
therapsids in the Late Permian of Malagasy.

THE MATERIAL

Locality and material

The specimen was collected in 1948 by Francois Tortochaux, a French geologist then working for the
Malagasian Petroleum Company. It consists of a partially crushed skull, coming from a locality whose exact
position is unknown, between Ranohira and Benenitra (south Malagasy). The skull was in a calcareous nodule
from the upper part of the psammitic sandstone constituting the latest part of the Lower Sakamena Formation
(Late Permian) (Text-fig. 1). The matrix has been removed by acid processing. The skull is nearly complete,
but dorsoventrally crushed with a broken skull roof (Text-fig. 2). The bone surface is weathered and sutures
are indistinct. The anterior part of the snout is slightly distorted by shear, and the posterolateral part of the
right temporal fenestra is preserved only as a film of bone. The total length of the specimen is 160 mm, and
the maximal width is 95 mm. The lower jaw is mostly complete and preserved in situ. The lateral bones of the
right ramus and of the posterior part of the left ramus are missing. In the palate, the pterygoids and right
basioccipital tuber are visible. The anteriormost part of the palate is not visible. The atlas, axis and three
succeeding vertebrae are also preserved.

Description

In dorsal view, the pre-orbital region is short in comparison with the post-orbital, even allowing for distortion
(Text-fig. 3a). The postorbitals are large and rise to form a low sagittal crest ending at the level of the pineal
foramen. Anteriorly, they are separated by a small shield-shaped preparietal. They constitute the anterior and

IPalacontologv, Vol. 34, Part 4, 1991, pp. 837-842.)

© The Palaeontological Association

838

PALAEONTOLOGY, VOLUME 34

text-fig. 1. Stratigraphical scale of the Sakamena
Formation, with locations of the different fossil levels.
1, clays and sandstone called 'Upper continental red
level’; 2, Septaria clays with interbeds of sandstone;
3, psammitic sandstone; 4, clays with gypsum; 5,
sandstone and conglomerate. FI, small amphibian
and reptile bones; F2, nodules with reptiles; F3,
shales with Glossopteris , Volt:ia...\ F4, lens of
sandstone with small amphibian and Saurichthys
remains; F5, nodules with fishes; F6, rare plant re-
mains. The dicynodont fossil comes from the level F2.

medial border of the temporal fenestrae, and all of the posterior edge of the orbital opening. The squamosals
are very large, forming the posterior and lateral borders of the temporal fenestrae. Because the skull roof is
damaged, the postfrontals are not visible, but their impression persists as a depression. They were elongated,
with a triangular shape, and seem to have contacted only the postorbitals. The left prefrontal persists as an
elongated bone bordering the anterosuperior edge of the orbital opening. The frontals seem to be large, but
their anterior limits cannot be determined. Likewise, the narinal area cannot be observed because of the
anterior distortion and loss of the snout. The measurements of the temporal fenestrae are as follows: left:
length, 58 mm; width, 30 mm; right: length, 59 mm; width, 31 mm.

In lateral view, the general outline of the skull is modified by the dorsoventral crushing (Text-fig. 3c). The
specimen is totally edentulous. The maxilla is large and extends vertically. The jugal is anteroposteriorly
elongated, forming the inferior and posterior edges of the orbital opening. The mandible is strong, with a thick
and high anterior part composed of the dentary. Posteriorly, the mandible is thinner. The articular is divided
into lateral and median articulatory condyles.

The ventral view is more difficult to observe (Text-fig. 3 b). The palate is largely damaged, and the postero-
lateral part of the skull is destroyed. The mandibular symphysis is strong and anteriorly extended. Conversely,
the posterior part of each ramus is narrow.

Identification

The short snout, elongate temporal fenestra, large laterally-expanded squamosal, and distinctive
articular confirm that this specimen is a dicynodont. Among the dicynodonts, this specimen can be
compared with the four main Permian genera : Dicynodon , Oudenodon, Diictodon and the rare
Kingoria (Cluver and Hotton 1981). Because of the lack of a caniniform tusk, it cannot belong to
Dicynodon. The absence of the distinctive square caniniform process indicates that the specimen
does not belong to the genus Diictodon. To distinguish between Oudenodon and Kingoria , it is
necessary to know the state of certain characters of the lower jaw, in particular whether a dorsal
dentary sulcus is present or not. The present specimen does not have the anterior dorsal surface of
the jaw ramus completely exposed, but the dorsal edge of the dentary which is visible is sharp-edged

MAZIN AND KING: DICYNODONT FROM MALAGASY

839

text-fig. 2. Oudenodon sakamenensis. Late Permian, south Malagasy. Type specimen, PVHR 288. a. Dorsal

view; b , ventral view.

as would be expected of the medial wall of a dorsal dentary sulcus, as present in Oudenodon. Another
feature which indicates an assignment to Oudenodon is the suggestion of a large palatal exposure of
the palatine bone. This would be small in Diictodon and Kingoria.

King (1988) lists three species of Oudenodon. The genotype is Oudenodon bainii Owen, 1860,
known from South Africa. O. grandis (Haughton 1917) is also from South Africa, while O.
luangwensis (Boonstra 1938) is known from Zambia (Keyser 1975).

The present specimen presents two features which distinguish it from the known species of
Oudenodon. First, the lateral dentary shelf is drawn up into a pronounced boss anteriorly which is
continued postero-ventrally as a sharp ridge (Text-fig. 3c). Secondly, the posterior edge of the
mandibular fenestra is drawn up into a blunt ridge oriented antero-ventrally, forming an anterior
shelf to the reflected lamina of the angular (Text-fig. 3c). On account of these features the present
specimen is assigned to a new species of Oudenodon.

SYSTEMATIC PALAEONTOLOGY
Suborder anomodontia Owen, 1859
Infraorder dicynodontia Owen, 1859
Superfamily pristerodontoidea Cluver and King, 1983
Family dicynodontidae Cluver and King, 1983
Subfamily cryptodontinae Owen, 1859
Genus Oudenodon Owen, 1860

Diagnosis. Given by King (1988).

840

PALAEONTOLOGY, VOLUME 34

mx

i

pal

spl

Pt

any

art

q

sq

text-fig. 3. Oudenodon sakamenensis, Late Permian, south Malagasy. Type specimen PVHR 288. a. Dorsal

view; b , ventral view; c, right lateral view.

MAZIN AND KING: DICYNODONT FROM MALAGASY

841

Oudenodon sakamenensis sp. nov.

Text-figures 2 and 3

Etymology. From Sakamena, the name of the Formation and of the river which crosses this area.

Holotype. Nearly complete skull, PVFIR 288, preserved at the Laboratoire de Paleontologie des Vertebres de
1’Universite Paris 6.

Type locality. Between Ranohira and Benenitra, South-western Malagasy.

Horizon. Top of the Lower Sakamena Formation, Late Permian.

Diagnosis. Medium to small member of the genus, skull longer than broad. Lower jaw with lateral
dentary shelf drawn up into a boss anteriorly and extending postero-ventrally as sharp-edged ridge;
reflected lamina of angular bearing a pronounced blunt shelf anteriorly.

PALAEOBIOGEOGRAPHICAL SIGNIFICANCE

Dicynodonts are known from South Africa, eastern Africa, Morocco, India, China, Antarctica,
Europe, USSR, and North and South America, but until now had not been reported from
Malagasy. Oudenodon is known from South Africa, Zambia and, now, Malagasy. In South African
localities, dicynodonts are very abundant fossils, and Oudenodon is a common genus from
Cistecephalus Zone localities, and even more so in the succeeding Daptocephalus Zone (King 1986).
It is therefore puzzling that these abundant therapsids should be found so rarely in the otherwise
reptile-rich fauna of Malagasy.

Two alternative explanations present themselves. First, that Malagasy was separated from South
and East Africa by some kind of barrier (sea, mountain) during the Late Permian, preventing easy
migration of dicynodonts to Malagasy. In this case, an impoverished dicynodont fauna, composed
of species very distinct from the mainland forms, would be expected. This explanation can be
dismissed for two reasons. First, it is known that Malagasy was part of the same landmass as Africa
during the Late Permian. There was no ocean barrier to migration, and no evidence of other kinds
of barrier. Secondly, although the present Malagasian dicynodont is specifically distinct from
mainland African forms, it is not particularly different, and no more different than is the Zambian
Oudenodon from the South African.

The second explanation to account for the rarity of therapsids in Malagasy is that the conditions
of fossilization there were very different from those in South Africa. This has been suggested by
Piveteau (1926), Currie (1981) and Carroll (1981). This would imply that the habitats in which
dicynodonts and other therapsids were abundant in South Africa were not those habitats which
have been sampled in the Malagasian sediments. There are some indications that some of the
Malagasian diapsids coming from the same locality ( Hovasaurus , Claudio saur us) were aquatic or
semi-aquatic (Currie 1981 ; Carroll 1981 ; de Buffrenil and Mazin 1989), and it may, therefore, be
that dicynodonts were occupying more terrestrial environments which are not sampled in these
Malagasian localities. By contrast. South African localities, where no similar small diapsid
assemblages are known, may represent the more terrestrial (or less aquatic) habitats. Further
sedimentological and palaeontological sampling of contemporaneous localities in Malagasy, and
functional anatomical study of the fossils they contain, is needed before this conclusion can be
substantiated.

Acknowledgements. The authors are very grateful to M. Francois Tortochaux and Claude Germain who
provided useful information about the Lower Sakamena Formation and the discovery of this specimen.

842

PALAEONTOLOGY, VOLUME 34

REFERENCES

boonstra, l. d. 1938. A report of some Karoo reptiles from the Luangwa Valley, Northern Rhodesia.
Quarterly Journal of the Geological Society of London, 94, 371-384.

buffrenil, v. df, and mazin, j. m. 1989. The bone histology of Claudiosaurus germaini (Reptilia,
Claudiosauridae) and the problem of pachyostosis in aquatic tetrapods. Historical Biology , 2, 311-322.

carroll, r. l. 1981. Plesiosaur ancestors from the Upper Permian of Madagascar. Philosophical Transaction
of the Royal Society of London, Series B , 293, 315-383.

cluver, m. a. and hotton, n., in 1981. The genera Dicynodon and Diictodon and their bearing on the
classification of the Dicynodontia. Annals of the South African Museum 83, 99-146.

cluver, m. a. and ring, G. m. 1983. A reassessment of the relationships of Permian Dicynodontia (Reptilia,
Therapsida) and a new classification of the Dicynodontia. Annals of the South African Museum, 91, 195-273.

Currie, p. j. 1981. Hovasaurus boulei , an aquatic eosuchian from the Upper Permian of Madagascar.
Palaeontologia Africana, 24, 99-168.

haughton, s. h. 1917. Investigations in South African fossil reptiles and Amphibia. Part 10. Descriptive
catalogue of the Anomodontia, with especial reference to the examples in the South African Museum. Annals
of the South African Museum 12, 127-174.

reyser, a. w. 1975. A re-evaluation of the cranial morphology and systematics of some tuskless Anomodontia.
Memoir of the Geological Society of South Africa, 67, 1-110.

ring, g. m. 1986. Distribution and diversity of dicynodonts: an introductory approach. Modern Geology, 10,
109-120.

ring, g. 1988. Anomodontia. Encyclopedia of Paleoherpetology, 17C, 1-174.

Owen, r. 1859. On the orders of fossil and recent Reptilia and their distribution in time. Report and Proceedings
of the British Association for the Advancement of Sciences (1858), 1859, 153-166.

owen, R. 1960. On some reptilian fossils from South Africa. Quarterly Journal of the Geological Association
of South Africa, 67, 1 110.

piveteau, j. 1926. Le Permien du Sud de Madagascar et sa faune de Vertebres. These de la Faculte des Sciences
de Paris, A, 1061, 55 179.

piveteau, j. 1957. Les reptiles permo-triasiques de Madagascar. Compte- Rendu du 3eme Congres de la
P.I.O.S.A., Tananarive, C, 229-232.

JEAN MICHEL MAZIN

Laboratoire de Paleontologie des Vertebres
Universite Paris 6
4 place Jussieu
F-75252 Paris Cedex 05

GILLIAN M. RING
The University Museum

Typescript received 14 May 1990 Parks Road

Revised typescript received 18 December 1990 Oxford OX1 3PW

ABBREVIATIONS USED IN TEXT-FIGURES

ang: angular; art: articular; b.o.t.: basioccipital tuber; d: dentary; f: frontal; FT: temporal fenestra; j: jugal;
1: lachrymal; l.d.s. : lateral dentary shelf ; mx: maxilla; ORB : orbital opening; pal: palatal; pi : pineal foramen;
po: postorbital; pp: postparietal ; prf : prefrontal; pt : pterygoid ; q: quadrate; qj : quadratojugal ; r : ridge; spl :
splenial; sq: squamosal.

AN ONTOGENETIC SEQUENCE OF COCCOLITHS
FROM THE LATE JURASSIC KIMMERIDGE CLAY

OF ENGLAND

by JEREMY R. YOUNG and PAUL R. BOWN

Abstract. The Kimmeridge Clay of Kimmeridge Bay, Dorset, UK, includes several stone bands which
electron microscopy shows are coccolith limestones dominated by the species Watznaueria fossacincta. The
assemblage includes not only fully grown coccoliths but also specimens at earlier growth stages, including
proto-coccolith rings. A complete ontogenetic sequence can thus be reconstructed. This interpretation is
supported by the occurrence of early growth stages inside coccospheres, i.e. intracellular coccoliths preserved
in the process of growth. The coccolith formation process appears to have been directly comparable to that
known in living coccolithophores. The early growth phases have previously been described as separate species;
these can now be put in synonymy.

Coccoliths are minute calcareous platelets produced by unicellular planktonic algae of the
division Prymnesiophyta. They are borne extracellularly as a coccosphere of several individual
coccoliths, but are formed intracellularly, as the product of an elaborate biomineralization process.
Studies of this process in living coccolithophores have highlighted the fact that it does not occur by
calcification of a pre-existing organic matrix, but rather is a growth process. Typically coccolith
growth involves first nucleation of a proto-coccolith ring of simple crystals, then closely regulated
growth of these crystals into the complex segments that constitute fully formed coccoliths
(Westbroek et al. 1984, 1989; Mann and Sparks 1988; van Emburg 1989; Young 1989). Since this
is a growth process with distinct developmental stages, individual coccoliths can be considered to
have an ontogenetic history. When this concept is contrasted with the alternative of regarding
coccoliths as formed objects, it is clear that it can provide an improved framework for the
interpretation of coccolith structure, morphological variation, classification and evolution.
However, since the concept has been derived from study of a few living species, nannofossil
palaeontologists have not been widely convinced of its relevance or applicability. We describe here,
for the first time, an ontogenetic sequence of this type in the fossil record, and use it to illustrate
the utility of such studies.

MATERIAL AND PALAEOENVIRONMENT

The Kimmeridge Clay Formation is a thick ( c . 500 m) organic-rich shale of Late Jurassic age
deposited over an extensive area of the NW European shelf including much of the North Sea and
southern England. The best known outcrops are in Kimmeridge Bay, Dorset, the stratotype of the
Kimmeridgian Stage. A remarkable feature of the formation is the presence of several bands of
laminated limestone almost entirely formed of coccospheres and isolated coccoliths of the species
Watznaueria fossacincta (Downie 1957; Noel 1973; Gallois and Medd 1979). The extensive
occurrence of coccospheres, the lamination of the sediment, and the absence of admixed clay are
almost certainly the result of rapid deposition in anoxic bottom waters, as also indicated by
geochemical and macrofaunal evidence (Myers and Wignall 1987; Oschmann 1988). The domi-
nation of the assemblage by a single species is partly a reflection of the low diversity of Late
Jurassic nannofloras but is also suggestive of a degree of ecological restriction. The combination of

IPalaeontology, Vol. 34, Part 4, 1991, pp. 843-850, 1 pl.|

© The Palaeontological Association

844

PALAEONTOLOGY, VOLUME 34

low diversity, good preservation and abundant coccospheres made these beds attractive for a study
of morphological variation within a single species. For this reason a special study was made of
coccoliths from one of the best developed limestone bands, the White Stone Band, within the
Pectinatites pectinatus Zone.

ONTOGENETIC SEQUENCE

A typical fracture surface of the White Stone Band is illustrated in Plate 1, fig. 1. This shows
abundant coccoliths and coccospheres of the species Watznaueria fossacincta (Black, 1971)
Bown in Bown and Cooper 1989. Rare specimens of species from a few other genera also occur
( P.m and Z.e on PI. 1, fig. 1). In addition, several coccoliths with broad central areas and narrow
rims are visible (arrows in PI. 1, fig. 1). These lack the well developed shields of W. fossacincta but
have structural similarities and have been regarded as separate species of the genus Watznaueria
(e.g. Grun and Zweih 1980). However, if a series of these forms from the open rings to typical W.
fossacincta specimens is assembled, it is apparent that there is continuous variation between them.
Two such sequences are shown in Plate 1, one in distal view (PI. 1, figs 3-5) and the other in
proximal view (PI. 1, figs 6-9). The continuity of variation is primary evidence that the variation is
intraspecific but the range of morphology is too great for preservational, ecophenotypic or
genotypic variation of mature forms. A simple alternative interpretation of this sequence is that it is
an ontogenetic growth sequence, and indeed it is closely comparable to growth sequences described
from living species such as Emiliania huxleyi (Westbroek et al. 1984, 1989; van Emburg 1989).
Further evidence is provided by the presence of broken coccospheres containing ring-like coccoliths
(PI. 1, figs 2 and 10). The occurrence of these coccoliths inside the coccospheres is unlikely to be
coincidental since several examples were seen in the electron microscope, and in the light microscope
single coccoliths were also often discernible inside entire coccospheres. For the purposes of
comparison, scanning electron micrographs were taken of broken coccospheres of cultured
Emiliania huxleyi (PI. 1, figs 1 1 and 12). The analogy with the fossil specimens is plain. So rather
than being a chance association, these ring-like coccoliths occurring within coccospheres are almost
certainly examples of fossilized intracellular coccoliths preserved due to death of the cell during the
process of coccolith formation. This in turn makes it only reasonable to interpret the series of
coccolith morphologies as an ontogenetic sequence of coccolith growth stages.

EXPLANATION OF PLATE 1

Scanning electron micrographs of Watznaueria specimens in the White Stone Band, and comparable specimens
of the modern coccolithophore Emiliania huxleyi.

Fig. 1 Typical fracture surface showing how rock is almost entirely composed of Watznaueria coccoliths, either
isolated or in coccospheres. Arrows, incompletely formed coccoliths. Coccoliths of other species: Z.e.,
Zeugrhabdotus erectus ; P.m., Polypodorhabdus madingleyensis. x 2500.

Fig. 2. Broken coccosphere with internal coccolith at early growth stage. The living cell would have occupied
the space within the coccosphere, and the incomplete coccolith would have been forming within it. x 6500.

Fig. 3. Very early growth stage/proto-coccolith ring, distal view, x 10200.

Fig. 4. Mid growth, distal/oblique view, x 1 1 500.

Fig. 5. Complete coccolith, distal view, x 6900.

Fig. 6. Very early growth/proto-coccolith ring, x 11 500.

Fig. 7. Early growth, proximal view, x 1 1 500.

Fig. 8. Mid growth, proximal view, x 9300.

Fig. 9. Complete coccolith, proximal view, x 6900.

Fig. 10. Broken coccosphere of Watznaueria , with internal proto-coccolith ring. This internal coccolith shows
the earliest growth stage observed, x 3200.

Fig. 1 1 Coccosphere of Emiliania huxleyi, grown in culture, showing internal proto-coccolith ring, x 6000.

Fig. 12. Coccosphere of E. huxleyi showing internal coccolith, x 5100.

PLATE 1

Mm?

YOUNG and BOWN, Watznaueria , Emiliania

846

PALAEONTOLOGY, VOLUME 34

text-fig. I . Growth stages of Watznaueria, based on a synthesis of observations on many specimens. Each
growth stage is illustrated in distal view (left), cross-section (centre) and proximal view (right). Shaded

portion of cross-sections is V-unit.

STRUCTURAL INTERPRETATION

An interpretative series of diagrams illustrating the growth sequence is given in Text-figure 1. Four
growth stages are illustrated, each in proximal view, cross-section, and distal view. Terminology
applied to the elements is explained in Text-figure 2. At the earliest observed stage the coccolith
consists of an elliptical proto-coccolith ring of sub-quadrate crystals with intervening very small
peg-like crystals. During subsequent growth these two sets of crystal units develop separately. They
are referred to as the V- and R-crystal units (Text-fig. 2).

YOUNG AND BOWN: JURASSIC COCCOLITH ONTOGENY

847

text-fig. 2. Terminology applied to the crystal units. The V-unit (shaded) has an approximately vertically
directed crystallographic c-axis, and so is dark in cross-polarized light in plan view. The R-unit has radially
directed c-axis and so is bright in cross-polarized light in plan view.

The larger, initially quadrate, R-crystal units develop into complexly shaped units, composed of
several elements, and forming most of the coccolith. The proximal shield elements are formed by
growth radially outward from the proto-coccolith ring. The elements have a slight initial counter-
clockwise kink possibly due to change in growth direction after initial impingement of the crystals
(PI. 1, figs 6-9; Text-fig. 1, left side). The distal shield elements show the same kink, clockwise
directed in this view, then growth with counter-clockwise precession (PI. 1, figs 3-5; Text-fig. 1, right
side). At the early growth stages the distal shield element directly overlies the proximal shield
element, and they are united by a thickened outer tube cycle (PI. 1, figs 2, 4, 7). Subsequent growth
separates the proximal and distal shield elements since there is about a 30° difference in radial
growth direction between them.

In proximal view the R-crystal units of the proto-coccolith ring can be seen to grow radially
inward as well as outward, closing the central area. These elements also grow upward forming
wedge-shaped inner tube elements (PI. 1, figs 3-5; Text-fig. 1, centre).

The interstitial peg-like V-crystal units of the proto-coccolith ring develop upwards only, forming
relatively simple crystals, composed of only one element. They outcrop on the distal surface of the
complete coccoliths with the appearance of small rectangular plates, but root back to the proximal
surface forming a core to the tube, and so are termed here tube-core elements.

The R- and V-crystal units remain distinct throughout ontogeny. They also are clearly separated
during diagenesis. Under conditions of diagenetic secondary calcification (as in this material)
different overgrowth crystal faces develop on the two types of units and there is no sign of them
fusing; by contrast the proximal and distal shield elements of the R-crystal unit often do fuse. Under
conditions of diagenetic dissolution the V-crystal units are often selectively etched, and may be
entirely removed; this morphotype has occasionally been regarded as a separate species Calolithus
martelae Noel, 1965. The consistent separation of the two crystal units indicates that they are
discrete, separately nucleated cycles, presumably with different crystallographic orientations. The
crystallographic orientation of the main units is readily determined by light microscope
observations, using cross-polarized light and a gypsum plate (as described by Romein 1979). The
c-axis is orientated along the length of the proximal shield elements, and so parallel to the base of
the coccolith. This is also reflected in the prismatic overgrowths on these elements and in their
termination by a zone of crystal faces (PI. 1, figs 2 and 9). The optical orientation of the V-crystals
is less easily determined, but careful light microscope observation of side views suggests the c-axes
are directed approximately perpendicular to the base of the coccolith, i.e. at 90° to the c-axes of the
main units. In plan view this is indicated by the tube-core cycle appearing as a dark ring in cross-
polarized light.

848

PALAEONTOLOGY, VOLUME 34

COMPARISON WITH LIVING SPECIES

Information on coccolith growth is available for a number of living species. Comparison of these
species suggests that all the heterococcolith rims show a similar pattern of development (Young
1989; Westbroek et al. 1989). Holococcoliths and the central area structures of some
heterococcoliths appear to develop in different ways, but Watznaueria coccoliths are straightforward
heterococcoliths, and so should be directly comparable with the living species. The principal points
of similarity between coccolith formation in the living species and their applicability to Watznaueria
are as follows:

1. Calcification is intracellular (PI. 1, figs 11 and 12). In W. fossacincta the partially formed
coccoliths inside coccospheres are striking evidence of this (PI. 1, figs 2 and 10).

2. Nucleation occurs as a single event - around the rim of an organic base-plate producing a
proto-coccolith ring of simple crystals. Proto-coccolith rings are figured here for W. fossacincta (PI.
1, figs 1, 3, 6, and 10). The original presence of an organic base-plate can only be inferred.

3. Crystal growth proceeds upward and laterally from the proto-coccolith ring so that it remains
on the proximal surface of the coccolith. This is precisely the pattern of growth seen here.

4. All the crystal units of a single cycle develop according to a single plan, although in some cases
to differing degree. Watznaueria is a typical coccolith in this respect.

5. Individual crystal units are of complex form but there are rather few separate sets of crystal
units. W. fossacincta is a fine example of this; superficial examination might suggest that the various
cycles of elements were all discrete crystallites but in fact most of them unite, to form the R-crystal
units, leaving only the V-crystals separate - as shown here and argued earlier (Sown 1987).

The most distinctive feature of the Watznaueria structure is the double proto-coccolith ring and
development of a tube-core cycle. This does not have an obvious analogue in the coccoliths of the
best studied species, Emiliania huxleyi, but does in several other species, for instance Coccolithus
pe/agicus (Young 1989; Steinmetz and Young unpublished data) and Pleurochrysis carterae (Outka
and Williams 1971). Thus the morphological development of this Mesozoic coccolith was directly
comparable to that of living species, and there is no reason to believe that qualitatively different
biochemical processes were involved.

TAXONOMIC IMPLICATIONS

The most immediate application of these observations is that they allow a simplification of
Mesozoic nannofossil taxonomy. A number of species of Watznaueria/ Ellipsagelosphaer a (a
subjective junior synonym) have been proposed based on the degree of opening of the central area
and on shield width. On the basis of the growth sequence worked out here, these forms can be seen
to be simply growth stages. They include: E. strigosa Grun and Zweili, 1980 - proto-coccolith ring;
E. tubulata Grun and Zweili, 1980 -early growth stage; and W. ovata Bukry, 1969 -late growth
stage.

Complications are introduced by the presence of forms with a disjunct bridge in the central area
( W. britannica , and its growth stages W. lucasii and W. reinhardtii ) and by an evolutionary trend
toward forms with entirely closed central areas (W. barnesae). With the clarification of the structure
provided by understanding the growth sequence, and the resultant reduction in number of taxa it
should, however, be possible to deal with the genus as a coherent lineage. This in turn should allow
meaningful description of its evolutionary development and allow the definition of biostratigraphi-
cally useful trends. This would be a considerable contribution to Mesozoic nannofossil
biostratigraphy since Watznaueria species are the most common component of nannofossil
assemblages from the Upper Jurassic to the end Cretaceous plankton extinctions.

Our observations are also significant for higher taxonomy and the relationships between
nannofossil families. Rim structure has long been recognised as a primary criterion of higher
classification (e.g. Perch-Nielsen 1971; Romein 1979; Bown 1987). However, without a clear
understanding of how rim structures develop during ontogeny, their description and interpretation

YOUNG AND BOWN: JURASSIC COCCOLITH ONTOGENY

X49

has been confused. The growth model provides such a framework, and highlights the importance
of identifying crystal unit cycles and their crystallographic orientation. In this specific case the
identification of a double proto-coccolith ring and determination that the tube cycle has a quite
different optical orientation to that of the main units permits a number of hypotheses as to the
relationships of Mesozoic coccoliths to be made. These are currently the subject of active
investigation.

PALAEOECOLOGICAL SIGNIFICANCE

Incomplete coccolith morphotypes have not been widely reported from the fossil record, even when
they have been described as separate species. Probably this is mainly because they are somewhat
inconspicuous in the light microscope or easily dismissed as preservational artefacts. If they are
specifically looked for they can be found in most coccolith assemblages. Nonetheless their
abundance in the White Stone Band does seem unusually high, for either fossil or living coccolith
assemblages. Culture work on the living species Emiliania huxleyi provides one possible
interpretation. Incomplete coccoliths are most common in samples from the logarithmic growth
phase of cultures grown in high nutrient media (Young unpublished data). An ecological analogue
for this situation might be bloom conditions. Such conditions have been independently suggested
for the pectinatus Zone of the Kimmeridge Clay, in order to explain the coincidence of high
productivity, monospecific assemblages and anoxic conditions (Gallois 1976). Possibly the high
occurrence of early growth stages may constitute a useful indicator of blooms.

CONCLUSION

Evidently coccolithogenesis occurred in the Jurassic by processes very similar to those acting today.
In consequence information on the process derived from study of living coccolithophores can be
applied to the study of fossil coccoliths. This can have useful applications in understanding
taxonomy and evolution, and possibly in making palaeoecological interpretations. More generally
it helps provide a palaeobiological perspective to coccolith research, which has been rather
dominated by biostratigraphy. Also, understanding the ontogenetic growth of coccoliths allows the
application of basic evolutionary concepts such as heterochrony to coccoliths. Since coccoliths have
one of the best fossil records of any group this should enrich evolutionary studies as well as
coccolithophore research.

Acknowledgements . This work was carried out during the tenureship of post-doctoral research fellowships
funded by Clyde Petroleum (J.R.Y.), and the Natural Environment Research Council (P. R.B). The electron
microscopy unit of The Natural History Museum are thanked for their assistance and expertise. Andy Gale,
Dave Ward, and Marigold White provided invaluable support at various stages. Paul Taylor, Dave Watkins,
and an anonymous referee made useful comments on the manuscript.

REFERENCES

black, m. 1971. Coccoliths of the Speeton Clay and Sutterby Marl. Proceedings of the Yorkshire Geological
Society , 38, 381-424.

bown, p. r. 1987. Taxonomy, evolution and biostratigraphy of Late Triassic-Early Jurassic calcareous
nannofossils. Special Papers in Palaeontology , 38, 1-118.

- and cooper, m. k. e. 1989. New calcareous nannofossil taxa from the Jurassic. Journal of
Micropalaeontology , 8, 91-96.

bukry, d. 1969. Upper Cretaceous coccoliths from Texas and Europe. University of Kansas Paleontological
Contributions, Article , 51, 1 -79.

downie, c. 1957. Microplankton from the Kimmeridge Clay. Quarterly Journal of the Geological Society of
London, 112, 413-434.

850

PALAEONTOLOGY, VOLUME 34

emburg, p. R. van 1989. Coccolith formation in Emiliania huxleyi. Unpublished Ph.D. thesis, University of
Leiden Geobiochemistry Unit.

gallois, r. w. 1976. Coccolith blooms in the Kimmeridge Clay and origin of North Sea oil. Nature , 259,
473-475.

- and medd, a. w. 1979. Coccolith-rich marker bands in the English Kimmeridge Clay. Geological
Magazine , 116, 247-260.

grun, w. and zweili, f. 1980. Das kalkige Nannoplankton der Dogger-Malm-Grenze im Berner Jura bei
Liesberg (Schweiz). Jahrbuch der Geologischen Bundesanstalt Wien , 123, 231-341.

mann, s. and sparks, n. h. c. 1988. Single crystalline nature of coccolith elements of the marine alga Emiliania
huxleyi as determined by electron diffraction and high resolution transmission electron microscopy.
Proceedings of the Royal Society of London , Series B , 234, 441 — 453.

myers, k. j. and wignall, p. b. 1987. Understanding Jurassic organic-rich mudrocks - new concepts using
gamma-ray spectrometry and palaeoecology. Examples from the Kimmeridge Clay of Dorset and the Jet
Rock of Yorkshire. 172-189. In leggett, j. k. and zuffa, g. g. (eds). Marine clastic sedimentology concepts
and case studies. Graham and Trotman, London, 209 pp.

noel, d. 1965. Coccolithes Jurassiques. Centre National de la Recherche Scientihque, Paris, 209 pp.

1973. Nannofossiles calcaires de sediments jurassiques hnement lamines. Bulletin du Museum National
d'Histoire Naturelle (3), 75, 95-155.

oschmann, w. 1988. Kimmeridge Clay sedimentation - a new cyclic model. Palaeogeography, Palaeoclimat-
ology, Palaeoecology, 65, 217-251.

outka, D. E. and williams, d. c. 1971. Sequential coccolith morphogenesis in Hymenomonas carterae. Journal
of Protozoology, 18, 285-297.

perch-nielsen, k. 1971. Dursicht tertiarer coccolithen. 939-980. In farinacci, a. (ed.). Proceedings Second
Planktonic Conference, Roma 1970, 2, 1-1369.

romein, a. l t. 1979. Lineages in early Palaeogene calcareous nannoplankton. Utrecht Micropalaeontological
Bulletins, 22. 1 231.

WESTBROEK, P., VAN DER WAL, P., BORMAN, A. H., DE VRIND, J. P. M., KOK, D., DE BRUIJN, W. C. and PARKER,
s. b. 1984. Mechanism of calcification in the marine alga Emiliania huxleyi. Philosophical Transactions of the
Royal Society of London, Series B, 304, 435 — 444.

- young, j. r. and linschooten, p. 1989. Coccolith production (biomineralisation) in the marine alga
Emiliania huxleyi. Journal of Protozoology, 36, 368-373.

young, J. R. 1989. Observations on heterococcolith rim structure and its relationship to developmental
processes. 1-20. In crux, j. a. and van heck, s. e. (eds). Nannofossils and their applications. Ellis Horwood,
Chichester, 356 pp.

JEREMY R. YOUNG

Department of Palaeontology
The Natural History Museum
Cromwell Road, London SW7 5BD

Typescript received 8 June 1990

Revised typescript received 28 September 1990

PAUL R. BOWN

Department of Geological Sciences
University College London
London WC1E 6BT

AMINO ACIDS FROM FOSSILS, FACIES AND

FINGERS

by DEREK WALTON and GORDON B. CURRY

Abstract. The possibility of introducing contaminating amino acids while preparing fossils and rocks for
analysis is a major problem for molecular palaeontology. The most important source of contamination is
human finger tips, which source approximately 100 times the concentration of amino acid than New Zealand
fossil shells. Latex gloves can also transmit appreciable quantities of modern day amino acids. However,
hierarchical clustering and principal component analysis reveal that the relative proportions of finger-tip amino
acids are constant to an extent which allows them to be readily distinguished from those for fossil shells and
their enclosing sediment. This approach therefore represents a useful method of checking for contamination
by finger tips in analyses of fossil molecules.

Amino acids are ubiquitous in fossils and rocks, and represent a potentially rich source of
information on a wide variety of geological topics (Schroeder and Bada 1976; Curry 1988). For
palaeontology, there is considerable potential in studying amino acids from fossil shells and
skeletons, for example in the determination of the isotopic composition of individual amino acids
which were originally synthesized during the life of the fossil organism. The major problem,
however, is that amino acids are so abundant and widely distributed in rocks, sediments and the
atmosphere, that there is considerable potential for contamination of fossil shells by diffusion from
surrounding sediments and rocks. It is partially with this consideration in mind that attention has
recently focused on intracrystalline molecules from brachiopod shells, the rationale being that
because such molecules are entombed within, rather than around, biocrystals (Collins et al. 1988),
they were situated within a protective microenvironment where they were much less prone to
contamination. Brachiopods contain a surprising number of different intracrystalline molecules,
including various proteins which have recently been sequenced (Curry et al. 1991). It is the
breakdown products of these proteins (i.e. amino acids and peptides) which offer the best
opportunity of recovering genuinely mdigeneous molecular information from fossils, and are
present in all groups of fossils as far back as the Cambrian (Curry 1988).

However, concentrating on intracrystalline molecules does not completely solve the problem of
contamination, because there is still the risk of introducing amino acids during sample preparation.
Extraneous amino acids could be added to geological samples during storage (from bacteria, for
example) or even by touching specimens with fingers, which are constantly shedding skin composed
of vast numbers of amino acids. The question of contamination by finger-tip amino acids was first
raised in 1965 (Hamilton 1965; Oro and Skewes 1965), but investigation of this phenomenon has
not kept pace with technological advances in amino acid analysis, which have increased
incrementally in sensitivity since this early, and often overlooked, work.

It was therefore essential to investigate both the abundance and the relative proportion of finger-
tip amino acids as a prelude to a detailed investigation of the molecular palaeontology of fossil shells
from New Zealand. It was hoped that identification and an understanding of the characteristics of
such amino acids would, when compared with data generated from fossils and sediments using an
identical high-sensitivity analysis system, allow the recognition of such contamination and hence
eradicate it as a potential distortion of our analyses.

| Palaeontology. Vol. 34, Part 4, 1991, pp. 851— 858.|

© The Palaeontological Association

852

PALAEONTOLOGY, VOLUME 34

GEOLOGICAL HISTORY OL SAMPLES

The mid-late Pleistocene sediments and brachiopod fossils investigated were collected from the
South Wanganui Basin, which is one of a number of sedimentary basins in the south west of present
day North Island, New Zealand. These basins were formed by tectonism related to the subduction
of the Pacific Plate below North Island, causing maximum subsidence during the early Miocene.
Since this time, there has been general tectonic emergence in these areas (Stern and Davey 1990),
indicating that the major control of sedimentation during the mid-late Pleistocene were glacially
moderated eustatic changes. Glaciation of South Island began during the Hautawan (c. 2-5 Ma) and
caused repeated submergence/emergence of the sediments which, coupled with the continuing
emergence of the region, resulted in a series of well preserved marine terraces, which have undergone
very little burial, along the Taranaki coast (Fleming 1953). It is from these terraces that both the
fossils and sediments were collected.

MATERIALS AND METHODS

Free amino acids from finger tips were collected from the fingers of members of the Department of
Geology and Applied Geology, University of Glasgow. Sample selection was completely random
with respect to the time of day, sex and age of the individuals. Citrate buffer, pH 2-2, was added to
the base of a sterile Cel-cult tissue culture plate well (500 /d of 100 mM), and mixed on a Luckman
(Model 802) suspension mixer to ensure complete coverage of the base. The index finger of each
individual was wiped with a tissue immediately prior to sampling, to remove traces of solid particles,
and then pressed into the well for 1 5 seconds. Nine samples were taken, including some from fingers
which had been washed with either Millipore Reverse Osmosis Water or soap and tap water
immediately prior to the test. One well was left untouched to provide a sample blank, and one was
touched with the finger in a latex glove that had been used for sample preparation and collection.
An aliquot (30 fiX) of each of the samples was applied to the sample frit of the analysis system.

Surface contaminants were removed from a variety of fossil brachiopods by immersion of the
shells in an aqueous solution of sodium hypochlorite (10% v/v), and scraping away the encrusting
epifauna with a steel needle. Intercrystalline molecules from the powdered or disaggregated shells
were destroyed by a second incubation with 10% (v/v) sodium hypochlorite. Amino acids were
recovered by decalcifying the shells in an aqueous solution (20% w/v) of EDTA (Ethylene-
diaminotriacetic acid, disodium salt) at pH 8. EDTA interferes with amino acid analysis, and was
therefore removed prior to the analysis of the macromolecular components from the shell by
dialysis or ultrafiltration using a 10 kDa cut-off membrane or filter. Amino acids were recovered
from sediments by mixing 30 g of dry sediment with 50 ml of ultra-pure water (MilliQ™) in an
airtight bottle held at 1 10 °C for 24 hours. When cool, the liquid was centrifuged (2 k.g.h.), filtered
and finally concentrated using a rotary evaporator (Howe Gyrovap).

Hydrolysis of the peptide bonds between amino acids in a polypeptide is a prerequisite of amino
acid analysis. Both manual and automatic vapour phase hydrolysis was employed in this study. For
manual hydrolysis, 20 //I of the samples, in newly pyrolysed (500 °C/4 hours) glass hydrolysis tubes,
were placed in an airtight flask with 500 //I of 6 N HC1, and heated to 165 °C for 60 minutes under
an Argon blanket (Dupont et al. 1989). After hydrolysis the samples were reconstituted in an
aqueous solution of EDTA (0025% w/v; tripotassium salt) and thoroughly mixed. Automated
hydrolysis, reproducing these conditions with 6N HC1 vapour phase hydrolysis, was performed by
the Applied Biosystems 420-H amino acid analyser. An aliquot (10-20 / A ) of solution was loaded
onto the amino acid analyser, and derivitized using PITC (Phenylisothiocyanate ; Heinrikson and
Meredith 1984), prior to identification using a linked narrow bore HPLC (High Performance Liquid
Chromatography) system optimized for PTC-amino acids. Amino acids were identified and
quantified relative to amino acid standards (Pierce standard H). All analyses were repeated at least
once to ensure reproducibility. Water and buffers used were also analysed to check for
contamination.

WALTON AND CURRY: FOSSIL AND CONTAMINANT AMINO ACIDS

853

Glycine

Neothyris
Sediments
Finger tips

text-fig. 1 . Pie charts showing the absolute proportions of selected amino acids (normalized to nmol/g) from
the shells of Plio-Pleistocene Neothyris brachiopods, sediments (from the Plio-Pleistocene South Wanganui
Basin of New Zealand; see Text-figure 3a for Formation details), and from finger tips (1 ml ss 1 g).

table 1. Results of amino acid analyses (nmol/g). Figures in brackets refer to sample size.

Neothyris (4) Finger-tips (8) Sediments (3)

mean

S.D.

mean

S.D.

mean

S.D.

Aspartic acid

0-659

0-450

18-347

15-170

0-000

0-000

Glutamic acid

0-161

0-058

16-020

11-470

0-310

0-197

Serine

0-168

0-089

38-772

44-526

1-091

0-303

Glycine

0-505

0143

56-21 1

19-670

5-978

0-768

Alanine

0-211

0164

22-337

7-383

5-221

1-349

Valine

0-105

0-057

9-323

3-800

3-200

0-910

ABSOLUTE ABUNDANCE OF AMINO ACIDS

In absence terms, the quantities of amino acids that can be transmitted by finger-tips are
considerably greater than those present in fossil shells and in sediments (Table 1 ; Text-fig. 1). Not
surprisingly, there is slight variation in the yields for individual amino acids in some samples, but
typically the levels for finger tips are 10 times higher than in sediments, and 100 times higher than
in fossils (for example, glycine, see Table 1). Such levels are probably reasonably realistic
assessments of the amounts of free amino acids which would be transferred to fossils or sediments
by touching them with wet fingers. As the samples were not hydrolysed, amino acids still attached
to one another in proteins from the skin would not have been quantified; small fragments of human
skin are constantly being shed, and there is no doubt that the potential levels of amino acids
transferred by finger contact would be greater than that measured here. Such contamination
occurring at a critical stage of sample preparations will clearly overwhelm the indigenous amino
acids.

854

PALAEONTOLOGY, VOLUME 34

Amino Acid

Amino Acid

text-fig. 2. a, relative proportions (Mole %) of the common amino acids from the finger tips of five individuals
from the first run of the experiment, b, relative proportion (Mole %) of the common amino acids for New

Zealand sediments.

WALTON AND CURRY: FOSSIL AND CONTAMINANT AMINO ACIDS

855

text-fig. 3. a, relative proportions (Mole %) of the amino acids from the shells of fossil Neothyris. Sample
250 19001 A shows the effect of incomplete removal of EDTA, a phenomenon which provides a check on the
preparation purity of the sample, b, absolute properties (un-normalized, the concentrations (pmol) are for
ammo acids in 30 /A of buffer) showing the potential levels of contamination from latex gloves in comparison

to finger tip data.

856

PALAEONTOLOGY, VOLUME 34

(///?) Sediments
( ) Finger

6^ Fossil

X Citrate Buffer

O Tap water

text-fig. 4. Single linkage cluster analysis (representing nearest neighbour groupings) for the relative
proportion data (Mole %), generated by DATADESK™.

RELATIVE ABUNDANCE OF AMINO ACIDS

Amino acid data are often presented in the form of mole percentages which overcomes the difficulty
of comparing samples of varying size. Text-figure 2a shows a remarkably consistent pattern in the
relative abundance of finger-tip amino acids, with all individuals being rich in either serine or
glycine. These results show some similarity with those of Oro and Skewes (1965), where serine is the
most common amino acid. In sediments, serine and glycine are also the most abundant (Text-fig.
2b), while glycine is also a major constituent of fossil Neothyris shells (Text-fig. 3a). The relative
proportions of amino acids are quite consistent within all three groups sampled (Text-figs 2a, 2b and
3a), to the extent that incomplete removal of EDTA from fossil samples becomes readily apparent
(Text-fig. 3a). Apart from this sample, the fossil Neothyris all have consistent amino acid profiles.

Superficially there are a number of common features between these three datasets, but such
subjective assessments are potentially misleading. To produce a more rigorous assessment of the
relative similarity between samples, hierarchical clustering analysis using the statistical programme
DATADESK™ on a Macintosh micro-computer was applied to the relative abundance data. The
resulting cluster diagram resolves the finger tip, sediment and fossil shell amino acid data into
discrete clusters (Text-fig. 4). The citrate buffer and tap water have been included as outgroups.

Principal component analysis (PCA) using DATADESK™ was also applied to the datasets
(Text-fig. 5), and this confirms the groupings of the cluster plot. A useful attribute of PCA is that
it is possible to use the eigen-vector values to determine which elements (in this case individual
amino acids) are most important in distinguishing between the different groups. Text-figure 5 shows

WALTON AND CURRY: FOSSIL AND CONTAMINANT AMINO ACIDS

857

-0.0 1.5 3.0

Second Principal Component

text-fig. 5. Scatter plot of the first two principal components for relative proportion data (Mole %), generated
by DATADESK™. These data do not take into account variations in the actual concentrations of the amino

acids.

that the clear resolution between sediments, fossils, and fingers is achieved along the first principal
component axis, which is aligned in the direction of maximum varibility and in this case contributes
26-9% of the total variation detected. Examining the eigen-vector values from this analysis, it is
clear that this differentiation primarily reflects high scores for serine (+0273), histidine ( + 0-276),
arginine ( — 0-258), proline ( — 0-275), tyrosine ( + 0-229), valine ( — 0-377), methionine ( — 0-205),
isoleucine ( — 0-323). leucine ( — 0-347) and phenylalanine ( — 0-345). The sign of these scores is also
informative, indicating that samples which have high scores on this eigen-vector (i.e. the fingers)
contain large amounts of serine, histidine and tyrosine, and relatively low yields of arginine, proline,
valine, methionine, isoleucine, leucine and phenylalanine.

DISCUSSION

In one respect it is not surprising to discover a clear differentiation between these three groups of
samples. Finger-tip amino acids are predominantly derived from human skin proteins. Amino acids
from sediments presumably come from a complex mixture of sources including the breakdown of
various life-forms which originally lived in or on the sediment and mobile amino acids carried in
by percolating ground waters. The fossil amino acids considered here have been extracted from
intracrystalline macromolecules; most published data from shells have also included a component
of intercrystalline macromolecules (e.g. Jope 1967a, 19676; Weiner et al. 1976), which are not
protected by shell crystallites, and hence may be easily degraded or contaminated. It is very
encouraging for the study of amino acids in these New Zealand successions that the finger-tip data
can be so readily distinguished from the fossil and sediment data, both in terms of absolute
abundance and mole percentages. Further samples of fossil shells have been analysed, and a similar
pattern of differentiation from finger-tip and sediment data is apparent.

In practical terms, the use of clustering and statistical approaches represents an important and

858

PALAEONTOLOGY, VOLUME 34

rapid method of checking for human contamination in geological samples. As the absolute
abundance of finger-tip amino acids is so much greater than that of fossils, any contamination
should produce an analysis which clusters with the former rather than the latter, and this is probably
also true for sediments although to a lesser extent.

The results of this investigation have also provided important information for sample preparation.
High levels of contamination from finger-tip amino acids are to be expected, but these experiments
have shown that even hands which have washed prior to handling samples could be a significant
source of contamination. Much more surprising, and potentially worrying, is the discovery that
even latex laboratory gloves can be a source of contamination amino acids (Text-fig. 3b). Latex
gloves are widely used in laboratories, but obviously they give a false sense of security with modern
high-sensitivity amino acid analysers capable of detecting a few picomoles (10“12 mole) of each
amino acid. For optimum recovery of uncontaminated indigenous amino acids from fossils and
sediments, it is necessary to minimize or completely eradicate any handling of the sample or of the
interior surfaces of any vessel involved in the release of the incarcerated amino acids.

Acknowledgements , We thank Dr M. Cusack for her comments on this paper, and the members of the
Department of Geology and Applied Geology for providing samples. A UK NERC studentship to D.W.
(GT4/89/GS/42), and funding from the Royal Society and NERC to G.B.C. are gratefully acknowledged.

REFERENCES

COLLINS, M. J., CURRY, G. B., QUINN, R., MUYZER, G., ZOMERDIJK, T. and WESTBROEK, P. 1988. Sero-taxonomy of
skeletal macromolecules in living terebratulid brachiopods. Historical Biology 1, 207-224.
curry, g. b. 1988. Amino acids and proteins from fossils. 20-33. In runnegar, b. and schopf, j. w. (eds).
Molecular evolution and the fossil record. Short Courses in Paleontology , 1. Paleontological Society,
Knoxville, USA, 167 pp.

— cusack, m.. endo, k., walton, D. and quinn, R. 1991. Intracrystalline molecules from brachiopod shells.
35-40. In suga, s. and nakahara, h. (eds). Mechanisms and phytogeny of mineralization in biological systems.
Springer-Verlag, Tokyo, 517 pp.

dupont, d. r., keim, p. s., chui, a., bello, r., bozzini, m., and Wilson, k. j. 1989. A comprehensive approach
to amino acid analysis. 284—294. In hugli, t. e. (ed. ). Techniques in protein chemistry. Academic Press, New
York, 357 pp.

Fleming, c. a. 1953. The geology of Wanganui Subdivision. New Zealand Geological Survey Bulletin , 52, 1-362.
Hamilton, p. b. 1965. Amino acids on hands. Nature, 205, 284—285.

heinrikson, r. L. and Meredith, s. c. 1984. Amino acid analysis by reverse phase high-performance liquid
chromatography: precolumn derivatization with phenylisothiocyanate. Analytical Biochemistry, 136, 65-74.
jope, m. 1 967c/. The protein of brachiopod shell-I. Amino acid composition and implied protein taxonomy.
Comparative Biochemistry and Physiology , 20, 209-224.

- 1967A The protein of brachiopod shell-II. Shell protein from fossil articulates: amino acid composition.
Comparative Biochemistry and Physiology , 20, 601-605.
oro, j. and skewes, h. b. 1965. Free amino acids on human fingers: the question of contamination in
microanalysis. Nature, 207, 1042-1045.

schroeder, r. a. and bada, J. L. 1976. A review of the geochemical applications of the amino acid racemization
reaction. Earth Science Reviews, 12, 340-391.

stern, t. a. and davey, f. j. 1990. Deep seismic expression of a foreland basin: Taranaki Basin, New Zealand.
Geology , 18, 979-982.

weiner, s., lowenstam, h. a. and hood, l. 1976. Characterisation of 80 million year old mollusk shell proteins.
Proceedings of the National Academy of Sciences of the USA, 73, 2541-2545.

DEREK WALTON and GORDON B. CURRY

Department of Geology and Applied Geology
University of Glasgow
Glasgow G12 8QQ, UK

Typescript received 13 September 1990
Revised typescript received 17 December 1990

BAJOCIAN STEPHANOCERATID AMMONITES
FROM THE BAKONY MOUNTAINS, HUNGARY

by A. GALACZ

Abstract. Ammonites are described from the Bajocian Otoites sauzei and Stephanoceras humphriesianum
Zones in a stratigraphically discontinuous red, pelagic limestone sequence. The fauna consists mainly of
phylloceratids and lytoceratids, thus showing a typical Mediterranean association. The Ammonitina are
almost exclusively represented by stephanoceratids, suggesting a part of the Sauzei Zone and two subzones of
the Humphriesianum Zone. The Sauzei Zone is indicated by a single specimen of Emileia sp. The upper part
of the Stephanoceras humphriesianum Subzone is suggested by Dorsetensia subtecta Buckman and associated
stephanoceratids: Stephanoceras triplex Weisert, S. scalare Weisert, S. mutabile (Quenstedt), and S. leoniae
Schmidtill and Krumbeck, while the basal part of the Teloceras blagdeni Subzone is identified from an
assemblage of Stephanoceras sturanii Pavia, S. ( Normannites ) orbignyi Buckman and S. (N.) sp. aff. fort is Pavia.
The stratigraphic representation in the sequence is rather incomplete. However, a description of its fauna may
help to evaluate similar assemblages recorded in other areas of the Mediterranean region.

Detailed studies on Jurassic ammonites from the Bakony Mountains (Transdanubian Hungary)
have resulted in the recognition of exceptionally rich Liassic faunas. These results contributed
substantially to our understanding of Mediterranean Lower Jurassic stratigraphy. Recent
investigations indicate a similar, important role for the Middle Jurassic faunas. The ammonites
described here came from a typical Mediterranean Jurassic section of red, nodular pelagic
limestones below an ammonite-free radiolarite. The fauna represents only a small part of the Middle
Jurassic Bajocian stage and consists mainly of phylloceratids and lytoceratids; the Ammonitina are
overwhelmingly represented by stephanoceratids.

Stephanoceratidae, that characteristic Bajocian group, are very incompletely known, especially in
the Mediterranean. In spite of the relatively large number of described species, the interpretation
of taxa and their stratigraphic distribution is far from clear. The aim of this paper is to make known
the taxonomy and stratigraphic occurrences of some stephanoceratids within a small, but
stratigraphically well-defined succession.

LOCALITY AND SUCCESSION

Kozoskut Ravine is a narrow valley at the southwestern margin of the Hajag Hills, near the village
of Harskut, in the Bakony Mountains (Text-fig. 1). The Jurassic sequence exposed here was first
mentioned by Noszky (1943), who gave a short description of the local stratigraphy. The
Hettangian Dachsteinkalk-type limestone is overlain by higher Lower Liassic, then Toarcian and
Bajocian ammonitic limestones, which are followed by a cherty radiolarite, then a complete Upper
Jurassic-Lower Cretaceous calcareous succession. Later, in a comprehensive treatment of the
Jurassic of Hungary, Noszky (1961) gave a similar summary, indicating some non-sequences in the
section.

The outcrop was visited during the Mediterranean Jurassic Colloquium in 1969. A detailed
description and sections outlining the Jurassic sequence was given by Fiilop et al. (1969, fig. 19) and
Fulop (1971, fig. 10). Konda (1970, pp. 192-193) has described the lithology and the microfacies
features of the Liassic and Middle Jurassic rocks. The biostratigraphic evaluation of the ammonite
faunas collected was made by Geczy (1971, 1976) for the Liassic, and a preliminary account of the
results of Middle Jurassic studies has also been published (Galacz 1976).

IPalaeontology, Vol. 34, Part 4, 1991, pp. 859-885, 5 pls.|

© The Palaeontological Association

860

PALAEONTOLOGY, VOLUME 34

text-fig. 1. Location of Kozoskut Ravine in the Bakony Mountains, Transdanubian Hungary.

These biostratigraphic evaluations made it clear that considerable hiatuses can be identified
within this sequence with a basically uniform lithology. The Liassic and the Bajocian is represented
here by the typical Mediterranean Rosso Ammonitico limestone, which is a thin, pelagic, nodular
carbonate with ferromanganese nodules and crusts. These sequences are common in the Bakony
Mountains and elsewhere in the Mediterranean region, and are generally regarded as accumulations
on elevated parts of submarine topographic highs in the Jurassic deep-sea Tethys.

In the Kozoskut sequence, the incomplete representation of the Liassic and Middle Jurassic is
demonstrated by the complete lack of the Sinemurian and Aalenian stages (see Text-fig. 2).
However, the documented stages are only partially represented, for the 119 cm thick Lower Jurassic
limestones yielded ammonites merely from the Tragophylloceras ibex Zone of the Pliensbachian and
the Hildoceras bifrons and Dumortieria levesquei Zones of the Toarcian. Similarly, the 133 cm thick
Bajocian series is represented only by the Otoites sauzei and the Stephanoceras humphriesianum
Zones.

Detailed stratigraphic evaluation of the Bajocian ammonites revealed the presence of further,
minor discontinuities. Bed 1 1, with a single Sauzei Zone ammonite species and only two specimens,
certainly represents only a portion of this zone. The limestone above Bed 1 1 gave a fauna with a
very rich Stephanoceras association, which can be correlated to a limited extent with the middle and
upper parts of the Humphriesianum Zone.

Jurassic successions of similar lithology and discontinuous biostratigraphic representation, where
the paracomformable beds contain only a few faunal horizons of certain stages, are common in the
Mediterranean pelagic realm. Thus descriptions of locally intermittent faunal sequences may help
to piece together a reliable picture of the ammonite successions for the wider area.

COMPOSITION AND CORRELATION OF THE BAJOCIAN FAUNA

Bed 1 1, the single layer of Otoites sauzei Zone age, gave only two ammonite specimens; therefore,
detailed evaluations were made only for the overlying ten beds belonging to the Stephanoceras
humphriesianum Zone.

In general aspect the fauna is typically pelagic. The megafauna is almost entirely represented by

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

861

i i
i i
i
i
i

1 ' i

! i
i i

c E -c

*0 CD N C

g- £ Q) -o 2

W Qj D Q)

CO O o 0 Q

S|>.«2

<u >.C <0 >

Ur'1- —

= Q-
> —

- £

D O) (D

"> -2 do
to <d

cocococococy^. — . — ^

(O CO
(D CD

CO

CO ■ -

CO

C C CD

<DCD<DCDCD<DCCO

O O O O O O

o o o o o o

c c c c c c

co co

E E

o O C <D

CL o

o o

CD (0

0) -c

— cl

E '

LU(/)C/)(/)(/)C/)C/)C/)C/)(/)

CO co
x: x:
CL Q.
CD CD

CO CO

x: x:

CL CL
CD CD

c o o

0.2! Z

red flaser limestone with
ferromanganese nodules

marly, nodular limestone
crinoidal limestone
massive limestone

text-fig. 2. The Lower and Middle Jurassic limestone sequence in the Kozoskut Ravine, with the range of the

Bajocian ammonites (S. = Sauzei).

ammonites: only a few belemnite rostra, a single bivalve ( Inoceramus sp. ), and one nautiloid were
collected. Benthonic elements are completely missing.

The composition of the ammonite assemblages is characteristically Mediterranean, with
phylloceratids dominant and lytoceratids common. The percentage distribution of the ammonite
fauna gave the following values by suborders: Phylloceratina = 54 % ; Lytoceratina = 13 % ;
Ammonitina = 33%. A closer analysis of the Ammonitina reveals that this is an almost pure
stephanoceratid fauna: there are eighty-seven Stephanoceras, Normannites and Stemmatoceras
specimens, but only three poorly preserved oppeliids and a single Dorsetensia. Other groups (e.g.
sphaeroceratids : Chondroceras , Sphaeroceras ; haploceratids : Strigoceras , Cadomoceras , Poecilo-
morphus) are conspicuous by their absence.

These features make the Kozoskut fauna exceptional among assemblages of similar age. The
recently monographed Sub-Mediterranean faunas from Digne, SE France (Pavia 1983), from the

GALACZ: BAJOC1AN AMMONITES FROM HUNGARY

862

PALAEONTOLOGY, VOLUME 34

Betic Cordilleras (Sandoval 1983) and from the Iberian Cordilleras (Fernandez Lopez 1985) are
similar in having common elements, but different in faunal composition. The only places where
similar faunas were recorded (but only partially documented) are within the s.s. Mediterranean
region: from the Alpi Feltrine, N. Italy (Della Bruna and Martire 1985) and from the Appennino
marchigiano. Central Italy (Cecca et al. 1986).

In the absence of non-stephanoceratid Ammonitina, a correlation with other Humphriesianum
Zone faunas is difficult. A further complication is that the subzonal subdivision of the Stephanoceras
humphriesianum Zone, as well as its successive faunal horizons, are incompletely known. Parsons
(1976) gave a comprehensive overview of the subzonal units, but in the last decade important new
results have been published. Some of the recently suggested subzonal divisions are given in Table
1 . This shows that the infrazonal units of the Bajocian Humphriesianum Zone are far from generally
agreed. While a subdivision into three or four subzones seems to be accepted, several ‘horizons’ are
suggested without exact inter-regional correlation.

PARSONS 1976
(adopted here)

PAVIA 1983

SANDOVAL 1983

Subfurcatum Zone

Niortense Zone

Leptosphinctes Zone

Blagdeni

Subzone

Humphriesianum

Subzone

Romani Subzone

Blagdeni
Subz

Dubium
Hor .

Humphriesianum

Subzone

Romani

Subz.

Paulu-
lus H.

Edouar-
diana H.

Blagdeni

Subzone

Humphriesianum

Subzone

Cycloides

Subzone

Sauzei Zone

Sauzei Zone

Sauzei Zone

DIETL, HAGER and

FERNANDEZ LOPEZ 1985

OHMERT 1988

SAUER 1984

Niortense Zone

Subfurcatum Zone

Niortense

Zone

Blagdeni

Blagdeni

Blagdeni

CD

Subzone

CD

Biohorizon

CD

Subzone

o

O

o

E_i

E

E

C

Humphriesianum

C

Scalare

C

Humphriesianum

Subzone

in

Biohorizon

in

Subz .

Umbilicum

CD

CD

CD

Horizon

SZ

sz

SZ

E

Romani Subzone

ZD

ZD

Nodosum

■r

Pinguis Subzone

Biohorizon

Pinguis Subzone

Sauzei Zone

Sauzei Zone

Sauzei Zone

table 1. Comparison of some recent subdivisionings of the Bajocian Stephanoceras humphriesianum Zone.

The Humphriesianum Zone fauna of the Kozoskut Ravine can be divided into two assemblages.
The lower one, from Bed 10 to Bed 4 contains stephanoceratids, Stephanoceras triplex , S. scalare,
S. mutahile , S. leoniae , and the associated Dorsetensia subtecta. The best match with this fauna is
that from Beds 355 to 339 of the Chaudon section, Digne area (Pavia 1983, Table III6). While some
of Pavia's determinations are reinterpreted here (see descriptions below), the basic composition of
stephanoceratids s.s. is very similar. Thus the lower seven beds of the Kozoskut Humphriesianum
Zone section can be correlated with the upper part of the Stephanoceras humphriesianum Subzone.

In Bed 3 new elements appear: Stephanoceras sturanii and Normannites spp. (see Text-fig. 2).
Comparable forms are known from the Digne area in the basal part of the Blagdeni Subzone
(‘ Dubium horizon’ of Pavia 1983, pp. 36-37). Accordingly, the topmost part of the Kozoskut
Bajocian belongs to the lower Blagdeni Subzone, and the whole sequence encompasses the boundary
beds of the two higher subzones of the Humphriesianum Zone.

SYSTEMATIC PALAEONTOLOGY

Deposition of specimens. All the ammonites studied are deposited in the Palaeontological
Collections of the Hungarian Geological Survey, Budapest. Measured and illustrated specimens
have inventory numbers.

Dimensions. Measurements are given in millimetres, in the following order and with the following
abbreviations: maximal preserved diameter (MDm); diameter (Dm); whorl height (Wh); whorl
breadth (Wb); and umbilical diameter (Ud). Numbers in parentheses refer to dimensions as a

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

863

percentage of the diameter. Pr indicates number of primary ribs, S the number of secondary ribs per
whorl.

Order ammonoidea Zittel, 1884
Suborder phylloceratina Arkell, 1950
Superfamily phyllocerataceae Zittel, 1884
Family phylloceratidae Zittel, 1884
Genus calliphylloceras Spath, 1927

Type species. Phylloceras disputabile (Zittel, 1868) = Ammonites tatricus Kudernatsch non Pusch, 1852, p. 4, pi.
1, figs 1-4, by original designation of Spath (1927-33, p. 24).

Calliphylloceras heterophylloides (Oppel, 1856)

Plate 1, fig. 4; Text-fig. 3a

1856 Ammonites heterophylloides Oppel, p. 493.
v 1871 Phylloceras heterophylloides Oppel; Neumayr, p. 331, pi. 15. fig. 1 a-c.

v 1878 Phylloceras heterophylloides Oppel; Bayle, pi. 32, figs 1, 2, 5-8.

1927 Phylloceras gr. de heterophylloides Oppel; Roman and Petouraud, p. 16, pi. 2, figs 9 and 10.
non 1937 Phylloceras aff. heterophylloides Oppel; Kakhadze, p. 126, pi. I, fig. 4.

71937 Phylloceras heterophylloides Oppel; Horwitz, p. 240, pi. 9, fig. 1.

1956 Calliphylloceras heterophylloides Oppel; Kakhadze and Zesashvili, p. 18, pi. 2, figs 2 and 3.
71973 Calliphylloceras cf. heterophylloides (Oppel); Myczinsky, p. 59, pi. 1, fig. 6.

1976 Calliphylloceras heterophylloides (Oppel); Joly, p. 213, and figures.

Material. Eleven measured, well preserved internal moulds and several fragments or strongly corroded
specimens.

Measurements.

MDm Dm

Wh

Wb

Ud

J9471

94-5 94-5

54-5

(57-5)

36 (38)

6 (6-5)

74

42

(56-5)

25 (33-5)

5-5 (7-5)

67-5

38

(56)

24 (35-5)

4-5 (6-5)

41 5

23

(55)

13-5 (32-5)

4 (10)

21

10

(47)

7 (33)

4 (19)

1 1

5

(45)

4-5 (41)

2-5 (22)

J9477

70-5 70-5

41

(58)

24 (34)

5 (7)

51

29

(57)

18 (35)

4 (8)

Description. Medium size ammonite. Most of the studied specimens are well preserved internal moulds, but
the body chamber is usually missing. The umbilicus is narrow, attaining only 5-7% on the outer whorls. The
whorl section is high-oval with steep umbilical wall, broadly arched, slightly compressed flanks and narrow,
highly arched venter. The inner whorls are inflated. The whorls are smooth, shell-sculpture cannot be seen. The
larger forms show seven, rarely eight constrictions per whorl. The constructions are slightly prorsiradiate and
straight, becoming curved near the venter, which they cross with a strong forward arch.

The suture line (Text-fig. 3a) is of characteristic Phylloceras type, with diphyllid first lateral saddle, triphy 1 1 id
second lateral saddle, and diphyllid then monophyllid auxiliary elements.

Remarks. The lectotype was designated by Joly (1976, p. 213, pi. 21, fig. I). This is a relatively large,
wholly septate specimen with partially preserved shell from the Upper Bajocian of St Vigor,
Calvados, France. Joly figures (1976, pi. 21, fig. 3) another specimen from the same locality as
'paratype'. This is the original of plate 42, figure 2 of Bayle (1878), and thus should be regarded
as a topotype. The Kozdskut specimens are readily comparable with those from St Vigor. The single
difference is in the umbilical width which is extremely small in the large lectotype. On the basis of
the detailed statistical studies of Joly (1976, pp. 215-223), C. heterophylloides can be easily

864

PALAEONTOLOGY, VOLUME 34

distinguished from the similar, but somewhat younger (i.e. uppermost Bajocian to Callovian),
C. disputabile (Zittel).

Distribution. The exact horizon of the type within the Upper Bajocian is uncertain. The literature mentions this
species usually from the Humphriesianum Zone, whereas the Madagascan material of Joly came from the
Upper Bathonian.

C. heterophylloides is one of the commonest species in the Kozoskut Bajocian, occurring in almost every bed
of the Humphriesianum Zone.

Genus holcophylloceras Spath, 1927

Type species. Ammonites zignodianus (d'Orbigny, 1848, p. 493, pi. 182), see Galacz 1980, p. 41.

Holcophylloceras zignodianum (d’Orbigny, 1848)

Plate 2, fig. 2; Text-fig. 3c

1848 Ammonites Zignodianus d’Orbigny, p. 493, pi. 182, figs 1-3.
v 1871 Phylloceras mediterraneum Neumayr, p. 340, pi. 17, figs 2-5.

1980 Holcophylloceras zignodianum (d’Orbigny); Galacz, p. 41, pi. 5, figs 4 and 5; pi. 6, fig. 1 ; pi. 7,
fig. 1 ; text-figs 30-32. (cum syn.).

Material. Twenty-five specimens in various states of preservation.

Measurements.

Mdm Dm Wh Wb

Ud

J9468 92

92 50 (52-5) 34 (37)

68-5 35 (51) 23 (33-5)

12(13)

10(14-5)

Description. Medium to large specimens; the figured specimen (J9468) is a well preserved internal mould. The
deep and narrow umbilicus has low, vertical walls, which form distinct shoulders with the slightly convex flank.
The venter is narrow and highly arched. The whorls are smooth, except for the characteristic constrictions,
which are prorsiradiate furrows with a sharp break just above the middle of the flanks. There are seven
constrictions per whorl. The figured specimen shows the end of the phragmocone at 95 mm diameter; no
specimen with an entire body chamber and aperture was found.

The suture line (Text-fig. 3c) shows diphyllic first lateral and triphyllic second lateral saddle and deep lobes.

Remarks. This species was discussed in detail recently (Galacz 1980, pp. 41 44), so this short
reference to newly collected specimens seems sufficient.

Distribution. H. zignodianum is a species of extremely wide stratigraphic range, appearing in the Bajocian
Otoites sauzei Zone and enduring up to the Upper Jurassic. In the Kozoskut sequence this is the commonest
ammonite, occurring in almost every bed of the Humphriesianum Zone.

EXPLANATION OF PLATE I

Fig. I Emileia sp. indet. J9469; Kozoskut; Bed 11, Otoites sauzei Zone; lateral view, x 1.

Fig. 2. Stephanoceras ( Stephanoceras ) mutabile (Quenstedt). J9496; Kozoskut; Bed 9, Stephanoceras
humphriesianum Subzone ; lateral view, x 1 .

Fig. 3. Stephanoceras (Stephanoceras) sp. J9488; Kozoskut; Bed 9, Stephanoceras humphriesianum Subzone;
lateral view, x 0-6.

Fig. 4. Calliphylloceras heterophylloides (Oppel). J9477, Kozoskut; Bed 2, Teloceras blagdeni Subzone; lateral
view, x0-75.

Fig. 5. Stephanoceras ( Stephanoceras ) sturanii Pavia. J9461; Kozoskut; Bed 3, Teloceras blagdeni Subzone;
lateral view of a specimen with distorted body chamber, x 1.

PLATE 1

GALACZ, Emileia, Stephanoceras, Calliphylloceras

866

PALAEONTOLOGY, VOLUME 34

Suborder lytoceratina Hyatt, 1889
Superfamily lytocerataceae Neumayr, 1875
Family lytoceratidae Neumayr, 1875
Genus nannolytoceras Buckman, 1905

Type species. Ammonites pygmaeus (d’Orbigny 1842-51, p. 391, pi. 129, figs 12 and 13), by original designation
of Buckman (1905, p. 151).

Nannolytoceras polyhelictum (Bockh, 1881)

Plate 3, fig. 2

v 1881 Lytoceras polyhelictum Bockh, p. 35, pi. 1, figs 2 and 3.

1964 Eurystomicercis polyhelictum (Bockh); Pugin, p. 42, pi. 3, fig. 7; text-fig. 8 (cum syn.).

1980 Nannolytocerns polyhelictum (Bockh); Galacz, p. 51, pi. 10, fig. 2; pi. 11, fig. 1 ; text-figs 39^4 1

(cum syn.).

Material. Four poorly preserved, fragmentary internal moulds.

Measurements.

M Dm

Dm

Wh

Wb

Ud

J9476

61

61

18

(29-5)

18

(29-5)

29

(47-5)

J9475

54

54

15

(28)

15

(28)

26-5

(49)

41

13-5

(33)

13-5

(33)

19-5

(47-5)

Description. The figured specimen is a small fragment. The umbilicus is wide, the whorl section high-oval,
becoming subcircular on the body chamber. The umbilical wall and the flanks are rounded, the venter is slightly
arched. The whorls are smooth, only constrictions are present. The number of constrictions is five on the
outermost whorl. They are projected and slightly curved, crossing the venter perpendicularly. The frontal
margin of the constrictions is inflated. No entire body chamber is preserved, thus the aperture cannot be seen.
Suture line is not visible.

Remarks. This species was discussed (Galacz 1980, p. 51) on the basis of bigger and better preserved
material. The Kozoskiit specimens match these previously studied forms well.

Distribution. The type (Bockh 1881, pi. 1. fig. 1) came from the Niortense Zone of the Mecsek Mountains,
southern Hungary, but from there and elsewhere, Humphriesianum Zone occurrences were also mentioned. The
specimens described here came from Beds 8, 4 and 2, i.e. from the middle to the upper part of the
Humphriesianum Zone.

Suborder ammonitina Hyatt, 1 889
Superfamily hildocerataceae Hyatt, 1867
Family sonniniidae Buckman, 1892
Genus dorsetensia Buckman, 1892

Type species. Ammonites Edouardianus (d’Orbigny 1844, p. 392, pi. 130. figs 3-5), by original designation of
Buckman (1892 p. 302).

Dorsetensia subtecta Buckman, 1892
Plate 2, fig. 4; Text-fig. 3b

v 1892 Dorsetensia subtecta Buckman, p. 309, pi. 54, figs 3-5; pi. 55, figs 1 and 2.

p. 1935 Dorsetensia subtecta Buckman; Dorn, p. 103, pi. 21, fig. 2; pi. 25, fig. 7; pi. 29, fig. 4; pi. 8, figs

9 and 10 only.

71937 Witchellia ( Dorsetensia ) subtecta Buckman; Horwitz, p. 254, pi. 10, fig. 4.

71937 Witchellia subtecta Buckman; Gillet, p. 67, text-fig. 49.

1943 Dorsetensia subtecta Buckman; Roche, p. 13, pi. 2, fig. 2.

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

867

1949

1951

1951

71961

1967

1968
non 1973

1977

1983

'.>1984

1985

Dorsetensia Thilense Maubeuge, p. 172, pi. 12; text-figs on p. 172.

Dorsetensia subtecta Buckman; Maubeuge, p. 34, pi. 12, fig. 2.

Dorsetensia cf. subtecta Buckman; Maubeuge, p. 36. pi. 10, fig. 3.

Dorsetensia sp. aff. thilense Maubeuge, p. 71, fig. on p. 72.

Dorsetensia subtecta Buckman; Kopik, p. 27, pi. 7, fig. 5; pi. 8, figs 1 and 2; pi. 9, fig. 1 ; text-
figs 13 and 14.

Dorsetensia liostraca subtecta Buckman; Huf, p. 103, pis 41 47.

Dorsetensia cf. D. subtexta [s/cl] Buckman; Imlay, p. 71, pi. 28, figs 1-7; pi. 29, fig. 7.
Dorsetensia cf. subtecta Buckman; Parsons, p. 212, pi. 13, fig. I.

Dorsetensia ( Dorsetensia ) subtecta Buckman; Pavia, p. 61, pi. 5, figs 3 and 8.

Dorsetensia subtecta Buckman; Dietl, Franz and Reis, p. 310, text-fig. 2/1.

Dorsetensia subtecta Buckman; Schlegelmilch, p. 66, pi. 20, fig. 3.

Material. A single, well preserved, but partly fragmentary internal mould.

Measurements.

MDm Dm

Wh

Wb

Ud

Wh/Wb

J9494 210 210

88 (42)

38 (18)

52 (25)

2-31

156

70-5 (45)

31 (20)

36 (23)

2-27

Description. A large specimen with a relatively wide, shallow umbilicus. The whorl section is high, compressed.
The umbilical wall is flat, oblique, forming a distinct shoulder with the flank. The lateral side is slightly convex,
and forms a ventro-lateral edge with the narrow, tectiform venter. The internal mould bears a blunt ventral
keel showing traces of the hollow keel of the shell. The preserved parts of the inner whorls have no traces of
ribbing. The specimen is septate up to 1 50 mm diameter, and has preserved body chamber of nearly half a
whorl. The terminal part of the last whorl, with the aperture, is missing.

The suture line is partially visible (Text-fig 3b). This shows finely serrated lobes and saddles as compared to
those usual in the genus. However, the broad lateral lobe indicates the characteristic Dorsetensia- type suture.

Remarks. D. subtecta is distinguished from the similar, large Dorsetensia by its relatively wide
umbilicus and high, flattened, nearly parallel flanks. Huf (1968, p. 103) gave a detailed discussion
on the variations. The whole group, with its size, style of ornament on the early whorls and
septicarinate venter shows affinities to Sonniniinae, and may be better placed in the genus Sonninites
Buckman (J. Callomon, pers. comm. 1986).

The Kozoskiit specimen is very similar to the large forms described by Maubeuge (1951, p. 34),
the closest form being the variety previously designated as D. thilense (Maubeuge 1949, p. 172).
Another very similar form is that described by Hoyermann (1917, p. 39) and refigured later by Huf
(1968, pi. 43, fig. lc; pi. 44, fig. I).

Distribution. D. subtecta , like other large Dorsetensia species, usually occurs in the lower part of the
Humphriesianum Zone. In the type area of Oborne, Dorset, Parsons (1976, pp. 121, 131) collected topotypes
in association with typical Romani Subzone ammonites. Higher occurrences within the Humphriesianum Zone
are also known (see e.g. Pavia 1983, p. 61). The Kozoskiit specimen came from Bed 6, i.e. from the higher part
of the Humphriesianum Subzone.

Superfamily stephanocerataceae Neumayr, 1875
Family otoitidae Mascke, 1907
Genus emileia Buckman, 1898

Type species. Ammonites brocchii (J. Sowerby, 1818, p. 233, pi. 202), by original designation of Buckman (1898,
p. 456).

Emileia sp. indet.

Plate 1, fig. I

Material. A single, fragmentary specimen (J9469), but relatively well preserved. Measurements cannot be
made.

868

PALAEONTOLOGY, VOLUME 34

text-fig. 3. Suture lines of a, Calliphylloceras heterophylloides (Oppel), J9477, Bed 1 ; B, Dorsetensia subtecta
Buckman, J9494, Bed 6; c, Holcophvlloceras zignodianum (d'Orbigny), J9486, Bed 1; d, Stephanoceras
( Stephanoceras ) scalare Weisert, J9472, Bed 10; E, Stephanoceras (Stephanoceras) triplex Weisert, J9497, Bed

10. Scale bars 10 mm.

Description. A small fragmentary inner part, with a recrystallized and manganese-coated partly preserved shell.
The specimen is cut by subsolution, which removed more than half the ammonite.

The umbilicus is very narrow, the whorl section is inflated with oblique, rounded umbilical wall and arched
flank and venter. The dense ribbing consists of blunt, straight, rectiradiate primary ribs reaching about one-
third of the whorl height. The secondaries arise from the slightly elongated tubercles at the termination of the
inner ribs. There are seventeen primary and fifty-nine secondary ribs on the 225° segment of the preserved outer
whorl. The specimen is septate throughout, thus chambered part and the body chamber have been broken off.
Only small parts of the suture line can be seen.

Remarks. The single, fragmentary specimen is insufficient for specific identification ; however, some
visible features suggest closer affinities. The very small umbilicus and the numerous, long primary

EXPLANATION OF PLATE 2

Fig. 1. Stephanoceras ( Stephanoceras ) leoniae Schmidtill and Krumbeck. J9489; Kozoskiit; Bed 9,
Stephanoceras humphriesianum Subzone; lateral view of a nearly complete specimen, x 1.

Fig. 2. Holcophylloceras zignodianum (d'Orbigny). J9468; Kozoskiit; Bed 1, Teloceras blagdeni Subzone;
lateral view, xO-7.

Fig. 3. Stephanoceras ( Stephanoceras ) mutabile (Quenstedt). J9486; Kozoskiit; Bed 5, Stephanoceras
humphriesianum Subzone; lateral view, x 1.

Fig. 4. Dorsetensia subtecta Buckman, J9494; Kozoskiit; Bed 6, Stephanoceras humphriesianum Subzone;
lateral view, x 0-6.

PLATE 2

GALACZ, Stephanoceras, Holcophylloceras, Dorsetensia

870

PALAEONTOLOGY, VOLUME 34

ribs closely resemble the inner whorls of Emileia greppini Maubeuge, 1961. This species is usually
represented by large specimens, but the type series (Greppin 1898, pis 1-3) also contain incomplete
forms, showing the characteristic juvenile features.

Distribution. The type of E. greppini, according to Greppin (1898, p. 31) was collected from the ‘Sauzei-
Schichten’ at Liestal, Switzerland. The Sauzei Zone age of these beds was supported recently by Ohmert et al.
(1982, pp. 153-154). Parsons (1974, p. 159) cited E. greppini as a species characteristic in the Sauzei Zone.

In the Kozoskut sequence the genus Emileia , even without closer identification, undoubtedly indicates the
presence of the Sauzei Zone. This representation confines to a single bed (Bed 1 1), because the underlying Bed
12 yielded Toarcian. ammonites, whereas Bed 10 above contained Humphriesianum Zone ammonites.

Family stephanoceratidae Neumayr, 1875
Subfamily stephanoceratinae Neumayr, 1875
Genus and subgenus stephanoceras Waagen, 1869

Type species. Ammonites Humphriesianus (J. de C. Sowerby, 1825, p. 161, pi. 500, fig. 1), by subsequent
designation of Buckman 1898, p. 454.

Stephanoceras ( Stephanoceras ) triplex Weisert, 1932
Plate 3, fig. I ; Text-fig. 3e

1932 Stephanoceras triplex Weisert, p. 52, pi. 16, fig. 1.
non 1961 Stephanoceras cf. triplex Weisert; Maubeuge, p. 108, fig. on p. 109.

v 1971 Stephanoceras ( Stephanoceras ) aff. triplex Weisert; Morton, p. 279, pi. 46, figs 1 and 2; pi. 47,
figs 1 and 2.

1985 Teloceras (?) triplex Weisert; Schlegelmilch, p. 78, pi. 27, fig. 7.

Material. Two relatively well preserved, but fragmentary specimens.

Measurements.

MDm

Dm

Wh

Wb

Ud

Pr

S

J9497

225

225

60 (26-5)

94 (42)

114(51)

—

—

152

39 (25-5)

?60 (? 39-5)

76 (50)

39

86

J9499

225

191

51 (27)

88 (46)

94 (52)

40

95

152

40 (26-5)

64 (42)

76 (50)

36

92

Description. Large, robust form with moderately wide umbilicus. The figured specimen is a well preserved
internal mould with fragmentary body chamber and manganese coating which has partly replaced the original
shell. The width of the umbilicus remains constant ( c . 50%) with growth. The whorl section is subcircular and
hardly varies during growth. The umbilical slope and the lateral side are convex, and the venter is evenly
rounded. The sculpture is a dense, regular ribbing. The primary ribs arise from the umbilical seam. The radial,
slightly backwardly arched primaries are rounded on the internal mould, but probably sharp on the test. The
primaries end in pointed, radially somewhat elongated, relatively small tubercles. These are situated at half of
the height in the inner whorls, while the position of the tubercle row is gradually displaced lower, to about one-
third the height of the body chamber. The number of primary ribs is fairly constant, i.e. 35^40 per whorl, but

explanation of plate 3

Fig. 1. Stephanoceras ( Stephanoceras ) triplex Weisert. J9497; Kozoskut; Bed 10, Stephanoceras
humphriesianum Subzone; lateral view, x 1.

Fig. 2. Nannolytoceras polyhelictum (Bockh.) J9475; Kozoskut; Bed 2, Teloceras blagdeni Subzone; lateral
view, x 1 .

PLATE 3

GALACZ, Stephanoceras, Nannolytoceras

872

PALAEONTOLOGY, VOLUME 34

somewhat increases on the outer whorls. Generally two secondary ribs arise from the tubercles, but eight to
ten triplications may appear on each whorl, and additional intercalatories, starting at the row of the tubercles,
also occur. The secondary ribs are rounded and projected slightly throughout the whorls and pass straight
across the venter. The length of the body chamber cannot be measured, because apertural parts are missing
in the studied material. The phragmocone ends at c. 150 mm diameter, and the longest preserved body chamber
fragment comprises nearly one whorl.

The suture line is characteristic (Text-fig. 3e), showing narrow saddles and rather wide lobes. The lateral lobe
is remarkably long and deeply incised. The umbilical elements are retracted.

Remarks. S. triplex belongs to the broadly-whorled Stephanoceras brodiaei group. The relatively
wide umbilicus, the generally biplicate ribbing, the rather small tubercles, and the extremely long
lateral lobe in the suture are all characteristic.

The species name "triplex' does not appear in the published work of Mascke (1907), in spite of
the reference made by Weisert (1932, p. 52). Thus Weisert should be regarded as the author of this
species, a procedure followed previously by others (e.g. Morton 1971, p. 279). The lectotype is the
original of plate 16, figure 1 in Weisert (1932). This is a smaller specimen (see Schlegelmilch 1985,
pi. 27, fig. 7), but nevertheless closely comparable with the inner whorls of the Kozoskut forms. The
number of the primaries is somewhat lower in the lectotype; on the other hand that form shows
some irregularities in its sculpture. The form described by Maubeuge (1961, pp. 108-109 as S. cf.
triplex) is not conspecific because of the wider whorls, somewhat narrower umbilicus and denser
secondary ribbing. The forms figures by Morton (1971, pi. 46, figs 1 and 2; pi. 47, figs 1 and 2),
though fragmentary and crushed, are near to the typical S. triplex.

Distribution. Stephanoceras (S.) triplex is a common and frequently recorded element in the Humphriesianum
Zone faunas. Its subzonal distribution is to be ascertained. The Kozoskut specimens came from Bed 10, i.e.
from the Humphriesianum Subzone.

Stephanoceras ( Stephanoceras ) scalare Weisert, 1932
Plate 4, fig. 4; Text-fig. 3d

v 1849 Ammonites Humphriesianus ; Quenstedt, p. 180, pi. 14, fig. 10.

1858 Ammonites Humphriesianus plicatissimus Quenstedt, p. 298, pi. 54, fig. 3.
p. 1887 Ammonites Humphriesianus ; Quenstedt, p. 531, pi. 65, fig. 15 (only).

1932 Stephanoceras scalare Mascke; Weisert, p. 143, pi. 16, fig. 2; text-figs 7 and 8.

1937 Stephanoceras scalare Mascke; Horwitz, p. 261, pi. 11, fig. 3.

71938 Stephanoceras scalare Mascke em. Weisert; Schmidtill and Krumbeck, p. 330, pi. 13, fig. 2; text-
fig. 5.

p. 1938 Stephanoceras auerbachense Schmidtill and Krumbeck, p. 339, pi. 13, fig. 5 (only).

1946 Cadomites Bigoti Munier-Chalmas; Gardet and Gerard, p. 34, pi. 8, fig. 1.

1961 Stephanoceras (Stephanoceras) scalare Mascke em. Weisert; Krymholz, p. 112, text-fig. 14.
non 1969 Stephanoceras cf. scalare Mascke; Maubeuge, p. 71, fig. on p. 71.

1973 Stephanoceras cf. scalare Mascke; Myczynski, p. 91, pi. 11, fig. 1; pi. 12, fig. 1.

1982 Stephanoceras (Stephanoceras) scalare Mascke em. Weisert; Azarian, p. 83, pi. 14, fig. 2; pi. 16,
fig. 4; pi. 1 7, fig. 5.

1983 Stephanoceras (Stephanoceras) scalare Mascke emend. Weisert; Sandoval, p. 235, pi. 16, fig. 1.

p. 1983 Stephanoceras ( Stephanoceras ) scalare Weisert (Mascke); Pavia, p. 89, pi. 12, fig. 5 (only).

71985 Stephanoceras (Stephanoceras) scalare Weisert; Rostovtsev, p. 141, pi. 39, fig. 1.

non 1985 Stephanoceras scalare Weisert; Fernandez Lopez, p. 283, pi. 29, fig. 1.

1985 Stephanoceras ( Stephanoceras ) scalare Loewe; Schlegelmilch, p. 70, pi. 22, fig. 4.

Material. Four specimens, two nearly entire internal moulds, one phragmocone and one body chamber
fragment.

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

873

Measurements.

MDm

Dm

Wh

Wb

Ud

Pr

S

J9484

270

225

45

(20)

40 (18)

141

(65)

—

—

141

34

(23)

? 30 (? 20-5)

81

(55)

47

—

J9498

250

235

45

(19)

?40 (717)

153

(65)

—

—

200

42

(21)

740 (720)

125

(62-5)

64

121

136

33

(24)

730 (722)

71

(52)

47

—

J9483

131

131

32-5

(25)

32 (24-5)

72

(55)

48

—

112

31

(27-5)

32 (28-5)

56-5

5 (50-5)

42

—

89

30

(33-5)

30 (33-5)

43 5

> (49)

37

—

70

21

(30)

21 (30)

29‘

5(42)

—

—

52

18

(34-5)

19 (36-5)

20

(38-5)

—

—

Description. Large, planulate, serpenticone form. The figured specimen is an almost complete, well preserved
internal mould with partial manganese coating, which replaces the original shell. The umbilicus is very wide
and gradually opens with individual growth : its width is c. 45-50 % in the inner, and 60-65 % in the outermost
whorls. The shape of the whorl section also varies with growth, being somewhat depressed-circular in the inner
whorls and compressed-oval on the body chamber. The umbilical slope is short and convex, the lateral sides
are also convex, becoming slightly flattened on the outer whorls. The venter is convex and wider in the inner
whorls. The ornament consists of dense ribbing. The inner ribs arise at the umbilical seam, run radially in the
inner, and become slightly backwardly arched in the outer whorls. The number of primary ribs increases with
growth: 30-35 in the inner whorls (at 50-70 mm diameter), and 60-65 in the outer whorls (at c. 190 mm
diameter). The primary ribs end in slightly elongated, small, pointed tubercles around the lower third of the
flanks. The long secondary ribs arise from the tubercles, are radial in the inner whorls and slightly prorsiradiate
on the body chamber. Here the number of the secondary ribs is smaller, two per primary. In the outermost
whorl simple ribs and intercalatories appear, arising freely at the level of the tubercles. The rounded
secondaries run straight across the venter. The body chamber is very long, beginning at 130-140 mm diameter
and consisting of more than one and a half whorls. The apertural part is missing in all specimens.

Entire suture line cannot be seen in either specimen. The clearly visible sutural portions show moderately
deep, relatively broad lateral, and moderately retracted umbilical lobes (Text-fig. 3d.).

Remarks. The synonymy of early citations of this form was discussed in detail by Pavia (1983, pp.
89-91). However, the origin of the species’ name is mysterious. In Weisert (1932, p. 143) the whole
name is ’ Stephanoceras scalare Mascke sp. 1903, em. Weisert 1931’. Even the 1903 date is strange,
because although this is the year when Mascke defended his thesis, the work itself was published
in 1907, so this is the valid citation date. On the other hand, the species’ name ‘ scalare ’ does not
appear in the publication. This is probably why subsequent authors used the S. scalare Mascke or
the S. scalare Weisert versions.

The photograph of the holotype was first published by Weisert (1932, pi. 16, fig. 2). Nevertheless,
Weisert (1932, p. 143) had at least one additional specimen, from which measurements were
recorded. These show that this specimen is probably not conspecific. More recently Pavia (1983, pi.
12, fig. 5) published a photograph of the plaster cast of the holotype and also recorded its
dimensions. However, the three other forms figured by Pavia (1983, pi. 12, figs 1, 3, 4) differ
significantly from the type in size and sculpture and thus are regarded here as non-conspecific.

Fragments of S. scalare were figured by Horwitz (1937) and by Myczynski (1973) from the
Pieniny Klippen Belt, Polish Carpathians, and by Gardet and Gerard (1946, pi. 8, fig. 1, as
‘ Cadomites Bigoti') from the Middle Atlas Range. One of the best descriptions of the species is by
Krymholz (1961, p. 112), who recorded it from the Northern Caucasus. Good specimens were
described by Azarian (1982) and a fragment of a very similar form was figured by Rostovtsev (1985).
Sandoval (1983, p. 235, pi. 16, fig. 1) recorded the species from the Betic Cordilleras. The identity
of the two specimens figured by Schmidtill and Krumbeck (1938, pi. 13, fig. 2; text-fig. 5; pi. 13, fig.
5 - the latter as Stephanoceras anerbachense) is doubtful because of their poor preservation. The
ammonite figured by Maubeuge (1969, pi. 71, No. 66959, as S. cf. scalare ) is a badly preserved
fragmentary specimen with coarse ribbing and short, bulky primaries.

874

PALAEONTOLOGY, VOLUME 34

Westermann and Riccardi (1979, p. 155) regarded, with question marks, S. scalare (and other
Stephcmoceras species) as synonyms of S. pyritosum (Quenstedt). They published a photograph of
the type of this latter species: from this it is clear that the number of the primary ribs remains low
throughout the whorls, whereas it is higher in all stages in typical Stephcmoceras scalare.

The features mentioned in the description (i.e. wide umbilicus, dense ribbing, whorl section
becoming compressed on the outermost whorl, etc.) clearly distinguish this form from other species
of the Stephanoceras humphriesianum group. S. humphriesianum itself is a rare ammonite, which can
be characterized by shorter inner ribs with clearly distinguished, rounded tubercles and regularly
three, or rarely four, secondaries. Another, similar form of this group is S. rhytus (Buckman 1921,
pi. 250). However, this species is characterized by strongly rectiradiate primaries and strongly
projected boundles of secondaries, and these result in a distinguishing, marked break in the
sculpture.

Distribution. S. scalare is apparently a widely distributed Stephcmoceras. The type probably came from the
Humphriesianum Zone of Bophngen, South Germany (see Pavia 1983), the other described forms are similarly
recorded from the Humphriesianum Zone. Parsons (1976, pp. 131, 134) mentioned S', cf. scalare from the
English Humphriesianum and Romani Subzones. Sandoval (1983, p. 236) found the species in the upper part
of the humphriesianum Zone. Thus the species occurs in wide geographic distribution from North Africa to the
Caucasus, both in Mediterranean and northwest European faunas, in the Humphriesianum Zone.

The Kozoskiit species came from Beds 10 and 4 of the Humphriesianum Subzone.

Stephanoceras ( Stephanoceras ) sp.

Plate 1, fig. 3.

Material. A single, well-preserved, entire, but somewhat corroded specimen.

Measurements.

MDm Dm Wh Wb Ud

Pr

J9488 169 159

130
107

35 (22) 38 (24)

30 (23) 34 (26)

27 (25) 37 (34-5)

99 (62)

73-5 (56-5)

57-5 (54) 41

Description. A relatively large serpenticone internal mould. The umbilicus is shallow and extremely wide,
especially on the outer whorls. The whorl section is subcircular, high-oval on the body chamber. The umbilical
wall is convex, the flanks are slightly rounded, and the venter is highly arched. Maximum width can be
measured around the lower third of the whorl height. The ornament is formed by dense ribbing. The primary
ribs are rounded and short, at about one-quarter their height they branch into three or rarely two secondaries.
The secondary ribs are straight, rounded and slightly rectiradiate on the body chamber. At 107 mm diameter
the number of primary ribs is forty-one, and here the primary : secondary rib ratio is 18:48 for a half whorl.
On the visible inner whorls the primary ribs are longer, prorsiradiate, and terminate in tubercles. The body
chamber occupies the entire last whorl, with a slightly expanding, smooth collar.

The entire suture line cannot be seen, but the visible parts show typical Stephanoceras- like divided saddles,
deep lateral lobe and strongly retracted umbilical lobe.

EXPLANATION OF PLATE 4

Fig. 1 . Stephanoceras ( Normannites ) sp. aft. fort is Pavia. J9480; Kozoskut; Bed 3, Teloceras blagdeni Subzone;
lateral view of a complete adult microconch, x 1 .

Figs 2 and 3. Stephanoceras ( Normannites ) orbignyi Buckman. J9463; Kozoskiit; Bed 3, Teloceras blagdeni
Subzone; 2, lateral and 3, ventral views of an adult microconch, x I.

Fig. 4. Stephanoceras ( Stephanoceras ) scalare Weisert. J9498; Kozoskiit; Bed 10, Stephanoceras
humphriesianum Subzone ; lateral view of an almost complete specimen, x 0-7.

PLATE 4

u\iwTuti

&9E

f,A; • -

GALACZ, Stephanoceras

876

PALAEONTOLOGY, VOLUME 34

Remarks. Because of incomplete preservation, this single specimen is insufficient for specific
determination. A similar form is that described by Pavia (1983, p. 97, pi. 14, fig. 6) from the
Humphriesianum Zone of Bayeaux. He assigned this specimen to Stephanoceras tenuicostatum;
however, this latter species is smaller, has distant and longer primary ribs with tubercles throughout
the internal mould of the body chamber. The individual features of the Normandy species are close
to those of the Kozoskiit form; nevertheless, designation of a new species would need further
material.

Distribution. The Kozoskiit specimen came from Bed 9, i.e. from the Humphriesianum Subzone. The
Normandy specimen was collected from the condensed Humphriesianum Zone of Bayeux.

Stephanoceras ( Stephanoceras ) leoniae Schmidtill and Krumbeck, 1938

Plate 2, fig. 1

1938 Stephanoceras leoniae Schmidtill and Krumbeck, p. 343, pi. 11, fig. 4.

non 1949 Stephanoceras leoniae Schmidtill and Krumbeck; Maubeuge, p. 173, pi. 13.

Material. Two incomplete, relatively well preserved specimens.

Measurements.

MDm

Dm

Wh

Wb

Ud

Pr

J9473

117

117

32-5

(27-5)

32 (27-5)

58 (49-5)

—

88-5

28-5

(32)

31 (35)

38 (43)

—

J9489

113

113

34

(30)

35 (31)

53 (47)

—

92

31

(33-5)

36 (39)

42 (45-5)

31

Description. A relatively small stephanoceratid. The figured specimen is a corroded internal cast with partial
manganese coating. The umbilicus is moderately wide, the whorl section is circular with convex umbilical side
and flanks, and with evenly arched venter. The sculpture is only partially visible. The straight, slightly
prorsiradiate primary ribs arise at the umbilical seam and their number somewhat increases on the outer whorl.
The primaries are slightly strengthened at the lower third of the flanks, without forming true tubercles. At this
level the slightly curved, generally radial, rounded secondary ribs appear. Their strength is variable; those
arising from the thickened primaries are stronger, whereas those starting freely are somewhat weaker. These
differences disappear towards the venter. The better preserved half of the penultimate whorl bears fifteen
primary and about forty-eight secondary ribs. The slightly contracted body chamber occupies two-fifths of the
preserved last whorl; the aperture is missing. The suture line cannot be seen.

Remarks. The type (Schmidtill and Krumbeck 1938, pi. 11, fig. 4) is an incomplete, poorly preserved
specimen of 90 mm diameter. The original description mentioned the dense ribbing and strong
secondaries as characteristic. The type specimen is wholly septate, but the excentric umbilical seam
indicates that the missing part was probably the body chamber. Thus this is a relatively small form,
comparable to the Kozoskiit specimens. The ribbing is also similar, showing the smoothed tubercles
mentioned in the original description.

The ammonite assigned to this species by Maubeuge (1949, pi. 13) differs in being more evolute,
compressed (its Wb: Wh ratio is 0-85, as against 11 of the type) and having lowly situated, slightly
elongated tubercles instead of real primary ribs.

Sandoval (1983, pp. 241-243) and also Schlegelmilch (1985, p. 72) regarded S. leoniae as a
synonym of Stephanoceras zogenreuthense , another species of Schmidtill and Krumbeck. However,
the type of this latter is different, having denser ribbing, stronger primaries and distant tubercles.

Distribution. The horizon of the type is uncertain within the ‘ Humphriesianum-Schichten ' of South Germany.
The Kozoskiit specimens came from Bed 9, i.e. from the Humphriesianum Subzone.

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

877

Stephanoceras ( Stephanoceras ) mutabile (Quenstedt, 1886)

1886
non 1909
1932
v 1971
71983

p. 1983
1985
1985

Plate 1, fig. 2; Plate 2, fig. 3

Ammonites Humphriesianus mutabilis Quenstedt, p. 537, pi. 66, fig. 5.

Stephanoceras mutabile Quenstedt; Lissajous, p. 181, pi. 6, fig. 14.

Stephanoceras mutabile Quenstedt; Weisert, p. 153, pi. 17, fig. 6; text-figs 15 and 16.
Stephanoceras (Stephanoceras) mutabile Quenstedt ; Morton, p. 273, pi. 40, figs 5-10; text-fig. 2a.
Stephanoceras ( Stephanoceras ) mutabile (Quenstedt); Sandoval, p. 238, pi. 17, fig. 3; text-fig.
99b.

Stephanoceras ( Stephanoceras ) gr. umbilicum (Quenstedt); Pavia, p. 100, pi. 12, fig. 2 (only).
Stephanoceras umbilicum (Quenstedt); Fernandez Lopez, p. 278, pi. 27, fig. 3; text-fig. 29h.
Stephanoceras ( Stephanoceras ) mutabile (Quenstedt); Schlegelmilch, p. 72, pi. 25, fig. 1.

Material. Two incomplete and fragmentary internal moulds.

Measurements.

MDnr

Dm

Wh

Wb

Ud

Pr/2

S/2

J9496

115

115

40 (35)

50 (43-5)

49

(43)

18

47

48

30 (35)

37 (43-5)

35

(41)

—

—

J9486

94

94

32 (34)

744 (747)

33-5

(35-5)

18

51

Description. Medium-sized forms. The two figured specimens are fragmentary, the bigger is a corroded internal
mould, the smaller is a similarly badly preserved specimen, with partial manganese coating. The umbilicus is
relatively narrow and deep, its width is constant during growth. The whorl section is subcircular, with steep,
slightly convex umbilical walls and widely arched venter. The sculpture consists of dense ribbing. The primary
ribs arise from the umbilical seam, are rounded on the internal mould, but sharp on the shell. The number of
the slightly rectiradiate, backwardly arched primaries is eighteen and sixteen on the preserved last and
penultimate half-whorl of the bigger specimen. The primary ribs end in sharp, circular tubercles at the lower
third of the flank. The tubercles are weaker on the internal mould. Two to three secondary ribs start from the
tubercles, i.e. there are forty-seven secondaries on the last half whorl. The secondary ribs are radial, only
slightly curving. The smaller specimen is wholly septate, the larger one shows a very short part of the body
chamber, which begins at c. 106 mm diameter. The suture line cannot be seen clearly.

Remarks. Quenstedt (1886-87, pi. 66, fig. 5) figured only the ventral view of the holotype; later
Weisert (1932, pi 17, fig. 6) published a photograph of the lateral side. More recently Schlegelmilch
published both views of the holotype ( 1985, pi. 25, fig. 1 ). These show clearly that its characteristics
are the relatively narrow umbilicus and the dense ribbing with low tubercle row (see also Morton
1971, p. 274). Pavia (1983, pp. 100-101) regarded S. kreter (Buckman), S. mutabile (Quenstedt) and
S. umbilicum (Quenstedt) as a single group, suggesting a transient position for S. mutabile , on the
basis of the umbilical width and the length of the primary ribs.

An additional feature of the species is its relatively small size. As Westermann and Riccardi
pointed out (1979, p. 158), a part of the last whorl of the holotype is body chamber. The Kozoskut
specimens, though fragmentary, correspond well with the figures of the holotype. These are also
small forms: one of the specimens shows the end of septation at 106 mm diameter.

S. mutabile is a commonly recorded but rarely illustrated species. The ammonite figured by
Lissajous (1907-1912, pi. 6, fig. 14) from the Parkinsoni Zone is clearly different, probably being
the inner whorls of a Cadomites sp. Good figures were published by Schmidtill and Krumbeck
(1938). Sandoval (1983, pi. 17, fig. 3) figured a specimen of strongly corroded outer whorls, thus
difficult to identify. However, the recorded dimensions and the cross-section show a closely allied
form.

Distribution. The holotype came from the middle part of the Humphriesianum Zone (see Morton 1971, p. 275,
and Pavia 1983, p. 102). Other records indicate the same stratigraphic level. Parsons (1976, p. 139) cited the
species as a characteristic form in the Humphriesianum Subzone. The Kozoskut specimens came from Beds 9
and 5 of the Humphriesianum Subzone.

878

PALAEONTOLOGY, VOLUME 34

Stephanoceras ( Steplianoceras ) sturanii Pavia, 1983
Plate 1, fig. 5

1983 Stephanoceras ( Stephanoceras ) sturanii Pavia, p. 94, pi. 13, figs 4 and 6.

Material. Two incomplete, fragmentary specimens (J9461-J9462) of poor preservation. Precise measurements
cannot be made.

Description. The figured specimen (J9461 ) is a fragmentary internal mould with subsolved inner whorls and an
incomplete, slightly distorted body chamber. The umbilicus is narrow and deep in the inner whorls, becoming
excentric and wide on the last whorl. The whorl section is subcircular with low, oblique umbilical wall rounding
into the convex whorl-side without shoulder. The venter is widely arched. The ribbing is extremely dense,
consisting of short, lamellar primaries, which end in radially elongated tubercles below the lower third of the
flanks. From the tubercles arise the secondaries, which are sharp on the flanks and rounded on the venter. Both
the primary and the secondary ribs are radial. On the preserved quarter of the last whorl there are seventeen
primary and forty-one secondary ribs.

The suture line cannot be seen clearly, only the strongly retracted umbilical elements are visible.

Remarks. The specimens closely resemble the recently designated species of Pavia. With its narrowly
coiled inner and middle whorls, excentric body chamber and very dense sculpture this is a species
from an incompletely known group of Mediterranean stephanoceratids. Another member of this
group is described above as Stephanoceras (S'.) sp.

Description. The holotype of this species came from the Blagdeni Subzone of Digne, SE France, and Pavia
(1983, pi. 13, fig. 4) described an additional specimen from the Hwnphriesianum Zone of Normandy. The
Kozoskut specimens came from Bed 3, i.e. the Blagdeni Subzone.

Subgenus normannites Munier-Chalmas, 1892

Type species. Normannites orbignyi Buckman, 1908 = Ammonites Braikenridgii d’Orbigny (non Sowerby),
1845, p. 400, pi. 135, figs 3 and' 4.'

Stephanoceras ( Normannites ) orbignyi Buckman, 1908
Plate 4, figs 2 and 3

v 1845 Ammonites Braikenridgii Sowerby; d’Orbigny, p. 400, pi. 135, figs 3 and 4 (only).

1908 Normannites Orbignyi Buckman, p. 146.

1927 Normannites orbignyi Buckman; Buckman, pi. 734, figs 1-3.

71937 Cadomites Orbignyi Buckman; Gillet, p. 86, text-fig. 62.

71937 Normannites Orbignyi Buckman; Horwitz, p. 265, pi. 12, fig. 2.
non 1939 Normannites Orbignyi Buckman; Roche, p. 219, pi. 1, fig. 5 a-b.

1954 Normannites (Normannites) orbignyi Buckman, and subspp. ; Westermann, pp. 135-152, pi. 5,
figs 3 and 4; pi. 6, figs 1, 3-5; pi. 7, figs 1-5; pi. 8, fig. 1; text-figs 35-44.
non 1971 Stephanoceras ( Normannites )? orbignyi Buckman; Morton, p. 282, pi. 51, figs 1 and 2.

non 1973 Normannites (Normannites) orbigny Buckman; Imlay, p. 82, pi. 41, figs 9, 10, 18, 20.

v 1978 Normannites orbignyi Buckman; Dietl et al., p. 10, text-fig. 3 a.

1983 Normannites (s.s.) orbignyi Buckman; Pavia, p. 142, pi. 27, figs 3 and 5.

1985 Normannites orbignyi Buckman; Fernandez Lopez, p. 321, pi. 34, fig. 1; text-fig. 36e.

1985 Stephanoceras (Normannites) orbignyi Buckman; Schlegelmilch, p. 73, pi. 25, fig. 3.

Material. A single, fragmentary internal mould of a damaged individual (J9463). Measurements would be
difficult to make and misleading.

Description. The size of the specimen is c. 65 mm, and only the last half whorl is well preserved. The specimen
is a small microconch with moderately wide umbilicus and somewhat depressed-oval whorl section. The well
preserved part, which is the last half segment of the body chamber is slightly contracted with decreasing width

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

879

and height. The umbilical wall is steep, arched, the flanks are convex and the venter is widely arched. The sharp,
prorsiradiate, slightly curved primary ribs arise at the umbilical seam. There are fifteen primaries on the last
half whorl. The primary ribs end in radially elongated, pointed tubercles near the middle of the whorl height,
giving rise to usually two secondary ribs. Some intercalatories also appear so the primary : secondary rib ratio
is 13:34 on the last half whorl. An important feature is that the level of the tubercles gradually sinks on the
body chamber. The secondaries are straight and cross the venter perpendicularly. Despite the bad preservation,
some parts of the ornament are visible on the inner whorls. At c. 20 mm diameter the estimated number of
primary ribs is twenty. The aperture cannot be seen properly, because the lappets are not preserved on the
internal mould. However, the lateral inflation probably represents the base of the lappet. The suture line cannot
be studied.

Remarks. Buckman (1908, p. 146) distinguished this species on the basis of the "Ammonites
BraikenridgiV figure of d’Orbigny. He published the photograph of the type in 1927 (in 1909-30,
pi. 734). Subsequently Westermann (1954, p. 136) named this specimen [unnecessarily] as the
‘neotype’.

Westermann (1954), on the basis of whorl section and rib density, split N. orbignyi into four
subspecies. However, these features are so variable in stephanoceratids in general, and Normannites
in particular, that N. orbignyi is better regarded as a species of wide morphological variability (see
also Westermann 1964). In this respect the Kozoskut specimen is a form with a slightly wider
umbilicus, but otherwise matching the type rather closely.

Of the rather common citations of N. orbignyi , the specimen of Horwitz (1937, pi. 12, fig. 2) is
a very poorly preserved ammonite, insufficient for specific identification. The description and
drawing of Gillet (1937, p. 86) is similarly insufficient to decide on specific identity. The specimen
figured by Roche (1939, pi. 1, fig. 5) is a big form with an extremely wide umbilicus, and thus
certainly not conspecific. The poorly preserved specimen of Morton ( 1971 , pi. 51 , figs 1 and 2), with
large, elevated tubercles and dense secondary ribbing is also different.

Distribution. The type came from the Epalxites hemera of the Humphriesianum Zone of Dorset, which
corresponds to the Humphriesianum Subzone (see Parsons 1976, p. 1 16). Other references mention the species
from the Blagdeni Subzone and from the basal Niortense Zone. The Kozoskiit specimen came from Bed 3, i.e.
from the Blagdeni Subzone.

Stephanoceras (Normannites) sp. aff. fort is Pavia, 1983
Plate 4, fig. 1

aff. 1983 Normannites fortis Pavia, p. 146, pi. 28, figs 2, 3, 6.

Material. A single, well preserved internal mould.

Measurements.

MDm

Dm

Wh

Wb

Ud

Pr

S

J9480 77

77

27 (35)

31 (40)

35(45)

35

68

63-5

20(31-5)

26 (41)

28 (44)

28

—

Description. A relatively large form. The umbilicus is moderately wide, opening slightly on the last half whorl.
The whorl section is depressed, rounded, becoming more depressed near the end of the body chamber by
contraction. The umbilical wall and the flanks are evenly convex, the ventral side widely arched. The sculpture
consists of regular ribbing. The straight, prorsiradiate primary ribs are rounded on the internal cast, begin on
the umbilical wall and reach the half of the flanks on the phragmocone, and the third on the body chamber,
respectively. At the endings they are strengthened without tubercles, and branch into secondaries. There are
thirty-five primaries on the outermost whorl, twenty-eight on the penultimate whorl and the same number on
the preceding ones. The secondary : primary rib ratio is 68 : 35 on the last whorl, i.e. some singular ribs run onto
the venter. The secondary ribs are straight and prorsiradiate, crossing the venter perpendicularly. The length
of the body chamber is four-fifth of a whorl, the peristome is partially preserved, showing a slightly flared
ventral ridge and short lateral lappets. Details of the suture line cannot be seen.

880

PALAEONTOLOGY, VOLUME 34

Remarks. Despite good preservation, identification with known Normannites species was apparently
not possible. Special features are the low number of primary ribs in the middle whorls, the relatively
numerous singular ribs on the outer whorl, and the complete lack of tubercles. This latter is
characteristic to the ‘ Parallites ’ of Westermann (1954). However, Parallites species show denser
ribbing and narrower umbilicus. Of the s.s. Normannites , N. vulgaricostatus pfaffi Westermann
(1954, p. 176, pi. 10, fig. 5) is similar, but this is smaller and has a narrower umbilicus and arched
primary ribs. Closest similarity appears to be with N.fortis, a species of Pavia (1983, p. 146) from
Digne. The most closely allied form is the paratype on plate 28, fig. 3 of Pavia. This differs from
the holotype (loc. cit., pi. 28, fig. 6) in having denser ribbing and undeveloped tubercles: features
just present in the form described here. A very similar specimen was figured by Fernandez Lopez
(. Normannites cf .fortis, 1985, pi. 33, fig. 3) from the Banski Subzone of the Niortense Zone from the
Iberian Cordilleras.

Distribution. All specimens of Normannites fortis Pavia came from the uppermost Humphriesianum Zone; the
Kozoskut specimen was yielded by Bed 3, the Blagdeni Subzone.

Genus stemmatoceras Mascke, 1907

Type species. Ammonites Humphriesianum coronatus Quenstedt, 1886, p. 539, pi. 66, fig. 1 = Stephanoceras
Frechi Renz, 1904, p. 77, by original designation of Mascke, 1907, p. 30.

Stemmatoceras frechi (Renz, 1904)

Plate 5, figs 1-3

1886 Ammonites Humphriesianus coronatus Quenstedt, p. 539, pi. 66, fig. 11.

1904 Stephanoceras Frechi Renz, p. 77.

p. 1932 Stemmatoceras coronation Quenstedt; Weisert, p. 159, pi. 18, fig. 4 (only).

1938 Stemmatoceras coronation Quenstedt; Schmidtill and Krumbeck, p. 345, pi. 12, fig. 2.

1939 Cadomites Quenstedti nomen novum Roche, p. 205.

non 1951 Stephanoceras quenstedti Roche; Maubeuge, p. 64, pi. 14, fig. 2.

71951 Stephanoceras aff. quenstedti Roche; Maubeuge, p. 64, pi. 5, fig. 4.

? 1967 Stemmatoceras (Stemmatoceras) aff. frechi (Renz); Seyed-Emami, p. 136, pi. 4, fig. 22; pi. 15, fig.
3 a — c.

non 1969 Stemmatoceras ad. frechi Renz; Pavia, p. 447, fig. 3/4.

non 1983 Stemmatoceras ( Stemmatoceras ) sp. cf. S. (Stm.) frechi Renz; Sandoval, p. 252, pi. 12, fig. 1;

text-fig. 99i.

non 1985 Stemmatoceras coronation (Quenstedt); Rostovtsev, p. 143, pi. 40, fig. 1 ; pi. 41, fig. 1.

1985 Teloceras frechi (Renz); Schlegelmilch, p. 77, pi. 27, fig. 6.

Material. Several well preserved specimens of which four were sufficient for detailed study.

Measurements

MDm Dm

Wh

Wb

Ud

Pr

S

J9482 145 131

29 (23)

758 (745)

68

(52)

35

87

112

29 (26)

56 (50)

55

(49)

29

84

83

24 (29)

46(55-5)

38

(46)

26

—

EXPLANATION OF PLATE 5

Figs 1-3. Stemmatoceras frechi (Renz). I and 2, J9459; Kozoskut; Bed 9, Stephanoceras humphriesianum
Subzone; dorsal and lateral views of a smaller specimen. 3, J9482; Kozoskut; Bed 4, Stephanoceras
humphriesianum Subzone; lateral view of an adult form. All x 1.

PLATE 5

GALACZ, Stemmatoceras

882

PALAEONTOLOGY, VOLUME 34

MDm

Dm

Wh

Wb

Ud

Pr

S

J9479

133

133

32

(24)

68 (51)

71 (53-5)

—

—

112

28

(25)

63 (56)

54-5 (48-5)

—

-

J9466

117

117

32

(24)

68 (51)

59 (50)

33

74

100

32

(32)

58 (58)

52-5 (52-5)

31

—

J9495

90

90

25

(28)

54 (60)

42 (47)

29

72

75

21-5 (29)

42 (56)

34-5 (41)

28

—

Description. Medium to large forms of evolute coiling, with deep, relatively wide umbilicus, which becomes
even wider with growth. The whorl section is depressed oval, the height: width ratio is c. 0-5 throughout the
whorls. The umbilical slope is steep, the flanks are low, and the venter is broadly arched. The sculpture consists
of relatively dense primary and secondary ribs. The primaries are long, radial, but sightly projected in the inner
whorls. In the middle whorls the primary ribs are somewhat sharper on the internal mould. The number of
primaries is nearly constant, being twenty-eight at 75 mm and thirty-five at 131 mm diameters, respectively.
The primary ribs end in sharp tubercles. The row of tubercles is situated above the middle of the flanks in the
internal and middle whorls, then lowers to middle position on the body chamber. Radial or slightly projected
secondary ribs arise from the tubercles. The secondary : primary rib ratio is 2-8 to 2-4 on the phragmocone and
2-2 on the body chamber. The specimens studied are incomplete, with the outer whorls partially preserved, so
the entire length of the body chamber and the aperture cannot be studied. One large specimen shows septation
at 133 mm largest preserved diameter, other specimens show the beginning of the body chamber at 95 and
1 1 5 mm.

Suture line cannot be clearly seen on either specimen, only some characteristic stephanoceratid portions are
visible.

Remarks. The date of designation of Stemmatoceras frechi by Renz is usually regarded as 1913 (see
e.g. Westermann and Riccardi 1979, p. 166); however, the first explicit reference to this form by this
name was made by Renz in 1904, p. 77. The status of Quenstedt’s form as the type of the genus
Stemmatoceras has been commonly neglected and thus most references are misinterpretations (see
synonymy). Weisert (1932), while refiguring the original specimen of Quenstedt, also illustrated
another form (pi. 18, fig. 1, text-fig 23) from Beuren, now in the collections of the Stuttgart Museum
(No. 22206), and this is a normal Stephanoceras s.s. Roche (1939, p. 205) recognized the difficulty
arising from the use of name 1 coronation' and so he introduced a new name, ‘ Cadomites
Quenstedti' , for the form of Quenstedt, which is thus an objective synonym. This name was applied
by Maubeuge (1951) for a serpenticone Stephanoceras s.s. (pi. 14, fig. 2), and for a badly preserved,
specifically undeterminable ammonite (pi. 5, fig. 4). The form figured by Rostovtsev from the
Transcaucasus (1985, pi. 40, fig. 1; pi. 41, fig. 1) is a robust stephanoceratid, close to Teloceras
acuticostatum.

Distribution. The type came from the Humphriesianwn Zone of Eningen, Swabia, and Kumm (1952) suggested
the species (by the name Stephanoceras coronation (Quenstedt) as a guide in the lowermost part of the
Humphriesianwn Zone (see Westermann 1967, p. 142). However, Sturani (1971, p. 50) stated, that
Stemmatoceras frechi has a long vertical range throughout the Humphriesianwn Zone, thus this species never
became widely used as subzonal index.

The Kozoskiit specimens came from Beds 9, 4, 3 and 2, i.e. from both the Humphriesianwn and Blagdeni
Subzones, thus confirming the extended range of the species.

REFERENCES

arkell, w. t. 1950. A classification of the Jurassic ammonites. Journal of Paleontology , 24. 354-364.
azarian, n. r. 1982. Jurassic ammonites of the Armenian S.S.R. Institut Geologicheskikh Nauk, Akademiya

Nauk Armyanskoi S.S.R., Yerevan, 191 pp. [In Russian],
bayle, M. 1878. Fossils principaux des terrains. Explication de la carte geologique de la France , 4, Atlas,

1-158 pis.

bockh, j. 1881 . Data to the knowledge on the Jurassic deposits of the Mecsek Mountains and adjacent areas.

Part 2. Palaeontology. Ertekezesek a Termeszettudomdnyok Korebol , 11 (9), 1-107. [In Hungarian].

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

883

buckman, s. s. 1887-1907. A monograph on the Inferior Oolite ammonites of the British Islands. Monograph
of the Palaeontographical Society , 1-456.

— 1898. On the grouping of some divisions of so-called ‘Jurassic’ time. Quarterly Journal of the Geological
Society of London , 54. 442 462.

1905. On certain genera and species of Lytoceratidae. Quarterly Journal of the Geological Society of
London, 61, 142.

— 1908. The genera of Stephanoceras and allies. Annals and Magazine of Natural History , 8, 145-149.

— 1909-1930. Yorkshire type ammonites , vols I, II, and Type ammonites , vols III— VII. Wesley and Son,
London and the Author, Thame, 790 pis with explanatory text.

cecca, f., cresta, s., pallini, g. and santantonio, m. 1986. Biostratigrafia ed ammoniti del Dogger-Malm di
Colle Tordina (Monti della Rossa, Appennino marchigiano). Bolletino del Servizio Geologico d' Italia, 104
(1983-1984), 177-203.

della bruna, g. and martire, l. 1985. La successione giurassica (Pliensbachiano - Kimmeridgiano) della Alpi
Feltrine (Belluno). Rivista Italiana di Palontologia e Stratigrafia , 91, 15-62.
dietl, G., flaig, R. and gluck, e. 1978. Zur Stratigraphie des Ober-Bajocium (Braunjura delta/epsilon
Grenzschichten) am Plettenberg bei Balingen, Wurttemberg. Stuttgarter Beitrcige zur Naturkunde , Serie B,
40, 1-16.

— franz, m. and reis, h. 1984. Das Mittel- und Oberbajocium im Gebiet der Wutach unter besonderer
Beriicksichtung der pinguis- Subzone, basale humphriesianum- Zone (Mitteljura, SW-Deutschland). Jahres-
bericht , Mitteilungen des Oberrheinische Geologische Verein , 66, 307-320.

— hager, h. and sauer, f 1984. Ein cycloides- Horizont (humphriesianum-Zone, Mittlerer Jura) im Gebiet
von Aalen/Ostalb. Jahreshefte der Gesellschaft fur Naturkunde in Wurttemberg, 139, 47-55.

dorn, p. 1935. Die Hammatoceraten, Sonninien, Ludwigien, Dorsetensien und Witchellien des stiddeutschen,
insbesondere frankische Doggers. Palaeontographica , Abteilung A, 82, 1-124.
fernandez-lopez, s. r. 1985. El Bajocense el la Cordillera Iberica. Thesis, Universidad Complutense de
Madrid.

fulop, J. 1971. Les formations jurassiques de la Hongrie. Annales Instituti Geologici Publici Hungarici, 54 (2),
31-62.

— geczy, B., konda, J. and nagy, E. 1969. Excursion geologique dans le Montagues Centrale de Transdanubie ,
Mecsek et de Villany. Hungarian Geological Survey, Budapest, 63 pp.

galacz, a. 1976. Bajocian (Middle Jurassic) sections from the Northern Bakony (Hungary). Annales
Universitatis Scientiarium Budapestinensis, Sectio Geologica, 18, 177-191.

— 1980. Bajocian and Bathonian ammonites of Gyenespuszta, Bakony Mts., Hungary. Geologica Hungarica ,
Series Palaeontologica , 39, 1-227.

gardet, g. and gerard, c. 1946. Contribution a l’etude paleontologique du Moyen Atlas septentrional: Lias
inferieur - Bathonien. Notes et Memoires du Service Geologique du Maroc , 64, 1-88.
geczy, b. 1971. The Pliensbachian of the Bakony Mountains. Acta Geologica Academiae Scientiarum
Hungaricae, 15, 117-125.

— 1976. Les ammonites du Carixien de la Montague du Bakony. Akademiai Kiado, Budapest, 223 pp.
gillet, s. 1937. Les ammonites du Bajocien d’Alsace et de Lorraine. Memoires du Service de la Carte

Geologique d' Alsace et de Lorraine , 5, 1 130.

greppin, e. 1898. Description des fossiles du Bajocien superieur des environs de Bale. Memoires de la Societe
Paleontologique Suisse , 25, 1-52.

horwitz, l. 1937. La faune et Page des couches a Posidonomyes (zones Pienine des Klippes. Karpates
polonaises). Bulletin du Service Geologique de Pologne, 8 (4), 1-99; 9 (1), 1-221.
hoyermann, t. 1917. Ueber Dorsetensia Buckman und Ammonites Romani Opp. Inaugural- Dissertation der
Universitdt zu Tubingen.

huf, w. 1968. Uber Sonninien und Dorsetensien aus dem Bajocium von Nordwestdeutschland. Beihefte zum
Geologischen Jahrbuch , 64, 1-126.

hyatt, a. 1867. The fossil Cephalopods of the Museum of Comparative Zoology. Bulletin of the Museum of
Comparative Zoology , Harvard University , 1, 71- 102.

— 1889. Genesis of the Arietidae. Smithsonian Contributions to Knowledge , 26 (No. 673), 1-238.

imlay, r. w. 1973. Middle Jurassic (Bajocian) ammonites from eastern Oregon. Geological Survey Professional
Paper, 756, 1-100.

joly, b. 1976. Les Phylloceratidae malgaches au Jurassique. Generalites sur les Phylloceratidae et quelques
Juraphyllitidae. Documents des Laboratoires de Geologie de la Faculte des Sciences de Lyon, 67, I 471

884

PALAEONTOLOGY, VOLUME 34

kakhadze, J. 1937. Les ammonites bajociennes de la Georgie occidentale. Bulletin de I'lnstitut Geologique de
la Georgie , 2 (2), 123-198 [In Russian].

— and zesashvili, v. I. 1956. Bajocian fauna from the valleys of the Kuban and its some affluents. Trudy
Geologicheskikh Instituty, Akademiya Nauk Georgicheskoi S.S.R. , Seriya Geologiya , 9 (2), 5-55. [In
Russian].

konda, j. 1970. Lithologische und Fazies-Untersuchung der Jura-Ablagerungen des Bakony-Gebirges.

Annales Instituti Geologici Publici Hungarici, 50 (2), 31-62.

KRYMHOLZ, G. j. 1961. Ammonites of the Lower and Middle Jurassic deposits of the Northern Caucasus.
University Press, Leningrad, 166 pp. [In Russian],

kudernatsch, J. 1852. Die Ammoniten von Swinitza. Abhandlungen der Kaiserliche und Konigliche
Geologischen Reichsanstalt , 1 (2), 1-16.

K.UMM, a. 1952. Der Mesozoikum in Niedersachsen : Der Dogger. Geologie und Lagerstatten der Sachsen, 2 (2),
329-509.

lissajous, m. 1907-1912. Jurassique Maconnais: fossiles caracteristiques et les plus repandus. Bulletin de la
Societe d'Histoire Naturelle de Macon, 3 (3-16), 1-126.
mascke, E. 1907. Die Stephanoceras-Verwandten in den Coronaten-Schichten von Norddeutschland.
Inaugural-Dissertation, Universitat zu Gottingen.

maubeuge, p. l. 1949. Notes paleontologiques sur quelques ammonites jurassique rares ou nouvelles de la
region frontiere Franco-Luxembourgeoise et la Lorraine centrale. Institut Grand-Ducal de Luxembourg,
Section des Sciences Naturelles, Physiques et Mathematiques , Archives, 8, 149-178.

— 1951. Les ammonites du Bajocien de la region frontiere franco-beige. Memoires de I'lnstitut Royal des
Sciences Naturelles du Belgique, 2 (No. 42), 1-104.

1961. Catalogue des ammonites du Jurassique inferieur et moyen (Hettangien a Bathonien) du Musee
cantonal de Bale-Campagne. Tdtigkeitsberichte der Naturforschenden Gesellschaft Baselland, 22, 1-164.

— 1969. Catalogue des ammonites du Jurassique inferieur et moyen (Hettangien a Bathonien) du Musea
cantonal de Bale-Campagne. Ill : Ammonites du Jurassique etranger. Tatigkeitsbericht der Naturforschenden
Gesellschaft Baselland , 26, 13-79.

morton, n. 1971. Some Bajocian ammonites from Western Scotland. Palaeontology, 14, 266-293.
munier-chalmas, e. c. p. a. 1892. Sur la possibility d’admettre un dimorphisme sexuel chez les Ammonitides.

Societe Geologique de la France, Comptes Rendus, 3 (20), 170-174.
myczynski, r. 1973. Middle Jurassic stratigraphy of the Branisko succession in the vicinity of Czorsztyn
(Pieniny Klippen Belt, Carpathians). Studia Geologica Polonica, 62, 1-122.
neumayr, m. 1871. Die Phylloceraten des Dogger und Malm. Jahrbuch der Kaiser lich-Kongilichen Geologischen
Reichsanstalt, 21, 297-354.

1875. Die Ammoniten der Kreide und die Systematik der Ammonitiden. Zeitschrift der Deutschen

Geologischen Gesellschaft, 27, 854—892.

noszky, j. 1943. Bericht iiber geologische Untersuchungen im Innengebiet des nordlichen Bakonygebirges.
Relationes Annuae Instituti Regii Hungarici Geologici . 1939—40, 253-261.

— Formations jurassique de la Hongrie. Annales Instituti Geologici Publici Hungarici, 49 (2), 481-501.
ohmert, w. 1988. The Humphriesianum Zone in the type area. Abstracts of Meeting on Bajocian Stratigraphy,

Piobbico, Italy, 1988.

— schatz, R. H. and stetter, l. 1982. Das Mittel-Bajocium von Lorrach. 1. Die sauzei-Zom. Jahreshefte
des Geologischen Landesamtes Baden- Wiirttemberg, 24, 131-165.

oppel, a. 1856-58. Die Juraformation Englands, Frankreichs und des sudwestlichen Deutschlands.

Wiirttembergische Naturwissenschaftliche Jahreshefte, 12-14, 1-857.
orbigny, A. c. v. d’ 1842—49. Paleontologie franfaise ; terrain jurassique. 1. Cephalopodes. Masson, Paris,
642 pp.

parsons, c. f. 1976. A stratigraphic revision of the humphriesianum / subfurcatum Zone rocks (Bajocian Stage,
Middle Jurassic) of southern England. Newsletters on Stratigraphy, 5, 114-142.

1977. A stratigraphic revision of the Scarborough Formation (Middle Jurassic) of north-east Yorkshire.
Proceedings of the Yorkshire Geological Society, 41, 203-222.
pavia, g. 1983. Ammoniti e biostratigrafie del Baiociano inferiore di Digne (Francia SE, Dip. Alpes-Haute-
Provence). Museo Regionale di Science Naturali di Torino, Monografie 2, 1-260.
pugin, L. 1964. Etude critique des Lytoceratina du Dogger. Memoires suisses de Paleontologie, 80, 1-67.
pusch, g. g. 1837. Polens Palaontologie. Schweizerbart’s Verlag, Stuttgart, 218 pp.

quenstedt, F. A. 1845-49. Petrefaktenkunde Deutschlands. 1. Die Cephalopoden. Fues, Tubingen, 380 pp., 36
pis.

GALACZ: BAJOCIAN AMMONITES FROM HUNGARY

885

quenstedt, F. A. 1858. Der Jura. Laupp, Tiibingen, 842 pp., 100 pis.

- 1886-87. Die Ammoniten des schwabischen Jura. Schweizerbart’s Verlag, Stuttgart, 1140 pp., 128 pis.
renz, c. 1904. Der Jura von Daghestan. Neues Jahrbuch fur Miner alogie , Geologie und Palaeontologies 2,

71-85.

roche, p. 1939. Aalenien et Bajocien du Maconnais et de quelques regions voisiens. Travaux du Laboratoire
de Geologie de la Faculte des Sciences de Lyon , 29, 1-355.

1943. Sur les couches dites a Ammonites Blagdeni du Mont d’Or Lyonnais. Travaux du Laboratoire de
Geologie de la Faculte des Sciences de Lyon , 30, 1-36.

roman, f. and petouraud, c. 1927. Etude sur la faune du Bajocien superieur du Mont d’Or Lyonnais (Ciret);

I. Cephalopodes. Travaux du Laboratoire de Geologie de la Faculte des Sciences de Lyon , 9, 1-55.
rostovtsev, k. o. 1985. The Jurassic deposits of the south part of the Transcaucasus. Nauka, Leningrad, 188 pp.
Sandoval, j. 1983. Bioestratigrafia y paleontologia (Stephanocerataceae y Perisphinctaceae) del Bajocense y
Bathoniense en las Cordilleras Beticas. Thesis Universidad de Granada.
schlegelmilch, r. 1985. Die Ammoniten des suddeutschen Doggers. G. Fischer, Stuttgart and New York, 284

pp.

schmidtill, e. and krumbeck, L. 1938. Die Coronaten-Schichten von Auerbach (Oberpfalz, Norbayern).

Zeitschrift der Deutschen Geologischen Gesellschaft , 90. 297-360.
seyed-emami, k. 1967. Zur Ammoniten-Fauna und Stratigraphie der Badamu-Kalke bei Kerman, Iran (Jura,
oberes Untertoarcium bis nrittleres Bajocium). Inaugural-Dissertation, Universitat zu Munchen.
sowerby, j. and sowerby, j. de c. 1812-1846. Mineral conchology. The Authors, London, 648 pis.
spath, l. f. 1927-1933. Revision of the Jurassic cephalopod fauna of Kachh (Cutch). Palaeontologica Indica ,
9 (2), 1-945.

sturani, c. 1971. Ammonites and stratigraphy of the ‘ Posidonia alpina ' beds of the Venetian Alps (Middle
Jurassic, mainly Bajocian). Memorie degli Istituti di Geologia e Mineralogie delTUniversitd di Padova , 27,
1-190.

waagen, w. 1869. Die Formenreihe des Ammonites subradiarus. Geognostisch-Palaontologische Beitrdge , 2 (2),
181-256.

weisert, k. 1932. Stephanoceras im Schwabischen Braunen Jura-delta. Palaeontographica , Abteilung A , 76,
121-195.

westermann, g. 1954. Monographic der Otoitidae (Ammonoidea). Beihefte zum Geologischen Jahrbuch , 15,
1-364.

- 1964. Sexual-Dimorphismus bei Ammonoideen und seine Bedeutung fur die Taxonomie der Otoitidae
(einschlieslich Sphaeroceratinae ; Ammonitina, M. Jura). Palaeontographica , Abteilung A , 124, 33-73.

- 1967. Jurassique moyen (Alpes exclues). Lexique Stratigraphique International , /, Europe , Allemagne , 5f2,
1-197.

- and riccardi, a. c. 1979. Middle Jurassic ammonoid fauna and biochronology of the Argentine-Chilean
Andes. Part II: Bajocian Stephanocerataceae. Palaeontographica , Abteilung A, 164, 85-188.

zittel, k. a. 1868. Palaontologische Notizen iiber Lias-, Jura- und Kreide-Schichten in den baierischen und
osterreichischen Alpen. Jahrbuch der Kaiserlich-Koniglichen Geologischen Reichsanstalt , 18, 599-610.

- 1884. Handbuch der pa/aeontologie, Abteilung I. Band 2. R. Oldenbourg Verlag, Munchen und Leipzig,
893 pp.

A. GALACZ

Department of Palaeontology
Eotvos L. University
Kun Bela ter 2
1083 Budapest, Flungary

Typescript received 11 April 1990
Revised typescript received 17 January 1991

THE ORDOVICIAN GRAPTOLITES AZYGOG RAPTUS
AND JISHOUG RAPTUS IN SCANDINAVIA AND

BRITAIN

by a. j. beckly and jorg maletz

Abstract. Two groups within Azygograptus are identified on the presence or absence of adpressed growth of
th l1 along the sicula. The second group has an unusual facies association and this is interpreted to reflect
adaptation to a shoreward, and possibly r-selective environment. Adaptation to such an environment is argued
to have the potential to affect proximal development and consequently the evolutionary origins of this group
are obscure. The subgenera previously erected are not used because of difficulties with the type species.
Features of the proximal development are the main characters used to distinguish the species, and all the
previously named Scandinavian and British species are redescribed except for A. coelebs for which no reliable
specimens can be found. Three new species are described, Azygograptus minutus, Jishougraptus novus and J.
lindholmae. Jishougraptus is recognized for the first time outside China, and the first specimens of Azygograptus
from South America are illustrated.

The genus Azygograptus was established by Nicholson ( 1875) with A. lapworthi, from the Skiddaw
Slates of the English Lake District, as the type species. The single-stiped morphology and simple
thecal form make the genus easily recognizable, but the differences between species are relatively
subtle and identification has been further complicated by the lack of modern descriptions for many
of the species from Scandinavia and Britain. This paper redescribes the type material for the species
from these areas, and three new species are described from Scandinavia, two of which are placed
in Jishougraptus. A summary of the other species that have been placed in Azygograptus , but which
are not described in this paper, is given, with a brief comment on their probable validity as species
and placement in Azygograptus.

Azygograptus has been recognized as one of the most characteristic genera of the Atlantic
Province during the Arenig (Skevington 1973), and Fortey (1984) suggested that the genus belonged
to the epiplanktonic fauna which allowed it to penetrate more shoreward than other graptolite
genera (Fortey and Owens 1987, p. 278; Chen and Yang 1987, fig. 2a). We examine the distribution
of the genus not only with regard to palaeogeography, but also with respect to facies association.
This provides a guide to the ecology of the genus from which its evolution and biostratigraphic
usefulness are reviewed.

DISTRIBUTION

Facies association

Biofacies association. In Wales Azygograptus is generally the first graptolite to occur in the
transgressive succession above the basal Arenig unconformity, where it commonly forms abundant,
monospecific assemblages (Fortey and Owens 1987; Beckly 1987). Locally Azygograptus is replaced
in this position by extensiform graptolites comparable to D. simulans (Beckly 1985; Zalasiewicz
1986) which share a number of characteristics with Azygograptus'. low thecal inclination,
metasicular origin of th l1 (Beckly 1985) and the production of synrhabdosomes (Zalasiewicz 1984).

The relationship of the Welsh occurrences to the trilobite communities identified by Fortey and
Owens (1978) also indicates a position low in a transgressive succession (Text-fig. 1). Azygograptus
typically occurs shortly above sandstone beds that either yield a fauna indicative of the Neseuretus

| Palaeontology, Vol. 34, Part 4, 1991, pp. 887-925, I pl.|

© The Palaeontological Association

888

PALAEONTOLOGY, VOLUME 34

cross- bedding

V

graded beds

o ferruginous
g oolifhs/pisolifhs

current ripple
lamination

\ synsedimentary
^ faults

• phosphatic
• nodules

parallel

lamination

deformed clasts
[synsedimentary ]

bioturbaf ion

text-fig. 1. Graptolite and trilobite distribution within the Arenig transgressive succession of North Wales.

Community (e.g. Bangor Area) or for which it is implied by the presence of Cruziana (cf. Fortey
and Morris 1982): in the Arenig area rare specimens of Azygograptus are associated with a
Neseuretus fauna. The raphiophorid community generally overlies the occurrences of Azygograptus
in North Wales and the associates of A. eivionicus in the Aberdaron Area (see A. eivionicus , p. 904)
suggest a position at the shoreward limit of this community.

These occurrences contrast with those of Azygograptus in Scandinavia where the genus is
generally associated with diverse graptolite faunas, which occasionally indicate the deep water
isograptid biofacies (cf. Fortey and Cocks 1986). However, A. eivionicus is the only species common
to both areas and in the Toyen Shale this species occurs just below the Orthoceras Limestone, which
has been taken as indicative of a regressive eustatic event (Fortey 1984). A. suecicus appears to have
a more restricted distribution and to have a positive association with I. gibberulus sensu Moberg,
which may reflect some degree of biofacies association.

In the English Lake District Azygograptus occurs in both low diversity to monospecific
assemblages and as a member of more diverse faunas. However, even among the more diverse

BECKLY AND MALETZ: ORDOVICIAN GR APTOLITES

889

faunas Azygograptus tends to occur as fairly abundant populations on certain bedding-planes from
which other species are absent or rare. In addition, adjacent graptolite assemblages are often slightly
unusual and difficult to correlate (e.g. Barf and Hodgson How Quarry), and may reflect a slightly
different biofacies. Interestingly, D. simulans , which appears to favour a similar facies to
Azygograptus in North Wales, also occurs at Barf. The only suggestion of association with a more
oceanic biofacies is the presence of a juvenile isograptid from Hodgson How Quarry, but this cannot
be taken as diagnostic and Azygograptus and isograptids are generally not associated in the Lake
District.

Lithofacies association. Azygograptus occurs mainly in three different lithofacies: (1) flaggy
sandstone facies (cf. Beckly 1987); (2) turbidite sandstone beds interbedded with siltstone; (3)
massive siltstone or mudstone.

I

I

i

+

o

CO

o

CD

o

cl

o

O

o

§-

E

c

"o

o>

to

o

O)

=c:

o

CL)

o

X)

03

o

C

jfr

CL)

CD

o

£

ex

o

"o

O

03

p

£ UJ

03

o

E

s

£

03

o

c

o

03

o

~o

o

CL

Cl

o

03

o

text-fig. 2. Graptolite distribution within the Toyen Shale Formation of Oslo. Lithological column after

Erdtmann (1965).

890

PALAEONTOLOGY, VOLUME 34

text-fig. 3. British Azygograptus- yielding localities, a. Lake District: Al, Robinson Mountain, Buttermere,
scree SW of Robinson Crags (NY 210170); A2, scree approximately 50 m N of Cairn at 703 m on ridge between
Hopegill Head and Whiteside (NY 17822225); A3, large scree, 970 m SSE of summit of Swinside, Lorton,
(locality 1046 of Jackson (1962)) (NY 179233); A4, Hodgson How Quarry, near Portinscale (NY 24372362);
A5, road-cutting 466 m SW of summit of Seat How, Thornwaite (locality 1032 of Jackson (1962)) (NY 210253);
A6, Barf, near Basenthwaite Lake (NY 217267); A7, Carlside Edge, Skiddaw (NY 254278); A8, 160 m at 160°
from summit of Ling Fell (NY 18012844); A9, scree approximately 50 m N of cairn at 703 m on ridge between
Hopegill Head and Whiteside (NY 17822225); A 10, Tom Rudd Beck west of Ling Fell. 100 m SW of Beckhouse

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

891

The flaggy sandstone facies comprises thinly interbedded and interlaminated fine sandstone and
siltstone. The sandstone laminae are often lenticular and show ripple cross-lamination, whilst
bioturbation is moderately abundant but is predominantly parallel to bedding. This is the typical
facies in which Azygograptus is found in North Wales and generally occurs between an underlying
cross-bedded sandstone unit and overlying parallel laminated siltstone (see Text-fig. 1). This, and
the nature of the facies suggest deposition in a shelf setting, probably below normal wave base
(Beckly 1987).

The other two lithofacies generally suggest a deeper water depositional environment, with the
exception of the silty mudstone in which Azygograptus occurs in the Aberdaron Area. This is
comparable to the flaggy sandstone facies in both its stratigraphic position and associated biofacies,
and is thought to have been deposited in a similar environment which was starved of coarser clastic
material (Beckly 1987).

The shales in which Azygograptus occurs in Scandinavia, are predominantly dark in colour, and
sandstone is generally absent. In the Toyen Shale Formation (Text-fig. 2), A. eivionicus occurs in the
slightly paler, greenish, Slemmestad Member, and the occurrences of A. suecicus in Scania are in
yellowish-green shale.

Though much of the material from the Lake District is derived from screes, an indication of the
lithofacies is provided by the slabs on which the specimens occur, and Azygograptus is
predominantly associated with sandier lithologies, and not massive mudstone. Since graptolites are
preserved on siltstone slabs the association is not a function of the less penetrative deformation in
sandy lithologies. In-situ occurrences further support this with the appearance of Azygograptus
locally associated with the incoming of sandy beds and laminae (A. W. A. Rushton pers. comm.).
The two populations collected from Hodgson How Quarry both occur in parallel laminated,
micaceous, fine sandstone, whilst the interbedded mudstone is virtually barren of graptolites.

The significance of the association with turbidite beds is less clear because the amount of
fragmentation appears inadequate for the material to have been transported by a turbulent flow.
Alternative explanations are:

1. Azygograptus was carried into the area of deposition by processes affecting the higher part of
the water column which have a functional relationship to the turbidite flows, but this still fails to
explain the absence of other graptolites.

2. Rapid burial by the sandstone beds allowed preservation on an oxic sea-floor, in an
environment not favoured by other graptolites.

3. The turbidite beds reflect a change to an environment which Azygograptus favoured, e.g.
increased turbidity of the water.

The common factor in all these explanations is a change in habitat to one which was favoured
by Azygograptus , and the low diversity of many associated faunas suggests the change was
disadvantagous to most other graptolite species.

Geographical distribution

The Azygograptus yielding localities of Britain and Scandinavia are shown in Text-figures 3 and 4
respectively, whilst the global distribution is shown in Text-figure 5. When plotted on the same
palaeogeographic reconstruction, Azygograptus shows a very similar distribution to the trilobite
genus Neseuretus (cf. Fortey and Morris 1982, fig. 2). The major difference is that Azygograptus is
not restricted to Gondwana but extends into Baltica. However, all occurrences lie above 30°
latitude, and a high latitude position is now thought to characterize the Atlantic graptolite province.

Cottage (NY 16382898). b, North Wales: Bl, Maen Gwenonwy (SH 20052596); B2, Rhiw (SH 22322847); B3,
Bryncroes (SH 23173150); B4, Nant (SH 297293); B5, Penrhyn Du (SH 31742653); B6, Castlcmarch Farm (SH
31452980); B7, Penrhyndeudraeth (SH 616401); B8, scree on northern slopes of Moel Llyfnant, Arenig (SH
80793616); B9, above Hafotty Ffilltirgerig, Arenig (SH 81613960); BIO, Bangor (SH58573I and SH 591726).
c, South Wales: Cl, Afon Ffinnant (SN 510198); C2, Whitesand Bay (SM 732274).

892

PALAEONTOLOGY, VOLUME 34

BECKLY AND MALETZ: ORDOVICIAN GR APTOLITES

893

text-fig. 5. Global distribution of Azygograptus. a, Present day geography, b. Lower Ordovician
reconstruction (after Fortey and Morris 1982). I. British occurrences (see Text-fig. 3) and SE Ireland
(Brenchley and Harper 1967); 2, Scandinavia (see Text-fig. 4); 3, Bohemia (Boucek 1973); 4, SW China (Mu
et al. 1979); 5, Gorny Altai (Obut and Sennikov 1984); 6, Middle East Baltic Area (Ulst 1976, pp. 214—221 ;
Paskevicius 1976, p. 135); 7, near Tarija, southern Bolivia (new material); 8, Cantabria, northern Spain

(Gutierrez Marco and Rodrigues 1987).

894

PALAEONTOLOGY, VOLUME 34

Ecology of Azygograptus

In both bio- and lithofacies associations, differences exist between the Scandinavian and British
representatives of the genus. However, this may be largely explained by the predominance of A.
lapworthi and A. eivionicus in Britain, compared to their virtual absence from Scandinavia. The
facies association and position within the succession of Azygograptus in North Wales support the
shoreward distribution advocated by Fortey and Owens (1987) but an epiplanktonic mode of life
should result in the presence of Azygograptus in all graptolite faunas of the Atlantic province. This
is not observed and therefore appears to be fairly good evidence that at least some species of
Azygograptus favoured a shoreward environment.

The distribution of extensiforms similar to D. simulans is comparable to that of Azygograptus and
both groups develop synrhabdosomes (Zalasiewicz 1984), suggesting these are also a feature with
functional significance to the particular environment. Though the function of synrhabdosomes is
uncertain, Zalasiewicz (1984) suggested they maximized breeding in temporary, highly favourable,
conditions, implying an /--selective environment. Adaptation to an unstable environment may also,
at least in part, explain the association of Azygograptus with sandstone beds in the Lake District.

At Barf (Locality A6 of Text-fig. 3), D. simulans is associated with A. ellesi, a species more
common in Scandinavia. However the Scandinavian occurrences, with the possible exception of A.
suecicus , do not display significant differences in distribution to the majority of other graptolites.
This may be a consequence of the different palaeogeographic position of Baltica (see Text-fig. 5b).

SYSTEMATIC PALAEONTOLOGY

Parameters used in descriptions. The measurements used in descriptions are summarized in Text-figure 6:

1. Basal free length of sicula (A on Text-fig. 6d). The distance along a line parallel to the dorsal sicula
margin, between the aperture and the final point of contact of the first theca.

2. Stipe Divergence Angle (/? on Text-fig. 6c). Angle between dorsal wall of sicula and dorsal wall of
proximal part of stipe.

text-fig. 6. Parameters used in descriptions. A, entire
rhabdosome; 4>5, <J>10, <J>15, stipe rotation by each of these
thecae, b, stipe detail showing stipe width and thecal
separation, c, sicula: x, distance to origin of th . 1 1 ; y,
sicular length ; /?, stipe divergence angle, d, detail of
sicular aperture: a, basal free length.

BECKLY AND MALETZ: ORDOVICIAN GR APTOLITES

895

3. Origin of th l1 is given as a fraction of the length of the sicula measured away from the apex (x/y on Text-
fig. 6c).

4. Thecal Separation/Density. Thecal density is given as a recalculated value in thecae/ 10 mm, based on
measurements of thecal separation: the distance between successive thecal apertures (see Text-fig. 6b).

5. Stipe curvature: Quantified as the degrees of rotation between lines tangential to the proximal and distal
dorsal stipe margin, measured at th 5, th 10, th 15 etc. (see Text-fig. 6a).

Repositories of material mentioned. The following abbreviations are used to indicate where material mentioned
is lodged: BU, University Museum, Birmingham; BGS, British Geological Survey, Keyworth; BM, British
Museum (Natural History), London; IC, Imperial College, London (teaching collection) ; LO, Palaeontological
Institute, University of Lund; NMW, National Museum of Wales, Cardiff; PMO, Paleontologisk Museum,
Oslo; SGU, Sveriges Geologiska Undersokning, Uppsala; SM, Sedgwick Museum, Cambridge; TUB,
Technische Universitat Berlin.

Order graptoloidea Lapworth, 1875
Family dichograptidae Lapworth, 1873
Subfamily azygograptinae Mu, 1950

Diagnosis. Dichograptids with single stipe originating from th l1. The sicula bears two processes on
aperture.

Discussion. Mu (1950) erected a monotypic family Azygograptidae containing Azygograptus , but
since then a number of other genera have been established for single-stiped species: Nicholsono-
graptus Boucek and Pribyl, 1953, Parazygograptus Kozlowski, 1954, Pseudazygograptus Mu, Lee
and Geh, 1960, Hemiholmograptus Hsu and Chao, 1976, and Jishougraptus Geh, 1988.

In all these genera the stipe is formed by th l1, except for Parazygograptus in which the stipe is
based on th l2 and is pendent. This genus is founded on a single isolated specimen of the type species
P. erraticus Kozlowski. In other respects this specimen is very similar to the co-occurring pendent
Didymograptus rozkowskae Kozlowski and could therefore represent a broken rhabdosome. For
this reason, Parazygograptus is tentatively regarded as a junior synonym of Didymograptus.

Nicholsonograptus was included within the Sinograptidae by Skevington (1966), Bulman (1970)
and Boucek (1973). Wang (1975) placed Nicholsonograptus in his new family Paradidymograptidae
though the necessity for this family is questionable. Hemiholmograptus is now regarded as a junior
synonym of Nicholsonograptus (Geh 1988). Pseudazygograptus Mu et al. , 1960, with A. incurvus as
the type species, was referred to the Virgellina by Fortey and Cooper (1986) on account of the
prominent virgellar structure illustrated by Finney (1980). However, this structure seems to be an
extended growth of the rutellum rather than a virgellar spine and inclusion of Pseudazygograptus
within the Dichograpidae is therefore preferred.

Genus azygograptus Nicholson and Lapworth, in Nicholson, 1875
Type species. Azygograptus lapworthi Nicholson, 1875.

Diagnosis. Rhabdosome of single, uniserial, declined stipe originating from th l1. Stipe either
straight or curved (causing stipe to be reclined in distal part). Thecae simple, elongate, conical and
inclined at low angle to the dorsal margin.

Discussion. Elies and Wood (1902, p. 93) recognized three informal groupings of Azygograptus
which Obut and Sennikov (1984) elevated to subgenera: A. ( Azygograptus ), A. ( Eoazygograptus ),
and A. ( Metazygograptus ). One of the main characters on which these groups were based was the
origin of th 1 1 and its growth relative to the sicula. Only two groups are recognized reliably by the
present study on the presence or absence of adpressed growth of th l1 and the sicula (see Text-fig.
7). However, the named subgenera have not been used because of difficulties with the type species.

The type species of the subgenus Metazygograptus is A. suecicus but the characters of the Elies

896

PALAEONTOLOGY, VOLUME 34

Group 1

Group 2

E F G

text-fig. 7. Groupings of Azygograptus on proximal development (all x 20). Group 1 - lacking adpressed
growth of th l1. Group 2 - with adpressed growth of th l1. Species illustrated are: a, A. eivionicus ; b, A.
lapworthi; c, A. hicksii; d, A. suecicus ; E, A. ellesi; F, A. minutus ; G, A. validus.

and Wood grouping (1902) were based on material from Barf, now thought to be A. ellesi. As a
result A. ( Metazygograptus ) is a subjective junior synonym of A. ( Azygograptus ) which is
characterized by the absence of adpressed growth on th l1 and the sicula.

The type species of A. (Eoazygograptus) is A. coelebs but no specimens adequate to confirm the
character of this species have been found. The published descriptions (Elies and Wood 1902) would
indicate placement in Group 2 (see Text-fig. 7), but until this can be confirmed the use of the
subgenus is unsupported.

Azygograptus lapworthi Nicholson, 1875.

PI. 1, fig. 2; Text-figs 10a-o, 11 a-e, 12

1875 Azygograptus lapworthi Nicholson, p. 270, pi. 7, fig. 2-2 c.

1898 Azygograptus lapworthi Nicholson; Elies, p. 513.

EXPLANATION OF PLATE 1

Figs 1 and 11. Azygograptus ellesi Monsen. 1, SGU 7554; Nipan, Jemtland, x 10. 11, PMO 118.576;

1 6-25 1 6-42 m interval of Toyen Section, Oslo, x20.

Fig. 2. Azygograptus lapworthi Nicholson. SM A55166a; Castlemarch Farm, North Wales, x20.

Figs 3, 4, 7, 8. Azygograptus eivionicus Elies. 3, NMW 85. 16G.94; Garth Point, Bangor, North Wales, x 15.

4, 7, 8, Rhiw, North Wales; 4, BM Q5884, x 15; 7, BM Q5896, x20; 8, SM A22601, x 10.

Fig. 5. Azygograptus suecicus Moberg. SGU 7895; Killerod, Scania, x 8.

Figs 6, 9, 10. Jishougraptus novus sp. nov. Paratypes; from 161 5—1 6-25 m interval of Toyen Section, Oslo,
x 20. 6, PMO 118.582-3. 9, PMO 1 18.593. 10, PMO 118.588.

Figs 7 and 8. Azygograptus eivionicus Elies. Rhiw, North Wales. 7, BM Q5896, x 20. 8, SM A22601, x 20.
Figs. 11. Azygograptus ellesi Monsen. PMO 1 18.576; 1 6-25—1 6-42 m interval of Toyen Section, Oslo, x20.
Fig. 12. Azygograptus minutus sp. nov. SGU 7556 [paratype]; Diabasbrottet Section at 7-09 m level,
Flunneberg Mountain, Vastergotland, Sweden, x 20.

Fig. 13. Azygograptus validus Tornquist. SGU 7557; Diabasbrottet Section at 4-74 m level. Hunneberg
Mountain, Vastergotland, Sweden, x 15.

PLATE 1

BECKLY and MALETZ, Azygograptus , Jishougraptus

898

PALAEONTOLOGY, VOLUME 34

1902 Azygograptus lapworthi Nicholson; Elies and Wood, p. 93, text-fig. 54; pi. 13, fig. 1 a-b.

1915 Azygograptus lapworthi Nicholson; Nicholas, pp. 112, 121.

1938 Azygograptus eivionicus Elies; Matley, p. 559.

1943 Azygograptus lapworthi Nicholson; Fearnsides and Davies, p. 253.

71979 Azygograptus lapworthi Nicholson; Mu et al., pp. 109-110, pi. 38, figs 3-7.

1984 Azygograptus lapworthi Nicholson; Zalasiewicz, p. 427, fig. 2 e-f.

1988 Azygograptus lapworthi Nicholson; Beckly, p. 325.

Diagnosis. Azygograptus with no adpressed growth of th l1 and sicula and basal free length of the
sicula commonly exceeding 0-2 mm. Prominent rutellum on sicular aperture in some populations.

Type material. The original slab figured by Nicholson from Dover's collection has not been traced. Differing
populations of Azygograptus are certainly present at the type locality, and there is no clear evidence on the level
from which the original material came. A considerable amount of material from Hodgson How is present in
many museum collections and the original slab may be amongst this.

Type locality. Hodgson How Quarry, Keswick, English Lake District (National Grid Reference NY 243236).
The succession comprises turbidite sandstone beds, some over 30 cm thick, interbedded with siltstone. Two
Azygograptus populations have been recovered during this study; one close to the base of the exposed section
and the other about 20 m above, both from parallel laminated, micaceous sandstone. These populations show
quite marked differences (see Text-figs 9 and 10) and further variations shown by some museum material
suggest at least a third population. The lower of the in situ populations is sufficiently different to the original
description to suggest it is unlikely to be topotype (see discussion).

Jackson ( 1962) claimed the fauna from Hodgson How Quarry indicated the nitidus biozone, whilst Eastwood
et al. (1968) claimed the assemblage was indicative of the Dichograptus biozone, a subdivision rejected by
Jackson (1962). Material collected during this study, which includes D. hirundo and a juvenile ‘ Isograptus',
indicates a higher horizon and probably the gibberulus biozone. Lower in situ population : BM Q5630-Q563 1 .
Higher in situ population: BM Q5108-Q51 17.

Other occurrences.

1. Carlside Edge (locality A7 of Text-fig. 3). A poorly preserved fauna in micaceous grey siltstone. BM
P721 1-P7212, P7218.

2. Castlemarch Farm, North Wales (SH 31452980) (locality B6 of Text-fig. 3). Sparse fauna, occasionally
preserved in partial relief. Only faunal element known. SM A55166; NMW 85.16G. 73-74.

3. Penrhyn Du (locality B5 of Text-fig. 3); L5 km south-southeast of Abersoch (NGR SH 31742653)
(Localities and t]i of Nicholas 1915, pi. 13). Material comes from the ‘Transition Beds’ between the St
Tudwal's Sandstone and Llanengan Mudstone (cf. Nicholas 1915). The lithofacies is typical of the 'flaggy
sandstone facies’ and the population forms a monospecific assemblage. NMW 85. 16G. 122-124.

4. Small Valley 400 m west of Bryncroes (SH 23173150) (locality B3 of Text-fig. 3) Abundant population
forming monospecific assemblage in 'flaggy' sandstone facies of Sarn Formation (Beckly 1988). BM
Q5876-Q5879.

5. Abandoned mine workings just north of Penrhyndeudraeth (SH 616401) (locality B7 of Text-fig. 3).
Termed Pant-y-wrach Beds by Fearnsides and Davies (1943), who mention a fauna comprising A. lapworthi
and Phyllograptus angustifolius . Only specimens of A. lapworthi have been examined and these occur in parallel
laminated coarse siltstone. All specimens have been collected from the spoil of the mine workings. NMW
8516G. 121 ; BGS Z1672; IC 1514.

6. Near Tarija, southern Bolivia. Few specimens as monospecific assemblage, preserved as silvery periderm
in grey silty mudstone. The succession comprises 4,500 m of shale with few sedimentary structures, and all
other graptolites found indicate a lower to middle Arenig age, TUB BOL 387P 01-05.

Description. Sicula 10-1 5 mm long and 0-2-0-35 mm wide at aperture excluding ventral lip (see Text-fig. 8).
Stipe originates at about 0-7 of sicular length and diverges immediately at an angle of about 135-140°.
Prominent basal free length to sicula up to 0-45 mm long and often greater than 0 2 mm.

Stipe moderately flexed with rotation by th 5 between 15° and 30° but ranging up to 55° by th 10 or th 15.
Rhabdosomes up to 30 mm long observed but 50 mm has been claimed (Elies 1898). Initial stipe width at first
theca about 0-5 mm or less but rapidly increases to fairly constant width 0-8-1 05 mm. Maximum widths in

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

899

Hodgson How Quarry (lower Population]

Hodgson How Quarry (A. lapworthi )

Hodgson How Quarry [GSM 46365]

Castlemarch Farm

Bryncroes

4 ii ^

Penrhyn Du

1| ]|X«

■ % “ ~ “

Penrhyndeudraeth

n r

c — J I ■ u pi

sicular apertures X 20

ji — M—

■ i

n 1 r

pJ »l »» ■ , i

i r

— p

U

n j-

0.1 0.2 0.3 0.4

basal free length of sicula

text-fig. 8. Sicular apertures and basal free length histograms for different populations of Azygograptus

lapworthi.

900

PALAEONTOLOGY, VOLUME 34

excess of 1 mm are more characteristic of longer stipes and may indicate a slow distal stipe expansion. Thecal
spacing 1-2-1 -6 mm; thecal density 6-9 th/10 mm; thecal inclination 10-12°; thecal overlap a little over one-
third. Thecal density and overlap generally increase over the first few thecae. One specimen from
Penrhyndeudraeth (Text-fig. 10g) has some thecae with apertures isolated by a notch about 0-2 mm deep.

Zalasiewicz (1984) described a synrhabdosome of A. lapworthi from Hodgson How. Another possible
synrhabdosome has been recognized during this study in the population from Bryncroes (Text-fig. 12). It
appears to comprise predominantly juvenile rhabdosomes and is more unidirectional than the example from
Hodgson How.

Discussion. The narrower stipe and higher thecal density serve to distinguish A. lapworthi from A.
hicksii (Text-fig. 15), but stipe characters are inadequate to distinguish it from A. eivionicus , with
only slightly greater stipe width in A. lapworthi. Distinction of A. lapworthi from A. eivionicus is
therefore dependent on features of proximal development and in particular the basal free length of
the sicula, which in A. eivionicus only rarely exceeds 0 2 mm (Text-fig. 9). A more prominent
rutellum (see Text-fig. 8) and shorter sicula also distinguish some populations of A , lapworthi but
these characters are not always developed. A suecicus is distinguished by the absence of a rutellum.

The majority of published descriptions of A. lapworthi (Nicholson 1875; Elies 1898; Elies and
Wood 1902) have been based on material from the type locality, and therefore synonymy is
assumed. Despite this direct comparison, there are quite significant differences in the values quoted,
particularly sicular length. The value given by Nicholson (1875) in the original description was
105 mm, but Elies (1898) and Elies and Wood (1902) quoted values of 1-26 mm and T5 mm
respectively.

These differences may be largely explained by the presence of more than one population at the
type locality, and that more than one form of the species may exist. The population from the base
of the exposed succession in Hodgson How Quarry has lower thecal separation (1-1 — 1 -3 mm) and
narrower stipes (0-4— 0-5 mm, max. 0-75 mm) than is typical of either A. eivionicus or A. lapworthi ,
though the longest only reaches th 7. The sicular characters are a mixture of those seen in A.
eivionicus and A. lapworthi (see Text-fig. 9) with the basal free length ranging above 0-2 mm, though
predominantly less, and a sicular length of about 1 mm. The population is more similar to A.
eivionicus but with some aspects of a transitional form.

The population from 20 m above is characterized by a short sicula, about TO mm long, significant
basal free length to the sicula and a very prominent rutellum (see Text-figs 8 and 9). However,
indications of at least a third population are provided by a slab in the BGS collections (GSM 46365)
in which the sicular length ranges up to T5 mm and there is only slight development of a rutellum,
but with a significant basal free length. The specimen is labelled A. lapworthi in Nicholson’s
handwriting.

These differences appear subtle but similar variation is shown by the occurrences in North Wales.
Though few specimens are available, the population from Castlemarch Farm shows the prominent
rutellum associated with a sicular length of about 1 mm. However the other three North Wales
populations lack the prominent rutellum, and have a sicular length ranging up to T5 mm. Since
both these forms appear to occur at the type locality, either could be the true type-population. The
original description is also inadequate to distinguish them, since it combines short sicular length
with a poorly developed rutellum. For these reasons the species is broadly defined on the basal free
length of the sicula only.

The only confirmed occurrence of A. lapworthi from outside Britain is the new material from
Bolivia (Text-fig. 1 1). The basal free length of the sicula clearly places this population in A. lapworthi
but the stipe width appears to be greater than in the British populations. The material described by
Mu el at. (1979) from China as A. lapworthi may also belong to the species. From translations
provided by Chen Xu, they quote a sicular length of T6 mm and for which a basal free length of
0-4 mm may be inferred. However, it is unclear from the description whether there is adpressed
growth of th 1 1 and the sicula; th l1 is described as growing ‘downward and outward’.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

901

(wuu) DinDis m6uai aajj )osog

X

X

<1

+

<

O

<

I 1 1 1

(uuuu) D)nDis jo qj6u0] aajj idsdq

cq

(UJUJ) D]nDIS JO Lj j6U9) 00JJ IDSDg

?UoS

??§ c

S,t ■£ ^

^ Q,

^ §- 6

— o

x

T3

O

x

bJJ

CO X

o3 j—
£ “

Ji c

^ o

O X .c

2 o3 ^
O

o

3

X O
^ X C

. 9; °

X x* ••-

<3 ^

C 3

■= t a

"5: a O
£ 3 a
20^

B. S
-S o

X

o3

£ 3

? £

CO CO

O X
C X

C/5 <-»-H

T3 W ^

• -h CO

S « a

c 73 2
5 o* x
°

CO c/5

£ c <*-

Sh ° c

S3 CO p

£ bij O
&c X X
? O 3
X)

n‘ e '£

X C to
O ^
t- ,>- ~
<0 — *,

^ co co

C <U

-5oS

ao'4= aj

C

J3 3 <*-
. a. o

I

O

X

G

a3

1 1
<*> §
<YC0

c ^ T3

Si ^ C

C3

C
&

o a 5

c/3 ^ a>

c> -5 £

73 s ~

C/3 ^

U. C/3 ^

d 2 o

o .5 s

|t3 1

8i^

Co CO ~

P 03

■a; 73 o
■< c <

•2 >3

g x

*> £ £

^ -S -a

co O O

<D 5-i *-h

X a3 <L>
O N

^ co
O. 3 73

CO o o

-> o co
o3
o

X

W)

c

o „

O o3

o

o G

c£ X ^

— J

> 03 O
co O C
o3 — *-
X o £

, Oh CO
«

3 o ^
O X G

X ^ 5
O ^ p

03 O £

<>-> s

^ S' E

a ^
a. >c

a

3

D-.td- _

2 to E

D- o

O I

O c

o. 3

U

■5 y3
— e

^ ^ -rt

•Bo?

03 ^ X
'G h/i ^

a
o

Oh

Ctf) ,2JJ

c t:

as

CO

C

• O *ss

o x "S

rr 03

H

Oh ^

UJ o •

H Oh ^

o X

C g
O O
o X

,||
S .2

-O T3

902

PALAEONTOLOGY, VOLUME 34

text-fig. 10. Azygograptus lapworthi Nicholson, a, b, BM H3365, Hodgson How Quarry, x 5. c, BM Q6331,
Hodgson How Quarry, lower population, x 5. D, BM Q6330, Hodgson How Quarry, lower population, x 5.
e. BM Q5876, Bryncroes x 20. f, BM Q5877, Bryncroes, x 5. G, BGS Z1672, Penrhyndeudraeth, x5 (detail
x 20). h, BM Q5879, Castlemarch Farm, x20. i, BM Q5877, Bryncroes, x 20. k, BM Q5110, Hodgson How
Quarry, higher population, x 20. l, BM Q6331. Hodgson How Quarry, lower population, x20. M, NMW
85.16G.124, Penrhyn Du, x 20. n, 1C 1514, Penrhyndeudraeth, x 20. o, NMW 85.16G.123. p, NMW
85. 16G. 122, different distal stipe morphologies, Penrhyn Du, x 5.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

903

text-fig. 1 I Azygograptus lapworthi Nicholson; from near Tarija, southern Bolivia, x 10. a, TUB BOL387P
01. b, TUB BOL387P 02. c, TUB BOL387P 03. d, TUB BOL387P 04 ( x 10). e, TUB BOL387P 05.

Azygograptus hicksii (Hopkinson in Hopkinson and Lapworth, 1875)

Text-fig. 13a-c

1875 Tetragraptus hicksii Hopkinson in Hopkinson and Lapworth, p. 651, pi. 33, fig. 12 a-d.

1902 Azygograptus hicksii (Hopkinson); Elies and Wood, p. 94, text-fig. 55 a-b\ pi. 13, fig. 2 a-e.
1987 Azygograptus hicksii (Hopkinson); Fortey and Owens, p. 275, text-fig. 1286-c.

Diagnosis. Azygograptus with larger stipe than other species; maximum stipe width 1-0-1 -5 mm,
thecal density 4-6 th/10 mm. Sicular aperture with elongate rutellum. No adpressed growth of th l1
and sicula.

Type material. Lectotype (SM A 1 738 1 ) ; selected as ‘type’ by Elies and Wood (1902). All other available
specimens (SM A 17379-94) are paratypes.

Type locality. Pwlluog, Whitesand Bay, near St David’s, Dyfed (SM 732274) (locality C2 of Text-fig. 3).
Specimens occur as carbonaceous films in fissile siltstone of Whitlandian age, Gymnostomix gibbsii Biozone
(Fortey and Owens 1987).

Other occurrences. Afon Ffinnant, east of Carmarthen, South Wales (SN 51061973 to SN 51032003) (locality
Cl of Text-fig. 3). Specimens collected from isolated outcrops of shale. Described by Fortey and Owens (1987)
and considered transient between A. eivionicus and A. hicksii. BM Q5172-Q5173.

text-fig. 12. Azygograptus lapworthi Nicholson, BM Q5878,
possible synrhabdosome of juvenile rhabdosomes, Bryn-
croes; a, x 5; b, x 10.

904

PALAEONTOLOGY, VOLUME 34

text-fig. 13. Azygograptus hicksii (Hopkinson); Whitesand Bay, South Wales, a, SM A17392, x 5. b, SM

A 17933, x 5. c, SM A 17933, x20.

Description. Sicula 1-3 — 1-5 mm long; 04 mm wide at aperture. Prominent rutellum up to 0-5 mm long, on
opposite side to stipe. Stipe originates at 0-7-08 of sicular length and diverges immediately at an angle of
130-140° leaving a basal free length to the sicula of up to 0-3 mm. Stipe slightly to moderately flexed with
20-30° rotation by th 5 and 40° rotation by th 10. Stipe width at th 1 is about 1 mm (0-8-1 -1 mm) and slight
expansion gives a distal stipe width of up to L5 mm. Thecal separation 1 -7-2-7 mm (4-6 th/10 mm) with little
variation along the stipe. Thecal inclination about 18°; thecal overlap not visible.

Discussion. The massive form of the stipe and elongate rutellum on sicular aperture make the species
readily distinguishable. Forty and Owens (1987) concluded that the size is unlikely to be the product
of deformation. The values given in the above description closely match those given by Fortey and
Owens (1987) except for sicular length which they give as reaching 2-1 mm, but no specimens of this
length have been observed in the type population during this study.

Apart from the transient population from Afon Ffinnant described by Fortey and Owens (1987),
the species has not been recognized other than at the type locality, and therefore on present evidence
is restricted to South Wales.

Azygograptus eivionicus Elies, 1922

Plate 1, figs 3, 4, 7, 8; Text-figs 16, 17a-p, 18f-o

1915 Azygograptus lapworthi Nicholson; Nicholas, 1915, p. 113.

1922 Azygograptus eivionicus Elies, p. 299, figs 1-3.

1932 Azygograptus suecicus Moberg; Matley, p. 261.

1987 Azygograptus eivionicus Elies; Fortey and Owens, p. 276, figs 128a and 129a-c.

Diagnosis. Basal free length of sicula consistently less than 0-2 mm. Slight to moderate development
of rutellum. No adpressed growth of th l1 and sicula.

Type material. Holotype, SM A 17372; Paratypes, SM A17373-A17374. Strachan (1971, p. 19) listed all these
specimens as syntypes, though in the original description a type specimen is clearly designated.

The type material is flattened, matted, fragmentary and strongly deformed. The three specimens figured by
Elies (1922) are shown in Text-figure 16 against her original drawings. The differences are most significant in
the case of the holotype, for which there is no evidence of the sicula indicated and it is possible that the
proximal end is not even reached.

There are very few proximal ends among the material and only one complete sicula (A 17373): 1 1 mm long

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

905

Toyen section
A. eivionicus

m -w

Bangor ( A eivionicus)

Rhiw & Maen Gwenonwy

Afon Ffinnant

sicular apertures x 20

1 ' • , r-

01 02 0.3 0.4

basal free length of sicula

text-fig. 14. Sicular aperture and basal free length in populations of Aigograplus hicksii (Hopkmson) and A.

eivionicus Elies.

with a basal free length of 0-08 mm. The basal free lengths in two other measurable siculae are 0 05 and 0- 1 mm.
None of these are associated with more than a couple of thecae and therefore all may be immature. In figuring
A 17373, Elies (1922) showed th l1 as originating in the mid-part of the sicula and growing adpressed to it for
a short distance. There is no evidence for the shoulder this produces and the stipe diverges abruptly at 0-8 of
the length. There is a slight inflation of the sicula approximately at its mid-point, but this is also developed on
the ventral side and may represent the expansion at the base of the metasicula. The stipe divergence angles
measurable on two specimens are 138° and 144° (no correction made for deformation).

906

PALAEONTOLOGY, VOLUME 34

Whitesand Bay ( A.hicksii)

l,lnl,JI 1 1

Hodgson How Quarry (A. lapworthi)

— ■ Ml 1 1

T

Garth Point Bangor ( A eivionicus)

T

T

T

k 5 6 7 8 9 th/IOmm 0.5 1.0 1.5 mm

thecal density maximal stipe width

text-fig. 15. Comparison of stipe characters between Azygograptus lapworthi Nicholson, A. eivionicus Elies

and A. hicksii (Hopkinson).

Stipe characters are difficult to quantify because of the deformation but are probably most reliable on the
holotype which lies almost parallel to the lineation. This has a stipe width of 0-6 mm and a thecal density of
7-4 th/10 mm. Stipes are gently flexed.

Type locality. Valley southeast of Nant, St Tudwal’s Peninsula, North Wales (SH 297253), Loc. p of Nicholas
(1915, pi. 13), in the Llanengan Mudstone.

Other occurrences.

1. Bangor, North Wales. Two localities on the Bangor foreshore were given in the original description of
A. eivionicus (Elies 1922): Garth Point (SH 585731) and University College Cliff (SH 591 17260). Populations
have been collected from two levels at Garth Point (SH 58467320 and SH 58547315), and that from University
College Cliff comes from the same part of the succession repeated by faulting. All the populations are very
similar and are treated as one. The fauna is monospecific, and occurs in ‘flaggy sandstone’ of the Maes y
Geirchen Sandstone Member of the Nant Ffrancon Formation. The only independent faunal control is an
endemic trilobite fauna from the sandstone beds towards the base of the succession (Beckly 1989), and the
presence of the Arenig-Llanvirn boundary a few hundred metres above. Material from Bangor Pier: NMW
85-16G-89-99; BM Q5880-Q5883, Q5895. Material from University College Cliff: BM Q5886.

2. Aberdaron Area, Western Llyn. Populations from Maen Gwenonwy (SH 20052596) (locality B1 of Text-
fig. 3) and Rhiw (SH 22322847) (locality B2 of Text-fig. 3) identified as A. eivionicus and A. suecicus respectively
by Elies (in Matley, 1932), but now considered to occur in the same stratigraphic unit, the Wig Member of the
Aberdaron Formation, and both to be A. eivionicus (Beckly 1988). The lithofacies is a massive silty mudstone,
and Azygograptus along with the trilobite Merlinia is very abundant whilst the rest of the fauna is only known
from rare specimens and comprises Hanchungolithus primitivus (Born), 1 Fur colit hus, and Expansograptus
praenuntius (Tornquist). The fauna is interpreted to come from the lower part of the Arenig succession and to

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

907

be overlain by proven Whitlandian and Fennian Faunas. The Merlinia is a transient form between M. selwynii
(Salter) and M. rhyacos Fortey and Owens (R. A. Fortey pers. comm.), and this would suggest a late
Moridunian to early Whitlandian age (cf. Fortey and Owens 1987). Material from Maen Gwenonwy: NMW
8516G100 103 ; BM Q5885. Material from Rhiw: NMW 85-16G- 103-108 ; BM Q5844, Q5871, Q5875,
Q5896-Q5897 ; SM A22601.

3. Afon Ffinant, South Wales (SN 51032003) (locality Cl of Text-fig. 3). Material described by Fortey and
Owens (1987), and occurring as monospecific assemblages in a succession of mterbcdded turbidite sandstone
beds and siltstone. A. eivionicus is the lowest graptolite to occur in the succession of the area, and is identified
as coming from the F. radix biozone at the base of the Whitlandian. BM Q5 1 74 — Q5 1 79 ; NMW 84T 7G- 105-1 07.

4. Hodgson How Quarry, English Lake District (NY 243236) (locality A4 of Text-fig. 3). Lower in situ
population discussed under A. lapworthi.

5. Tom Rudd Beck, west of Ling Fell, English Lake District (NY 16382898) (locally A10 of Text-fig. 3).
Specimens preserved as rusty films in fairly homogeneous dark grey micaceous, fissile siltstone. BGS Rx2207,
Rx2210-Rx221 1 .

6. Ling Fell, English Lake District (NY 18012844) (locality A8 of Text-fig. 3). Specimens preserved as
chlorite lined moulds in crenulated mudstone. BGS Rx 1 69 1 Rxl692.

7. Toyen Shale, Oslo. Abundant specimens from an interval just below the Orthoceras Limestone between
1919 and 19-75 m (Text-fig. 2). Graptolite abundance is reduced, compared to the darker lithology of the
Galgeberg Member, but the fauna is still diverse and includes Pseudotrigonograptus ensiformis, Holograptus
diffusus , Xiphograptus cypselo , X. svalbardensis , Expansograptus abditus , E. distinctus, Tetagraptus taraxacum ,
and Dichograptus sp. PMO 118-601-610.

Description. Sicula 1 1 3—1 -62 mm long. Stipe originates at 0-7-0-9 along sicular length and diverges at an angle

text-fig. 16. Type specimens of Azygograptus eivionicus Elies, a, Elles’s original drawings (1922) quoted as
x 5. b, same specimens redrawn at x 10. 1, SM A17372 (holotype); 2, SM A17377; 3, SM A17374.

908

PALAEONTOLOGY, VOLUME 34

between 120° and 160°. Basal free length of sicula 01-0-2 mm. Development of rutellum variable but is readily
apparent in the majority of specimens. Stipe moderately flexed with up to 50° rotation by th 10 or th 15. Stipe
width and thecal separation both vary over the first five thecae. Thecal separation falls from as much as 2-5 mm
(4 th/10 mm) to the more typical value of 1-5-T7 (6-7 th/10 mm); stipe width increases from 0-4-0-6 mm at
the aperture of th l1 to 0-7-0-8 mm more distally. These changes result mainly from increasing overlap. Thecal
inclination 8-13°. Thecae fairly straight, conical with slight ventral concave curvature. Thecal overlap about
two-fifths.

Discussion. Because of the poor preservation of the type material, the population from Garth Point,

text-fig. 17. Azygograptus eivionicus Elies from North Wales (a-g. Bangor; h-p, Rhiw, Western Llyn). a, BM
Q5881, x 5. b, BM Q5880, x 5. c, BM Q5880, x 5. d, BM Q5883, x 5. e, BM Q5882, x 5. f, BM Q5880, x 20.
G, BM Q5880, x 20. n. BM Q5871, x 5. I, BM Q5872, x 5. k, BM Q5874, x 5. l, NMW 85. 16G. 103, x 20.
m, BGS A22601 , x20. n. BM Q5873, x20. o, BM Q5884, x 5. p, BM Q5875, x 5.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

909

Bangor was selected as the basis for the description because that, and University College Cliff,
Bangor, were the only other occurrences accurately located in the original description of the species.

The population from the Aberdaron area is similar to that from Bangor in all characters except
sicular length, which is more restricted (11-1 -4 mm), and in a slightly less pronounced rutellum.

The population from South Wales described by Fortey and Owens (1987) is closely comparable
to the populations from North Wales, the only significant difference being the form of the rutellum
which is slightly longer and turns away from the aperture more sharply (see Text-fig. 14). The
populations from the Lake District are sparse and only the population from Hodgson How Quarry
shows significant differences from the North Wales material with more slender stipes and shorter
sicula.

The specimens from the Toyen Shale formation are also closely comparable, but with a slightly
more slender sicula, and a less prominent basal free length and rutellum (see Text-fig. 14). These
features make them similar to A. suecicus from which it is distinguished by the presence of a
rutellum. The population may be considered transient. Fortey and Owens (1987) suggested
specimens described as A. suecicus by Mu et al. ( 1979) are A. eivionicus , but the description indicates
adpressed growth of th l1 and sicula and therefore they may be excluded from either of these
species.

A. eivionicus is distinguished from A. hicksii by its narrower stipe width and smaller thecal
separation, and from A. suecicus by the presence by a rutellum. Distinction from A. lapworthi is
more subtle, and is mainly dependent on the basal free length of the sicula which only rarely exceeds
02 mm in A. eivionicus , and then only slightly (compare Text-figs 9 and 14). Sicular length ranges
above that seen in A. lapworthi , particularly in the Bangor populations (see Text-fig 9).

Azygograptus suecicus Moberg, 1892
PI. I fig. 5; Text-fig. 18a-e, p

1892 Azygograptus suecicus Moberg, p. 342, pi. 8 figs l and 2.

1904 Azygograptus suecicus Moberg; Tornquist, p. 27, pl. 4 figs 6-11.

Diagnosis. Azygograptus with very short basal free length to sicula ( < 0-2 mm); no rutellum, and no
adpressed growth of th 1 1 and sicula.

Type material. The specimen figured by Moberg (1892, pl. 8, fig. 1) was selected as the lectotype by Boucek
(1973). However Boucek (1973, p. 107) stated that the type specimens were in the collection of the Geological
Institute of Lund, whilst they are in fact held in the type collection of the Sveriges Geologiska Undersokning.
There is therefore some doubt that Boucek examined the type specimens, and unfortunately the lectotype he
selected is poorly preserved.

Type locality. Flagabro-Killerod, Scania from just below Komstad Limestone. A translation of the original
description of the locality was given by Lindholm (1985, pp. 39-41). The species occurs in the uppermost part
of the Toyen Shale together with Isograptus gibberulus (sensu Moberg, 1892), Maeandrograptus schmalenseei,
Didymograptus mobergi , Expansograptus hirundo , E. cf. abditus , Pseudotrigonograptus ensiformis , Kinnegraptus
kinnekullensis (sensu Williams and Stevens 1988, non Skoglund 1961). The overlying Komstad limestone
belongs to the Megistcispis ( Megistaspis ) simon trilobite zone (Nielsen 1985). The graptolites indicate the I. v.
lunatus or I. v. victoriae biozone of the Australian zonation and the Scandinavian D. hirundo biozone.

Other occurrences. Though the species has been listed from a number of localities in Scandinavia the majority
are probably A. e/lesi. Only one occurrence outside the type locality has been reliably identified. This is at
Graptolittdalen, Slemmestad, Oslo Region (PMO 118-570), and is a single slab with specimens in the silvery
carbonaceous preservation typical of the area. Indirect evidence which supports the occurrence is that
Isograptus gibberulus sensu Moberg is also present at Slemmestad (Spjeldnaes 1953, pl. I , fig. 5) and is also only
known from the Moberg locality in Scania.

Description. Sicula 11—1-3 mm long and curves away from the stipe at the aperture which is 0-25-0-3 mm wide.

910

PALAEONTOLOGY, VOLUME 34

text-fig. 18 a-e, p. Azygograptus suecicus Moberg, t'rom Killerod, Scania; A, SGU 5247, lectotype, x5; B,
SGU 5248, paratype, x5; c, SGU 7895, drawing in latex, x 5; D, SGU 7896, proximal part in relief, distal
part preserved as mould, x 5; e, SGU 5247, same slab as holotype, preserved as mould, x 5; p, SGU 7895,
detail of sicula, x 20. f-m, A. eivionicus Elies, from Toyen Section, Oslo, 1 9-43—1 9-55 nr interval; f, PMO
118.608, x 5; G, PMO 118.603-2, x 5; h, PMO 1 18.604, x 5; i, PMO 1 18.603-1, detail of sicula, x20;k,
PMO 1 18.604, detail of sicula x20;l, PMO 1 18 . 603-1, x 5 ; m, PMO 1 18 . 605, x 5. n-o, A. eivionicus Elies,
from Toyen Section, Oslo, 1919-29 m interval ; n, PMO 118.602, x 5;o, PMO 118.602, x 20.

Sicular aperture is symmetrical without rutellum. Stipe originates low on the sicula and diverges abruptly at
an angle of 140-150°, leaving a basal free length to the sicula of 01-0-2 mm. Stipe moderately flexed with up
to 35° rotation by th 5. Thecae are slender, with stipe width at aperture of first theca 0-45-0-5 mm; distal stipe
width 0-7-0-8 mm. Thecal separation 1 -4— 1 -7 mm (6-7 th/10 mm), but the first theca appears somewhat longer.
Thecal inclination 14—18°.

Discussion. A. suecicus is easily distinguished from other members of the genus by the simple form
of the sicular aperture (see Text-fig. 14) associated with the lack of adpressed growth of th l1 and
the sicula.

Azygograptus validus Tornquist, 1904
Plate I, fig. 13; Text-fig. 19a-l

1904 Azygograptus validus Moberg MS.; Tornquist, p. 27, pi. 4, figs 12-14.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

911

Diagnosis. Azygograptus with higher thecal inclination (approximately 25°) than other species.
Obvious adpressed growth of sicula and th l1 with origin of th l1 very high on the sicula.

Type material. LO I755T, LO 1756t; additional topotype SGU 7555-7557, 7889-7892.

Type locality. Diabasbrottet, Hunneberg Mountain (previously called Mossebo) (see Lindholm and Maletz
1989). Moderately abundant specimens, predominantly occurring as silvery organic films on black shale.
Topotype specimens came from the interval 4-74-4-89 m of the section in Erdtmann et at. (1987, p. 112).
Associated with T. phyllograptoides.

Other occurrences.

1. Krapperup boring in Scania (Lindholm 1981); 6 specimens at the 96-45 m level, together with D.
undulatus in the T. phyllograptoides biozone.

2. Robinson Mountain, Lake District (locality A I of Text-fig. 3); solitary, very poorly preserved specimen
in medium grey mudstone (BGS Rx 1273-Rx 1274); associated with Acrograptus filiformis.

3. Toyen Shale, Toyen Underground Station, City of Oslo, Norway; single, poorly preserved specimen
(PMO 118-571) from the interval 7-65-7-85 m (Text-fig. 2), associated with T. phyllograptoides.

Description. Sicula long and slender; 2 0-2-3 mm long; 0 3-0-4 mm wide at aperture. A short nema is sometimes
present. Th l1 originates very high on the sicula, possibly from the prosicula, though this is irresolvable from
available material. The stipe turns away at about 120-150° leaving 0-3-0-4 mm basal free length to sicula.
Thecae slender, slowly expanding; overlap about one-half. Thecal inclination 24-26°. Stipe characters uniform
with largest specimen reaching th 14. : thecal density 11-12 th/10 mm; stipe width 0-5-0-8 mm. Significant stipe
curvature, stipe becoming reclined in some larger specimens.

Discussion. Tornquist (1904), in describing a few small specimens from Mossebo, remarked on the
strong resemblance to Azygograptus coelebs. A. validus differs from the published descriptions of A.
coelebs (see below) in possessing a much longer sicula and shorter thecae. Thecal inclination serves
to distinguish it from other species.

Azygograptus ellesi Monsen, 1937
Plate 1, figs 1 and 11; Text-fig. 20a-s
1898 Azygograptus suecicus Moberg; Elies p. 514, fig. 29.

1902 Azygograptus suecicus Moberg; Elies and Wood p. 95, text-fig. 56; pi. 3, fig, 3 a-b.

1937 Azygograptus ellesi Monsen, p. 205, pi. 5, figs 21, 28, 730.

1937 Azygograptus cf. suecicus Moberg; Monsen, p. 204, pi. 5, figs 2 and 20.

1973 Azygograptus suecicus Moberg; Boucek, p. 107, pi. 19, figs 1^4; text-fig. 33 a-g.

Emended diagnosis. Azygograptus with long sicula (1 -8-2-4 mm); origin of th 1 1 in distal half of
sicula and adpressed growth of th 1 1 and sicula till close to similar aperture. Little overlap in
proximal thecae.

Type material. PMO K 0288. Slab contains two specimens figured by Monsen (1937, pi. 5, figs 21 and 28) both
of which were described as a ‘holotype’ in the plate explanation. The specimen represented in plate 5 figure
21 is indicated as the holotype on the label for the slab and is here adopted as the holotype; the other specimen
is certainly conspecific and is regarded as a paratype. The specimen figured on plate 5, figure 30 is a distal
fragment on a slab from the P. angustifolius elongatus biozone at Ensjo, Oslo region. Though there are several
specimens of A. ellesi on the slab, the figured specimen is only a distal fragment and thecal spacing and
inclination suggest it may not belong to A. ellesi.

Type locality. Toyen shale at Slemmestad, Oslo Region. The specimens come from the P. angustifolius
elongatus biozone and are associated with Phyllograptus s.s , Didymograptus ‘ Iprotoindentus , Expansogratus sp.
and Tetragraptus cf. serra.

912

PALAEONTOLOGY, VOLUME 34

text-fig. 19. A-D, f-h, Azygograplus validus Tornquist from Diabasbrottet, Hunneberg Mountain,
Vastergotland, Sweden; a, SGU 7890B, 4-83 m level, x 5; b, SGU 7556, 4-89 m level, x 5; c, SGU 7557-2,
4-74 m level, x 5; d, SGU 7557-1, drawing of mould, 4-74 m level, x 5; f, SGU 7892, 4-79 m level, x 10; G,
SGU 7889B-2, 4-79 m level, x 10; h, SGU 7889B-1. 4-79 m level, x 10. E, ?A. validus , BGS 1273, Robinson
Mountain, Lake District, England, x 5. I, K, A. validus from Mossebo (now Diabasbrottet), Hunneberg
Mountain, Vastergotland, Sweden; i, paratype, LO 1756/, x 10; k, holotype, LO 1755T, x 10. l, A. validus ,
PMO 118.571, 7-65-7-85 m level, Toyen Section, Oslo, Norway, x 10. m-u, A. minutus sp. nov. from
Diabasbrottet, Hunneberg Mountain, Vastergotland, Sweden, x 10. M, SGU 7564, paratype, 709 m level; N,
SGU 7593, paratype, 7-09 m level; o, SGU 7565, paratype, 7 09 m level; p, SGU 7560, holotype, 7-09 m level;
Q, SGU 7562, paratype, 7-09 m level ; r, SGU 7559, paratype, 7-09 m level ; s, SGU 7893, paratype, 7 09 m level;
t, SGU 7894, 8-29 m level; u, SGU 7561, paratype, 7-09 m level; Remarks: Diabasbrottet section as figured

in Erdtmann et al. (1987).

BECKLY AND MALETZ: ORDOVICIAN GR APTOLITES

913

<!v

text-fig. 20. Azygograptus ellesi Monsen. a-d. Barf, Lake District, England ; a, BM H948, x 5 ; b, BM P7216,
x 5; c-d, BM P2010, two specimens on same slab, x 20. e— G, Slemmestad, Norway; E, PMO K 0187, paratype,
x 5; f, PMO K 0288, paratype x 5; G, PMO K 0288 holotype, x 5. h, m-n, p-r, from 1 6-25— 1 6-42 m level in
Toyen Section, Oslo, Norway ; h, PMO 118. 578, x 5 ; M, PMO 118. 575, x 5 ; N, PMO 118. 576, x 5 ; p, PMO
118.576, x 20; Q, PMO 1 18.579, x 5; R, PMO 118.959, x 5. i, o, SGU 7554, drawing of mould, Nipan,
Jemtland; i, x5; o, x20. k, PMO K 0658, figured as A. cf. suecicus by Monsen (1937, pi. 5, fig. 20),
Stensbergstrafe, Oslo, Norway, x 5. L, s, Ensjo, Norway, from slab with stipe fragment of A. ellesi figured by

Monsen (1937, pi. 5, fig. 30), PMO K 0657, x 5.

Other occurrences.

1. Toyen Underground Station, City of Oslo, Norway; several specimens from the section described by
Erdtmann (1965). The specimens from the intervals 1 5-35—1 5-95 m and 1 6-25—1 6-42 m, within the upper part
of the P. a. elongatus biozone, and are associated with Pseudophyllograptus augustifolius , Tetragraptus bigsbyi
askerensis , Expansograptus grandis and Acrograptus sp.

2. Nipan, Jemtland; single relief specimen (SGU 7554) on a loose slab with no associated fauna; a nearby
outcrop has yield a fauna indicating the P. densus biozone, or higher parts of the Toyen Shale.

3. Barf, English Lake District (NY 217267) (locality A6 of Text-fig. 3); fairly numerous specimens preserved
in partial relief as limomtic internal moulds in monospecific assemblages or associated with Pseudo-
phyllograptus, but the locality has yielded a very diverse fauna indicative of the D. nitidus biozone. BM
P7216, P2010, G948.

Description. Sicula straight and slender, 1 -8-2-4 mm long; 0 25 mm wide in relief specimens, a little wider
(0-3 mm) in flattened specimens and with both dorsal and ventral processes on the aperture. Th l1 originates

914

PALAEONTOLOGY, VOLUME 34

at 0-6-0-75 of the sicular length and grows down along sicula until OT-O-2 mm above the sicular aperture.
Declined stipe very slender, generally straight but may be slightly curved in some specimens. Stipe width
0-45-0-7 mm and is nearly constant throughout rhabdosome, with the only increase over the first couple of
thecae. Thecal overlap one-quarter; thecal inclination 6-10°. Sicular aperture to aperture th 1: 1 -9-2-2 mm,
distal thecal separation: 1 7—1 -5 mm (6-5— 8-5 th/10 mm).

Discussion. The only measurements that can be made on the holotype are the length of th 1 (2-2 mm)
and stipe width at the aperture of th 1 (045 mm). Both are consistent with the description given
above.

Until now the majority of azygograptids from Scandinavia have been identified as A. suecicus , but
the origin of th I apparent from the relief material shows this to be incorrect. The specimens
described as A. cf. suecicus by Monsen (1937) also clearly belong to A. ellesi and one specimen is
refigured here (Text-fig. 20k).

The proximal development (see Text-fig. 20c, D) and stipe width (045-0-8 mm) of the specimens
from Barf, English Lake District, are comparable to A. ellesi , but the sicula is somewhat shorter
( 1 -2—1-5 mm). Thecal density is also slightly higher in the distal part of the stipe (7-9 th/10 mm) of
the British specimens, which appears to result from the slightly greater thecal overlap. The
specimens figured as A. suecicus by Elies and Wood (1902. pi. 13, fig. 3 a-b) are also from Barf, for
which they note the adpressed growth of th l1 and the sicula.

The specimens described as A. suecicus by Boucek (1973) from the Tetragraptus reclinatus
abbreviatus zone of Bohemia, possess a relatively long sicula and are described as having downward
growth of the initial part of th 1. The proximal development cannot be discerned on the photos, but
the specimens are tentatively included in A. ellesi.

Azygograptus minutus sp. nov.

Plate 1, fig. 12; Text-fig. 19m-u

Origin of name. In recognition of the small size of the species.

Diagnosis. Azygograptus with origin of th l1 high on sicula and adpressed growth until close to
sicular aperture associated with low thecal inclination and high thecal density (9-10 th/mm).

Type material. Holotype: SGU 7560. Paratypes: SGU 7558-7559; 7561-7565.

Type locality. Vastergdtland, north-eastern edge of Hunncberg Mountain, Diabasbrottet. A few specimens in
association with Paradelograptus kinnegraptoides, Expansograptus vacillans and E. holmi on a single bedding
plane in the Didymograptus balticus biozone, at 7-09 m in the section of Erdtmann et al. (1987, p. 1 12).

Other occurrences. None known.

Description. Slender and small species with longest specimen only 6 mm long which reaches th 7. All specimens
may therefore be immature. Sicula 1-6-1 -9 mm long with apertural width 0-25-0-3 mm. Th 1 originates high on
the sicula and grows down adpressed for the majority of the sicular length before flexing away at an angle of
120-150°. The basal free length to the sicula is 0-2-04 mm and the sicular aperture has symmetrical, small,
dorsal and ventral lips. Thecae slender and long, thecal inclination 10-15° and thecal overlap about a half.
Thecal density 9-10 th/10 mm. Stipe is slightly flexed and reaches a maximum width of 0-6 mm; no variation
along stipe has been recognized.

Discussion. A. minutus is characterized by its high thecal density. The only species with which it is
likely to be confused is A. validus which also has a high origin of th l1. The main differences are
the lower inclination of thecae in A. minutus and shorter sicula.

BECK.LY AND MALETZ: ORDOVICIAN GRAPTOLITES

915

Other species which have been placed in Azygograptus are as follows:

1. A. canadensis Ruedeman, 1947, p. 357, pi. 58, figs 9-11. Glenogle Shale, Glenogle, Canada. Belongs to
Pseudazygograp tus .

2. A. coelebs Lapworth, 1880, p. 159, pi. 5 fig. 16 a-c. This species has only been described from the Skiddaw
Slate Group of the Cross Fell Inlier, Northern England, though it has been identified from Algeria (Whitman
1971). The original description was based on material from Ellergil, and the species has been recorded from
Milburn Beck (Shotton 1935). The associated fauna, which includes biserial and pendent graptolites, indicates
a Llanvirn age. The validity of the species is in doubt because only the two specimens (SM A 180 1 6 and BM
1027) figured by Elies and Wood (1902, pi. 13 fig. 4 a-b) appear to exist and neither is adequate to characterize
the species or confirm the generic placement.

3. A. flexilis Chen and Xia, in Mu et al. 1979, pp 1 10-1 11, pi. 38, figs 13 and 14. Described from the A. suecicus
biozone of Southwest China. The description states that th 1 grows down beside the sicula and on this feature
it may be grouped with A. ellesi. It differs from A. ellesi in having a shorter sicula ( 1-3 vs 1 -8-2-4 mm). No
description is given here because the original material has not been examined. The only reason this may not
be a valid species is synonymy with another Chinese species.

4. A. fluitans Geh, in Mu et al. 1979, p. 112, pi. 39, figs 1-5. Described from the A. suecicus biozone of
Southwest China. No material of this species has been examined but it may be synonymous with A. undulatus
(Chen Xu, pers. comm.).

5. A. groenwalli Monsen, 1937, p. 206, pi. 5, fig. 29. From the Pseudophyllograptus densus bio^one, Ensjo, Oslo
Region, Norway. Only known from the type specimen which is almost certainly a broken specimen of the co-
occurring Didymograptus cf. minutus. The species is therefore considered a junior synonym of D. cf. minutus.

6. A. hexianensis Jiao, 1984, p. 623, text-fig. 2; pi. 2, figs 31 and 32. A. suecicus biozone of Hexian, Anhui. This
species is difficult to interpret from illustrations but appears to belong to group 1. Occurs with A. undulatus ,
A. suecicus and A. lapworthi.

7. A. novozealandicus Skwarko, 1962, p. 224, text-fig. 4; fig. 6. South side of Peel Range, northwest Nelson,
New Zealand, Lower Gisbornian. From the drawing alone this may be a fragment of a dichograptid, though
the description states every specimen has a long nema implying a proximal end is present. However, positive
identification with Azygograptus is not possible.

8. A. (?) oelandicus Bulman, 1936, pp. 46-48, pi. 2, figs 16-29; text-fig. 17. An indeterminable fragment; sicula
unknown. The species was transferred to the Dendroidea as Parvitubus oelandicus by Skevington (1963, pp.
47-51, figs 62-72).

9. A. prolixus Keble and Benson, 1929, fig. 14. Northwest Nelson, New Zealand, Lower Aorere Series,
Douglas zone, Cobb subzone, ?upper Arenig-Llanvirn.

10. A. saltaensis Loss, 1949. From the drawing this may be a fragment of an expansograptid, but the original
material would need to be examined to confirm this.

11. A. (?) simplex Ruedemann. 1908, pp. 258-260, text-figs 163 171 ; pi. 14, fig. 10. A very curious form with
a huge sicula and only one associated theca. Certainly not closely related to Azygograptus.

12. A. undulatus Chen and Xia, in Wang 1974, p. 157, pi. 68. fig. 4. This species has only been recognized from
the A. suecicus biozone in Southwest China. It displays adpressed growth of th 1 and has a high origin on sicula
suggesting comparison with A. validus or A. minutus. Probably a valid species. The sicular length is less than
in A. validus or A. minutus , but the taxonomic importance of pro-thecal folds is uncertain. A. fluitans and A.
undulatus spinosus may be junior synonyms of A. undulatus (Chen Xu, pers. comm.). This species has also been
described from northern Spain (Gutierrez Marco and Rodriguez 1987).

13. A. undulatus spinosus Chen, in Mu et al. 1979, p. 112, pi. 38, figs 24 and 25. Described from the A. suecicus
biozone of Southwest China. Probably synonymous with A. undulatus (Chen Xu, pers. comm.).

14. A. (?) walcotti Gurley, 1896, pp. 69, 92. Lower Dicellograptus biozone, Stockport, New York, USA. The
species has never been redescribed but was figured by Ruedeman (1947, pi. 58, fig. 8). Its relationships remain
unclear, but it might belong to Pseudazygograptus.

916

PALAEONTOLOGY, VOLUME 34

Genus jishougraptus Geh, 1988
Type species. Jishougraptus mui Geh, 1988.

Emended diagnosis. Single deflexed, uniserial stipe, dorsal margin undulate but without folds, thecae
long and thin, tending to leptograptid type but not forming bundles. Thecae simple to elaborate.

Discussion. The only change made from the original diagnosis is the inclusion of simple thecae as
well as elaborate. Neither of the species described below show the ventral geniculation or elaborate
thecal aperture of the type species. The only difference from Azygograptus is the elongate, thin form
of the thecae with greater overlap.

Jishougraptus novus sp. nov.

Plate 1, figs 6, 9, 10; Text-fig. 21a-n

Origin of name. Novus = new, in reference to the fact that this is the first species to show sigmoidally curved
thecae.

Diagnosis. Jishougraptus with simple cylindrical thecae, and elongate sicula ( 1 -8— 2-4 mm).

Type material. Holotype: PMO 118-585, preserved as a mould in dark shale. Paratypes: PMO 118-580-584,
118-586-594.

Type locality. Toyen Underground station. City of Oslo, Norway. Abundant specimens in the interval
16-15-16-25 m in the Toyen section (see Text-fig. 2). The specimens are associated with Tetragraptus bigsbyi
divergens , Pseudophyllograptus augustifolius, Expansograptus extensus linearis , E. decens, F. grandis and
Acrograptus sp. Most specimens are preserved in relief, with only a few preserved as flattened films.

Other occurrences. Slemmestad, section 2b of Spjeldnaes (1985). PMO 1 17-553, 555. A few poor specimens
collected by B.-D. Erdtmann on an excursion in 1988. The species occurs with Oslograptus peculiaris and T.
bigsbyi divergens.

Description. Sicula l-8-2-4mm long and very slender. Short nema, up to 0-4 mm long, sometimes detectable
but apex of sicula often poorly defined. Th l1 originates at half of sicular length and is adpressed until
0-1-0-15 mm above sicular aperture, where it diverges at 140-160°. Thecae are simple, slender and cylindrical
with slight sigmoidal curvature. Prothecal folds not developed but distinct 'ridges’ present at origins of thecae
in relief specimens. Thecal separation 1-5-2-3 mm (4- 5-6-5 th/10 mm) and relatively constant along
rhabdosome. Stipe width at th 1 0-35-0-5 mm; distally 0-7-0-8 mm.

Discussion. J. novus is very similar to J. lindholmae and distal stipe fragments of the two species are
virtually indistinguishable. However proximal development does show significant differences: J.
novus has a sicula almost twice as long on average and much shorter th l1.

Jishougraptus lindholmae sp. nov.

Text-fig. 21p-y

1981 Azygograptus n. sp., Lindholm, p. 15, fig. 6a— b.

Origin of name. Named after Miss Kristina Lindholm (Lund University) who first recognized and described
the species, though she did not name it.

Diagnosis. Jishougraptus with small sicula (10-1 -5 mm) and simple, cylindrical thecae.

Type material. Holotype: PMO 118-595; flattened specimen in dark shale. Slab is crowded with other

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

917

text-fig. 21. Jishougraptus novus sp. nov. and J. lindholmae sp. nov. from the Toyen Section, Oslo, Norway.
a— o, J. novus , c is the holotype, the rest are paratypes, from 1615 16-25 m interval, Toyen Section; a, PMO
118.582-2, x 10; b, PMO fl 8 . 588, x 10; c, PMO 1 18.585, x 5; d, PMO 1 18.592-1, x 5; E, PMO 118.590,
x 5; f, PMO 1 18.589, x 5;g, PMO 118.593-2, x 5;h,PMO 1 18.582-1, x 5;i, PMO 118.581, x5;k, PMO
118.592-2, x 5; l, PMO 118.583, x 5; m, PMO 118.583, x 5; n, PMO 1 18.580, x 5; o, PMO 118.592-2,
x 10. p y, J. lindholmae , q is the holotype, the rest are paratypes, from 16-81 -1710 m interval, Toyen Section;
p, PMO 118.599-3, x 10; q, PMO 118.595, x 10; r, PMO 118.600, x 10; s, PMO 118.597, x 5; t, PMO
1 18.599, x 5; u, PMO 1 18.596, x 5; v, PMO 118.597, x 10; w, PMO 118.600, x 10; x, PMO 1 18.600, x 5;

y, PMO 1 18.596, x 5.

918

PALAEONTOLOGY, VOLUME 34

fragments of the same species. Paratypes: PMO 1 18-596-600; mainly flattened specimens in dark shale. A few
incomplete specimens are in very low relief.

Type locality. Toyen Underground Station, City of Oslo, Norway, from the Toyen Shale (16-81—17-10 m)
within the P. augustifolius elongatus biozone (see Text-fig. 2). Abundant fragmentary specimens on slabs of
black mudstone. Usually preserved as flattened films of silvery periderm, but a few specimens are preserved in
low relief. The associated fauna comprises P. angustifolius elongatus , Expansograptus sp., Tetragraptus
isograptoides , T. cf. serra , T. reclinatus, Pendeograptus cf. pendens , Janograptus sp., Acrograptus gracilis and
‘ ITrichograptus .

Other occurrence. Krapperup Core in Scania (Lindholm 1981), from the zone of P. augustifolius elongatus.

Description. Small, slender sicula: 10-1-5 mm long; 0-25 mm wide at aperture. Short nema occasionally present
to 0-8 mm long, but sicula apex narrow and poorly differentiated from nema. Near apex of sicula distinct
corrugations apparent, probably indicating prosicula. Th 1 1 originates mid-way along the sicula and grows
down adpressed to it. Thecae are very long, slender and slightly sigmoidally curved. Thecal overlap about one-
half proximally and increases slowly distally. Thecal spacing: more than 3 mm proximally, 1-4— 1-6 mm
(6-7 th/mm) distally; thecal inclination: 5-8°. The stipe width increases slowly from 0-3-0-4 mm proximally to
0-6-0-8 mm at about th 7.

Discussion. Stipe characters distinguish J. lindholmae and J. novus from other species but do not
allow distinction between them. This is only possible on features of proximal development; J.
lindholmae has a sicula only half as long on average, and longer th 1 than J. novus.

The material from the Krapperup core in Scania (Lindholm 1981) is closely comparable to that
described above but the description of the former would suggest some slight differences.
Constrictions at the level of the aperture of the preceeding theca and the development of true
prothecal folds are not seen in the Toyen specimens but such features could be the product of
preservation. The prothecal folds in figure 6b of Lindholm (1981) are weakly developed and are
thought to be induced by the origin of the next theca: at the point of origin of a theca a small ridge
is usually visible and might be over-interpreted as a prothecal fold. In flattened material there is no
evidence of prothecal folds (Lindholm 1981, fig. 6b). Constrictions of the thecal apertures could be
the product of differential compaction of the apertures in otherwise relief material.

BIOSTRATI GRAPHIC USE

Azygograptus may be considered a good biostratigraphic indicator of the Arenig Series since species
described from the Llanvirn (e.g. A. coelebs) cannot be confirmed as belonging to the genus, if valid
species. However, its use in the biostratigraphic subdivision of the Series has proved more limited.
Elies (1922) proposed a lineage from A. lapworthi through A. eivionicus to A. suecicus, but provided
little evidence to support this conclusion. In China, a zone of the Ningkuoan has locally been
referred to as the A. suecicus biozone. However, this appears to reflect an abundance of the genus
at the given level rather than a limited range for the named species. Due to the difficulty in reliably
identifying species from descriptions and illustrations alone, the following discussion concentrates
on the evidence from Britain and Scandinavia. Unless otherwise stated, the biozone refers to the
local stratigraphy from which the specimens under discussion have come (cf. difference in British
and Scandinavian hirundo biozone (Text-fig. 22)).

The major problem in using Azygograptus for biostratigraphy is that it displays a number of
characters which are undesirable for this purpose. The absence of Azygograptus from the Pacific
‘province’ precludes its direct use for international correlation, and the nature of its distribution
suggests at least the possibility of geographically isolated populations even within the Atlantic
‘province’. However, it is its facies controlled distribution, which often causes the genus to be
isolated from other biostratigraphic control, that makes the possibility of using the genus for
correlation desirable.

BECKLY AND MALETZ: ORDOVICIAN GR APTOLITES

919

ENGLAND + WALES

1/1 “ <D

3t..^ g|

SCANDINAVIA

O (/>

\L w

o

eivion
. lapwo
hicksi
suecic
. validu
e lie s i
mmut
lindhc
novus

o o

< N
CL

<b

O

CO

Q

-S?

ci

23

C

3

<<<<<<<->->

CO

UJ

Z

o

Ya 3

-c

<c

oc

LU

Yq2

z

<

c

o

Q

O

o

H

CO

UJ

Yal

z

z

LU

LL

-c

10

3

k-

03

CO

,3

3

I

1—

cr

o

Ca3

CO

e

Qj

-Q

Hodgson

• •

Slemmestad/

O

£

-o'

0

01

<3

CO

How Quarry

•

•

Killerod

T0yen

19.19-19.75

Q

hirundo:

i

i

Ca2

10

-Q

^ Whitesand
Bay

Afon

•

•

•

• ••

CO

3

a

c

CO

3

a

Co

c

o

Cal

Z

<

Q

<3

Ol

CO

3

m | Ffinnant
l-Barf

Teryen

16.15-16.25

1

8

!q

Qj

CO

3

1

Ch2

Z

<

_l

i—

23

-*w

C

cj

3

.23

-Q

Q

CO

3

Co

c:

D

Chi

I

£

N

Afon

Ffinnant

•

CL

Be 4

^cT

O

-*— ? Bangor

•

co

U;

Maen

Gwenony

•

•

?Nipan — ►

3

CO

Q

o

co

3

CO

e

QJ

Be 3

z

<

10

o

o

o

t

CO

3

N

cu

CL

23

CL

Be 2

z

$

"S

•

Diabas- —
brottet

”"i"

*

O

D

baltic.

Bel

3

o

"c

Q

D.

validus

cc

o

?

Robinson 2
Mt.

Tieiyen

7 65-7 85 mj*-
Hunneberg

CL

CL

O

T

appr

2

cu

CO

5

•

T, phyllo-
graptoides

La 3

text-fig. 22. Stratigraphic distribution of Azygograptus and Jishougraptus in Britain and Scandinavia with a

correlation between the two areas.

920

PALAEONTOLOGY, VOLUME 34

Objective evidence for the succession of species in Azygograptus is very limited. The Diabasrottet
section on the Hunneberg Mountain has A. minutes overlying A. validus, and the Toyen section has
yielded three species of Azygograptus and two of Jishougraptus but these come from a very small
proportion of the overall succession. The biostratigraphic use of the genus is therefore largely
dependent on ranges inferred by independent correlation of isolated occurrences (see Text-fig. 22).

The oldest species identified in the Toyen section is A. validus and here, as elsewhere in
Scandinavia, it is only known from the T. phyllograptoides biozone. The overlying Scandinavian
biozone, D. balticus , has also yielded a single species only, A. minutus , which is unknown outside
its type locality.

Since the international correlation of the base of the Moridunian Stage (cf. Fortey and Owens
1987) is uncertain, the single specimen of A. validus from Robinson Mountain may be the only
representative of Azygograptus in Britain from strata equivalent to the lowest two biozones of the
Arenig in Scandinavia. The biostratigraphic use of the genus is therefore difficult to evaluate over
this interval and must await evidence from a wider area.

A. ellesi appears to range through the densus (Nipan) and augustifolius elongatus (Slemmestad)
biozones, and the British occurrence at Barf may reasonably be considered to come from the same
time interval, though the difference in sicular length does reduce the confidence in the identification
of the two populations as conspecific. The species of Jishougraptus that occur with A. ellesi at Toyen
are both known only from the P. augustifolius elongatus biozone. The genus, based on the evidence
of the type species, ranges somewhat higher.

Given the uncertainty in the correlation of the British Stages (Moridunian and Whitlandian) with
the Scandinavian succession, the precise ranges of contemporary British species are unclear. A.
eivionicus is associated with a Moridunian trilobite fauna at Arennig (Zalasiewicz 1986) and may
range as high as the gibberulus biozone (Hodgson How Quarry). There is little to constrain
independently the age of the occurrence at Bangor, but it may well be approximately contemporary
with other occurrences in Wales and close to the Moridunian /Whitlandian boundary. A. eivionicus
therefore appears to be a long ranging species of little biostratigraphic potential.

This contrasts with the species to which A. eivionicus may be related on the evidence of transient
populations. A. hicksii is unknown outside South Wales and appears to be restricted to the
Gymnostomix gibbsii biozone (late Whitlandian). A. lapworthi has not been found associated with
any independent age control outside the type locality, where it may be placed in the gibberulus
biozone on the basis of an isograptid associated with D. hirundo. Circumstantial evidence for the
North Wales occurrences is compatible with a Fennian age and at present the species may be
considered restricted to this interval in Britain. However, the material from Bolivia may extend the
range downward. A suecicus in Scandinavia is a similarly rare form and appears to be restricted to
the Scandinavian D. hirundo biozone, and may occur contemporary with A. lapworthi.

In summary, neither Azygograptus , nor the related genus Jishougraptus , have species that provide
useful or reliable correlative indicators between Scandinavia and Britain. Interestingly the species
that do provide some hints of use are A. ellesi and A. validus which have a less restricted biofacies
association. The restriction of all species of Group 1, apart from A. eivionicus , to either Scandinavia
or Britain may indicate that the Tornquist Sea was a partial barrier to migration. For this reason
the British species may prove to have more potential for correlation with other peri-Gonwannan
localities.

EVOLUTION OF AZYGOGRAPTUS

Azygograptus forms part of the adaptive radiation of the Dichograptidae that took place in the
early Ordovician, and a possible origin of the genus is during the upper Tremadoc regression (cf.
Erdtmann 1986; Fortey 1984). However, the evolutionary position of Azygograptus in this radiation
is less clear, particularly with regard to whether the genus represents a mono- or polyphyletic
grouping.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

921

As illustrated in Text-figure 7, two morphological groups can be recognized in Azygograptus on
the presence (Group 2) or absence (Group 1) of adpressed growth of th 1 1 and the sicula.

Fortey and Cooper (1986) proposed a phylogenetic classification for graptoloids on the principle
that the earlier a growth decision in the development of the rhabdosome the greater its evolutionary
and taxonomic significance. Within this classification Azygograptus was taken as separating from
Expansograptus by suppression of dichotomy dl. However the growth decision series proposed
(Fortey and Cooper 1986, table 1 ) has nine decisions which preceed that affecting dichotomy dl, and
the majority of these concern the form of the sicula and its relationship to th l1. The concept of
growth decisions therefore implies a polyphyletic origin for Azygograptus , with suppression of
dichotomy dl having occurred in at least two separate lineages.

An alternative way of examining the origin of Azygograptus is to consider the mechanisms by
which the genus may have originated and the ancestors indicated. Skevington (1966) suggested that
a dimorphic population with both biramous and uniramous forms would provide a suitable
ancestor for Azygograptus. Support for this model is provided by a population of Kiaerograptus
(Rushton 1981) which does show variation in the number of stipes. However in this population, as
in the single, questionable specimen of Par azygograptus, the first theca of a second stipe is present,
probable th I b

The presence of a vestigal stipe can be explained by failure of th l2 to be dicalycal in an ancestor
with isograptid development. The origin of the second stipe in isograptid development is therefore
not strictly the result of a dicalycal theca, but of the opposite direction of growth of th l2 and th 1 1 .
Similarly, from a dicalycal theca the new thecae grow in opposite directions. The azygograptid
morphology can therefore be derived from an ancestor with either isograptid or artus development
by the loss of the ability for a new theca to develop in the opposite direction to that from which it
buds, and this automatically causes failure of dichotomy dl.

From the arguments presented earlier on the ecology of Azygograptus the more restricted ecology
of the species in group 1 may indicate that they are better adapted to an /--selective environment.
The selective pressures of such an environment are ideally suited to the formation of isolated
marginal populations that may be founded on only a few individuals. Since all the species of group
I grade into A. eivionicus to a greater or lesser extent, they may almost be regarded as end members
in an extremely variable species, and in this respect the differences between the species are much less
marked than in group 2. Therefore, despite the apparently greater number of species in group 1 the
evolutionary pool represented by group 2 is probably much greater.

A possibly more important aspect of the peculiar ecology of the species in group 1 is that
adaptation to an /'-selective environment could have affected proximal development. Such an
environment favours rapid development and the morphology of azygograptids can be explained as
a response to this pressure in that is maximizes stipe length for a given number of budding events
(see Text-fig. 23):

oceanic

text-fig. 23. Comparison of rhabdosome form for the same number of budding events between
Pseudisograptus (a) and Azygograptus (b). 1 1 thecae present in each.

922

PALAEONTOLOGY, VOLUME 34

1. abrupt metasicular origin of th 1 1 maximizes the stipe length provided by the first theca;

2. suppression of dichotomy dl removes ‘doubling-back’ and overlap;

3. low thecal inclination with limited overlap maximizes stipe length for a given number of
thecae.

An effect of the above is to maximize the separation of thecal apertures and therefore a selective
pressure favouring this could be an explanation for the morphology, but the implications for
proximal development remain the same.

No firm conclusion can therefore be reached on the evolutionary origins of Azygograptus. The
oldest known species is A. validus and an origin from a extensiform didymograptid with isograptid
development is attractive for this species (see Text-fig. 24). Once the single stipe morphology had
originated the other members of the genus could have evolved from it by progressive adaptation to
an /--selective environment. Alternatively, the species of group 1, and possibly A. ellesi , could have
evolved separately from an extensiform with artus development such as D. simulans , with which
they appear to be geographically and ecologically associated. Therefore, the same selective pressures
that may have caused convergence in rhabdosome morphology for a polyphyletic group, could also
have produced differences in proximal development in a monophyletic group.

text-fig. 24. Alternative extensiform ancestors for Azygograptus species.

Acknowledgements. Many people have kindly helped in the preparation of this paper and in making museum
material available. Our particular thanks go to Drs R. A. Fortey, B.-D. Erdtmann and A. W. A. Rushton for
their encouragement and valuable comments for the improvement of the manuscript. A three-year postgraduate
award made by the Trustees of the British Museum (Natural History) is gratefully acknowledged, during which
time the majority of the work by one of the authors (A.J.B.) was done. Part of the work by J. M. was supported
by Deutsche Forschungsgemeinschaft project Er 96/6-1.

REFERENCES

beckly, a. j. 1985. The Arenig Series in North Wales. Unpublished Ph.D. thesis. University of London.
1987. Basin development in North Wales during the Arenig. Geological Journal , 22. 19-30.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

923

1988. The Stratigraphy of the Arenig Series in the Aberdaron to Sarn area. Western Llyn, North Wales.

Geological Journal , 23, 321-337.

1989. A new Arenig trilohite fauna from the Bangor area. North Wales. Bulletin of the British Museum
(Natural History), (Geology), 45, 1-20.

BOUCEK, B. 1973. Lower Ordovician graptolites of Bohemia. Akademia, Publishing House of the Czechoslovak
Academy of Sciences, Prague, 164 pp, 24 pis.

— and pribyl, a. 1953 [for 1951], Taxonomy and phylogeny of some Ordovician graptolites. Bulletin
international de V Academie tcheque des Sciences 1951 , 52, 1-17.

brenchley, p. j. and harper, j. c. 1967. Lower Ordovician shelly and graptolitic faunas from south-eastern
Ireland. Proceedings of the Royal Irish Academy , 65, 385-390.
bulman, o. m. b. 1936. On the graptolites prepared by Holm (7). The graptolite fauna of the Lower Orthoceras
limestone of Halludden, Oland, and its bearing on the evolution of the Lower Ordovician graptolites. Arkiv
for Zoologi, 28A (17), 1-107.

1970. Graptolithina. VI -VI 63. In teichert, c. (ed.). Treatise on invertebrate paleontology. Part V.

Graptolithina , with sections on Enteropneusta and Pterobranchia. Geological Society of America and
University of Kansas Press, Boulder, Colarado and Lawrence, Kansas, xxxii+163 pp.
chen xu and yang daquan, 1987. Early Ordovician graptolite faunal ditferentiation and correlation based on
the material from Yushan, NE Jiangxi. Advances in Science of China, Earth Sciences, 2. 129-139.
eastwood, t., hollingworth, s. e., rose, w. c. c. and trotter, f. m. 1968. Geology of the country around
Cockermouth and Caldbeck. Memoirs of the Geological Survey of Great Britain (England and Wales),
London, x + 298 pp.

elles, G. l. 1898. The graptolite fauna of the Skiddaw Slates. Quarterly Journal of the Geological Society of
London, 54. 463-539.

1922. A new Azygograptus from North Wales. Geological Magazine, 59, 299-301.

— and wood, e. m. r. 1902. A monograph of British graptolites. Part 2. Monograph of the Palaeontographical
Society, 56 (265), i-xxviii + 55-102, pis V-XIII.

erdtmann, B.-D. 1965. Outline stratigraphy of graptolite-bearing 3b (Lower Ordovician) strata in the Oslo
region, Norway. Investigations of a section at Toyen, Oslo. Norsk Geo/ogisk Tiddsskrift, 45, 481-547.

1986. Early Ordovician cycles and their bearing on punctuations in early nematoporoid (planktic)
graptolite evolution. In walliser, o. h. (ed.). Global Bio-events. Lecture Notes in Earth Sciences, 8. 139-152,
Springer Verlag, Heidelberg- Berlin, vii + 422 pp.

— maletz, J. and gutierrez marco, J. c. 1987. The new Early Ordovician (Hunneberg Stage) graptolite
genus Paradelograptus (Kinnegratidae), its phylogeny and biostratigraphy. Palaontologische Zeitschrift, 61,
109-131.

fearnsides, w. g. and davies, w. (1943). The geology of Deudraeth, the country between Traeth Mawr and
Traeth Bach. Merioneth. Quarterly Journal of the Geological Society of London, 99, 247-276.
finney, s. c. 1980. Thamnograptid, dichograptid and abrograptid graptolites from the Middle Ordovician
Athens Shale of Alabama. Journal of Paleontology, 54, 1184-1208.
fortey, R. a. 1984. Global earlier Ordovician transgressions and regressions and their biological implications.
37-50. In bruton, d. L. (ed.). Aspects of the Ordovician System. Palaeontological Contributions from the
University of Oslo, No. 295, Universitetsforlaget, Oslo, 228 pp.

— and cocks, L. R. m. 1986. Faunal criteria in the positioning of some Palaeozoic microcontinents. Journal
of the Geological Society of London, 143, 151-160.

and cooper. R. a. 1986. A phylogenetic classification of the graptoloids. Palaeontology, 29, 631-654.

— and morris, s. f. 1982. The Ordovician trilobite Neseuretus from Saudi Arabia, and the palaeogeography
of the Neseuretus fauna related to Gondwanaland in the earlier Ordovician. Bulletin of the British Museum
(Natural History), (Geology), 36, 63-75.

— and OWENS, r. m. 1978. Early Ordovician (Arenig) stratigraphy and faunas of the Carmarthen district.
South West Wales. Bulletin of the British Museum (Natural History), (Geology), 30, 225-294.

1987. The Arenig Series in South Wales. Bulletin of the British Museum (Natural History),
(Geology), 41. 69-307.

geh mei-ju 1988. New material of Azygograptidae and its significance. Acta Palaeontologica Sinica, 23,
618-628.

Gurley, r. r. 1896. North American graptolites, new species and vertical ranges. Journal of Geology, 4,
63-102, 291-311.

gutierrez marco, j. c. and rodriguez, l. 1987. Discovery of Arenig graptolites in the Rioseco thrust-sheet
(Cantabrian Zone, N of Spain). Cuadero de Laboratorio Xeoloxico de Luxe, 11, 209-220.

924

PALAEONTOLOGY, VOLUME 34

hopkinson, j. and lapworth, c. 1875. Descriptions of the graptolites of the Arenig and Llandeilo rocks of
St. David’s Quarterly Journal of the Geological Society of London, 31. 631-672.
hsu, c and chao, y. t. 1976. The evolution of the family Sinograptidae. Acta Geologica Sinica, 1976,
121-140.

jackson, d. e. 1962. Graptolite zones in the Skiddaw Group in Cumberland, England. Journal of Paleontology,
36. 300-313.

jiao shi-ding 1984. Graptolites from the lower Ordovician Sinianpan Formation of Hexian, Anhui. Acta
Palaeontologica Sinica, 23. 618-628.

keble, r. a. and benson, w. n. 1929. Ordovician graptolites from northwest Nelson. Transactions of the New
Zealand Institute , 49, 840-863.

kozlowski, r. 1954. Sur la structure de certain Dichograptides. Acta Geologica Polonica , 4, 423-444.
lapworth, c. 1873. On an improved classification of the Rhabdophora. Geological Magazine, 10. 500-504,
555-560.

1880. On New British Graptolites. Annals and Magazine of Natural History , 5, 149-177.
lindholm, k. 1981. A preliminary report on the uppermost Tremadocian - lower Middle Arenigian
stratigraphy of the Krapperup 1 drilling core, southern Sweden. Unpublished undergraduate thesis, Lund
University.

— 1985. Field excursion guide to Scania, Sweden. Third International Graptolite Conference, Copenhagen,
1985 , 46 pp.

— and maletz, j. 1989. Intraspecific variation and relationships of some Lower Ordovician species of the
dichograptid Clonograptus. Palaeontology, 32, 711-743.

loss, r. 1949. Sur la presence de schistes graptolitiferes dans l’Ordovicien inferieur de la frontiere Argentine-
Bolivia. Comptes Rendues de la Societe Geologique de France, 7-8, 140-142.
maletz, j. 1987. Biostratigraphie und Graptolithenfauna im Unteren Ordovizium des Hunneberges in
Vastergotland (Westliches Zentralschweden). Unpublished thesis, Institut und Museum fur Geologie und
Palaontologie der Georg-August-Universitat zu Gottingen.
matley, c. a. 1932. The Geology of the Country around Mynydd Rhiw and Sarn, South-western Lleyn,
Carnarvonshire. Quarterly Journal of the Geological Society of London, 88, 238-273.

— 1938. The Geology of the Country around Pwllheli. Llanbedrog and Madryn, South-west Carnarvonshire.
Quarterly Journal of the Geological Society of London, 94, 555-606.

moberg. J. c. 1892. Om nagra nya graptoliter fran Skanes Undre Graptolitskiffer. Geologiska Foreningens i
Stockholm Forhandlingar , 14. 339-350.

monsen, a. 1937. Die Graptolithenfauna im unteren Didymograptusschiefer (Phyllograptusschiefer)
Norwegens. Norsk Geologisk Tiddskrift. 16, 57-266.
mu, a. t. 1950. On the evolution and classification of the graptolites. Geological Review, 15, 171-183.

— geh, m. y., chen, x., ni, Y. n. and lin, Y. k. 1979. Lower Ordovician graptolites of southwest China.
Palaeontologica Sinica. 156B (13), 1-192, pis. 1-48. [In Chinese with English abstract].

— lee, c. k., and geh, m. y. 1960. Ordovician Graptolites from Xinjiang (Sinkiang). Acta Palaeontologica
Sinica, 8, 28-39.

Nicholas, T. c. 1915. The geology of St. Tudwal's Peninsula. Quarterly Journal of the Geological Society of
London, 71, 83-143.

nicholson, h. a. 1875. On a new genus and some new species of graptolites from the Skiddaw Slates. Annals
and Magazine of Natural History, 16, 269-272.

nielsen, a. t. 1985. Stratigraphy, depositional environment and palaeoecology of the Lower Ordovician
Komstad Limestone, southern Scandinavia. Unpublished thesis. Institute of Historical Geology and
Palaeontology, University of Copenhagen.

obut, a. m. and sennikov, n. v. 1984. Graptolites and zonal subdivisions of Lower Ordovician of the Gorny
Altai. 53-105. In kanygin, a. v. (ed.). Lower Ordovician Stratigraphy and Fauna of Mountain Altai. Trudy
Sibirskogo Nauchno-IssledovateFskogo Institute/ Geologii, Geofiziki i Mineral ' nogo Syr'ya, 565, 1-122. [In
Russian],

paskevictus, j. 1976. On the graptolite faunas of the Arenigian and Llanvirnian boundary beds of the Jelgava
depression. Middle East Baltic area. 135-139. In kaljo, d. and koren, t. (eds). Graptolites and stratigraphy.
Academy of Sciences of the Estonian SSR Institut of Geology, Tallinn, 256 pp. [In Russian],
ruedemann, r. 1908. Graptolites of New York. Part II. Graptolites of higher beds. New York State Museum
Memoir, 11, 583 pp., 31 pi.

— 1947. Graptolites of North America. Geological Society of America , Memoir , 19, x + 652 pp.
rushton, a. w. a. 1981. A polymorphic graptolite from concealed Tremadoc rocks of England. Geological

Magazine, 118, 615-622.

BECKLY AND MALETZ: ORDOVICIAN GRAPTOLITES

925

shotton, F. w. 1935. The stratigraphy and tectonics of the Cross Fell Inlier. Quarterly Journal of the Geological
Society of London, 91, 639-700.

skevington, d. 1963. Graptolites from the Ontikan limestones, (Ordovician) of Oland, Sweden. II.
Graptoloidea and Graptovermida. Bulletin of the Geological Institutions of the University of Uppsala, 43,

1-74.

- 1966. The morphology and systematics of ‘ Didymograptus' fasciculatus Nicholson, 1869. Geological
Magazine, 103. 487-497.

- 1973. Ordovician Graptolites. 27-35. In hallam, a. (ed . ). Atlas of palaeobiogeography. Elsevier Scientific
Publishing Company, Amsterdam, 531 pp.

skoglund, R. 1961. Kinnegraptus, a new graptolite genus from the Lower Didymograptus Shale of
Vastergotland, Central Sweden. Bulletin of the Geological Institutions of the University of Uppsala, 40,
389-400.

skwarko, s. k. 1962. Graptolites of Cobb River -Mount Arthur area, north-west Nelson, New Zealand.

Transactions of the Royal Society of New Zealand, (Geology), 1, 215-247.
spjeldnaes, n. 1953. The Middle Ordovician of the Oslo Region, Norway. 3. Graptolites dating the beds below
the Middle Ordovician. Norsk Geologisk Tiddskrift, 31, 171 -184.

- 1985. Excursion guide to Oslo Region. Third International Graptolite Conference, Copenhagen, 1985.
strachan, i. 1971. A synoptic supplement to ‘A Monograph of British Graptolites’ by Miss G. L. Elies and

Miss E. M. R. Wood. Monograph of the Palaeontographical Society, 125 (529), 1 130.
tornquist. s. l. 1901. Researches into the graptolites of the lower zones of the Scanian and Vestrogothian
Phyllo-Tetragraptus beds. Part I. Lunds Universitets Arsskift , 37, 26 pp., 3 pis.

- 1904. Researches into the graptolites of the lower zones of the Scanian Vestrogothian Phyllo-Tetra-
graptus beds. Part II. Lunds Universitets Arsskrift, 40, 1-29.

ulst, r. 1976. Stratigraphic significance of the late Tremadocian and the Arenigian graptolites of the Middle
East Baltic area. 214-221. In kaljo, d. and koren, t. (eds). Graptolites and Stratigraphy. Academy of
Sciences of the Estonian SSR Institute of Geology, Tallinn, 256 pp. [In Russian with English summary],
wang, xiao-feng 1975. On the classification and evolution of Paradidymograptus and Nicholsonograptus. Acta
Geologica Sinica, 1, 6-13. [In Chinese],

wang, yi kang (ed.) 1974. A handbook of the stratigraphy and palaeontology in Southwest China. Nanjing
Institute of Geology and Palaeontology, Academia Sinica Science Press, Peking, 454 pp. [In Chinese].
williams, s. h. and stevens, r. k. 1988. Early Ordovician (Arenig) graptolites of the Cow Head Group, western
Newfoundland, Canada. Palaeontographica Canadiana , 5, 1-167.
zalasiewicz, j. a. 1984. Dichograptid synrhabdosomes from the Arenig of Britain. Palaeontology , 27,
425-429.

1986. Graptolites from the type Arenig Series. Geological Magazine, 123, 537-544.

ANDY BECKLY

Department of Palaeontology
British Museum (Natural History)
London SW7 5BD. UK

Present Address:

BP Exploration
301 St Vincent Street
Glasgow G2 5DD, UK

JORG MALETZ

Institute of Geology and Palaeontology
Technical University of Berlin
EB 10, Ernst-Reuter-Platz 1
D-1000 Berlin 10, Germany

Typescript received 6 March 1990

Revised typescript received 15 November 1990

A NEW RHYNCHOSAUR FROM THE UPPER
TRIASSIC OF WEST TEXAS, AND THE
BIOCHRONOLOGY OF LATE TRIASSIC
RHYNCHOSAURS

by ADRIAN P. hunt and SPENCER G. LUCAS

Abstract. A new rhynchosaur (Reptilia, Diapsida), Otischalkia elderae, from the lower Dockum Group of
West Texas is represented by two premaxillary fragments, two femora, and two humeri. The lower Dockum
is early late Carnian ( Paleorhinus biochron) in age. Other Late Triassic rhynchosaurs are: indeterminate
hyperodapedontines from the Popo Agie Formation of Wyoming, Isalo II beds of Madagascar and an
undescribed stratigraphic unit in Tanzania; Scaphonyx from the Wolfville Formation of Nova Scotia, Santa
Maria, and Caturrita Formations of Brazil and the Ischigualasto Formation of Argentina; Hyperodapedon
from the Maleri Formation of India and the Lossiemouth Sandstone Formation of Scotland. These
occurrences of Late Triassic rhynchosaurs are of early late Carnian age. Late Triassic rhynchosaur
distribution is largely controlled by the facies that are preserved at different localities. Rhynchosaurs are
common in more terrestrial faunas that contain few or no phytosaurs, whereas rhynchosaurs are rare in
semiaquatic faunas dominated by phytosaurs.

Rhynchosaurs are an order of primitive archosauromorph reptiles whose fossils are found in
Upper Triassic strata on all continents except Antarctica and Australasia (Benton 1987; Text-fig.
1) as well as in Lower and Middle Triassic strata (Benton 1990). Several authors have mentioned
the presence of rhynchosaurs in the Late Triassic of North America (Baird 1964; Elder 1978;
Chatterjee 1980; Benton 19836, 1987; Murry 1989; Parrish 1 989a, 19896), but none of these
specimens has been adequately described.

In 1976, Sankar Chatterjee identified a fragment of a rhynchosaur premaxilla (TMM 31185-92;
PI. 1, figs 2 and 3) in the collections of the Texas Memorial Museum from the lower Dockum Group
near Otis Chalk, Texas (Elder 1978; Chatterjee 1980). Subsequently, Elder (1978) identified two
humeri and two femora from a nearby locality, and another premaxillary fragment from the same
locality, as pertaining to a rhynchosaur. Long (in Murry 1989) recently suggested that only the
premaxillary fragments pertain to a rhynchosaur and that the postcranial bones represent a large
trilophosaurid.

All but one of the putative rhynchosaur specimens were collected by workers employed by the
Federal Works Progress Administration from 1940-1941 from the lower Dockum Group in
Howard County in West Texas (Elder 1978; Murry 1989).

The putative rhynchosaur specimens include a 65 mm-long left premaxillary fragment (TMM
31185-92) which is very similar to rhynchosaur premaxillae in having the overall shape of a
flattened, tapering cylinder, in having a flat internal margin in its distal half where the two
premaxillae meet (Benton 1983u, p. 618) and in having longitudinal striations along its distal
portion (Benton 1983«, p. 618). Thus, we consider that TMM 31 185-92 is a rhynchosaur premaxilla.
TMM 31 185-93 is a 17 mm-long fragment also from quarry 3A, and it represents a portion of a
premaxillary indistinguishable from TMM 31 185-92.

All the putative rhynchosaur postcranial remains derive from quarry I and include two left
humeri (TMM 31025-263, PI. 1, fig. 1, Text-fig. 2; TMM 31025-262, PI. 1, figs 4 and 5). These

(Palaeontology, Vol. 34, Part 4, 1991, pp. 927-938, 1 pl.|

© The Palaeontological Association

928

PALAEONTOLOGY, VOLUME 34

text-fig. 1. Distribution of Late Triassic rhynchosaurs : 1, Popo Agie Formation, Wyoming; 2, Dockum
Group, Texas; 3, Wolfville Formation, Nova Scotia; 4, Lossiemouth Sandstone Formation, Scotland; 5,
Ischigualasto Formation, Argentina; 6, Santa Maria and Calurrita Formations, Brazil; 7, unnamed unit,
Tanzania; 8, Isalo II Beds, Madagascar; 9, Maleri and Tiki Formations, India.

humeri are very similar to rhynchosaur humeri (e.g. Huene 1929, pis 5, 6, 8; Fluene 1942, pi. 34, fig.
4<7-c; Chatterjee 1974, fig. 22 a-d~, Benton 1983/), fig. 49) in being very robust, with expanded ends
linked by a narrow shaft with a proximal end consisting of a broad posterior plate and a narrow
anteroventral deltopectoral crest, and with a distal articular surface divided into a large anterior
capitellum and a large posterior trochlea. Thus, we consider that these humeri represent a
rhynchosaur.

TMM 31025-266 is a reptilian femur which is very similar to rhynchosaur femora (Huene, 1938,
pi. 10, la-f; Chatterjee 1974, fig. 25 a-b\ Benton 19836, figs 33 b-e and 51 a-f) particularly in being
robust, with moderately expanded ends, and in having a marked intertrochanteric fossa between
two raised ridges on the proximal end, and in lacking a fourth trochanter. Thus, we consider that
TMM 31025-266 represents a rhynchosaur. TMM 31025-264 appears to represent a deformed left
femur of similar morphology.

In conclusion, it is apparent that a rhynchosaur is present in the lower Dockum Group of
Howard County, Texas which is distinct from other Late Triassic rhynchosaurs and represents a

HUNT AND LUCAS: RHYNCHOSAUR FROM TEXAS

929

new taxon. The aims of this paper are: (1) to name and describe the new rhynchosaur from Texas;
(2) to review the age relationships of all Late Triassic rhynchosaurs; and (3) to consider briefly their
palaeoecology. TMM refers to the Texas Memorial Museum, Austin, Texas.

SYSTEMATIC PALAEONTOLOGY

Class reptilia Laurenti, 1768
Subclass diapsida Osborn, 1903
Order rhynchosauria (Gervais, 1859) Osborn, 1903
Family rhynchosauridae Huxley, 1887
Genus otischalkia gen. nov.

Derivation of name. After the abandoned settlement of Otis Chalk, Howard County, Texas (Gregory 1945, fig.

1).

Type and only species. Otischalkia elderae sp. nov.

Type locality. Quarry 1, site 3, near Otis Chalk, Howard County, Texas (Elder 1978, fig. 7a-b).

Horizon. Lower Dockum Group (undivided), late Carnian (late Triassic), Paleorhinus biochron (Hunt and
Lucas 1990).

Diagnosis. As for species.

Otischalkia elderae sp. nov.

Plate 1, figs 1-7; Text-fig. 2a-c

1978 ‘ Dockumensia beckorum ’ Elder (unpublished), p. 66, figs 16a-b, I7a-b; 18a-c; PI. 2, figs 1a-d

and 2a-d, PI. 3, figs 1a-d, 2a-b, 3.

1980 Rhynchosaur, Chatterjee, p. 58.

1986 Rhynchosaur, Murry, p. 120.

1989 Rhynchosaur, Murry, p. 108.

1990 Rhynchosaur, Benton, p. 84.

Derivation of name. For Ruth L. Elder who identified much of this rhynchosaur material.

Holotype. TMM 31025-263; left humerus (PI. 1, fig. 1).

Referred specimens. TMM 31025-262, left humerus (PI. 1, figs 4 and 5); TMM 31025-266, left femur (PI. 1, figs
6 and 7); TMM 31025-264, left femur; TMM 31 185-92, fragment of left premaxilla (PI. 1 , figs 2 and 3); TMM
31185-93, tip of premaxilla.

Diagnosis. Otischalkia differs from all rhynchosaurs except Stenaulorhynchus in having a hook on
the supinator process of the humerus, and from Stenaulorhynchus in having a narrower proximal
than distal end to the humerus, a much smaller and more discrete deltopectoral crest, and a larger
capitellum than trochlea on the distal humerus.

Description. The left premaxillary fragment identified by Chatterjee from quarry 3A (TMM 31185-92; PI. 1,
figs 2 and 3) was described above. TMM 31185-93 is broken at both ends, but only a few millimetres are
probably missing from the tapering distal end. TMM 31 185-93 also tapers proximally.

The holotype humerus of Otischalkia , TMM 31025-263 (PI. 1, fig. 1; text-fig. 2a-c), and the referred
humerus TMM 31025-262 (PI. 1, figs 4 and 5), are very similar in morphology. These humeri are relatively short
and robust with greatly expanded ends, with an angle of torsion between the ends of about 55°, and a narrow
shaft linking them. The proximal end consists of a broad posterior plate and a relatively broad anteroventral
deltopectoral crest. The shaft is narrow and elliptical in cross-section. Like Hyperodapedon (Benton 19836, p.

930

PALAEONTOLOGY, VOLUME 34

668), the distal end has a large anterior capitellum for articulation with the radius and a posterior trochlea for
the ulna. The supinator crest above the ectepicondylar groove has a hook-like projection.

The femur of Otischalkia (PI. 1, figs 2 and 3) is robust, with a relatively straight shaft, a well-developed
intertrochanteric fossa and no fourth trochanter. TMM 31025-266 (PI. 1, figs 6 and 7) is a robust femur with
a relatively straight shaft. Proximally, the internal trochanter forms a broad projection, separated from a large
ovoid articular surface by an intertrochanteric fossa. The distal articular surface is divided into two distinct
condyles, both of which may have been for the tibia (Benton 19836).

Discussion. The Otischalkia humeri differ greatly from the elongate gracile humerus of
Trilophosaurus (Gregory 1945, PI. 27, figs 1-3; contra Long, in Murry 1989). Although rhynchosaur
humeri are conservative in their morphology, the Texas specimens differ from the humeri of other
rhynchosaurs. The Texas rhynchosaur differs from Hyperodapedon gordoni in having a distal
expansion of the humerus that is as wide as the proximal end, and from all Late Triassic
rhynchosaurs in having a hook-like process on the supinator crest. The morphology of the
supinator process in Otischalkia is very like that of the Middle Triassic Stenaulorhynchus which also
has a hook on the supinator process (Huene 1938, pi. 7, fig. 3 a-d). However, the humerus of
Stenaulorhynchus differs from the Texas humeri in having a greatly expanded proximal end which
is wider than the distal end, a much longer and narrow shaft between the expanded ends and in
having a much larger capitellum than trochlea on the distal end (Huene 1938, pi. 7, fig. 3a-d; PI.
1, figs 4, 5; Text-fig. 2a-c).

TMM 31025-266 is more gracile than femora of Hyperodapedon (Chatterjee 1974, fig. 25 a-b;
Benton 1983a, fig. 33 6-c) and apparently differs from all other rhynchosaurs in having an internal
trochanter that extends almost to the proximal articular surface, although this feature may be partly
due to poor preparation. TMM 31025-266 is very different in morphology from Trilophosaurus
femora (contra Long in Murry 1989) in: (1) being more robust; (2) not being strongly,
longitudinally sigmoidal; and (3) in having less offset distal condyles (Gregory 1945, pi. 29, figs 5
and 6).

Lacking cranial material, it is not clear what subfamily this taxon belongs to. But, as all Late
Triassic rhynchosaurs apparently pertain to the Hyperodapodontinae, we tentatively assign the
Texas taxon to this subfamily.

LATE TRIASSIC RHYNCHOSAUR TAXONOMY AND DISTRIBUTION

Western United States. Hotton (in Parrish 1989a, p. 362; 19896, p. 237; Hotton oral comm., 1990)
reports finding fragmentary remains of a rhynchosaur from the Popo Agie Formation, south of the
Big Horn Mountains in north-central Wyoming. Parrish (1989a, p. 362; 19896, p. 237) reported the
presence of rhynchosaur bones from the Placerias quarry of northeastern Arizona in the lower
Chinle Formation. R. A. Long (oral comm. 1990) believes that these specimens represent a large
trilophosaur, but, pending publication of these specimens, we tentatively assume that they represent
a rhynchosaur. Murry (1989) reported a ‘?rhynchosaur ' in the Norian Redonda Formation in east-
central New Mexico, but there is no evidence of any animal similar to a rhynchosaur in this
assemblage (Hunt 1991).

EXPLANATION OF PLATE 1

Fig. I. Otischalkia elderae gen. et sp. nov.; holotype left humerus (TMM 31025-263); dorsal view, x 0 65.
Figs 2 and 3. Otischalkia elderae gen. et sp. nov. ; left premaxillary fragment (TMM 31 185-92). 2, lateral view.
3, mesial view. Both x T9.

Figs 4 and 5. Otischalkia elderae gen. et sp. nov.; left humerus (TMM 31025-262). 4, dorsal view. 5, ventral
view. Both x 0-65.

Figs 6 and 7. Otischalkia elderae gen. et sp. nov.; left femur (TMM 31025-266). 6, posterior view. 7, anterior
view. Both x 0 65.

PLATE 1

HUNT and LUCAS, Otischalkia elderae

932

PALAEONTOLOGY, VOLUME 34

tr c

text-fig. 2. Otischalkia elderae gen. et sp. nov.; holotype left humerus (TMM 31025-263). a, posteroventral
view, b, anterodorsal view, c, anteroventral view. Abbreviations: c, capitellum; d, deltopectoral crest; sp.

supinator process; tr, trochlea.

Canada. Baird (1964) first mentioned rhynchosaur specimens from the Wolfville Formation of
Nova Scotia, but they have never been adequately described. The presence of one longitudinal
groove in the maxilla of the Wolfville specimens and the absence of lingual teeth on the maxilla
(Chatterjee 1980) places them in the subfamily Hyperodapedontinae (Chatterjee 1969, 1974; Benton
19836, 1987). The only member of this subfamily that lacks lingual dentition on the mandible
(Chatterjee 1980), and a single tooth row on the dentary (Benton 1990), like the Wolfville taxon,
is Scaphonyx. Thus, pending formal description, we assign the Wolfville specimens to Scaphonyx sp.

Brazil. Woodward (1907) named the rhynchosaur Scaphonyx fischeri from the Santa Maria
Formation of Brazil. Fluene (1926) erected several other taxa of rhynchosaurs from the Santa Maria
but these have all been synonymised subsequently with S. fischeri (Fluene 1942; Sill 1970). Scaphonyx
is restricted to the 1 Scaphonyx assemblage zone’ of the upper Santa Maria Formation (Barberena
et al. 1985). Barberena et al. (1985) also reported the discovery of a probable new species of
Scaphonyx in the Caturrita Formation which overlies the Santa Maria, but the specimens have not
been described.

Argentina. Sill (1970) described specimens from the Ischigualasto Formation of western Argentina
as Scaphonyx sanjuanensis. S. sanjuanensis is only slightly different from the better known S. fischeri
and may be synonymous (Benton 1987).

HUNT AND LUCAS: RHYNCHOSAUR FROM TEXAS

933

Scotland. All rhynchosaur specimens from the Lossiemouth Sandstone Formation of eastern
Scotland are referrable to the hyperodapedontine Hyper odape don gordoni , which has been the
subject of many articles, the last of which is an extensive monograph by Benton (19836).

Morocco. Dutuit (1976) described Acrodenta irerhi from the middle of the Argana Formation in
Morocco as a possible rhynchosaur. but this taxon is now considered to be captorhinid (Dutuit pers.
comm. 1990).

Tanzania. Boonstra ( 1953) named two species of rhynchosaur, Scaphonyx stockleyi and S. africanus,
from an unknown locality in the Tunduru district of what is now Tanzania. Chatterjee (1980)
erected the genus Supradapedon for S. stockleyi. We agree with Benton (19836, p. 712) that
Supradapedon could be a large Scaphonyx or Hyperodapedon , and we refer Supradapedon to
Hyperodapedontinae indeterminate. The holotype of Scaphonyx africanus , which is the proximal
third of a weathered femur, is indeterminate, and thus this taxon is a nomen dubium.

Madagascar. Buffetaut (1983) named Isalorhynchus genovefae for a rhynchosaur maxilla from the
Isalo II Beds which lacks lingual dentition. It is characteristic of the Hyperodapedontinae that they
lack lingual teeth on the maxilla (Benton 19836). The Madagascan genus contains only one species
whose holotype is not diagnostic below the subfamily level (Benton 1987).

India. Huxley (1869) referred rhynchosaur tooth plates from the Maleri Formation to
Hyperodapedon gordoni. Feistmantel (1880) mentioned the occurrence of rhynchosaurs in the Tiki
Formation. Lydekker (1881) named these specimens Hyperodapedon huxleyi, and Huene (1938)
erected the genus Paradapedon for this species. Benton (19836) concluded that the Indian
rhynchosaur should be considered a species of Hyperodapedon , H. huxleyi.

PALAEOECOLOGY OF LATE TRIASSIC RHYNCHOSAURS

Rhynchosaurs had a cosmopolitan distribution during the Late Triassic (Text-fig. 1), but their
relative abundance varies considerably between faunas. Faunas rich in rhynchosaurs (Santa Maria,
Ischigualasto, Lossiemouth Formations) all occur in Gondwanaland with the exception of the
Lossiemouth, and contemporaneous faunas depauperate in rhynchosaurs (lower Dockum, Popo
Agie, Chinle, Wolfville Formations) all occur in Laurasia. This distribution corresponds to two
broad palaeobotanical provinces, characterized in the south by the Dicroidium flora and in the north
by a ‘cycadophyte ’/fern /conifer flora (Tucker and Benton 1982; Benton 1983a; Olsen and Galton
1984). However, it should be noted that, except for Lossiemouth, in Laurasia, faunas rich in
semiaquatic phytosaurs and metoposaurs, represent different facies than the southern Gond-
wanaland faunas dominated by terrestrial animals (Benton and Walker 1985). Thus, we would
expect the semiaquatic and more terrestrial facies to produce different floras. Thus, although we
recognise that there are two different provinces in the Late Triassic, we believe that many differences
between the northern and southern faunas are ecological in nature as previously suggested (e.g.
Benton 1983a: Benton and Walker 1985). In strong support of this notion, we note the recent
recognition of characteristic elements of the southern faunas in Laurasia. These include the
occurrence of Scaphonyx in the Wolfville Formation of Nova Scotia and other rhynchosaurs in
Wyoming, Texas, and Scotland, the dicynodont Ischigualastia in the Santa Rosa Formation of New
Mexico (Lucas and Hunt 1991), the ictidosaur Pachygenelus in the Cooper Formation of Texas
(Chatterjee 1983), and a southern-Gondwana-aspect fauna in Virginia (Olsen 19896).

BIOCHRONOLOGY OF LATE TRIASSIC RHYNCHOSAURS

Prior to the 1970s, several of the rhynchosaur-bearing formations, principally those in South
America, that we now consider to be Late Triassic in age, were assigned to the Middle Triassic

934

PALAEONTOLOGY. VOLUME 34

(Romer 1962, 1966, 1970). However, now all the rhynchosaur-bearing formations reviewed above
are considered to be late Carnian in age (e.g. Olsen and Sues 1986; Text-fig. 3) or possibly early
Norian (Benton and Walker 1985). Our review of the evidence for the age of Late Triassic
rhynchosaurs indicate they are all of probable late Carnian age.

Two of the rhynchosaur records discussed above are from stratigraphic units with poor age
control. The hyperodapedontine from Tanzania was collected from beds of unknown age (Boonstra
1953), which have subsequently been assigned to the ?Upper Triassic because of the presence of the
rhynchosaur (Chatterjee 1980). The Madagascan hyperodapedontine is from the Isalo II beds
(middle Isalo Group; Buffetaut 1983) which have been considered Rhaetic (on the basis of
phytosaurs and metoposaurs) to Liassic (on the basis of ammonites) in age (Besaire and Collignon
I960; Guth 1963; Westphal 1970; Dutuit 1978). However, the presence of unidentifiable phytosaurs
and metoposaurs in the lower part of the unit gives no greater age precision than Late Triassic.
Buffetaut (1983) suggests that the rhynchosaur from the basal beds of Isalo II is a rhynchosaurine
and thus of Middle Triassic (Ladinian) age. Furthermore, the Isalo rhynchosaur is a
hyperodapedontine (Benton 1987) and thus is probably of Late Triassic age, which is in agreement
with other faunal evidence for the age of these beds.

Four of the remaining occurrences of Late Triassic rhynchosaurs are from faunas that contain
the phytosaur Paleorhinus (Popo Agie, Chinle, lower Dockum and Maleri). Paleorhinus is
important to Late Triassic biochronology because it is the only terrestrial vertebrate also known
from marine strata and thus provides a means of correlating terrestrial faunas directly to marine
stages (Huene 19396; Hunt and Lucas 1991). The Paleorhinus specimen from marine strata is from
the Opponitzer Beds near Lunz in Austria which are of Tuvalian (late Carnian) age (Janoscheck and
Matura 1980; Zapfe written comm. 1989). Thus, in the absence of any conflicting data we assign
all Paleorhinus- bearing strata to the Paleorhinus biochron of late, but not latest, Carnian age (Hunt
and Lucas 1990). Pollen data, where available, support these age assignments (Dunay 1972; Dunay
and Fisher 1979; Kumaran and Matheswari 1980; Litwin 1986). This means the rhynchosaur
occurrences in the Popo Agie, lower Dockum, lower Chinle, and Maleri are of late Carnian age.

The faunas of the Ischigualasto Formation of Argentina and the upper Santa Maria Formation
of Brazil share several taxa in common, including Scaphonyx, the aetosaur Aetosauroides
(Casamiquela 1960; Barberena et at. 1985), and the dinosaur Staurikosaurus (Colbert 1970;
Brinkman and Sues 1987). Faunal differences between these two units include the presence of

HUNT AND LUCAS: RHYNCHOSAUR FROM TEXAS

935

herrerasaurs and an ornithischian in the Ischigualasto and more specimens of therapsids in the
Argentinian than Brazilian strata (Barberena et al. 1985). These differences are probably ecological
in origin and thus there appears to be no basis for the widely held belief that the Santa Maria as
a whole is slightly older than the Ischigualasto (e.g. Sill 1969; Benton 1988). The Caturrita
Formation, which overlies the Santa Maria Formation in Brazil, shares Proterochampsa , Scaphonyx,
and Exaeretodon with the Ischigualasto (Barberena et al. 1985). It appears that the Caturrita
represents a more Ischigualasto-like facies than the upper Santa Maria, and that both are age
equivalents of parts of the Ischigualasto Formation of Argentina (Barberena et al. 1985, fig. 8).
These formations are usually considered ?middle-late Carnian in age (e.g. Benton 1988). There are
however, problems in correlating these faunas with those of Laurasia and northern Gondwanaland
because of marked ecological differences. The northern faunas, which are dominated by semi-
aquatic tetrapods, have been referred to the metoposaur/phytosaur empire and the southern
faunas, which are dominated by terrestrial tetrapods, to the rhynchosaur/diademodontid empire
(Tucker and Benton 1982; Benton 1983a). However, w'e believe that there is growing evidence that
the Brazilian and Argentinian faunas, containing rhynchosaurs, are of late Carnian age. Lucas and
Hunt (1991) recently recognised the dicynodont Ischigualastia from post -Paleorhinus biochron
strata of late Carnian age in New Mexico. This taxon is otherwise restricted to the Ischigualasto
Formation. Late Carnian (post -Paleorhinus biochron) strata in the American Southwest also
include staurikosaurid dinosaurs similar to Staurikosaurus from the Ischigualasto and Santa Maria
Formations (Murry and Long 1989; Hunt and Lucas 1989; Hunt 1990). In addition, we recognise
the rhynchosaur Scaphonyx in the Wolfville Formation of Nova Scotia which is also late Carnian
in age (Olsen 1989a). Scaphonyx also occurs in the Ischigualasto, Santa Maria, and Caturrita
Formations. Thus, we consider that these South American formations are of late Carnian age, based
on the evidence of the distribution of Ischigualastia. Scaphonyx. and staurikosaurids.

However, to refine the age of the Ischigualasto, Santa Maria, and Caturrita Formations, we
must further consider the age of the Wolfville Formation. The Wolfville fauna (Baird and Olsen
1983; Olsen 1989a) includes taxa known elsewhere from the Paleorhinus biochron in Texas
(‘ Metoposaurus" bakeri) and from post- Paleorhinus biochron, late Carnian strata of the American
Southwest (cf. Stagonolepis). There are no pollen (Traverse 1983) or other data to constrain further
the age of the Wolfville. However, as the Ischigualastia occurrence in New Mexico is in undoubtedly
post- Paleorhinus strata, we believe that it is probable that all Scaphonyx- bearing strata (Wolfville,
Ischigualasto, Santa Maria, Caturrita) are of post -Paleorhinus biochron age.

The age of the Lossiemouth Sandstone Formation of Scotland has been the subject of debate
(Benton and Walker 1985; Olsen and Sues 1986). At the generic level, the Lossiemouth shares
Hyperodapedon with the Maleri Formation (Benton 19836) of the Paleorhinus biochron and
Stagonolepis with post- Paleorhinus, late Carnian strata of the American Southwest (Walker 1961 ;
Murry and Long 1989). In addition, the Lossiemouth shares the procolophonid Leptopleuron with
the Wolfville (Baird and Olsen 1983), possibly at the specific level (Olsen and Sues 1986). We thus
tentatively refer the Lossiemouth to a post- Paleorhinus biochron (late Carnian) age based on the
procolophonid correlation with the Wolfville.

In conclusion, all Late Triassic hyperodapedontine rhynchosaurs that can be dated with any
accuracy are of late Carnian age (Text-fig. 3). The faunas that contain rhynchosaurs are both of
Paleorhinus-biochron and post.- Paleorhinus-biochron age. Scaphonyx is restricted to post-
Paleorhinus biochron (Wolfville, Santa Maria, Caturrita, Ischigualasto) strata whereas Hyper-
odapedon occurs in both Paleorhinus- biochron (Maleri) and post- Paleorhinus-age strata (Lossie-
mouth).

Acknowledgements. We thank Wann Langston Jr and Tim Rowe for permission to study specimens in their
care, H. Zapfe for information about the age of strata in Austria, M. J. Benton and two anonymous reviewers
for many helpful suggestions and the New Mexico Museum of Natural History for support.

936

PALAEONTOLOGY, VOLUME 34

REFERENCES

barberena, m. c. 1977. Bioestratigrafia preliminar da Formapao Santa Maria. Pesquisas , 7, 111-129.
barberena, M. c., araujo, d. c. and lavina, E. l. 1985. Late Permian and Triassic tetrapods of southern Brazil.
National Geographic Research , 1. 5-20.

baird, D. 1964. Rhynchosaurs in the Late Triassic of Nova Scotia. Geological Society of America Special Paper ,
73, 107.

— and olsen, p. e. 1983. Late Triassic herpetofauna from the Wolfville Fm. of the Minas Basin (Fundy
•Basin) Nova Scotia, Canada. Geological Society of America Abstracts with Programs , 15, 122.

benton, m. j. 1983n. Dinosaur success in the Triassic: a noncompetitive ecological model. Quarterly Reviews
of Biology, 58, 29-55.

— 1983^>. The Triassic reptile Hyperodapedon from Elgin: functional morphology and relationships.
Philosophical Transactions of the Royal Society of London. Series B, 302, 605-717.

— 1986. The Late Triassic tetrapod extinction events. 303-320. In padian, k. (ed.). The beginning of the Age
of Dinosaurs : Faunal change across the Triassic-Jurassic boundary . Cambridge University Press, Cambridge,
378 pp.

— 1987. The phylogeny of rhynchosaurs (Reptilia ; Diapsida; Triassic), and two new species. 18-23. In
Currie, p. J. and roster, e. h. (eds). Fourth Symposium on Mesozoic terrestrial ecosystems ; short papers ;
revised edition. Tyrrell Museum of Paleontology, Drumheller, 328 pp.

1988. The origin of dinosaurs. Modern Geology , 13, 41-56.

— 1990. The species of Rhynchosaurus , a rhynchosaur (Reptilia, Diapsida) from the Middle Triassic of
England. Philosophical Transactions of the Royal Society of London. Series B , 328, 213-306.

— and walker, a. d. 1985. Palaeoecology, taphonomy, and dating of Permo-Triassic reptiles from Elgin,
north-east Scotland. Palaeontology , 28, 207-234.

besaire, h. and COLLIGNON, m. 1960. Madagascar. Lexique Stratigraphique International , 4, 1-190.
boonstra, L. d. 1953. A note on some rhynchosaurian remains from Tanganyika territory. Annals of the South
African Museum , 12, 1—4.

brinkman, D. b. and sues, h.-d. 1987. A staurikosaurid dinosaur from the Upper Triassic Ischigualasto
Formation of Argentina and the relationships of the Staurikosauridae. Palaeontology , 30, 493-503.
buffetaut, e. 1983. Isalorhynchus genovefae, n. g. n. sp. (Reptilia, Rhynchocephalia), un nouveau
Rhynchosaure du Trias de Madagascar. Neues Jahrbuch fur Geologie und Palaontologie. Monatshefte ,
1983, 465-480.

Carroll, R. L., belt, E. s., dineley, D. L., baird, D. and MCGREGOR, D. c. 1972. Vertebrate paleontology of
eastern Canada. International Geological Congress, Montreal, 113 pp.
casamiquela, r. m. 1960. Noticia preliminar sobre dos nuevos estagonolepoideos Argentmos. Ameghiniana , 2,
3-9.

chatterjee, s. 1969. Rhynchosaurs in time and space. Proceedings of the Geological Society of London, 1658,
486-492.

1974. A rhynchosaur from the Upper Triassic Maleri Formation of India. Philosophical Transactions of
the Royal Society of London, Series B, 267, 209-261.

1980. The evolution of rhynchosaurs. Memoires de la Societe Geologique de France, 139. 57-65.

1983. An ictidosaur fossil from North America. Science, 220, 1151-1153.

Colbert, E. H. 1952. A pseudosuchian reptile from Arizona. Bulletin of the American Museum of Natural
History. 99, 565-592.

-1970. A saurischian dinosaur from the Triassic of Brazil. American Museum Novitates, 2405, 1-39.
dunay, r. e. 1972. The palynology of the Triassic Dockum Group of Texas, and its application to stratigraphic
problems of the Dockum Group. Unpublished Ph.D. thesis, Pennsylvania State University.

— and fisher, m. 1979. Palynology of the Dockum Group (Upper Triassic), Texas, Reviews of Palaeobotany
and Palynology, 28, 61-92.

dutuit, j.-m. 1976. II est probable que les Rhynchocephales sont represents dans la faune du Trias marocain.
Comptes Rendus hebdomadaires des Seances de TAcademie des Sciences. Series D, 283, 483-486.

1978. Description de quelques fragments osseux provenant de la region de Folakara (Trias superieur
malgache). Bulletin du Museum National cTHistoire Naturelle (Paris). Serie 4, 516, 79-89.
elder, r. l. 1978. Paleontology and paleoecology of the Dockum Group, Upper Triassic, Texas. Unpublished
M.Sc. thesis. University of Texas.

feistmantel, o. 1880. Palaeontological notes from Karharbi and South Rewa coalfields. Records of the
Geological Survey of India, 13, 189.

HUNT AND LUCAS: RHYNCHOSAUR FROM TEXAS

937

GREGORY, J. T. 1945. Osteology and relationships of Trilophosaurus. University of Texas Publication , 4401,
273-359.

1962. The genera of phytosaurs. American Journal of Science , 260, 652-690.

guth, c. 1963. Au sujet de restes de reptiles de Madagascar. Comptes Rendus Hebdomadaires des Seances de
T Academie des Sciences , 256, 2661-2663.

huene, f. von. 1929. Uber Rhynchosaurier und andere Reptilien aus den Gondwana-ablagerungen
Siidamerikas. Geologische und Palaontologische Abhandlungen , 17, 1-62.

— 1938. Stenaulorhynchus, ein Rhynchosauride der Ostafikanischen Obertrias. Nova Acta Leopoldina, 6,
83-121.

- 1939a. Die Verwandtschaftsgeschichte der Rhynchosauriden. Physis , 14, 499-523.

1939h. Ein primitiv Phytosaurier in der jungeren nordostalpin Trias. Zeitschrift fur M ineralogie. Geologie

und Paldontologie , 1939, 139-144.

— 1942. Die fossilen Reptilien des sudamerikanischen Gondwanalandes, C. H. Beck, Munich, 332 pp.

1956. Paldontologie und Phylogenie der niederen Tetrapoden. Gustav Fischer, Jena, 716 pp.

hunt, a. p. 1990. The early evolutionary diversification of dinosaurs in the western United States: they began
with a bang not a whimper. American Association of Petroleum Geologists Bulletin , 74, 1329.

— 1991. Tetrapod fauna of the Redonda Formation (Upper Triassic), east-central New Mexico. New
Mexico Geology , 12, 18.

— and lucas, s. G. 1989. Late Triassic vertebrate localities in New Mexico. 72-101. In lucas, s. g. and
HUNT, a. p. (eds). Dawn of the Age of Dinosaurs in the American Southwest. New Mexico Museum of
Natural History, Albuquerque, 414 pp.

— and lucas, s. G. 1991 . The Paleorhinus biochron and the correlation of the nonmarine Upper Triassic of
Pangaea. Palaeontology. 34, 487-501.

huxley, t. h. 1859. Postscript to: On the sandstones of Morayshire (Elgin etc.) containing reptilian remains;
and their relations to the Old Red Sandstone of that country by Murchison, R. I. Quarterly Journal of the
Geological Society of London , 15, 435-436.

— 1869. On Hyper odape don. Quarterly Journal of the Geological Society of London. 25, 138-152.
janoscheck, w. R. and matura, A. 1980. Austria. 1-88. In anonymous (ed.). Geology of the European countries.

Dunod, Paris, 433 pp.

krebs, b. 1965. Ticinosuchus ferox nov. gen. nov. sp. Ein neuer Pseudosuchier aus der Trias des Monte San
Giorgio. Schweizerische Palaontologische Abhandlungen , 81, 1-140.
kumaran, k. p. n. and matheswari, h. k. 1980. Upper Triassic spore dispersae from the Tiki Formation 2:
Miospores from the Janal Nala Section, south Gondwana basin, India. Palaeontographica , Abteilung B, 173,
26-84.

litwin, r. j. 1986. The palynostratigraphy and age of the Chinle and Moenave formations, southwestern USA.

Unpublished Ph.D. thesis, Pennsylvania State University.
lucas, s. G. and hunt, a. p. 1991 A Late Triassic dicynodont from New Mexico and its biochronological
significance. Journal of Paleontology, in review.

lydekker, r. 1881. Note on some Gondwana vertebrates. Records of the Geological Survey of India, 15,
174-178.

murry, p. a. 1986. Vertebrate paleontology of the Dockum Group, western Texas and eastern New Mexico.
109-137. In padian, k. (ed.). The beginning of the age of dinosaurs : Faunal change across the Triassic- Jurassic
Boundary. Cambridge University Press, Cambridge, 378 pp.

1989. Geology and paleontology of the Dockum Formation (Upper Triassic), West Texas and eastern
New Mexico. 102-144. In lucas, s. g. and hunt, a. p. (eds). Dawn of the age of dinosaurs in the American
Southwest. New Mexico Museum of Natural History, Albuquerque, 414 pp.

— and long, r. a. 1989. Geology and paleontology of the Chinle Formation, Petrified Forest National Park
and vicinity, Arizona and a discussion of vertebrate fossils of the southwestern Upper Triassic. 29-64. In
lucas, S. G. and hunt, A. p. (eds). Dawn of the age of dinosaurs in the American Southeast. New Mexico
Museum of Natural History, Albuquerque, 414 pp.

olsen, p. e. 1988. Paleontology and paleoecology of the Newark Supergroup (early Mesozoic, eastern North
America). 185-230. In manspeizer, w. (ed.). Triassic-Jurassic rifting , continental breakup, and origin of the
Atlantic Ocean and passive margins. Volume 1. Elsevier, New York, 523 pp.

1989a. Stop 11.1: Carrs Brook near Lower Economy, NS. 149-150. In olsen, p. e., schlische, r. w. and
GORE, p. w. (eds). Tectonic , depositional, and paleoecological history of early Mesozoic rift basins, eastern
North America. American Geophysical Union, Washington, 174 pp.

I989h. Tomahawk tetrapod quarry. 54. In olsen, p. e., schlische, r. w. and gore, p. w. (eds). Tectonic ,

938

PALAEONTOLOGY, VOLUME 34

depositional, and paleoeco/ogical history of early Mesozoic rift basins , eastern Northern America. American
Geophysical Union, Washington, 174 pp.

— and galton, p. m 1984. A review of the reptile and amphibian assemblages from the Storm berg of
southern Africa, with special emphasis on the footprints and the age of the Stormberg. Palaeontologia
Africana , 25, 87-110.

— and sues, h.-d. 1986. Correlation of continental Late Triassic and Early Jurassic sediments and patterns
of the Triassic- Jurassic tetrapod transition. 321-351. In padian, k. ( ed . ) . Beginning of the age of dinosaurs :
faunal change across the Triassic- Jurassic boundary. Cambridge University Press, Cambridge, 378 pp.

parrish, j. M. 1989m Late Triassic tetrapods of the North American Southwest. 360-374. In lucas, s. g. and
hunt, A. p. (eds). Dawn of the age of dinosaurs in the American Southwest. New Mexico Museum of Natural
History, Albuquerque, 414 pp.

19896. Vertebrate paleoecology of the Chinle Formation (Late Triassic) of the southwestern United
States. Palaeogeography, Palaeoclimatologv , Palaeoecology , 72. 227-247.
romer, a. s. 1962. La evolucion explosiva de los rhynchosaurios del Triasico. Revista Museo Argentino de
Ciencas Naturales Bernardino Rivadavia. Zoologica , 8, 1-14.

— 1966. Vertebrate Paleontology. 3rd edn. University of Chicago Press, Chicago, 468 pp.

— 1970. Tetrapod vertebrates and Gondwanaland. 111-124. In anonymous (ed.). Second Gondwana
Symposium: Proceedings and Papers. Council for Scientific and Industrial Research. Pretoria, 689 pp.

sill, w. d. 1969. The tetrapod-bearing continental Triassic sediments of South America. American Journal of
Science , 267, 805-821.

1970. Scaphonyx sanjuanensis , nuevo rincosaurio (Reptilia) de la Formacion Ischigualasto, Triasico de

San Juan, Argentina. Ameghiniana , 7, 341-354.

— 1971m Functional morphology of the rhynchosaur skull. Forma et Functio , 4, 303-318.

— 19716. Implicaciones estratigraficas y ecologicas de los rincosaurios. Revista Asociacion Geologica
Argentina , 26, 163-168.

traverse, a. 1983. Fundy Group palynoflorules on the northern margin of the Bay of Fundy, New Brunswick,
Can. Geological Society of America Abstracts with Programs , 15, 122.
tucker, M. E. and benton, M. j. 1982. Triassic environments, climates and reptile evolution. Palaeogeography ,
Palaeoclimatologv , Palaeoecology , 40, 361-379.

walker, a. d. 1961. Triassic reptiles from the Elgin area: Stagono/epis , Dasygnathus and their allies.

Philosophical Transactions of the Royal Society of London , Series B , 248. 103-204.
westphal, f. 1970. Phytosaurier-Hautplatten aus der Trias von Madagaskar. Neues Jahrbuch fur Geologie und
Palaontologie , Monatshefte , 10, 632-638.

williston, s. w. 1904. Notice of some new reptiles from the Upper Trias of Wyoming. Journal of Geology, 12,
688-697.

woodward, a. s. 1907. On some fossil reptile bones from the state of Rio Grande do Sul. Revista do Museo
Paulista, 7, 46-57.

a. p. hunt and s. G. lucas

New Mexico Museum of Natural History
1801 Mountain Road, NW
Albuquerque, New Mexico 87104, USA

Typescript received 20 March 1990
Revised typescript received 4 October 1990

ANATOMY, PATTERNS OF OCCURRENCE, AND
NATURE OF THE CONULARIID SCHOTT

by HE YO VAN ITEN

Abstract. Conulariid specimens that terminate adapically in a transverse wall, or schott, have been interpreted
in several ways. These interpretations have been based more on analogy with extant organisms than on
evaluation of fossil material. Currently favoured interpretations are that (1) schott-bearing specimens represent
individuals that were severed, in life, by currents ; (2) schott-bearing specimens were free-swimming individuals ;
(3) schott-bearing specimens represent non-injured, sessile individuals that retracted the apical part of their soft
body toward the oral end, as part of their normal life history. Examination of the test cavity of conulariid
specimens that do not terminate in a schott, and analysis of the frequency of occurrence of schott-bearing
specimens in low- versus high-energy sedimentary deposits, indicate that the most likely interpretation of
schott-bearing specimens is that they represent living individuals severed by currents. Proponents of a
scyphozoan affinity for conulariids have generally interpreted schott-bearing specimens as conulariid medusae,
one of the alternatives not favoured by the present analysis. In spite of this, the interpretation of schott-bearing
specimens as severed individuals is consistent with the hypothesis that conulariids and scyphozoans were
closely related.

Previous discussions of the mode of life and life history of conulariids have presented compelling
evidence that conulariid specimens terminating in a narrow apex (Text-fig. 1a) were sessile animals,
either free-standing and attached at the apex to shell material or other substrates, or embedded in
massive sponges or bryozoans (e.g. Sinclair 1948; Finks 1955, 1960; Rooke and Carew 1983;
Babcock et al. 1987; Harland and Pickerill 1987; Van Iten 1991a). Yet disagreement persists as to
the nature of conulariid specimens whose apical end, now truncated some distance above the apex,
terminates in a more or less smooth, generally convex transverse wall (Text-fig. 1b). This structure,
referred to by Kiderlen (1937) as the schott (the German word for bulkhead), was originally
interpreted as one of a series of internal septae, homologous to the septae of nautiloid cephalopods
(Hall 1847). Although schott-bearing specimens have been found that exhibit one or two additional
schotts above the terminal one (Babcock and Feldmann 1986a; see also below), it is now agreed that
the conulariid schott and nautiloid septa are not homologous structures (e.g. Werner 1966, 1967;
Bischoff 1978; Grasshoff 1984; Babcock and Feldmann 1986a).

More recent discussions of the nature of the conulariid schott have focused on one or more of
the following three alternatives: (1) the schott was a cicatrix, produced in response to severance of
the body by currents (Werner 1966, 1967); (2) the schott was an autotomy scar, produced when the
adult conulariid detached from its apical end and assumed a free-swimming mode of life (e.g.
Kiderlen 1937; Boucek 1939; Termier and Termier 1949, 1953; Moore and Harrington 1956;
Chapman 1966; Kozlowski 1968; Bischoff 1978; Grasshoff 1984); (3) the schott was a regular
growth feature, produced upon attainment of a certain age (Sinclair 1948; Babcock and Feldmann
1986a). With the exceptions of Termier and Termier (1949, 1953) and Kozlowski (1968), advocates
of the hypothesis that the conulariid schott was produced in response to assumption of a free-
swimming mode of life interpret schott-bearing conulariids as similar to medusae of scyphozoan
cnidarians, a group that has been widely regarded as the most likely candidate for an extant nearest
relative of conulariids (e.g. Kiderlen 1937; Boucek 1939; Moore and Harrington 1956; Chapman
1966; Werner 1966, 1967; Bischoff 1978; Grasshoff 1984; Van Iten 19916). Their interpretation of

|Palaeontology, Vol. 34, Part 4, 1991, pp. 939-954. |

© The Palaeontological Association

940

PALAEONTOLOGY, VOLUME 34

text-fig. 1. A, Archaeoconularia granulata (Hall); AMNH 791; Middle Ordovician (Trenton Group); New
York, USA; two specimens, both broken just above the apex and with their apical end (arrows) situated
immediately adjacent to a straight-shelled nautiloid (probably Endoceras proteiforme Hall), x 0-88. b,
? Conularia sp. ; UMMP 259; Upper Devonian (English River Siltstone); southeast Iowa, USA; crumpled
specimen terminating in a prominent, outwardly convex schott, x 1.

the conulariid schott, here designated the alternation of generations hypothesis, is based on analogy
with scyphozoans exhibiting alternating polypoid and medusoid life stages.

Werner (1966, 1967), himself a proponent of a scyphozoan affinity for conulariids, proposed that
schott-bearing specimens represent polypoid individuals that were severed by currents, in life, and
survived to heal their damaged apical end as they lay on their side. His interpretation, here
designated the severance by currents hypothesis, is based on the observation that conulariid tests
are often extremely thin (with tests 30 or more cm long but commonly less than 0-01 cm thick; Van
Iten in prep.), and on analogy with coronatid scyphozoan polyps whose theca and soft parts have
been severed experimentally. As documented by Werner (1966, 1967) and Chapman and Werner
(1972), coronatid polyps that survive severance cover exposed soft tissues by producing a thin,
smooth sheet of periderm that is similar in microstructure and gross morphology to the conulariid
schott.

The third alternative, offered with essentially no support, suggests that when conulariids reached
a certain age, they retracted the apical part of their soft body toward the aperture, producing a
schott to seal off the portion of the test cavity just abandoned. This interpretation will be referred
to here as the apical retraction hypothesis. (A similar hypothesis, proposed by Babcock and
Feldman (1984, p. 17), is that conulariids were pseudoplanktonic organisms and that ‘at certain
points in their life cycle, sections of the periderm at the apex divided along the last-formed
transverse wall [schott] and fell to the ocean floor’.)

The present paper seeks to evaluate the foregoing interpretations of the conulariid schott through
analysis of its occurrence within conulariids, and through analysis of patterns of occurrence of
schott-bearing specimens with respect to sedimentary facies. A critical prediction of the hypothesis
that the conulariid schott was a regular growth feature, produced upon attainment of a certain age
(apical retraction hypothesis), is that specimens that preserve their apex (pointed specimens) and
belong to taxa that have yielded schott-bearing specimens should exhibit one or more schotts within
the test cavity (provided, of course, that such specimens were old enough to form schotts).

VAN ITEN: CONULARIID SCHOTT

941

Two key predictions of the severance by currents hypothesis are that (1) specimens preserving the
apex (and therefore not severed) should not exhibit internal schotts; and (2) percentages of schott-
bearing specimens in samples of conulariids should vary significantly, depending on the
sedimentology of the deposits hosting the samples. As noted by previous workers (e.g. Barrande
1867; Boucek 1928; Sinclair 1948; Moore et al. 1952; Havlicek and Vanek 1966), conulariids occur
in large numbers in a variety of marine strata. These include strata that were deposited under
conditions of extremely low physical energy (e.g. dark, laminated, deep-shelf shales and lime
mudstones), and strata deposited under conditions of relatively high physical energy (e.g. shallow-
shelf quartz sandstones and lime grainstones). If schott formation was associated with severance by
currents, then not only should schotts be absent in the test cavity of specimens that were not severed
in life, but proportions of schott-bearing specimens should be significantly higher in samples from
high-energy deposits than in samples from low-energy deposits (all other factors being equal).

This pattern of occurrence would not be expected were schott formation associated with
assumption of a free-swimming mode of life (or, alternatively, if conulariids were pseudoplanktonic
organisms that sometimes dropped to the sea floor). This is because schott formation would have
been mediated principally by the life history of the developing organism, rather than by physical
characteristics of the surrounding water, and because the chances of free-swimming (or
pseudoplanktonic) conulariids being preserved in fine grained, low energy deposits, where
conulariids commonly occur in great abundance (e.g. Sinclair 1948; Moore et al. 1952) and have
been found preserved in situ (e.g. Rooke and Carew 1983), would have been at least as good as the
chances of their being preserved in relatively coarse grained, higher energy deposits. In short, the
alternation of generations hypothesis predicts that proportions of schott-bearing specimens in low-
versus high-energy deposits should not be significantly different.

Details of the anatomy of the conulariid schott and schott-bearing conulariids are covered in a
Ph.D. dissertation by Sinclair (1948), but this work has not been published. To rectify this deficiency
in the conulariid literature, and to present additional information having a potential bearing on
alternative interpretations of the cause(s) of schott formation, data on the anatomy of the conulariid
schott will be presented in the following section.

Schott-bearing specimens are currently known from five of the twenty-one recognized conulariid
genera, namely Anaconularia Sinclair, Archaeoconularia Boucek, Conularia Miller, Metaconularia
Foerste, and Paraconularia Sinclair. The present study is based on direct examination of
approximately 1 100 specimens of these genera, supplemented by published data on approximately
500 additional specimens. Information on the location of specimens and explanations of repository
abbreviations used in the remainder of this paper are presented in the appendices.

ANATOMY OF THE SCHOTT AND SCHOTT-BEARING CONULARIIDS

The conulariid schott is built of numerous, extremely thin (about 1 //m), apatitic lamellae,
alternatively dense and vacuity-rich, that parallel its surface (Text-fig. 2a). It consists of a transverse
portion, or wall, and a longitudinal portion, or sleeve, that extends along the inner surface of the
faces proper, toward the aperture. Tracing of schott lamellae in sectioned specimens preserved near
their aperture revealed that the sleeve probably extended all the way to the aperture. The wall/sleeve
boundary tends to be irregular (Text-fig. 2b), and the surface defined by this boundary may be
oriented more or less perpendicular to the test’s long axis or inclined at a high angle to it. The wall
may be adapically convex, more or less planar, or adaperturally convex. Its surface may be smooth
or finely wrinkled, with wrinkles usually concentrated along the wall/sleeve boundary and oriented
at various angles to it. In some cases, the wall exhibits a small dimple or dimple-bearing
protuberance, usually situated at or near its centre.

As noted by previous investigators (e.g. Barrande 1867; Slater 1907), the distance between the
schott and the former apex (estimated by extending the corners of the test to their point of
intersection) varies between specimens. For example, in Conularia trentonensis Hall (Middle to
Upper Ordovician, North America), characterized in life by a maximum test length of at least 10 cm

942

PALAEONTOLOGY, VOLUME 34

text-fig. 2. a, Conularia trentonensis Hall; UMMP 7000; Middle Ordovician (Trenton Group); New York,
USA; scanning electron photomicrograph of part of a longitudinal section through a terminal schott, showing
the finely laminate microstructure, the wall-sleeve junction (arrow labelled ‘a’; the sleeve is the portion above
the arrow), and the contact between the sleeve and the face proper (arrow labelled ‘b’) x 250. b, C. trentonensis
Hall; UMMP 66012; Middle Ordovician (Trenton Group); New York, USA; oblique view of part of the
external surface of a terminal schott with a relatively prominent, dimple-bearing protuberance; the wall-
sleeve boundary (arrows) is irregular and associated with fine wrinkling of the wall, x 22.

(Van Iten in prep.), schotts occur at test widths ranging from about 1-5 to 6-0 mm (N = 29, mean
= 3-6 mm, S.D. = 11 mm), or from about 10 to 40 mm above the apex.

The cross-sectional geometry of schott-bearing specimens may be more or less rectangular, or
strongly rhombic, in some cases to such an extent that members of one pair of opposing corners
nearly touch each other (Kiderlen 1937; Sinclair 1948). In specimens showing a strongly rhombic
cross section, the schott itself is often smoothly curved and shows no evidence of compactional
distortion. This suggests that rhombic specimens terminating in a non-compacted schott exhibited
a rhombic cross-sectional geometry while alive (Sinclair 1948).

Although most schott-bearing specimens documented by previous workers (e.g. Barrande 1867;
Boucek 1928; Kowalski 1935) exhibit only one schott, some investigators (e.g. Barrande 1867;
Slater 1907) maintained that conulariids usually produced multiple schotts. This interpretation was
based on the observation that the distance between the schott and the former apex varies between
specimens (presumably reflecting post-mortem break-up of specimens), and on the discovery of
specimens exhibiting more than one schott. However, present examination of the test cavity of 207
specimens terminating in a schott, including 49 specimens preserving test material (e.g. Text-fig. 1b)
and 158 specimens preserved as sandstone steinkerns (Appendix 1), yielded only two specimens.

VAN ITEN: CONULARIID SCHOTT

943

USNM 373992 ( Conularia quichua Ulrich) and AMNH 42316 (C. trentonensis Hall), having more
than one schott (both of these specimens have two schotts, with the second, internal schott situated
several millimetres above the terminal one). Previous authors (Steinmann and Doderlein 1890;
Slater 1907; Kiderlen 1937; Sinclair 1948) have collectively reported six specimens bearing multiple
schotts. Five of these specimens have two schotts, and one of them (Steinmann and Doderlein 1890)
has three schotts. These six specimens constitute a very small fraction of the total number (estimated
by the present author to be at least 500) of schott-bearing specimens reported by previous authors.
Most of these specimens are preserved as sandstone steinkerns or show extensive exfoliation, thus
revealing their test cavity. Neither these nor any of the specimens here examined (Appendix 1 )
exhibit a schott at their apertural end, where one might expect to find a schott (at least occasionally)
had conulariids normally produced multiple schotts and undergone break-up. In short, it appears
that conulariids almost always produced only one schott.

In specimens having multiple schotts, the sleeve of the schott situated closest to the aperture rests
on the inner (adaxial) surface of the sleeve of its neighbour (Sinclair 1948). Together with the
observation that schotts are lamellar, with sleeves that extended all the way to the aperture, this
suggests that the sequence of formation of multiple schotts was adapertural in direction, and that
schotts were produced by centripetal accretion of whole lamellae, along the entire length of the soft
body.

OCCURRENCE OF SCHOTTS IN THE TEST CAVITY OF POINTED SPECIMENS

To determine if schotts occur in the test cavity of conulariid specimens that do not terminate in a
schott (but that belong to species or genera that have yielded schott-bearing specimens),
observations were made on specimens of Archaeoconularia granulata (Hall) (Text-fig. 1a; N — 2),
Metaconularia manni (Roy) (N = 12), Conularia splendida Billings (N = 9), C. trentonensis Hall
(N = 115), and several species of Paraeonularia (N = 15) (Appendix 2). Except for certain
Paraconularia (Appendix 2), all of the 153 specimens in this sample terminate at a test width of
1 mm or less, well below usual levels of occurrence of terminal schotts in conspecific or congeneric,
schott-bearing specimens. (Some of the Paraconularia terminate at a test width of up to about 5 mm,
but this is still below levels of occurrence of terminal schotts in schott-bearing Paraconularia
specimens here examined.) Most of the specimens (including all Archaeoconularia granulata and
Metaconularia manni , and nearly all Conularia trentonensis) are flattened, having been preserved in
shale or lime mudstone. For many of these specimens, the presence or absence of internal schotts
had to be ascertained by checking for localized deformation of the faces (the test cavity of the
remaining specimens has been revealed through exfoliation or sectioning). Two of the species
chosen here, Conularia splendida and C. trentonensis , have yielded schott-bearing specimens. As
shown in Text-figure 3, pointed Conularia specimens for which reliable estimates of original test
length can be obtained fall within or close to the size range of schott-bearing, conspecific specimens
whose original test length likewise can reliably be estimated.

None of the specimens in this sample exhibits internal schotts. A similar result was obtained by
Barrande (1867), who found that schotts did not occur in the test cavity of several hundred
Bohemian specimens, all belonging to genera that have yielded schott-bearing specimens but
without a terminal schott themselves. Given that many of the specimens here examined are similar
in size to conspecific specimens terminating in a schott (suggesting that the specimens were similar
in age), these results tend to falsify the hypothesis that schott production was a regular feature of
conulariid ontogeny.

DISTRIBUTION OF SCHOTT-BEARING SPECIMENS RELATIVE TO
SEDIMENTARY FACIES

Background

As indicated by studies of the stratigraphy and sedimentology of strata known to bear conulariids
(e.g. Branson 1944; Lowenstam 1957; Svoboda 1966; Belt et al. 1967; Scoffin 1971; Lane 1973;

944

PALAEONTOLOGY, VOLUME 34

Specimen length (mm)

text-fig. 3. a. Histogram showing the estimated original test lengths (as measured along the midlines) of 12
specimens of Conularia splendida Billings (Upper Ordovician, northeast Iowa, USA), including 9 pointed
specimens (open bars; SUI 49979/3, 6151 1-61518) and 3 specimens terminating in a schott (solid bars; AC
1-1448, SUI 61519-61520). b. Histogram showing the estimated original test lengths (as measured along the
midlines) of 6 specimens of C. trentonensis Hall (Middle to Upper Ordovician, USA), including 3 pointed
specimens (open bars; AMNH 29649, SUI 61506, UWGM WW4001) and 3 specimens terminating in a schott
(solid bars; ROM 28324, SUI 61505, UMMP 66012). All of the specimens preserve (or come very close to
preserving) their aperture. For all specimens, the position of the apex was estimated by extending the corners

to their point of intersection.

Ramsbottom 1973; Bender 1974; Babin et at. 1976; Titus and Cameron 1976; Watkins 1978; Titus
1982, 1986; Shaver et al. 1986; Williams and Telford 1986; Harland and Pickerill 1982, 1987),
conulariids occur in marine deposits spanning a more or less continuous spectrum of shallow
nearshore to deep offshore facies. Based on levels of physical energy at the time of deposition, these
facies can be interpreted as belonging to one of three general depositional regimes. Arranged in
order of increasing energy, these are: (1) sheltered nearshore or deep, distal shelf or basinal waters
that were characterized by extremely low physical energy; (2) moderately deep, mesial shelf waters,
situated below fair-weather wave base but subject to episodes of relatively high current energy
during storms; and (3) shallow, proximal shelf waters, situated closer to fair-weather wave base than
Regime 2 settings and subject to more frequent and/or intense episodes of bottom turbulence.

VAN ITEN: CONULARIID SCHOTT

945

Depositional Regime 1 is represented primarily by dark (grey to black), laminated shales, lime
mudstones, and muddy siltstones, commonly with benthic macrofaunas that are sparse and of low
diversity. As indicated by faunal and sedimentological evidence and by stratigraphic relationships
(e.g. Belt et al. 1967; Lane 1973 ; Mikulic et al. 1985a, 19856; Watkins 1978 ; Harland and Pickerill
1987), Regime 1 sediments were generally deposited in sheltered lagoons or embayments, or on the
distal portions of marine shelves or in cratonic shale basins, below storm wave base.

Conulariids in at least some Regime 1 deposits (e.g. cementstones in the Calciferous Sandstone
Group, Lower Carboniferous, Scotland; Belt et al. 1967) occur in probable life clusters. In these
clusters, which consist of anywhere from 2 to 20 or more specimens, some or all of the specimens
converge adapically on a common centre (e.g. Text-fig. 4a; see also discussions and illustrations in
Slater (1907), Ruedemann (1925), Sinclair ( 1940), Lane (1973), and Babcock and Feldmann ( 1986a,
19866))- In the remainder of this discussion, such clusters will be referred to as radial clusters.

Depositional Regime 2 is represented primarily by light- to dark-coloured, mud-dominated shelf
limestones (mudstones and wackestones), with minor amounts of bioclastic limestone, and by
interbedded shales and fine sandstones. Faunas in these deposits are more diverse than faunas in
Regime 1 deposits, with greater numbers of ‘normal’ shelf benthos such as articulate brachiopods,
bryozoans, and echinoderms. Like conulariids in Regime I deposits, conulariids in Regime 2
deposits may be clustered; however. Regime 2 clusters almost never show a radial arrangement
(indicating that they have been disrupted; see below). Although Regime 2 deposits contain
substantial amounts of mud-sized sediment, suggesting deposition under conditions of low physical
energy, episodes of relatively high energy are indicated by the presence of tempestite horizons (e.g.
Titus 1982, 1986; Bakush and Carozzi 1986), characterized by sharp (scoured) lower contacts and,
in the case of carbonates, packstone to grainstone fabrics (commonly with intraclasts and
substantial fragmentation and abrasion of skeletal grains). Clastic analogues of Regime 2 carbonate
deposits generally consist of alternating thin shales and fine- to medium-grained, argillaceous or
micaceous sandstones (e.g. Havhcek and Vanek 1966; Bender 1974).

Depositional Regime 3 is represented by thick, laterally extensive deposits of clean, fine- to
coarse-grained quartz sandstone, and by carbonate deposits consisting mostly of mud-free,
bioclastic limestone, with minor amounts of shale and/or muddy limestone. As indicated by
sedimentological and stratigraphic evidence (e.g. Atherton et al. 1975; Johnson 1980), the bulk of
these units was deposited under conditions of relatively high energy, probably associated with
storms.

Clearly, the chances of live conulariids being subjected to current energy of sufficient magnitude
to cause breakage would have been greater in Regime 2 or 3 environments than in Regime 1
environments. Thus, if schott formation was associated with repair of a severed apical end, one
might expect to observe significantly greater proportions of schott-bearing specimens in Regime 2
and 3 samples than in samples from Regime 1 . Since Regimes 2 and 3 were both characterized by
episodes of relatively high physical energy (the distinction between these two regimes involving the
frequency and intensity of such episodes), differences in the proportion of schott-bearing specimens
between samples from Regimes 2 and 3 might not be significant.

Observations

Data on the frequency of occurrence of schott-bearing conulariid specimens in samples from the
three depositional regimes outlined above are presented in Table 1. Schott-bearing specimens are
extremely rare or absent (only 2 of 688 specimens examined) in samples from Regime 1 deposits,
but they are common in samples from Regimes 2 (41 of 75 specimens examined) and 3 (162 of 810
specimens examined). Application of the chi-square statistic indicates that differences in the
proportion of schott-bearing specimens between Regime 1 (2/668) and Regimes 2 (41/75) and 3
(162/810), respectively, are highly significant, with the chances that these differences are due to
random sampling error being substantially less than 1 in 1000. Similar high confidence levels apply
to comparisons involving conspecific specimens or congeneric specimens of very similar species only
(see Table 1). (Due to factors associated with sample size, comparisons involving conspecific or

946

PALAEONTOLOGY, VOLUME 34

A

B

text-fig. 4. a, Conularia splendida Billings; SUI 49979; Upper Ordovician (Maquoketa Formation); northeast
Iowa, USA; line drawing of a radial cluster consisting of eight specimens; the arrows indicate the directions
in which the apical ends point ; scale bar = 2 cm. b, C. trentonensis Hall ; BMS E10807 ; Ordovician ; New York,
USA; line drawing of three specimens preserved on a slab of coarse, skeletal (brachiopod/
bryozoan/echinoderm) lime grainstone; two of the specimens terminate in a schott (labelled ‘S'). The three
specimens are interpreted as members of a single, formerly radial cluster that underwent disruption and

transport (see discussion in text).

congeneric specimens only are most appropriately evaluated using Fisher’s Exact Test, discussed in
detail by Bliss (1967).)

These results are reinforced by semi-quantitative observations on additional samples in the
conulariid literature. Kowalski (1935, p. 291), discussing a sample of approximately 20 specimens
of Archaeoconularia pyramidata (Hoeninghausen) from the Upper May (‘ Conularia ’) Sandstone
(Upper Ordovician), northwest France, states that ‘many specimens’ terminate in a schott. Boucek
(1928, p. 78), who examined approximately 200 specimens of A. consobrina (Barrande) from Middle
Ordovician quartz sandstones in the Barrandian Basin, Bohemia, reports that 'the apex of the shell
is almost always missing, and one finds in its place a hemispherical [schott].’ Barrande (1867, p. 15),
commenting on Barrandian Basin conulariids in general (a sample consisting of several thousand
specimens), notes that 'in most species, especially those that [occur] in shales, we see conulariids
terminating in a sharp point ... [but] five of our species, preserved in quartzite..., show, in many
cases, a [schott].’ Similarly, Slater’s (1907) data on the anatomy of British conulariids indicate that
whereas conulariids from shallow- water, open shelf limestones (e.g. the Silurian Wenlock Limestone

VAN ITEN: CONULARIID SCHOTT

947

Species

Unit and Age

Locality

Host Lithology

N

Ns

Np

DeposilionaJ Regime 1

Archaeoconularia

(1 )A. exquisita**

Middle Ordovician (1)

Bohemia

Dark shale

200

0

?n

Conularia

(2 )C. trentonensis

Collingwood Shale (U. Ordo., 2)

Ontario

Black, laminated shale

40

2

38

(2 )C. trenlonensis++

Tetreauville Fm. (M. Ordo., 3)

Quebec

Black, laminated lime mudstone

100

0

?(*>

(2 )C. trentonensis

Brandon Bridge Fm. (Silur. , 4)

Wisconsin

Grey, laminated shale

159

0

113

Metaconularia

(3 )M. aspersa

Lower Ludlow Shale (Silur., 5)

England

Grey-brown, laminated shale

3

0

2

(3 )M . bihneata**

Liten Fm. (Silur., 6)

Bohemia

Black, laminated calcareous shale

17

0

?

(3 )M . manni

Lecthaylus Shale (Silur.)

Illinois

Grey, laminated shale

13

0

12

Paracomdana

(4 )P. chagnnensis

Chagrin Shale (U. Devon.)

Ohio

Grey, laminated shale

13

0

2

(4 )P. chesterensis*

Edwardsville Mbr., Muldraugh Fm.
(Miss., 7)

Indiana

Grey, laminated shalcy siltstonc

53

0

V)

(4 )P. tenuis*

Cementstone Facies, Calciferous
Sandstone Gp. (Miss., 8)

Scotland

Grey, laminated lime mudstone

70

0

?(*)

Depositional Regime 2
Conularia

(5)C. sowerbyu

Wenlock Limestone (Silur., 9)

England

Lime mudstone/wackcstone

25

14

0

(6 )C. subcarbonaria

Keokuk Limestone (Miss., 10)

Illinois

Lime wackestone/packstone/grainstonc

10

5

0

(6 )C. subcarbonaria

Ramp Creek Mbr., Harrodsburg
Limestone (Miss., 1 1)

Indiana

Lime wackcslone/packstone

1

0

0

(7)C. trentonensis

Verulam Fm. (M. Ordo., 2)

Ontario

Lime wackestonc/packstonc/grainstonc

11

5

0

(7)C. trentonensis

Denley, Kings Falls, Sugar River
Fms. (M. Ordo., 12)

New York

Lime wackcstonc/packstonc/grainstonc

1 1

6

0

(7)C. trentonensis

Sherwood Mbr., Dunlcith Fm.
(M. Ordo., 13)

Iowa

Lime wackestonc/packstonc

8

2

0

Metaconularia

(8 )M. calden

Verulam (Cobourg) Fm.
(M. Ordo., 2)

Ontario

Lime wackestonc/packstone/grainsionc

3

3

0

(8 )M . divisa

Dubuque Fm. (U. Ordo., 13)

Iowa

Lime mudstone/wackcstone

5

5

0

(8 )M. sp.

Lower Lindsay Fm. (M. Ordo., 14)

Ontario

Lime wackcslone/packstone

1

1

0

Depositional Regime 3
Anaconularia

(9 )A. anomala

Drabov Quartzite (M. Ordo., 1)

Bohemia

Fine to medium, micaceous quartz
sandstone

760

140

0

Archaeoconularia

(10)4. consobrina

Drabov Quartzite (M. Ordo., 1 )

Bohemia

Fine to medium, micaceous quart/
sandstone

10

5

0

(10)4. pyramidata

Upper May (Conularia) Sandstone
(U. Ordo.. 15)

Brittany

Fine to coarse quartz sandstone

20

10

0

Conularia

( 1 1 )C. subcarbonaria

Burlington Limestone (Miss., 16)

Missouri

Coarse, skeletal lime grainstonc

1 1

3

0

(12)C. trentonensis

Ordovician

New York

Coarse, skeletal lime grainstonc

3

2

0

Paracomdana

(13 )P. sp.

Burlington Limestone (Miss., 10)

Iowa

Coarse, skeletal lime grainstonc

5

2

0

(13)/>. sp.

Glen Dean Limestone (Miss., 1 1)

Indiana

Coarse, skeletal lime grainstonc

1

0

0

table 1. Frequency of occurrence of schott-bearing conulariid specimens in samples from the three
depositional regimes outlined in the text. Lettered symbols are as follows: N = number of specimens; Ns =
number of specimens terminating in a schott; Np = number of pointed specimens (specimens tapering to a
width of 1 mm or less and lacking a terminal schott). Sources of numerical data taken from the literature are
as follows: ** = Boucek (1928); + + = Sinclair (1948); * = Lane (1973); + = Slater (1907). An asterisk
following a question mark in the Np column indicates that the percentage of pointed specimens is reported to
be large. Numbers following the geologic age of the host rock unit refer to one of the following
sedimentological and/or stratigraphic studies: 1, Havhcek and Vanek (1966); 2, Brett and Brookfield (1984);
3, Flarland and Pickerill ( 1982); 4, Mikulic et al. (1985a, 1985fi); 5, Watkins ( 1978) ; 6, Svoboda (1966); 7, Lane
( 1973) ; 8, Belt et al. (1967) ; 9, Scoffin (1971); 10, Atherton et al. (1975); 11, Shaver et al. (1986); 12, Titus (1982,
1986) or Titus and Cameron (1976); 13, Bakush and Carozzi (1986); 14, Williams and Telford (1986); 15,
Babin et al. (1976); 16, Branson (1944). See Appendix 3 for information on specimen locations. Numbers in
parentheses preceding species names designate samples used in statistical analysis. Conspecific specimens from
different localities but similar deposits were combined into one sample. In some cases, congeneric specimens
from the same depositional setting were also combined to form a single sample (to compensate for small
numbers of specimens and/or restriction of species to a single depositional setting). Statistical analysis was
conducted on the following sample pairs: 1^4 vs 5-8; 1-4 vs 9 13; 1 vs 10; 2 vs 5; 2 vs 6; 2 vs 7; 2 vs 1 1 ; 2
vs 12; 3 vs 8; 4 vs 13.

948

PALAEONTOLOGY, VOLUME 34

(Scoffin 1971) and the Carboniferous Limestone (Ramsbottom 1973)) often terminate in a schott,
conulariids from dark, graptolitic slates and shales, deposited in settings further offshore (Rayner
1967), do not exhibit a schott but commonly taper to a point.

Discussion

The foregoing results falsify the hypothesis that schott-bearing conulariids were similar to
scyphozoan medusa, but they are consistent with the hypothesis that schott-bearing conulariids
represent living individuals severed by currents. Additional lines of evidence suggest that patterns
of occurrence observed here are not the result of other causes, for example mass mortality of very
young conulariids (killed before they had the chance to form schotts), or inhibition of schott
formation by other factors of the local environment (e.g. low dissolved oxygen content). Although
many of the conulariids examined here have been broken well below the aperture (making it
impossible to obtain reliable measurements of the original size of these specimens), those Regime
I specimens that are more or less intact are generally similar in size to conspecific or congeneric,
schott-bearing specimens from Regime 2 and 3 deposits, and thus it seems unlikely that low
frequency of occurrence of schott-bearing specimens in Regime 1 samples is due to unusually early
death. This conclusion is reinforced by data in Barrande (1867) indicating that Bohemian
Archaeoconularia (Table 1) from dark shales (Regime 1) are about as large as morphologically
similar, schott-bearing Archaeoconularia collected from quartz sandstones (Regimes 2 and 3).
Together with the observation that Regime 1 specimens here examined show no other evidence of
abnormal development (e.g. departure from patterns of test ornamentation exhibited by specimens
from Regime 2 or 3 deposits), these observations also tend to rule out the hypothesis that low
frequency of occurrence of schott-bearing specimens in Regime 1 deposits is due to inhibition of
schott formation by unfavourable environmental conditions.

Data obtained at finer scales of observation lend additional support to the hypothesis that schott
formation was prompted by severance. For example, the one Regime 1 sample here encountered
that contains schott-bearing specimens ( Conularia trentonensis Hall, Collingwood Shale; 2 schott-
bearing specimens out of 40 specimens sampled) appears to dilTer from several other Regime 1
deposits in lacking radial clusters. This suggests that Collingwood conulariids were subjected to
some degree of disturbance, possibly associated with storm events affecting adjacent, shallower-
water shelf environments (Brett and Brookfield 1984).

Evidence of disturbance is exhibited by many of the conulariids from Regimes 2 and 3. Conulariid
steinkerns consisting of quartz sandstone often contain large amounts of medium to coarse sand
(see also Kowalski 1935), suggesting that the horizons that yielded these specimens were deposited
under conditions of relatively high energy. Similarly, most of the schott-bearing specimens from
Regime 2 carbonates occur in a wackestone or packstone matrix rich in disarticulated echinoderm
ossicles and disarticulated and/or broken trilobites and brachiopods, suggesting that they were
collected from tempestite horizons. At least five of the schott-bearing specimens examined here
occur in non-radial clusters (e.g. Text-fig. 4b). The three-specimen cluster shown in Text-figure 4b,
which includes two specimens terminating in a schott, occurs in coarse, bioclastic limestone.
Conulariids in the units from which this and similar clusters were collected are so rare (Van Iten in
prep.) that it seems extremely unlikely that they are artefacts of concentration by currents or
scavengers. Rather, they seem best interpreted as former life clusters. Coupled with sedimentological
evidence indicating current action, the present non-radial arrangement of specimens in these clusters
suggests that they were physically disrupted. As expected, schott-bearing specimens do not occur in
currently known radial clusters (Van Iten in prep.), which have undergone little or no disruption.

CONCLUSION

Analyses of the anatomy and patterns of occurrence of the conulariid schott show that schotts do
not occur in the test cavity of specimens that lack a terminal schott and are preserved close to the
apex, and that schott-bearing conulariids occur in significantly higher proportions in samples from

VAN ITEN: CONULARIID SCHOTT

949

high-energy deposits than in samples from low-energy deposits. These results corroborate the
hypothesis that schott-bearing conulariids represent living individuals severed by currents, but they
are inconsistent with other previously suggested interpretations of schott-bearing conulariids. The
occurrence of specimens having more than one schott, interpreted by some investigators (e.g. Hall
1847) as evidence that schott formation was a regular feature of conulariid ontogeny, is not
necessarily inconsistent with the hypothesis that schott-bearing conulariids were severed in life. As
an example, severed individuals may have been susceptible to infection or necrosis, leading to
additional episodes of schott formation.

Certainly, previously offered interpretations of schott-bearing conulariids do not exhaust the full
array of interpretations that might reasonably be suggested by analogy with living organisms.
However, none of these interpretations (e.g. obligatory decollation of the apical end, encystment)
appears to carry with it the expectation that schott-bearing conulariids should occur in significantly
higher proportions in high-energy deposits than in low-energy deposits.

Although schott-bearing specimens probably were not medusa-like, as held by Kiderlen (1937)
and several other proponents of a scyphozoan affinity for conulariids, this does not necessarily mean
that conulariids did not produce medusae (through means other than direct transformation of test-
bearing, polypoid individuals; e.g. Van Iten 1989, 1991 b). Moreover, even if conulariids did not
produce medusae, this alone would not mean that conulariids and scyphozoans were not closely
related to each other, since some scyphozoans lack a medusa (Hyman 1940).

Acknowledgments. This study is based on part of a Ph.D. dissertation written in the Museum of Paleontology,
the University of Michigan, Ann Arbor. 1 thank the members of my dissertation committee, D. J. Eernisse,
D. C. Fisher, K. C. Lohmann, G. R. Smith, and B. H. Wilkinson, for comments on earlier versions of the
manuscript. Review comments by S. Conway Morris greatly improved the submitted version. Permission to
examine or borrow museum specimens was given by T. E. Bolton, F. J. Collier, R. C. Eng, D. C. Fisher, J.
Golden, J. Hannibal, A. Horowitz, E. Landing, R. Laub, R. Norbey, D. Rudkin, R. Titus, J. Waddington,
and K. Westphal. Data on the anatomy of conulariids in collections of the British Museum (Natural History)
and the University of Birmingham were obtained by B. S. Beall and R. S. Cox. Data on the distribution of
conulariids in the Trenton Group (Middle Ordovician) of New York were supplied by R. Titus, who also
donated 12 specimens of Conularia trentonensis Hall to the University of Michigan Museum of Paleontology.
Several other conulariid specimens examined here were collected by A. J. Gerk and C. O. Levorson. T. Van
Iten printed the photographs, and R. Carr helped prepare the table. This study was supported by funding from
the Department of Geological Sciences and the Horace H. Rackham School of Graduate Studies at the
University of Michigan, Sigma Xi, NSF Research Grant BNS-8521097 to D. C. Fisher, and R. J. and H. B.
Van Iten. The scanning electron microscope used in this study was acquired under Grant #BSR-83-14092 from
the National Science Foundation.

REFERENCES

atherton, e., collinson, c. and lineback, j. a. 1975. Mississippian system. 123-163. In willman, h. b.,

ATHERTON, E., BUSCHBACH, T. C., COLLINSON, C., FRYE, J. C., HOPKINS, M. E., LINEBACK, I. A. and SIMON, J. A.
Handbook of Illinois stratigraphy. Illinois State Geological Survey Bulletin , 95, 1-261.

BABCOCK, L. E. and feldmann, R. M. 1984. Mysterious fossils. Earth Science , 37, 16-17.

— 1986a. Devonian and Mississippian conulariids of North America. Part A. General description and
Conularia. Annals of Carnegie Museum, 55, 349-410.

— 1986/). Devonian and Mississippian conulariids of North America. Part B. Paraconularia,
Reticulaconularia new genus, and organisms rejected from Conulariida. Annals of Carnegie Museum, 55,
41 1^479.

— wilson, m. t. and suarez-riglos, m. 1987. Devonian conulariids of Bolivia. National Geographic
Research, 3, 210-231.

BABIN, C., ARNAUD, A., BLAISE, J., CAVET, P., CHAUVEL, P., DEUNFF, J., HENRY, J.-L., LARDEUX, H., MELOU, M., NION,

J., Paris, f., plaine, J., quete, Y. and robardet, m. 1976. The Ordovician of the Armorican Massif (France).
359-385. In bassett, m. g. (ed.). The Ordovician System. University of Wales Press and National Museum
of Wales, Cardiff, 696 pp.

950

PALAEONTOLOGY, VOLUME 34

bakush, s. h. and carozzi, a. v. 1986. Subtidal storm-influenced carbonate ramp model: Galena Group
(Middle Ordovician) along the Mississippi River (Iowa, Wisconsin, Illinois, and Missouri), U.S.A. Archives
des Sciences, 39, 141-183.

barrande, j. 1867. Systeme Silurien du centre de la Boheme. Premiere Partie. Tome 3. Classe des mollusques ,
ordre des pteropodes. Charles Bellmann, Prague, iii-xv + 179 pp.
belt, e. s., freshney, e. c. and reed, w. a. 1967. Sedimentology of Carboniferous Cementstone Facies, British
Isles and eastern Canada. Journal of Geology , 75, 71 1-722.
bender, f. 1974. Geology of Jordan. Contributions to the Regional Geology of the Earth, Supplementary
Edition of Volume 7. Gebriider Borntraeger, Berlin, 196 pp.
bischoff, G. c. o. 1978. Internal structures of conulariid tests and their functional significance, with special
reference to Circoconulariina n. suborder (Cnidaria, Scyphozoa). Senckenbergiana Lethaia, 59, 275-327 .
bliss, c. i. 1967. Statistics in biology , Volume 1. McGraw-Hill, New York, vii-xiv + 558 pp.
boucek, b. 1928. Revision des conulaires Paleozoi'ques de la Boheme. Palaeontographica Bohemiae, 11, 60-108.

1939. Conularida. A I 1 3— A 131. In schindewolf, o. h. (ed.). Handbuch der Paldozoologie, Band 2 A.
Gebriider Borntraeger, Berlin, 131 pp.

branson, E. B. 1944. The geology of Missouri. University of Missouri Studies, 19, 1-535.
brett, c. e. and Brookfield, m. e. 1984. Morphology, faunas and genesis of Ordovician hardgrounds from
southern Ontario, Canada. Palaeogeography, Palaeoclimatology, Palaeoeco/ogy, 46, 233-290.
chapman, d. 1966. Evolution of the scyphistoma. 51-75. In rees, w. j. (ed.). The Cnidaria and their evolution.
Zoological Society of London. Symposia, 16, 449 pp.

— and werner, b. 1972. Structure of a solitary and colonial species of Stephanoscyphus (Scyphozoa,
Coronata) with observations on periderm repair. Helgoldnder Wissenschaftliche Meeresuntersuchungen , 23,
393-421.

finks, it. M. 1955. Conularia in a sponge from the West Texas Permian. Journal of Paleontology, 29, 831-836.

I960. Late Paleozoic sponge faunas of the Texas region. The siliceous sponges. Bulletin of the American
Museum of Natural History , 120, 7-160.

grasshoff, m. 1984. Cnidarian phylogeny - a biomechanical approach. Palaeontographica Americana, 54,
127-135.

hall, j. 1847. Palaeontology of New- York. Volume 1. Containing descriptions of the organic remains of the
Lower Division of the New- York System (equivalent of the Middle Silurian rocks of Europe). C. Van
Benthuysen, Albany, ix-xxix + 338 pp.

harland, t. l. and pickerill, r. k. 1982. A review of Middle Ordovician sedimentation in the Saint Lawrence
Lowland, eastern Canada. Geological Journal, 17, 135-156.

1987. Epizoic Schizocrania sp. from the Ordovician Trenton Group of Quebec, with comments on
mode of life of conulariids. Journal of Paleontology, 61, 844-849.
havlicek, v. and vanek, j. 1966. The biostratigraphy of the Ordovician of Bohemia. Sbornik Geologickych
Ved. P8, 7-69.

hyman, L. H. 1940. The invertebrates : Protozoa through Ctenophora. McGraw-Hill, New York, vii-xii + 726 pp.
Johnson, H. D. 1980. Shallow siliclastic seas. 207-257 . In reading, h. g. (ed.). Sedimentary environments and
facies. Elsevier, New York, v-xi + 557_ pp.

kiderlen, h. 1937. Die Conularien. Uber Bau und Leben der ersten Scyphozoa. Neues Jahrbuch fur
Minera/ogie , Geologie und Paldontologie , Abteilung B, 77, 113-169.
kowalski, j. 1935. Les Conulaires. Quelques observations sur leur structure anatomique. Societe des Sciences
Naturelles de I'Ouest de France , Bulletin, Serie 5, 5, 281-293.
kozlowski, r. 1968. Nouvelles observations sur les conulaires. Acta Palaeontographica Polonica, 13, 497-531.
lane, n. g. 1973. Paleontology and paleoecology of the Crawfordsville fossil site (Upper Osagian: Indiana).
University of California Press, Berkeley, v-i- 141 pp.

lowenstam, h. a. 1957. Niagaran reefs in the Great Lakes area. 215-248. In ladd, h. s. (ed.). Treatise on
marine ecology and paleoecology. Volume 2. Paleoecology. Geological Society of America Memoir, 67,
v-x+ 1077 pp.

mikulic, d. g., briggs, d. e. g. and kluessendorf, j. 1985«. A Silurian soft-bodied biota. Science , 228,
715-717.

1 985/?. A new exceptionally preserved biota from the Lower Silurian of Wisconsin, USA.
75-85. In Whittington, h. b. and conway morris, s. (eds). Extraordinary fossil biotas: their ecological and
evolutionary significance. Philosophical Transactions of the Royal Society of London, Series B, 331, 1-192.
moore, r. c., lalicker, c. G. and fisher, a. g. 1952. Invertebrate fossils. McGraw-Hill, New York, v-xiii +
766 pp.

VAN ITEN: CONULARIID SCHOTT

951

— and Harrington, h. j. 1956. Scyphozoa. F27-F38. In moore, r. c. (ed . ). Treatise on invertebrate
paleontology. Part F. Coelenterata. Geological Society of America and University of Kansas Press, Boulder,
Colorado and Lawrence, Kansas, 498 pp.

ramsbottom, w. H. c. 1973. Transgressions and regressions in the Dinantian: a new synthesis of British
Dinantian stratigraphy. Proceedings of the Yorkshire Geological Society , 39, 567-607.
rayner, D. H. 1967. The stratigraphy of the British Isles. Cambridge University Press, Cambridge, iii-x-T
453 pp.

rooke, h. G. and carew, j. l. 1983. New light on a poorly understood fossil group. Geological Society of
America , Abstracts with Programs, 15, 53.

ruedemann, r. 1925. The Utica and Lorraine Formations of New York. Part 2. Systematic paleontology. New
York State Museum Bulletin, 262, 3- 171.

scoffin, t. p. 1971. The conditions of growth of the Wenlock reefs of Shropshire (England). Sediment ology,
17, 173-219.

SHAVER, R. H.. AULT, C. H., BURGER. A. M., CARR, D. D., DROSTE, J. B., EGGERT, D. L., GRAY, H. H., HARPER, D.,
HASENMUELLER, N. R., HASENMUELLER. W. A., HOROWITZ, A. S., HUTCHISON, H. C., KEITH, B. D.. KELLER, S. J.,
patton, j. b., rexroad, c. b. and wiER, c. e. 1986. Compendium of Paleozoic rock-unit stratigraphy in
Indiana -a revision. Department of Natural Resources (Indian), Geological Survey Bulletin, 59. Geological
Survey of Indiana, Bloomington, 203 pp.

Sinclair, g. w. 1940. A discussion of the genus Metaconularia with descriptions of new species. Royal Society
of Canada, Section IV, Transactions, Series 3, 34, 101 121.

— 1948. The biology of the Conularida. Unpublished Ph.D. thesis, McGill University.

slater, i. l. 1907. A monograph of the British Conulariae. Monograph of the Paleontographical Society, 61
(295), 1-40.

steinmann, g. and doderlein, l. 1890. Elemente der Palaontologie. Verlag von Wilhelm Engelmann, Leipzig.
svoboda, j. 1966. Regional geology of Czechoslovakia. Part I. The Bohemian Massif. Geological Survey of
Czechoslovakia, Prague, 668 pp.

termier, h. and termier, g. 1949. Position systematique et biologie des Conulaires. Revue Scientifique, 86,
711-722.

1953. Les Conulariides. 1006-1013. In piveteau, j. (ed.). Traite de paleontologie. Tome 3.
Onychophores , arthropodes, echinodermes, stomocordes. Masson et Cie, Paris, 1063 pp.
titus, r. 1982. Fossil communities of the Middle Trenton Group (Ordovician) of New York State. Journal of
Paleontology. 56, 477-485.

— 1986. Fossil communities of the Upper Trenton Group (Ordovician) of New York State. Journal of
Paleontology. 60, 805-824.

— and cameron, b. 1976. Fossil communities of the Lower Trenton Group (Ordovician) of central and
northwestern New York State. Journal of Paleontology, 50, 1209 1225.

van iten, h. 1989. Relic conulariid soft parts and their bearing on conulariid affinities. Geological Society of
America, Abstracts with Programs, 21, 256.

— 1991 a. Repositories of and additional comments on sponge specimens containing molds of Paraconularia.
Journal of Paleontology, 65, 333-335.

— 1991 b. Evolutionary affinities of conulariids. 145-155. In simonetta, a. m. and conway morris, s. (eds).
The early evolution of Metazoa and the significance of problematic taxa. Cambridge University Press,
Cambridge, ix + 296 pp.

watkins, R. 1978. Bivalve ecology in a Silurian shelf environment. Lethaia , 11, 41-56.
werner, b. 1966. Stephanoscyphus (Scyphozoa, Coronatae) und seine directe Abstammung von den fossilen
Conulata. Helgoldnder Wissenschaftliche Meeresuntersuchungen , 13, 317-347.

— 1967. Stephanoscyphus Allman (Scyphozoa, Coronatae), ein rezenter Vertreter der Conulata?
Palaontologische Zeitschrift , 41, 137 153.

williams, d. a. and telford, p. g. 1986. Paleozoic geology of the Ottawa area. Geological Association of
Canada, Mineralogical Association of Canada, Canadian Geophysical Union, Joint Annual Meeting,
Ottawa 1986, Field Trip 8: Guidebook, 25 pp.

HEYO VAN ITEN
Museum of Paleontology
University of Michigan
Ann Arbor, MI 48109, USA

Typescript received 8 May 1990
Revised typescript received 13 August 1990

952

PALAEONTOLOGY, VOLUME 34

APPENDICES

The following three appendices list conulariid specimens used as data in the present study. Abbreviations of
repositories housing these specimens are as follows: AC, Augustana College, Rock Island, Illinois; AMNH,
American Museum of Natural History, New York; BMNH. British Museum (Natural History), London; BM-
UW, Burke Museum, University of Washington, Seattle; BMS, Buffalo Museum of Science, Buffalo; BUGM,
Birmingham University Geology Museum, Birmingham; CM, Carnegie Museum of Natural History,
Pittsburgh; CMNH, Cleveland Museum of Natural History, Cleveland; FMNH, Field Museum of Natural
History, Chicago; GSC, Geological Survey of Canada, Ottawa; ISGS/ISM, Illinois State Geological
Survey/Illinois State Museum, Champaign/Urbana ; IUPC, Indiana University, Paleontological Collections,
Bloomington; NYSM, New York State Museum and Science Service, Albany; SUE State University of Iowa,
Iowa City; UMMP, University of Michigan Museum of Paleontology, Ann Arbor; UMMP*, University of
Montana Museum of Paleontology, Missoula; USNM, United States National Museum, Washington;
UWGM. University of Wisconsin Geological Museum, Madison.

APPENDIX 1

List of conulariid specimens terminating in a schott and examined for the presence of schotts within the test
cavity. Species followed by an asterisk are preserved as sandstone steinkerns. N = number of specimens
examined.

Species

Repository

N

Anaconularia anomala* (Barrande)

MCZ

144

Archaeoconularia consobrina* (Barrande)

MCZ

5

A. clubia (Sinclair)

ROM 18889

1

A. pyramidata* (Hoeninghausen)

FMNH PE294, PE379-382;

GSC 87183; ROM 22289, 22532

9

Conularia multicostata Meek and Worthen

AC 14160, 14164

2

C. quichua Ulrich

USNM 373992

1

C. sp.

UMMP 259

1

C. sp.

ISGS/ISM 4434

1

C. sp.

SUI 62615a

1

C. splendida Billings

SUI 61519-61520

2

C. subcarbonaria Meek and Worthen

FMNH UC 18494, UC19647;
ISGS/ISM 2609; MCZ 27951,
27954-27955

7

C. trentonensis Hall

AMNH 29650, 42316; BMS 10804;
NYSM 3492; ROM 67, 23278,
23737, 23738, 24007, 28324, 24917;
SUI 61502; UMMP 66012-66016;
UMPC W1991

25

Metaconularia calderi Sinclair

GSC 9794-9795

2

M. sp.

ROM 87DR

1

M. divisi Sinclair

SUI 53089, 62678-62679

3

Paraconularia byblis (White)

UMMP 2167

2

VAN ITEN: CONULARIID SCHOTT

953

APPENDIX 2

List of conulariid specimens not terminating in a schott and examined for the presence of schotts within the
test cavity. The letter P following a specimen number or numbers indicates that the corresponding specimen
or specimens taper to a test width of 1 mm or less.

Species

Repository

Archaeoconularia granulata (Hall)
Conularia splendida Billings

C. trentonensis Hall

Metaconularia manni (Roy)

Paraconularia chagrinensis Babcock and Feldmann
P. chesterensis (Worthen)

P. recurvatus Babcock and Feldmann
P. subulata (Hall)

P. yochelsoni Babcock and Feldmann

AMNH 7 9 1 ( P ) (2 specimens)

SUI 499793(P) (1 specimen), SUI
6151 1 —6 1 5 1 8( P) (8 specimens)

AMNH 29649(P) (1 specimen); SUI 61506 (1
specimen); UWGM WW4001(P) (113
specimens)

FMNH PE6252-6256(P), PE101 32(P),
PE23674-23675(P), unnumbered(P) (12
specimens)

CMNH 1788(P), 6633 (3 specimens)

GSC 49383 (1 specimen)

USNM 409806 (1 specimen)

CM 34521, 34524 (2 specimens); CMNH 1788,
6633 (3 specimens); UMMP* 561 3A/I7106,
5628NC/MI7106 (2 specimens)

UMMP 45499(P), 45500(P) (3 specimens)

954

PALAEONTOLOGY, VOLUME 34

APPENDIX 3

Specimen numbers for samples listed in Table I.

Species

Repository

Depositional Regime 1

Conularia

C. trentonensis Hall (Collingwood Fm.)

C. trentonensis Hall (Brandon Bridge Fm.)
Metaconularia

M. aspersa (Slater)

M. manni (Roy)

Paraconularia

P. chagrinensis Babcock and Feldman

Depositional Regime 2

Conularia

C. sowerbyii (Slater)

C. subcarbonaria (Meek and Worthen)

C. trentonensis Hall (Verulam Fm.)

C. trentonensis Hall (Denley Fm.)

C. trentonensis Hall (Sherwood Mbr.)
Metaconularia

M. calderi Sinclair

M. divisa Sinclair
M. sp.

Depositional Regime 3

Anaconularia

A. anomala (Barrande)
Archaeoconularia

A. consobrina (Barrande)

A. pyramidata (Hoeninghausen)

Conularia

C. subcarbonaria Meek and Worthen
C. trentonensis Hall
Paraconularia
P. sp.

ROM 27274 (40 specimens)

UWGM WW4001 (159 specimens)

BMNH G4603, G5373, G12879 (3 specimens)
FMNH PE6252-6256, PE10132,
PE23674-23975, FMNH unnumbered
(13 specimens)

CMNH 1247, 1272, 1427, 1622, 1674, 1788,
1818, 4030, 4292, 6633, 6717, 6807-6808
(13 specimens)

BMNH 6275, 6327, 10043, 11795,

17500-17505, unnumbered (14 specimens);
UBGM ‘Holcroft Pteropoda’ (10 specimens)
FMNH UC 18494, UC19647, FMNH
unnumbered (8 specimens); ISGS/ISM 2609
(1 specimen); ISM 2609 (1 specimen); IUPC
17482 (1 specimen)

ROM 67, 23738 (512T), 24917 (240U) (9
specimens); GSC 1725-1726 (2 specimens)
UMMP 66012-66022 (1 1 specimens)

SUI 61502-61505 (8 specimens)

GSC 9794-9795 (2 specimens); Sinclair (1940,
pi. 3, figs 3-5; 1 specimen)

BM-UW (2 specimens); SUI 53089,
62678-62679 (3 specimens); UMPC W1994 (3
specimens); Sinclair (1948, pi. 8, figs 12-14; 1
specimen)

ROM 87DR (1 specimen)

MCZ (760 specimens)

MCZ (10 specimens)

BMNH 3408-3409 (9 specimens); FMNH
PE294, PE379-382, FMNH unnumbered (7
specimens); GSC 87180 (1 specimen); ROM
22289, 22532 (3 specimens)

MCZ 27946-27956 (II specimens)

BMS El 0807 (3 specimens)

USNM 57164, 99502 (4 specimens); ISGS
(ISM) 10742 (4291) (1 specimen); IUPC 6071
( 1 specimen)

HENSONELLA DINA RICA, AN ORIGINALLY
CALCITIC EARLY CRETACEOUS
DASYCLADACEAN ALGA

by M. D. SIMMONS, D. EMERY and N. A. H. PICKARD

Abstract. Hensonella dinarica (Radoicic) is an Early Cretaceous, Tethyan microfossil, considered to be either
a dasycladacean alga or a problematicum. This is because its preservation in yellowish radial calcite is highly
unusual for dasycladacean algae, which are typically preserved in drusy calcite replacing the original aragonite.
Petrological, cathodolummescence and chemical microprobe studies of specimens of H. dinarica , principally
from the Middle East, suggest that the original mineralogy of this microfossil was calcitic. The morphology
of this species is suggestive of a dasycladacean alga, and so it is considered to be an originally calcitic
dasycladacean alga, an unusual phenomenon.

Hensonella dinarica (Radoicic) is a microfossil which is often found in shallow water. Early
Cretaceous (Hauterivian-late Aptian), Tethyan sediments, particularly those of the Middle East
and Mediterranean regions (Bassoullet et al. 1978; Simmons 1990). The species is most typical of
the Barremian-Aptian period in southern Tethys. Although originally considered to be a
dasycladacean alga ( Salpingoporella dinarica Radoicic, 1959), other authors have questioned this
taxonomic assignment (usually with reference to Hensonella cylindrical Elliott, 1960, junior synonym
of S. dinarica ), because of the unusual preservation of specimens of this species. In thin-section it
is typically preserved in fibrous, radial calcite with a yellowish tint and has a dark, micritic, inner
layer to the calcitic wall (Text-fig. 1a-c). Dasycladacean algae usually have an originally aragonitic
mineralogy which is replaced by drusy calcite during diagenesis (Wray 1977). The preservation of
H. dinarica suggests an originally calcitic mineralogy. In order to determine its original mineralogy,
conventional transmitted light and cathodoluminescence petrography have been employed, together
with microchemical techniques using an electron microprobe.

TECHNIQUES AND RESULTS

The main specimens examined in this study are from the Early Cretaceous Kahmah Group (Glennie et al. 1974)
of the Oman Mountains, notably the Late Barremian and Early Aptian Kharaib and Shuaiba Formations. For
further details of associated fauna, flora and sedimentology see Simmons and Hart (1987), Hughes Clarke
(1988) and Simmons (1990).

Topotype specimens of S', dinarica from the Early Cretaceous sediments of the Dinarides, Yugoslavia, were
kindly made available by Dr M. A. Conrad. The type specimens of H. cylindrica from the Early Cretaceous
sediments of the Middle East were studied at the British Museum (Natural History), and additional material
from the Middle East and the Mediterranean was studied from the reference collections of the BP Research
Centre, Sunbury-on-Thames.

In addition to conventional transmitted light petrography, cathodoluminescence petrography and
microchcmical techniques were used to ascertain the present and original mineralogy of H. dinarica. Polished
thin-sections containing the microfossil were examined firstly in cathodoluminescent light (Marshall 1987) for
evidence of diagenetic alteration (Popp et al. 1986). Neither the algae nor surrounding calcite matrix and
cement luminesced in any of the samples examined. This indicates that the algae had not been replaced by
luminescent calcite, but it is not unequivocal evidence of a primary calcite mineralogy.

The same samples examined in cathodoluminescent light were analysed on an electron microprobe for

| Palaeontology, Vol. 34, Part 4, 1991, pp. 955-961. |

© The Palaeontological Association

956

PALAEONTOLOGY, VOLUME 34

text-fig. 1. a-c, Hensonella dinarica (Radoicic). a, WG23; Early Aptian, Shuaiba Formation; Oman;
transverse section showing inner micritic wall and radial calcitic structure, x 80. b, WM105; Hauter-
ivian-Barremian, Kharaib Formation; Oman; transverse section showing radial structure and dasycl-
adacean verticils, x 80. c, RR019440/A309; Barremian-Aptian ; Zeta Plain, Yugoslavia; oblique transverse
section of a topotype of Salpingoporella dinarica Radoicic, showing dasycladacean verticils and faint radial
calcitic structure, x 60. Specimens WG23 and WM105 are held in the Biostratigraphy Reference Collection,
BP Research Centre, Sunbury-on-Thames, UK; specimen RR018440/A309 is held in the collection of

Petroconsultants, Geneva, Switzerland.

residual strontium and magnesium. High concentrations of strontium normally indicate a precursor aragonite
mineralogy (Sandberg 1985). Modern aragonitic algae contain up to several hundred to a few thousand ppm
strontium. High concentrations of magnesium, which may be sited in microdolomites (Lohmann and Meyers
1977; Meyers 1978) after diagenetic alteration, are indicative of an originally high magnesium calcite
mineralogy.

The microprobe analyses of specimens of H. dinarica were obtained on a Cameca Camebax Microbeam
system, operating with an accelerating voltage of 20 kV and beam current of 10 nA and a count time of 30
sec. The beam was rastered over a 7-5 /um square to reduce volatilization of CO.,. Twenty-three spot analyses
were made of three samples. Four spot analyses were also made of the calcite cement. In addition to the spot
analyses, a line scan was made across one alga and through the surrounding cement. Table 1 summarizes the
microprobe results.

The analyses show the alga to be composed of low magnesium calcite (i.e. < 4 mole% MgC03). There is no
evidence for high concentrations of strontium or magnesium in any portion of the alga. It is impossible to rule
out completely a high magnesium calcite precursor, because high magnesium calcite can lose magnesium
without any discernible petrographic change (Friedmann 1964; Towe and Hemleben 1976). Moreover, the
magnesium levels recorded in H. dinarica (Tabic 1) are not atypical of those quoted for original high

SIMMONS ET AL.: CRETACEOUS DASYCLADACEAN ALGA

957

table 1 . Magnesium and strontium content of algae and cement from microprobe spot analyses.

mole % MgCO, Sr (ppm)

N

Specimen 1

MO

508

9

Specimen 2

1.14

431

9

Specimen 3

105

321

5

Average

1 14

437

23

Calcite cement

116

523

4

magnesium calcite cements which have converted to diagenetic low magnesium calcite (e.g. see values quoted
in Lochmann and Meyers 1977; Davies 1977; Marshall and Ashton 1980; Videtich 1985; Sailer 1986;
Carpenter and Lohmann 1989; Mazzullo et al. 1990). However, the algae lack microdolomite inclusions which
are commonly formed during the stabilization of high magnesium calcite (cf. Lohmann and Meyers 1977;
Davies 1977; Meyers 1978; Carpenter and Lohmann 1989). Combined with the lack of replacement by drusy
calcite and albeit equivocal luminescence evidence, this strongly suggests that the alga had an originally calcitic
and probably low magnesium calcite mineralogy. The radial fibrous calcite of the thallus and its dark inner
micritic layer are hence interpreted as representing the primary mineralogy of H. dincirica.

DISCUSSION

Previous interpretations

Hensonella dinarica was first described by Radoicic ( 1959), from Early Cretaceous sediments of the
Dinarides, Yugoslavia, in the genus Salpingoporella , and considered to be a dasycladacean alga.
Independently, Elliott (1960) described H. cylindrica from the Early Cretaceous of the Middle East,
and from its preservation (see below) classified it as a problematicum, probably a scaphopod. Most
subsequent workers, including Elliott (1968; pers. comm. 1988), have agreed that H. cylindrica is a
junior synonym of S. dinarica. Johnson (1969) argued that the distinctive preservation of H.
cylindrica is not seen in S. dinarica and kept the two species separate: H. cylindrica being a
problematicum, S. dinarica a dasycladacean alga. However, Elliott (1968, p. 77) still expressed
doubts about its algal origin: ‘In conclusion, I consider this organism is best classified as a
problematicum’. In contrast to this, Bassoullet et al. (1978), Conrad (pers. comm. 1988) and
Radoicic (pers. comm. 1988) are in no doubt that S. dinarica and H. cylindrica are synonyms and
refer to a dasycladacean alga.

Luperto Sinni and Masse (1982, 1984, 1986) placed S. dinarica in synonymy with H. cylindrica ,
but under the generic name Hensonella in recognition of its unusual preservation. Like Elliott
(1968), they suggested (although did not demonstrate) that H. dinarica was originally calcitic.
Furthermore, they considered it to be originally high magnesium calcite, although admitting that
this was difficult to determine on petrological characteristics alone. The microprobe results
presented herein suggest that the species was originally formed of low magnesium calcite. Luperto
Sinni and Masse (1982) suggested that the species might be typical of inner shelf environments with
high salinities, ‘continental influence’ and fluctuations in the Mg/Ca ratio. In support of this, they
noted that Jaffrezo and Renard (1979) suggested that Zergabriella embergeri (Bouroullec and
Deloffre) and Clypeina jurassica Favre, species also considered by Luperto Sinni and Masse (1982)
to be originally calcitic, were typical of similar inner shelf conditions. However, our studies suggest
that H. dinarica can be found in a variety of shelf environments and is not particularly indicative
of hypersaline inner shelf conditions.

Morphological characteristics

The reasons for the uncertainty over the taxonomic position of H. dinarica centre around the
crystalline structure of the thallus wall as alluded to by Radoicic (1959), and more fully discussed
by Elliott (1960). Three features are noticeable: (i) the wall appears to be formed of radial, fibrous

958

PALAEONTOLOGY, VOLUME 34

calcite, (ii) this has a yellowish or honey coloured tint, (iii) a dark, inner micritic layer to the thallus
is typically present. These features are unusual for a dasycladacean alga, and were used by Elliott
(1960, 1968) to argue for a non-algal origin for this fossil. Typical Salpingoporella , and other
dasycladaceans, have a thallus wall composed of drusy, unstructured calcite, replacing the original
aragonite. Elliott (1968) was the first to suggest that the preservation of H. dinarica related to an
original, organically formed, calcite structure. Luperto Sinni and Masse (1982) also agreed with this
interpretation, while Bassoullet et al. (1978) and Radoicic (1959) placed little significance in the form
of preservation, arguing that the radial structure originates from diagenetic alteration. Our results
confirm the assertion of Elliott (1968) that H. dinarica was originally calcitic.

Examination of the type figures of S. dinarica and H. cylindrica strongly suggests that the two
species are synonymous. This is confirmed by comparison of topotypes of S. dinarica and the
syntypes of H. cylindrica. As stated by Elliott (1968, p. 76) 'Slight differences in the two authors’
descriptions can be reconciled by examination of large sets of specimens’.

Specimens described as S. dinarica and H. cylindrica show considerable variability, not only in
the nature of the wall, but also in terms of morphology. Some topotype specimens of S. dinarica
clearly show a fibrous, radial wall, whilst in others this is poorly developed. The topotypes do not
show a particularly strong yellowish tint, as do the syntypes of H. cylindrica. More importantly, the
topotype specimens of S. dinarica (and indeed the holotype illustrations) resemble dasycladacean
algae, especially the genus Salpingoporella. Features typical of dasycladacean algae are not clear in
the syntypes of H. cylindrica , but as noted and illustrated by Elliott (1968), they do occur ‘not
uncommonly’ in Middle eastern specimens referred to H. cylindrica. This is confirmed by
examination of specimens from the Kahmah Group of the Oman Mountains. Much of the
morphological variation associated with this species can be attributed to preservation problems
(erosion, etc.), and variability with depth and orientation of thin section.

Of particular note is the presence of 'pores’ which are regularly spaced, alternating in position
through progressively higher levels in the thallus. These can be interpreted as verticils of lateral
branches (see Elliott 1968, plate 22, fig. 2; Text-fig. lc). Overall the morphology is closely
comparable with that of 5. muehlbergii (Lorenz), the type species of Salpingoporella. The thallus is
unsegmented with a wide axial hollow and only primary branches are present. The dark inner
micritic layer to the thallus wall (thickness 0 012-0 018 mm) can be interpreted as being the
preserved organic membrane lining the main axis of the thallus. Elliott (1968) noted that crushed
specimens are still held together by this layer, suggesting that it had an original organic nature with
some flexibility.

Other records of algae which show all the features of dasycladaceans, but are considered to be
originally calcitic include the Visean ‘dasycladacean’ alga, Koninckopora (Wright 1981). This fossil
is preserved in very fine-grained, sometimes acicular calcite. Like H. dinarica , Koninckopora often
displays a micritic lining to the thallus wall. Other possible examples include the Late Jurassic
species Salpingoporella sel/ii (Crescenti) and Griphoporellal minima Nikler and Sokac, both
sometimes preserved in fine radial calcite. Elliott (1963) considered that Pseudovermiporella Elliott,
a questionable Permian dasycladacean, was of calatic origin. Some species of the Late Jurassic-Early
Cretaceous dasycladacean Zergabriella embergeri display fibro-radial walls with a yellowish tint and
a micritic inner layer to the thallus wall, as in H. dinarica (see illustrations in Granier 1989). Further
research is needed to establish how well developed is this feature in this species, and if it relates to
an original calcitic mineralogy as suggested by Luperto Sinni and Masse (1982). These authors also
considered Clypeina jurassica to be originally calcitic, and (Luperto Sinni and Masse 1984, 1986)
placed Salpingoporella urladanasi Conrad, Peybernes and Radoicic in the genus Hensonella on
account of its similar preservation to H. dinarica , but differing in its larger dimensions.

Thus whilst the majority of fossil dasycladaceans were originally aragonitic, there may have been
a few taxa which were originally calcitic. Pending further research to establish the number of calcitic
‘dasycladaceans’, and the reasons for their unusual mineralogy, no new suprageneric taxonomic
group is established here. Interestingly, the majority of occurrences of possible originally calcitic
‘dasycladaceans’ is within the Late Jurassic Early Cretaceous and Early Carboniferous. These were

SIMMONS ET AL.: CRETACEOUS D AS YCLADACEAN ALGA

959

periods when, according to Sandberg (1983), non-skeletal carbonates were dominated by calcite
rather than aragonite (‘ aragonite inhibiting episodes ’). These may have been brought about by plate-
tectonically influenced oscillations in the vapour pressure of C02, which in turn correlate with other
known global oscillations in eustacy and climate. However, it should be stated that originally
aragonitic dasycladaceans are also abundant during these periods, and it is uncertain whether the
occurrence of originally calcitic dasycladaceans can simply be attributed to an ‘aragonite inhibiting’
process.

It is worth noting that other mineral phases besides aragonite are known within dasycladaceans.
For example, calcium oxalate has been recorded in reproductive discs of Acetabularia (Elliott 1979).

CONCLUSIONS

The preservation of H. dinarica relates to a primary organic calcitic mineralogy, although the
morphology of the species is in keeping with the dasycladacean genus Salpingoporella Pia. Such an
original mineralogy is almost unique in the family Dasycladaceae where primary mineralogy is
normally aragonitic (Wray 1977). In fossils this is usually replaced by drusy calcite. The genus
Hensonella Elliott is therefore retained to identify originally calcitic homeomorphs of Salpingo-
porella. Further research is in progress which may lead to the identification of other originally
calcitic ‘dasycladacean’ taxa. If further taxa are identified, it may be necessary to erect a new
suprageneric group to accommodate such forms (e.g. ‘ Hensonellaceae ’), or to amend the diagnosis
of the Dasycladaceae.

The recognition that H. dinarica was originally calcitic, not aragonitic, has important implications
for diagenetic studies of the carbonates in which this microfossil occurs (e.g. the oil-bearing Shuaiba
Formation of the Middle East). Because it was originally calcitic it would not have been leached
during meteoric diagenesis, resulting in possible underestimations of the amount of meteoric
diagenesis in previous studies.

Acknowledgements . The authors wish to thank Drs G. F. Elliott, M. A. Conrad, R. Radoicic and Prof. F. T.
Banner for their helpful advice and for providing access to specimens which have been studied during the
course of this work. Dr S. Kearns assisted with the microprobe analysis. Dr F. Barattolo also provided
helpful advice and Drs R. W. Jones, I. D. Somerville, D. Edwards, and two anonymous reviewers suggested
improvements to the manuscript. Permission to publish this paper has kindly been granted by BP Research.

REFERENCES

bassoullet, j. p., bernier, p., conrad, m. a., deloffre, r. and jaffrezo, m. 1978. Les algues dasycladales du
Jurassique et du Cretace. Geobios Memoire Special , 2, I -330.
carpenter, s. J. and lohmann, k. c. 1989. KlsO and K13C variations in Late Devonian marine cements from
the Golden Spike and Nevis reefs, Alberta, Canada, Journal of Sedimentary Petrology , 59, 792-814.
davies, G. R. 1977. Former magnesian calcite and aragonite submarine cements in Upper Palaeozoic reefs of
the Canadian Arctic. A summary. Geology , 5, 11-15.

elliott, G. f. 1960. Fossil calcareous algal floras of the Middle East, with a note on a Cretaceous
problematicum, Hensonella cylindrical gen. et sp. nov. Quarterly Journal of the Geological Society of London ,
115,217-232.

— 1963. Problematical microfossils from the Cretaceous and Palaeocene of the Middle East. Palaeontology ,
6. 46 — 48.

1968. Permian to Palaeocene calcareous algae (Dasycladaceae) of the Middle East. Bulletin of the British
Museum ( Natural History), Geology, 4, 1 111.

1979. Taxonomy and opercular function of the Jurassic alga Stichoporella. Palaeontology , 22, 407-412.
Friedmann, G. M. 1964. Early diagenesis and lithification in carbonate sediments. Journal of Sedimentary
Petrology , 34, 777-813.

960

PALAEONTOLOGY, VOLUME 34

GLENNIE, K. W., BOUEF, M. G. A., HUGHES CLARICE, M. W., MOODY STUART, J., PILAAR, W. F. H and REINHARDT, B. M.

1974. Geology of the Oman Mountains. Verhandelingen van het Koninklijk Nederlands Geologisch
Mijnbouwkundig Genootschap , 31 (3 volumes), 423 pp.

granier, b. 1989. Zergabriella , un nouveau genre d’algue dasycladale du Portlandien-Valanginien. Revue de
Micropaleontologie , 32. 126-133.

hughes clarre, m. w. 1988. Stratigraphy and rock-unit nomenclature in the oil producing area of interior
Oman. Journal of Petroleum Geology , 11, 5-60.

jaffrezo, m. and renard, m. 1979. Elements en traces de calcaires a Dasycladales et Charophvtes. Bulletin des
Centres de Recherches Exploration- Production Elf- Aquitaine, 3, 639-649.

Johnson, j. h. 1969. A review of the Lower Cretaceous algae. Professional Contributions of the Colorado School
of Mines , 6. 1-180.

lohmann, k. c. and meyers, w. j. 1977. Microdolomite inclusions in cloudy prismatic calcites: a proposed
criterion for former high magnesian calcites. Journal of Sedimentary Petrology , 47, 1078-1088.
luperto sinni, e. and masse, j. p. 1982. Contributo della paleoecologia alia paleogeografia della parte
meridionale della piattaforma Apula nel Cretaceo inferiore. Geologica Romana , 21, 859-877.

— 1984. Donnees nouvelies sur la micropaleontologie et la stratigraphie de la partie basale du ‘Calcare
di Bari’ (Cretace inferieur) dans la region des Murges (Italie meridionale). Rivista Indiana di Paleontologia
e Stratigrafia , 90, 331-374.

— 1986. Donnees nouvelies sur la stratigraphie des Calcaires de plateforme du Cretace inferieur du
Gargano (Italie meridionale). Rivista Italiana di Paleontologia e Stratigrafia , 92, 33-66.
marshall, d. j. 1987 Cathodoluminescence of geological materials. Allen and Unwin, London, 172 pp.
marshall, j. d. and ashton, m. 1980. Isotopic and trace element evidence for submarine lithification of
hardgrounds in the Jurassic of eastern England. Sediment ology, 27, 271-289.
mazzullo, s. J., bischoff, w. d. and lobitzer, h. 1990. Diagenesis of radiaxial fibrous calcites in a
subunconformity, shallow-burial setting: Upper Triassic and Liassic, Northern Calcareous Alps, Austria.
Sediment ology, 37, 407-425.

meyers, m. j. 1978. Carbonate cements: their regional distribution and interpretation in Mississippian
limestones of Southwestern New Mexico. Sedimentology , 25, 371 — 400.
popp, b. n., anderson, t. f. and Sandberg, p. a. 1986. Textural, elemental and isotopic variations among
constituents in Middle Devonian Limestones, North America. Journal of Sedimentary Petrology , 56,
715-727.

radoicic, R. 1959. Salpingoporella dinarica nov. sp. dans les sediments Cretaces Inferieurs des Dinarides.
Geoldski Glasnik, 3, 33-42.

saller, a. h. 1986. Radiaxial calcite in Lower Miocene strata, subsurface Enewetak Atoll. Journal of
Sedimentary Petrology, 56, 743-762.

Sandberg, p. a. 1983. An oscillating trend in Phanerozoic non-skeletal carbonate mineralogy. Nature, 305,
19-22.

— 1985. Aragonite cements and their occurrence in ancient limestones. Special Publication of the Society of
Economic Paleontologists and Mineralogists, 36, 33-57.

simmons, m. d. 1990. Aspects of the micropalaeontology and stratigraphy of Cretaceous shelf carbonates from
the Oman Mountains. Unpublished Ph.D. thesis. Polytechnic South West, Plymouth, United Kingdom.

— and hart, m. b. 1987. The biostratigraphy and microfacies of the early to mid-Cretaceous carbonates of
Wadi Mi’aidin, central Oman Mountains. 176-207. In hart, m. b. (ed.). Micropalaeontology of carbonate
environments. Ellis Horwood, Chichester, 296 pp.

towe, k. m. and hemleben, c. 1976. Diagenesis of magnesian calcite: evidence from Miliolacean foraminifera.
Geology, 4, 337-339.

videtich, p. e. 1985. Electron microprobe study of Mg distribution in recent Mg calcites and recrystallized
equivalents from the Pleistocene and Tertiary. Journal of Sedimentary Petrology , 55, 421-429.
wray, j. l. 1977. Calcareous algae. Elsevier, Amsterdam, 185 pp.

wright, v. p. 1981. Ultrastructure and early diagenesis of the Visean alga Koninckopora. Palaeontology, 24,
185-193.

M. D. SIMMONS

Exploration Technology Branch
BP Research Centre
Sunbury on Thames
Middlesex TW16 7LN, UK

SIMMONS ET AL.\ CRETACEOUS DASYCLADACEAN ALGA

961

D. EMERY

BP Exploration
Britannic House
Moor Lane

London EC2Y 9BU, UK

N. A. H. PICKARD

Grant Institute of Geology
University of Edinburgh
West Mains Road
Edinburgh EH9 3JW, UK

Present address:

Department of Geology
University College

Typescript received 9 August 1990 Bclfield

Revised typescript received 22 March 1991 Dublin 4, Eire

THE RELATIONSHIP OF THE MESOZOIC BIVALVE
ATRETA TO THE DIMYIDAE

by P. HODGES

Abstract. Material recently collected from the Lower Lias (Lower Jurassic) of South Wales shows evidence
of dimyarian musculature in Atreta intusstriata (Emmrich) proving that the previous assignment of Atreta to
the monomyarian Plicatulidae Watson, 1930, by Cox (1964) is incorrect. The genus is therefore reassigned to
the Dimyidae Fischer, 1886. A lectotype and paralectotypes are designated from material located in the
Emmrich collection at the Geiseltal Museum, Halle, Germany.

Etallon (1862) erected Atreta and described the main diagnostic features, namely two projecting
teeth, dimyarian musculature and anastomosing riblets in the right valve. Five species were listed
as belonging to this genus: Ostrea blandina d'Orbigny, 1849, Spondylus dichotomus Buvignier, 1852,
Plicatula striatissima Quenstedt, 1858, Crania humbertina Buvignier, 1852, and Atreta imbricata sp.
nov. The last two were described, but none was figured and no type species selected. Rollier (1917)
later invalidly selected Atreta jurensis Etallon, 1862, as type species.

The designation of Ostrea blandina as type species was made by Cox (1964), who considered
Etallon’s generic diagnosis extremely doubtful due to the generally poor preservation of the material
available. He noted also that several authors after Etallon had failed to observe any musculature
in species of this genus. Although he admitted that the main characteristics were allied to the
Dimyidae, he nevertheless placed Atreta in the Plicatulidae due to his doubts about the presence of
dimyarian musculature.

RELATIONSHIP OF THE DIMYIDAE TO OTHER BIVALVE FAMILIES

The Dimyidae comprises species which are suborbicular in outline, cemented by the right valve and
with dimyarian musculature. Thiele (1935), Newell in Moore (1969) and Vokes (1980) placed the
Dimyidae in the superfamily Pectinacea, which is otherwise made up entirely of monomyarian
families. Newell's classification is as follows:

Order pterioida Newell, 1965
Suborder pteriina Newell, 1965
Superfamily pectinacea Rafinesque, 1815

Families pterinopectinidae Newell, 1938; leiopectinidae Krasilova, 1959; aviculopectinidae Meek
and Hayden, 1864; deltopectinidae Dickins, 1957; pseudomonotidae Newell, 1938;
posidoniidae Freeh, 1909; oxytomidae Ichikawa, 1958; entoliidae Korobkov, 1960;
pectinidae Rafinesque, 1815; monotidae Fischer, 1887; buchiidae Cox, 1953; plicatulidae
Watson, 1930; spondylidae Gray, 1826; terquemiidae Cox, 1964; dimyidae Fischer, 1886.

Nevesskaya et at. ( 1971 ) proposed a classification based on an analysis of shell structure and soft
parts (gills and stomach) in extant bivalves. All are filter feeders, cemented or byssally attached.
Their classification was:

Order pectinoida Adams and Adams, 1857
Superfamily dimyacea Fischer, 1886
Family dimyidae Fischer, 1886

|Palaeontology, Vol. 34, Part 4, 1991, pp. 963-970, 1 pi. |

© The Palaeontological Association

964

PALAEONTOLOGY, VOLUME 34

Superfamily spondylacea Gray, 1826

Families terquemiidae Cox, 1964; plicatulidae Watson, 1930; spondylidae Gray, 1826.

Yonge (1975) noted major resemblances between the mantle, shell and viscero-pedal mass in
extant species of the Dimyidae and Plicatulidae and proposed the following classification:

Superfamily plicatulacea Yonge, 1975

Families dimyidae Fischer, 1886; plicatulidae Watson, 1930.

Waller (1978) disagreed with Yonge's interpretation of the ligament in the Dimyidae and
Plicatulidae and offered a new interpretation of primary ligament, shell structure and soft tissue in
extant species of the Pteriomorphia. Utilizing cladistic analysis a revised classification was
proposed, which is adopted here:

Order ostreoida Ferussac, 1 822
Suborder ostreina Ferussac, 1822
Superfamily dimyacea Fischer, 1886
Superfamily plicatulacea Watson, 1930
Superfamily ostreacea Ralinesque, 1815

SYSTEMATIC PALAEONTOLOGY

Repositories of specimens. NMW, National Museum of Wales, Cardiff; BMNFI, British Museum
(Natural History), London; BGS, British Geological Survey, Keyworth; BCM, Bristol City
Museum, Bristol; GM, Geiseltal Museum, Halle, Germany.

Superfamily dimyacea Fischer, 1886
Family dimyidae Fischer, 1886
Genus atreta Etallon, 1862

Synonyms. Diploschiza Conrad, 1866; Cyclostreon Eichwald, 1868; Dimyopsis Bittner, 1895; (fide Cox 1964).

Type species. Ostrea blandina d'Orbigny, 1849, p. 375, from the Oxfordian of France; subsequent designation
of Cox 1964, p. 45.

Range. Early Rhaetian to Campanian of Europe and North America.

Diagnosis. Small, suborbicular, often slightly oblique; right valve attached by the greater part of its
surface, internally shallowly concave, bordered by a raised ridge, inner shell layer generally missing
exposing inner surface of outer shell layer with divaricating or anastomosing riblets ending in some
species as transverse crenulations of the raised ridge; left valve almost flat with commarginal
lamellose ornament, internal of valve showing suboval dimyarian musculature, resilium pit small,
crura short and thin.

Atreta intusstriata (Emmrich, 1853)

Plate 1, figs 1 — 10

+ . 1851 cf. Ostrea placunoides Munster; Schafhautl, p. 413, pi. 7, fig. la-c.
v* 1 853 Ostrea intusstriata Emmrich, p. 377.

. 1855 Spondylus liasinus Terquem, p. 327, pi. 23, fig. la d.

.1861 Ostrea interstriata Emmrich; Moore, p. 501, pi. 16, fig. 25.

HODGES: MESOZOIC BIVALVE ATRETA

965

.1864 Plicatula intus-striata (Emmrich); Dumortier, p. 74, pi. 1, figs 13-16.

. 1865 Plicatula liasina (Terquem); Terquem and Piette, p. 107, pi. 13, figs 11 13.

. 1865 Plicatula lotharingiae Terquem and Piette, p 109, pi. 13, figs 14 and 15.

. 1865 Plicatula parkinsoni Bronn; Terquem and Piette, p. 108, pi. 13, fig. 16.

. 1866 Plicatula intusstriata (Emmrich); Capellini, p. 484, pi. 6, fig. 12.

. 1867 Plicatula intusstriata (Emmrich); Stefano, p. 147, pi. 6, figs 35 and 36.

1876 Plicatula liasina (Terquem); Tate and Blake, p. 369.

1884 Spondylus liassinus (Terquem); Simpson, p. 178.

1884 Spondylus intus-striatus Tate; Simpson, p. 178.

1893 Plicatula intusstriata (Emmrich); Greco, p. 128.

1907 Plicatula intusstriata (Emmrich); Joly, p. 23.

1916 Dimyopsis ( Plicatula ) intusstriata (Emmrich); Goetel, p. 158.

1917 Plicatula (Atreta) intusstriata (Emmrich); Rollier, p. 536.

1917 Plicatula ( Atreta ) lotharingiae (Terquem and Piette); Rollier, p. 536.

1917 Plicatula ( Atreta ) liasina (Terquem); Rollier, p. 536.

1917 Plicatula (Atreta) ambigua Rollier, p. 537.

1929 Plicatula ( Dimyopsis ) intusstriata (Emmrich); Lanquine, p. 60.

. 1933 Dimyodon intus-striatus (Emmrich); Arkell, pi. 29, fig. 5.

1936 Plicatula intusstriata (Emmrich); Joly, p. 95.

Lectotype. Designated herein; GM MLU2/GM/EIS . 1, (plaster cast NMW 90.16G.2) attached right valve
from the ‘Gervillienbildung’ (Rhaetian) of Kossen, Austria.

Dimensions. Height 1 1 mm, length 9 mm, anterior length 4 mm.

Paralectotypes. GM MLU1/GM/HS . 1-10, (plaster cast NMW 90.16G.1) from the ‘Gervillienschichten’
(Rhaetian) of Sonntagshorn mountain, Bavaria, Germany; GM MLU3/GM/EIS, (plaster cast NMW
90.16G.3) from the ‘Gervillienbildung’ (Rhaetian) of Eipelgraben bei Staudach, Bavaria, Germany; GM
MLU2/GM/HS .2, details as lectotype.

Material. NMW 15.I31.G9; NMW 20.349.G10; NMW 20.349.G16; NMW 20.362.G47; NMW
22.345.G384; NMW 22.345.G387-390; NMW 22.345.G392; NMW 22. 345. G40 1-409; NMW
46.311.G12; NMW 65.345.GR26; NMW 67.121.G88; NMW 67 . 121 . G 1 38 ; NMW 83. 22G. 145-152;
NMW 83.22G. 174; NMW 83. 22G. 176-182; NMW 90. 16G. 1-3; BMNH L77133 ; BMNH L77176; BMNH
L77194-L77195; BMNH L 1 8 1 04 ; BGS Zdl4; BGS 312; BGS 91845-91863; BCM 9964 (9 specimens).

Stratigraphical range. Early Rhaetian - early Sinemurian (resupinatum Subzone)

Geographical range. United Kingdom (Moore 1861 ; Tate and Blake 1876; Simpson 1884; Arkell 1933; Hodges
1987), France (Terquem 1855; Dumortier 1864; Terquem and Piette 1865), Belgium (Joly 1907, 1936),
Germany (Goetel 1916), Austria (Schafhautl 1851), Italy (Capellini 1866; Stefano 1867; Greco 1893)

Description. Externally oyster-like, generally small, suborbicular to subovate in outline, often oblique,
inequivalve, inequilateral, hinge line straight to slightly convex; right valve attached by most of its surface area,
hinge area rarely preserved; left valve rather flattened often following the contours of the underlying attached
valve, with external ornament of commarginal imbrications, valve margins closed.

Internally, right valve shallowly concave with raised ridge around the shell margin; inner shell layer missing
exposing inner surface of outer shell layer with ornament of numerous divaricating anastomosing fine riblets,
number increases with size, with up to four orders of bifurcation, originating at the umbo and terminating at
the margin of the raised ridge, giving a crenulated appearance on the crest of the raised ridge. No hinge
structures or muscle scars seen. Left valve has dimyarian musculature, anterior adductor muscle scar suboval
in outline, posterior adductor slightly larger, suborbicular in outline and bilobed. muscle scars symmetrically
positioned at about one third of the height below hinge line. Hinge structures not seen.

Remarks. Emmrich’s (1853, p. 376) description made no direct reference to any type material. He
mentioned only that his species was widespread in the ‘Gervillienbildung’ (Rhaetian) of the
Austrian Alps and was found at Partenkirchen, Kreuth and Sonntagshorn mountain. His main

966

PALAEONTOLOGY, VOLUME 34

collection housed at the Geiseltal Museum, Martin-Luther Universitat, Wittenberg, Halle,
Germany, contains three limestone blocks collected from Sonntagshorn, Kossen and a locality near
Staudach (pers. comm. Dr G. Krumbiegel). One block (GM MLU1/GM/HS) has one syntype, a
second (GM MLU2/GM/HS) two syntypes, and the last (GM MLU3/GM/HS) has ten specimens
encrusted on ' Ostrea haidingeriana ’ Emmrich.

Schafhautl (1851) described and figured three right valves of this species from the Rhaetian of the
southern Bavarian Alps, and compared them with the Muschelkalk species Ostrea placunoides
Munster; he did not erect a name. His specimens were housed in the Bayerische Staatsammlung fur
Palaontologie und Historische Geologie, Munich but were destroyed during the Second World
War. Emmrich (1853) made reference to Schafhautl’s description when describing material he
collected from the Rhaetian of the Austrian Alps which he named Ostrea intusstriata. Although he
produced no figure of his species, his description mentions quite clearly the anastomosing riblets on
the inside of the attached right valve and their termination on the raised ridge at the shell margin.
There can be little doubt, from the features he described and from his reference to Schafhautl’s
description and figures, that his material is conspecific with specimens common in the Rhaetian and
Lower Lias of Europe.

In Terquem’s (1855) description of Spondylus liasinus mention is made of two cardinal teeth and
muscle impressions below the beaks. A further mention of two cardinal teeth was made by Terquem
and Piette (1865) when describing their Plicatula lotharingiae.

The shape of the muscle scars in this species and in particular the bilobed posterior adductor show
a close resemblance to those seen in the extant species Basiliomya goreaui Bayer, 1971 and Dimya
corrugata Hedley, 1902. Both these species belong to the Dimyidae; they were discussed in full by
Yonge (1978). This provides further evidence of the close relationship between Atreta intusstriata
and extant species of the Dimyidae.

Mode of life. It is presumed that Atreta was an epifaunal filter-feeder which lived permanently
attached to the shells of other molluscs. It is found both in the marginal and off-shore facies of the
Lower Jurassic. It has been observed encrusting the bivalves Gryphaea arcuata , Pinna (Pinna)
similis , Plagiostoma giganteum and Antiquilima succincta. Large shells such as Plagiostoma can
exhibit encrusted Atreta to a density of up to twenty per 0-01 sq. m. It is often found in association
with other encrusting bivalves such as species of Liostrea.

EXPLANATION OF PLATE 1

Figs 1-10. Atreta intusstriata (Emmrich). 1, lectotype GM MLU2/GM/HS . I , (plaster cast NMW 90 . 16G . 2),
internal of right valve; ‘Gervillienbildung’ (Rhaetian); Kossen, Austria. 2, paralectotype GM
MLU2/GM/HS .2; details as in 1. 3, NMW 83.22G. 182; internal of right valve attached to Plagiostoma
giganteum Sowerby, showing anastomosing riblets; bucklandi Zone, Lower Lias; between Cwm Nash and
Nash Point, South Glamorgan, South Wales. 4, paralectotype GM MLU1/GM/HS, (plaster cast NMW
90.16G.1); internal of right valve; ' Gervillienschichten ’ (Rhaetian); Sonntagshorn Mountain, Bavaria,
Germany. 5, NMW 83 .22G .182; external view of left valve attached in life position, details as in 3. 6, NMW
83.22G. 178; internal of right valve attached to Plagiostoma giganteum Sowerby; bucklandi Zone, Lower
Lias; F6 km E of Nash Point, South Glamorgan, South Wales. 7, paralectotypes GM MLU3/GM/HS . 1-3,
(plaster cast NMW 90.16G.3); internal of right valves attached to "Ostrea haidingeriana ’ Emmrich;
'Gervillienbildung’ (Rhaetian); Eipelgraben bei Staudach, Bavaria, Germany. 8, paralectotypes GM
MLU3/GM/HS .4-5; detail as in 7. 9, paralectotype GM MLU3/GM/HS.5; detail as in 7. 10a, NMW
83.22G. 174; silicified left valve internal showing dimyarian musculature; semicostatum Zone, Lower Lias;
temporary excavation 400 m N of Lord Motor Company site, Bridgend, Mid Glamorgan, South Wales. 106,
Detail as in lOu; dimyarian muscle scars highlighted for clarity, showing anterior adductor and bilobate
posterior adductor. Magnifications: 1-8, x2;9, x3; 10 a, b, x 4.

All specimens were coated with ammonium chloride prior to photographing, with the exception of figure 10.

PLATE I

HODGES, Atreta

968

PALAEONTOLOGY, VOLUME 34

DISCUSSION

The evidence presented here of dimyarian musculature in Atreta confirms the observations of
Etallon (1862). The assignment of this genus to the monomyarian Plicatulidae by Cox (1964) is
therefore incorrect. Atreta is transferred to the Dimyidae based on the diagnostic characteristics
described and also on the close similarity with musculature in modern species of this family. The
stratigraphical range of the Dimyidae is thus extended back to the early Rhaetian.

Although an entire specimen with conjoined valves has been observed (NMW.83.22G. 182), the
left valve of members of this genus is rarely seen. Therefore they are known almost exclusively by
the attached right valves, which due to preservation generally lack any evidence either of
musculature or dentition, almost certainly a result of the loss of an aragonitic shell layer. This
opinion has been given previously by Eudes-Deslongchamps (1860) and Cox (1964). Stenzel (1964)
showed that the prodissoconch of the oyster Crassostrea virginica (Gmelin) is aragonitic, whereas
the adults are mainly calcitic. This adds further credence to the possibility of shell layer loss in
Atreta.

Atreta intusstriata (Emmrich) reached its highest population density in the Rhaetian of Europe
where it was often used as a stratigraphical indicator. On the evidence of this species alone Lower
Lias beds have been incorrectly designated as Rhaetian.

Acknowledgements. I thank the Ford Motor Company, Bridgend for permission to collect during excavations
at their main factory site. Dr G. Krumbiegel kindly supplied plaster casts of syntypes in the Emmrich collection
at the Geiseltal Museum, Halle. I thank Dr M. G. Bassett, Dr R. M. Owens and Mrs K. Bryant of the National
Museum of Wales, Dr N. J. Morris and Mr R. Cleevely of the British Museum (Natural History), Dr H. C.
Ivimey-Cook of the British Geological Survey and Dr P. R. Crowther of the Bristol City Museum for access
to material in the collections of their institutions.

REFERENCES

adams, H. and adams, a. 1853-58. The genera of recent Mollusca. J. Van Voorst, London, 1 (1853-54), 484 pp. ;
2 (1854-58), 661 pp.; 3 (1858), 136 pis.

arkell, w. j. 1933. The Jurassic System in Great Britain. Clarendon Press. Oxford, xii + 681 pp., 41 pis.
bayer, f. m. 1971. New and unusual mollusks collected by R/V John Elliot Pillsbury and R/V Gerda in the
tropical western Atlantic. Bulletin of Marine Science , Miami , 21, 111-236.
buvignier, A. 1852. Statistique geologique , miner alogique, mineralurgique et paleontologique du departement de
la Meuse. J. B. Bailliere, Paris, New York, 694 pp.. Atlas, 52 pp., 32 pis.
capellini, G. 1866. I fossili infraliassici dei dintorni del golfo della Spezia. Memoria Accademia della Scienze
dell' Istituto di Bologna ; Classe Scienze Fisiche, Series 2 , 5, 413-M86, pis 1-6.
cox, l. R. 1953. Lower Cretaceous Gastropoda, Lamellibranchia and Annelida from Alexander Island
(Falkland Island Dependencies). Scientific report of the Falkland Island Dependencies Survey , 4, 1-14, 2 pis.

1964. Notes concerning the taxonomy and nomenclature of fossil Bivalvia (mainly Mesozoic). Proceedings
of the Malacological Society of London , 36, 39^48, pi. 1.
dickens, J. m. 1957. Lower Permian Pelecypods and Gastropods from the Carnarvon Basin, Western Australia.

Bulletin of the Australian Bureau of Mineral Resources, 41, 1-74, 10 pis.
dumortier, E. 1864. Etudes paleontologiques sur les depots jurassiques du bassin du Rhone. (1), Infra-Lias.
Berard, Paris, 187 pp., 30 pis.

emmrich, a. 1853. Geognostiche Beobachtungen aus den ostlichen bayerischen und den angranznden
osterreichischen Alpen. Jalubuch der Kaiserlich-Koniglichen Geologischen Reichsanstalt, 4(2), 326-94.
etallon, a. 1862. Etudes paleontologiques sur les terrains Jurassiques du Haut-Jura. Monographic de l’etage
Corallien. 1. Vertebres, Articules, Mollusques. Memoires de la Societe d' Emulation Departement Doubs,
Paris , Series 3, 6, 1-250.

eudes-deslongchamps, J. a. 1860. Essai sur les Plicatules fossiles des terrains du Calvados et sur quelques
autres genres voisins ou demembres de ces coquilles. Memoires de la Societe Linneenne de Normandie, 2,
1-28.

HODGES: MESOZOIC BIVALVE ATRETA

969

ferussac, a. e. de 1822. Tableaux systematiques des animaux mollusques. A. Bertrand, Paris and London,

111 pp.

fischer, p. h. 1880-87. Manuel de conchyliologie et de paleontologie conchyliologique. F. Savy, Paris, xxv + 1369
pp., 23 pis.

frech, f. 1909. Die Leitfossilien der Werfner Schichten und Nachtrage zur Fauna des Muschelkalks, der
Cassianer und Raibler Schichten sowie des Rhat und Dachsteindolomits (Hauptdolomit). Resultate der
Wissenschaftlichen Erforschung des Balatonsees , 1, 1-200, 16 pis.
goetel, w. 1916. Die rhatische Stufe und der unterste Lias der subtatrischen Zone in der Tatra. Bulletin
International de T Academie des Sciences de Cracovie. Series A , Sciences et Mathematiques , 1-222, pis 7-12.
gray, j. e 1826. On a recent species of the genus Hinnita of Defiance and some observations on the shell of
the monomyaires of Lamarck. Annals of Philosophy , 12, 103-106.
greco, b. 1893. II Lias Inferiore nel circondario di Rossano Calabro. Atti della Societa Toscana di Science
naturali , 13, 55-180, pis 1-7.

hedley, c. 1902. Thetis Expedition Mollusca. Part 1. Brachiopoda and Pelecypoda. Memoirs of the Australian
Museum , 4, 287-324.

hodges, p. 1987. Lower Lias (Lower Jurassic) Bivcdvia from South Wales and adjacent areas. Unpublished Ph.D.
thesis. University of Wales.

ichikawa, k. 1958. Zur Taxionomia und Phylogenie der triadschen 'Pteriidae’ (Lamellibranchia).
Palaeontographica , Series A. 3, 131-212, pis 21 24.

joly, h. 1907. Les fossiles du Jurassique de la Belgique avec description Stratigraphique de chaque etage. 1.
Infra-Lias. Memoires du Musee Roved d'Histoire naturelle de Belgique , 5, 1-156, pis 1-5.

1936. Les fossiles du Jurassiques de la Belgique avec description stratigraphique de chaque etage. 2. Lias
Inferieur. Memoires du Musee Royal d'Histoire naturelle de Belgique , 79, 1-244, pis 1-3.
korobkov, i. a. 1960. 83. In eberzin, a. (ed. ). Osnovy paleontologii. Moskva, 3, 300 pp., 44 pis. [In Russian],
krasilova, i. n. 1959. A new family Leiopectinidae from the Upper Silurian and Lower Devonian of the north
eastern part of the Balkhash Region. Paleontologicheskiy Zhurnal , 2, 41-46, 1 pi. [In Russian].
lanquine, A. 1929. Le Lias et le Jurassique des chaines provenqales. Bulletin de la Carte geologique de la France ,
32, 1-385, pis 1-12.

meek, f. b. and hayden, F. v. 1864. Palaeontology of the Upper Missouri invertebrates. Smithsonian
Contributions to Knowledge , 172, 14, art. 5, 136 pp., 5 pis.
moore, c. 1861. On the Zones of the Lower Lias and the Avicula contorta Zone. Quarterly Journal of the
Geological Society of London, 17, 483-516, pis 15 and 16.
moore, R. c. (ed.). 1969. Treatise on invertebrate paleontology. Part N. Mollusca 6, Bivcdvia. Geological Society
of America and the University of Kansas Press, 1 and 2, xxxviii + 952 pp.; iv + 953-1224 pp.
nevesskaya, l. a., scarlato, o. a., starobogatov, Ya. I. and eberzin, a. g. 1971. New ideas on bivalve
systematics. Translated from Paleontologicheskiy Zhurnal 5(2), 3-20; Paleontological Journal no. 2, 141-155.
Newell, n. d. 1938. Late Paleozoic pelecypods: Pectinacea. Kansas State Geological Survey, Publications. 10,
no. 1, 123 pp., 20 pis.

1965. Classification of the Bivalvia. American Museum Novitates, No. 2206, 1-25.
orbigny, A. D. d\ 1849. Prodrome de paleontologie stratigraphique universelle des animaux mollusques et
rayonnes faisant suite au cours elementaire de paleontologie et de geologie stratigraphiques. V. Masson, Paris,
394 pp.

quenstedt, f. a. 1856-58. Der Jura. H. Laupp, Tubingen. 1. (1856), 1 -208, pis 1-24; 2. (1856), 209-368, pis
25-49; 3. (1856), 369-576, pis 50-72; 4. (1857), 577-823, pis 73-100; 5. (1858).
rafinesque, c. s. 1815. Analyse de la nature. Published by the author, Palermo, 225 pp.
rollier, l. 1917. Fossiles nouveaux ou peu connus des terrains secondaires (Mesozoiques) du Jura et
des contrees environnantes. Memoires de la Societe Paleontologique Suisse, Nos 6 and 7. 42, 501-696,
pis 33-40.

schafhautl, f. 1851. Uber einige neue Petrefakten des Stidbayern’chen Vorgebirges. Neues Jahrbuch fur
Mineralogie, Geologie und Palaontologie, for 1851, 407^42 1 , pi. 7.

Simpson, M. 1884. The fossils of the Yorkshire Lias described from nature with a carefully measured section of
the strata and the fossils peculiar to each. Forth, Whitby, xxiv + 256 pp.
stefano, G. di. 1886-87. Sul Lias Inferiore di Taormina e de suoi dintorni. Giornale di Science Naturali ed
Economiche di Palermo, 18, 46-184, pis 1-4.

stenzel, h. b. 1964. Oysters: Composition of the larval shell. Science, 145, no. 3628, 155-156.

tate, r. and blake, j. f. 1876. The Yorkshire Lias. J. van Voorst, London, viii + 475 pp., 19 pis.

terquem, o. 1855. Paleontologie de l'etage inferieur de la formation Liasique de la province de Luxembourg,

970

PALAEONTOLOGY, VOLUME 34

Grande-Duche (Hollande) et de Hettange, du departement de la Moselle. Memoirs de la Societe Geologique
de France , Series 2 , 3, 219-343, pis 3-26.

— and piette, e. 1865. Le Lias Inferieur de l’Est de la France, comprenant la Meurthe, la Moselle, le Grand-
Duche de Luxembourg, la Belgique et la Meuse. Memoires de la Societe Geologique de France , Series 2, 8.
1-175, pis 1-18.

thiele, J. 1935. Handbuch der systematischen Weichtierkunde. G. Fischer, Jena, 2, 779-1154, figs 784-897.
vokes, H. E. 1980. Genera of the Bivalvia ; A systematic and bibliographic catalogue (revised and updated).

Paleontological Research Institution, Ithaca, New York, xxvii + 307 pp.
waller, t. r. 1978. Morphology, morphoclines and a new classification of the Pteriomorphia (Mollusca:
Bivalvia). Philosophical Transactions of the Royal Society of London , Series B , 284, 345-365.

Watson, H. 1930. On the anatomy and affinities of Plicatula. Proceedings of the Mcdacological Society of
London , 19, 25-30, pi. 5.

yonge, c. m. 1975. The status of the Plicatulidae and the Dimyidae in relation to the superfamily Pectinacea
(Mollusca: Bivalvia). Journal of Zoology, 176, 545-553.

1978. On the Dimyidae (Mollusca: Bivalvia) with special reference to Dimya corrugata Hedley and
Basiliomya goreaui Bayer. Journal of Mollusc an Studies , 44, 357-375.

P. HODGES

c/o Department of Geology
National Museum of Wales

Typescript received 2 March 1990 Cathays Park

Revised typescript received 14 May 1991 Cardiff CF1 3NP, UK

THE FIRST MESOZOIC PSEUDOSCORPION, FROM
CRETACEOUS CANADIAN AMBER

by WOLFGANG SCHAWALLER

Abstract. The first Mesozoic pseudoscorpion (Arachnida, Pseudoscorpionida) from Cretaceous (Campanian)
amber collected in southern Alberta, Canada, is described. The specimen is a deutonymph and belongs to the
family Chernetidae, but further taxonomic assignment is impossible. Some observations and deductions on its
palaeobiology are made.

Pseudoscorpions (Arachnida, Pseudoscorpionida) are extremely rare as fossils, mainly because
of their small size and the weak sclerotization of their cuticle. Nearly all fossil pseudoscorpions
previously described came from Tertiary ambers of various ages and provenance, and all were
assigned to modern families or even genera. Recently, three extremely well-preserved fossil
fragments were described from Middle Devonian sediments near Gilboa, New York (Shear et al.
1989), belonging to a single species of a new family. In this enormous gap in the fossil record,
between Eocene (45 Ma) and Middle Devonian (380 Ma), nothing is known about the evolution of
this species-rich arachnid order (Chamberlin 1931; Beier 1932a, 1932 b). Here the first record of a
Mesozoic pseudoscorpion is reported. It is preserved in good condition and allows morphological
documentation from several sides by a special technique (Schlee and Glockner 1978). It is embedded
in Cretaceous Canadian amber, dated as Campanian, 70-85 Ma (McAlpine and Martin 1969). The
fossil belongs to the Recent family Chernetidae. A further taxonomic assignment is impossible
mainly because the specimen is a deutonymph : the second of four post-hatching stadia in
pseudoscorpions.

McAlpine and Martin (1969) gave a general account of the Canadian amber and mentioned
several arthropod inclusions. Some spiders are under examination (Selden, personal com-
munication). Another Cretaceous amber, the Lebanese amber (Aptian or Neocomian, 110-140
Ma), has aroused considerable interest (Schlee and Dietrich 1970). It contains a few spiders
(unpublished material in the amber collection of the natural History Museum, Stuttgart, Germany)
and also a pseudoscorpion: Whalley (1980), in a paper on neuropteran insects from Lebanese
amber, incidentally mentioned a single pseudoscorpion among spiders and mites. Nothing is shown
about its location, morphology or taxonomy.

GEOLOGICAL SETTING

The amber piece with the pseudoscorpion was collected by R. Mussieux, R. Solkoski and L. Strong,
staff of the Provincial Museum of Alberta, at Grassy Lake, Southern Alberta, Canada, in 1976.
Precise coordinates of the site are on file at the Royal Tyrrell Museum of Palaeontology and are
available to qualified investigators on request. This locality was not listed by McAlpine and Martin
(1969) among more than thirty localities in Canada and Alaska, where Cretaceous amber has been
found. The primary amber deposits in Canada were dated as Campanian. When undisturbed, most
Canadian amber lies along bedding planes in coal or carbonaceous sediments. However, most
amber particles are found in secondary positions washed ashore in rivers and lakes.

(Palaeontology, Vol. 34, Pari 4, 1991, pp. 971— 976. |

© The Palaeontological Association

972

PALAEONTOLOGY, VOLUME 34

MATERIAL AND METHODS

Preservation. The specimen is incomplete: the tibia and chela of the right pedipalp are lacking, the
right pedipalp femur is somewhat damaged and the left pedipalp chela is broken in its basal part,
the tips of the left tarsi I, II, IV are cut off, and the abdominal segments are wrinkled, particularly
on the end of the abdomen.

From the wrinkled abdomen and because of some accompanying fungi mycelia it is concluded
that the pseudoscorpion was embedded in the tree resin after death or that only the moulted cuticle
was fossilized. Preservation is the same as in those amber fossils which have been heated artificially
or naturally (Schlee, personal communication). Amber and its included fossils could have been
heated in their primary or secondary fossil sites by different events.

Methods. The amber piece was embedded in polystyrol resin. After solidification, this amber-
polystyrol-block was cut and polished under water to avoid heating. After morphological
documentation under the microscope in a definitive plane, this technique was repeated for studying
other views of the fossil.

DISCUSSION

Relationships

The fossil has two trichobothria laterally on the movable pedipalp finger and is therefore a
deutonymph (assuming Recent conditions can be applied to the Cretaceous). A satisfactory
classification of a juvenile pseudoscorpion is nearly impossible; furthermore, most of the Recent
genera are fixed only by typological characters and not by synapomorphies. The fossil itself shows
no unique character which would justify a new genus, so I avoid assigning this Cretaceous
pseudoscorpion to a known or new genus or species.

The monotarsate legs, the structure of the chelicerae and the form of the carapace point without
doubt to the Cheliferoidea (Atemnidae, Chernetidae, Cheliferidae, Withiidae). The Atemnidae can
be excluded here, because all members of this family have a basally inserted tactile seta on tarsus
IV (this seta is clearly absent in the fossil). Most of the Cheliferidae/Withiidae have eyes on the
carapace and have distinctly more slender pedipalps than has the fossil; therefore the Cretaceous
specimen is placed in the remaining Chernetidae. However, Recent chernetids usually have a
poisonous tooth only on the movable pedipalp finger (the fossil has such a tooth also on the fixed
finger). Accessory teeth on the pedipalp fingers are sometimes absent also in Recent deutonymphs
of the Chernetidae, so perhaps adults of the Cretaceous species had such teeth.

The trichobothriotaxy of the pedipalp chela of the fossil seems incomplete: in Recent
deutonymphs of the Chernetidae, the fixed finger has three trichobothria laterally (Mahnert 1982).
But the insertion of the trichobothria is difficult to recognize in this amber specimen and,
furthermore, some fungal mycelia are impeding visibility, so this irregularity should not be weighed
too heavily.

Palaeobiology

Two further fossils are embedded in the same piece of amber (Text-fig. 1b), a mite and a juvenile
insect larva or a collembolan. They are, together with the pseudoscorpion, members of the same
Cretaceous biocoenosis, and not of a caenocoenosis as in other fossil sites apart from amber. These
accompanying animal groups are typical inhabitants of Recent pseudoscorpion biotopes.

The fossil has a galea on its chelicera which proves that spinning behaviour existed in
pseudoscorpions during the Cretaceous. So, the present deutonymph very probably spun a silken
moulting chamber. Spinning activity of pseudoscorpions is already known from the Middle
Devonian (Shear et al. 1989). Recent pseudoscorpions spin not only moulting chambers but also
capsules for their eggs and for hibernation. The Cretaceous pseudoscorpion was able to groom itself

SCHAWALLER: CRETACEOUS PSEUDOSCORPION

text-fig. 1. a, chernetid pseudoscorpion (deutonymph) from Cretaceous Canadian amber; body length (with
chelicerae), 1 -35 mm. b, mite (left) and juvenile insect larva or collembolan (right) from the same piece of amber
as the pseudoscorpion; body length of the mite about 04 mm.

■ \X-

974

PALAEONTOLOGY, VOLUME 34

by the cheliceral serrulae. The animal, as a predator, used poison in grasping its prey, shown by the
palpal venom teeth. The situation of the palpal trichobothria point to the same mode of orientation
as in Recent species. The granulate cuticle (as in most of the Recent Chernetidae and Cheliferidae)
may suggest a drier biotope; possibly this Cretaceous species lived in or under the bark of the tree
which produced the amber.

SYSTEMATIC PALAEONTOLOGY

Order pseudoscorpionida Banks, 1895
Family chernetidae Chamberlin, 1931
chernetid gen. et sp. indet.

Text-figs 1-3

Material. A single specimen in Cretaceous Canadian amber from Grassy Lake. Alberta. Deposited in Tyrrell
Museum of Palaeontology, Drumheller, Alberta, Canada.

Description. Body length (with chelicerae) T35 mm. Cuticle of the body and of the appendages with granules.
Carapace without eyes lenses and without eyes spots. Maximal width: 0-59 mm, median length: 0-58 mm.
Carapace strongly narrowed in front, surface with two weak cross-furrows and basally with three short
longitudinal furrows. Surface with uniform granulation and with irregularly inserted dentate bristles.

Abdomen with ten visible segments. At least the first tergites separated medially. Last tergite with two long
tactile setae, tergites otherwise with dentate bristles (as on carapace). Number and position of all tergite bristles
and details of the sternites not determinable.

Chelicerae have fixed and movable fingers without inner teeth. Movable finger with galea distally, its detailed

text-fig. 2. Chernetid pseudoscorpion (deutonymph) from Cretaceous Canadian amber, a, dorsal view of
carapace and abdomen, b, chelicerae with galea, tactile setae and serrulae, x L66 in relation to scale bar.

c, leg IV.

SCHAWALLER: CRETACEOUS PSEUDOSCORPION

975

text-fig. 3. Chernetid pseudoscorpion (deutonymph) from Cretaceous Canadian amber, a, left pedipalp
trochanter, femur and tibia from dorsal view, b, left pedipalp chela from dorsal view, c, left pedipalp fingers
from lateral view with teeth and recognizable trichobothriotaxy.

structure not visible. Basis of chelicera with at least four setae, their insertions covered by carapace. Movable
finger with subgaleal seta distally. Serrula exterior with at least seven lamellae, structure of the serrula interior
and of the flagellum not determinable.

Pedipalps: femur 049 mm long and 0-24 mm wide, tibia 038 mm and 019 mm, and chela 0-83 mm and 0-28
mm. Granulation on medial and lateral side of femur and tibia distinct, not so prominent on the chela and very
faint towards the tip of the fingers. Surface with dentate bristles irregularly inserted, fingers with acute bristles
distally. Trichobothriotaxy cannot be completely documented: movable finger with two trichobothria laterally
(deutonymph), fixed finger with one trichobothrium laterally near basis and probably with a further
trichobothrium in the middle; trichobothria on the medial side not visible. Fixed finger with twenty-six uniform
and acute teeth, movable finger with thirty; both fingers with a distinct poisonous tooth; no accessory teeth.

All legs monotarsate. Tibia I 0 27 mm, tarsus I with claws 0-35 mm, tibia IV 0-43 mm, tarsus IV with claws
0-37 mm long. Tarsus I not modified ; all claws simple, not dentate. Bristles on tibia and tarsus laterally dentate,
medially acute; subterminal seta simple, acute; tibia and tarsus without tactile seta. Details of the coxae not
visible without grinding away the legs.

Acknowledgements . I am very grateful to W. A. Shear (Hampden-Sydney College, Virginia), who enabled me
to examine this interesting fossil and who revised the manuscript. V. Mahnert (Geneva) and W. B. Muchmore
(Rochester, New York) made helpful comments on the drawings. Thanks are also due to A. Neumann
(Drumheller, Alberta) for the loan of the specimen and to D. Schlee (Stuttgart) for general discussion
concerning amber inclusions and phylogeny, and for taking the photographs. Anonymous referees helped with
linguistic matters.

976

PALAEONTOLOGY, VOLUME 34

REFERENCES

beier, m. 1932m Pseudoscorpionidea. E Subord. Chthoniinea et Neobisiinea. Das Tierreich, 57, 1-258. Walter
de Gruyter and Co., Berlin.

— 1932 b. Pseudoscorpionidea. II. Subord. Cheliferinea. Das Tierreich , 58, 1-294. Walter de Gruyter and
Co., Berlin.

Chamberlin, J. c. 1931. The arachnid order Chelonethida. Stanford University Publications , Biological
Sciences , 7, 1-284.

mahnert, v. 1982. Die Pseudoskorpione (Arachnida) Kenyas V. Chernetidae. Revue Suisse de Zoologie , 89,

mcalpine, j. f. and martin, j. E. h. 1969. Canadian amber - a paleontological treasure-chest. Canadian
Entomologist , 101, 819-838.

schlee, d. and dietrich, h.-g. 1970. Insektenfiihrender Bernstein aus der Unterkreide des Libanon. Neues
Jahrbuch fur Geologie und Paldontologie, Monatshefte , 1970, 40-50.

— and glockner, w. 1978. Bernstein. Stuttgarter Beitrdge fur Naturkunde, Serie C, 8, 1-72.
shear, w. a., schawaller, w. and bonamo, p. 1989. Record of Palaeozoic pseudoscorpions. Nature, 341,

whalley, p. e. s. 1980. Neuroptera (Insecta) in amber from the Lower Cretaceous of Lebanon. Bulletin of the
British Museum ( Natural History ), (Geology), 33, 157-164.

691-712.

527-529.

Typescript received 21 September 1990
Revised typescript received 9 November 1990

WOLFGANG SCHAWALLER
Staatliches Museum fur Naturkunde
Rosenstein 1, D-7000 Stuttgart 1
Germany

INDEX

Pages 1-240 are contained in Part 1 ; pages 241-501 in Part 2; pages 503-749 in Part 3; pages 751-981 in Part 4.

A

Actinocamax verus, 707

‘ Actinocamax' lundgreni, 734; ex gr. lundgreni, 735
Actinopeltis tejoensis sp. nov., 348
Adelograptus antiquus, 35; alius sp. nov., 30 \filiformis
sp. nov., 36; cf. A. tenellus , 33; sp. A, 35
Adrain, J. M., Chatterton, B. D. E. and Cocks,
L. R. M. A new species of machaeridian from the
Silurian of Podolia, USSR, with a review of the
Turrilepadidae, 637
Agetograptus primus , 676; sp., 678
Algae: Cretaceous, 955
Almucidaris durhami sp. nov., 632
Ambitisporites dilutus, 594, 615
Ambitisporitesl vavrdovii, 594
Amino acids: contamination during analysis, 851
Ammonites: Jurassic, Hungary, 859
Amphibians: Permian, Brazil, 561
Amplexopora dnestrense sp. nov., 91
Ananterodonta oretanica , 128
Anaphragma gwyndyense sp. nov., 92
cf. ? Angolasaurus sp., 660

Antarctica: Cretaceous echinoid, 629; Jurassic teu-
thids, 169

Antronestheria kilmaluagenesis sp. nov., 534; prae-
cursor sp. nov., 536
Aorograptus victoriae , 24
Arachnids: Devonian, USA, 241
Araneograptus murrayi, 294
Archaeozonotriletes chulus , 618
Artemopyra brevicosta sp. nov., 612; sp. A, 613
Aire t a intusstriata , 964
Attercopus fimbriunguis , 256

Azygograptus eivionicus , 904; ellesi, 911; hicksii , 903;
lapworthi , 896; minutus sp. nov., 914; suecicus, 909;
validus, 910

B

Babin, C. and Gutierrez-Marco, J.-C. Middle Ordo-
vician bivalves from Spain and their phyletic and
palaeogeographic significance, 109
Babinka prima , 128

Baker, P. G. Morphology and shell microstructure of
Cretaceous thecideidine brachiopods and their
bearing on thecideidine phylogeny, 815
IBatostoma sp., 90

Beckly, A. J. and Maletz, J. The Ordovician grap-

tolites Azygograptus and Jishougraptus in Scan-
dinavia and Britain, 887
Belemnellocamax ex gr. grossouvrei , 731
Belemnitella praecursor , 740; cf. praecursor , 744;

propinqua , 736; sp., 745
Belemnites: Cretaceous, England, 695
Bivalves: Jurassic, Wales, 963; Ordovician, Spain,
109; role of predation in evolution of cementation,
455; Silurian, Sweden, 219
Blake, D. B. and Zinsmeister, W. J. A new marsupiate
cidaroid echinoid from the Maastrichtian of Ant-
arctica, 629

Bonamo, P. M. See Selden, P. A., Shear, W. A. and
Bonamo, P. M.

Bown, P. R. See Young, J. R. and Bown, P R.
Brachiopods: Cretaceous thecideidines, 815; Ordo-
vician, Sweden, 195 ; Silurian, Ireland and Scotland,
439; immunological relationships, 785
Brazil: Permian fishes and amphibians, 561
Brazilichthys macrognathus sp. nov., 563
Bryozoans: Ordovician, Wales, 77
Burgess, N. D. Silurian cryptospores and miospores
from the type Llandovery area, south-west Wales,
575

Burgess, N. D. and Richardson, J. B. Silurian crypto-
spores and miospores from the type Wenlock area,
Shropshire, England, 601

Buttler, C. J. A new upper Ordovician bryozoan
fauna from the Slade and Redhill Beds, South
Wales, 77

C

Cadomia britannica , 122
Calliphylloceras heterophylloides , 863
Cambrian: skeletal problematica, China, 357; trilo-
bite ultrastructure, Sweden, 205
Cambroclaves : Cambrian, China, 357
Canada: Cretaceous pseudoscorpion, 971 ; Devonian
plant, 149; Ordovician graptolites, 1
Cardiolaria beirensis, 1 1 6
Ceramoporella distinct a, 100

Chaloner, W. G. See Gensel, P. G., Chaloner, W. G.
and Forbes, W. H.

Chatterton, B. D. E. See Adrain, J. M., Chatterton,
B. D. E. and Cocks, L. R. M.

Chen Menge See Conway Morris, S. and Chen Menge

978

INDEX

Chen Pei-Ji and Hudson, J. D. The conchostracan
fauna of the Great Estuarine Group, Middle
Jurassic, Scotland, 515
Chernetid gen. et sp. nov., 974
China: Cambrian skeletal problematica, 357
Christensen, W. K. Belemnites from the Coniacian to
Lower Campanian chalks of Norfolk and southern
England, 695

Clarkson, E. N. K. See Scrutton, C. T. and Clarkson,
E. N. K.

Clonoclimacograptus retroversus , 674
Clonograptus ( Clonograptus ) magnus sp. nov., 307
Clonograptus cf. norvegicus , 305
Clonograptus ? sp. A, 39; sp. B, 42; sp. C, 42
Coccoliths: Jurassic, England, 843
Cocks, L. R. M. See Adrain, J. M., Chatterton,
B. D. E. and Cocks, L. R. M.

Collbatothuria danieli sp. nov., 71
Collins, M., Curry, G. B., Muyzer, G., Quinn, R.,
Shanjm Xu, Westbroek, P. and Ewing, S. Immuno-
logical investigations of relationships within the
terebratulid brachiopods, 785
Colpocoryphe grandis , 338; aff. rouaulti, 337; cf.

thorali conjugens, 338; sp., 342
Colpocoryphe ? sp. indet., 340
Conulariids: schotts, 939

Conway Morris, S. and Chen Menge. Cambroclaves
and paracarinachitids, early skeletal problematica
from the Lower Cambrian of South China, 357
Corals: Ordovician, Scotland, 179
Cox, C. B. The Pangaea dicynodont Rechnisaurus
and the comparative biostratigraphy of Triassic
dicynodont faunas, 767

Cox, C. B. and Hutchinson, P. Fishes and amphibians
from the Late Permian Pedra de Fogo Formation
of northern Brazil, 561
Coxiconcha britannica , 129

Cretaceous: alga, 955; belemnites, England, 695;
echinoid, Antarctica, 629; lizard eggshells, Spain,
237 ; mosasaurs, Niger, 653 ; pseudoscorpion,
Canada, 971; thecideidine brachipods, 815
Crustaceans: Jurassic conchostracans, Scotland, 515
cf. Ctenodonta escosurae, 1 1 1

Curry, G. B. See Collins, M., Curry, G. B.,
Muyzer, G., Quinn, R., Shanjin Xu, Westbroek, P.
and Ewing, S.

Curry, G. B. See Walton, D. and Curry, G. B.
Cyrtodontula sp., 124

D

Dales, R. C. See Kendrick, P., Edwards, D. and
Dales, R. C.

Dalingwater, J. E., Hutchinson, S. J., Mutvei, H. and
Siveter, D. J. Cuticular ultrastructure of the trilo-
bite Ellipsocephalus polytomus from the Middle
Cambrian of Gland, Sweden, 205
Deiradoclavus trigonus sp. nov., 374

Dekayia cf. crenulata, 83; pengawsensis sp. nov., 80
Deltaclavus graneus sp. nov., 378
Dendrostracus hebridesensis sp. nov., 530
Devonian: arachnids, USA, 241; fish, 399; plant,
Canada, 149; plant, Wales, 751
Dichograptid gen. et sp. indet. 1, 315
Dicynodont: Permian, Malagasy, 837; Triassic, East
Africa, 767

Didymograptus sp. 1, 313
Diversification: trilobites, 461
Dorsetensia subtecta , 866

Doyle, E. N., Hoey, A. N. and Harper, D. A. T. The
rhynchonellide brachiopod Eocoelia from the
Upper Landovery of Ireland and Scotland, 439
Doyle, P. Teuthid cephalopods from the Upper
Jurassic of Antarctica, 169
Dulcineaia manchegana sp. nov., 134
Dyadospora murusattenuata , 614; cf. murusattenuata ,
592; murusdensa, 614

E

East Africa: Triassic dicynodont, 767
Ecchosis pulchribothrium sp. nov., 275
Echinoid : Cretaceous, Antarctica, 629
Edwards, D. See Kendrick, P., Edwards, D. and
Dales, R. C.

Ekaterodonta hesperica sp. nov., 118
Ellipsocephalus polytomus , 205
Emery, D. See Simmons, M. D., Emery, D. and
Pickard, N. A. H.

Emileia sp. indet., 867

England : Cretaceous belemnites, 695 ; Devonian fish,
399 ; Jurassic coccoliths, 843 ; Ordovician graptolites,
887; Silurian spores, 601
Eocoelia curtisi , 448
Eridotrypa sp., 91

Euestheria trotternishensis sp. nov., 524
Evans, S. E. See Milner, A. R. and Evans, S. E.
Ewing, S. See Collins, M., Curry, G. B., Muyzer, G.,
Quinn, R., Shanjin Xu, Westbroek, P. and Ewing, S.

F

Fibrestheria puncta sp. nov., 526
Fish: Devonian, England and Wales, 399; Permian,
Brazil, 561
Fistulipora sp., 99

Forbes, W. H. See Gensel, P. G„ Chaloner, W. G.
and Forbes, W. H.

Foote, M. Morphologic patterns of diversification:
examples from trilobites, 461

G

Galacz, A. Bajocian stephanoceratid ammonites from
the Bakony Mountains, Hungary, 859

INDEX

979

Gallemi, J. See Smith, A. B. and Gallemi, J.

Gensel, P. G., Chaloner, W. G. and Forbes, W. H.
Spongiophyton from the late Lower Devonian of
New Brunswick and Quebec, Canada, 149
Glyptarcal lustianica , 126
Glyptograptus aff. incertus , 673; sinuatus, 673
Goniophora ( Cosmogoniophora ) sp., 123
Gonioteuthis granulata , 715; granulataquadrata , 716;
quadra ta, 718; westfalica , 713; westfalicangran-
ulata , 715

Goronyosaurus sp., 655

Graptolites: feeding strategies, 797; Ordovician,
Canada, 1 ; Ordovician, Scandinavia and Britain,
887; Ordovician, southern Scandinavia, 283; Sil-
urian, Sweden, 671

Gutierrez-Marco, J.-C. See Babin, C. and Gutierrez-
Marco, J.-C.

H

Halisaurus sp., 660

Hallopora cf. elegantula, 88; peculiaris , 86
Harper, D. A. T. See Doyle, E. N., Hoey, A. N. and
Harper, D. A. T.

Harper, E. M. The role of predation in the evolution
of cementation in bivalves, 455
Hensonella dinarica , 955
Heterotrypa sladei sp. nov., 79
Hilate cryptospore type 1, 613
Hispanaediscus verrucatus , 610; wenlockensis sp. nov.,
611

cf. Hispanaediscus sp. A, 611

Hodges, P. The relationship of the Mesozoic bivalve
Atreta to the Dimyidae, 963
Hoey, A. N. See Doyle, E. N., Hoey, A. N. and
Harper, D. A. T.

Holcophylloceras zignodianum , 864
Holmer, L. E. The taxonomy and shell characteristics
of a new elkaniid brachiopod from the Ashgill of
Sweden, 195

Holothurians : Triassic, Spain, 49

Hudson, J. D. See Chen Pei-Ji and Hudson, J. D.

Hungary: Jurassic ammonites, 859

Hunnegraptus copiosus sp. nov., 299; robustus sp.

nov., 303; tjernviki sp. nov., 302
Hunt, A. P. and Lucas, S. G. The Paleorhinus
biochron and the correlation of the non-marine
Upper Triassic of Pangaea, 487
Hunt, A. P. and Lucas, S. G. A new rhynchosaur
from the Upper Triassic of West Texas, and the
biochronology of Late Triassic rhynchosaurs, 927
Hutchinson, P. See Cox, C. B. and Hutchinson, P.
Hutchinson, S. J. See Dalingwater, J. E., Hutchinson,
S. J., Mutvei, H. and Siveter, D. J.

I

Igdamanosaurus aegvptiacus , 659
Ireland: Silurian brachiopods, 439

J

Jishougraptus lindholmae sp. nov., 916; novas sp. nov.,
916

Jurassic: ammonites, Hungary, 859; bivalve, Wales,
963; coccoliths, England, 843; conchostracans,
Scotland, 515; teuthids, Antarctica, 169; theropod,
Portugal, 503

K

Kendrick, P., Edwards, D. and Dales, R. C. Novel
ultrastructure in water-conducting cells of the
Lower Devonian plant Sennicaulis hippocrepifor-
mis , 751

Kiaerograptus bulmani , 17; magnus sp. nov., 16;
pritchardi , 9; supremus sp. nov., 292; undulatus sp.
nov., 14

cf. Kiaerograptus taylori, 13
Kilbuchophyllia discoidea sp nov., 191
King, G. M. See Mazin, J. M. and King, G. M.
Kohring, R. Lizard egg shells from the Lower
Cretaceous of Cuenca Province, Spain, 237
Kuker sella borealis , 104

L

Laevolancis divellomedium , 607 ; plicata sp. nov., 607
Leioclema orbicularis , 83
Lichenalia cf. concentrica, 104

Liljedahl, L. Contrasting feeding strategies in bivalves
from the Silurian of Gotland, 219
Lindholm, K. Ordovician graptolites from the Early
Hunneberg of southern Scandinavia, 283
Lingham-Soliar, T. Mosasaurs from the Upper
Cretaceous of Niger, 653
Lisboasaurus estesi , 504
Lizard: eggshells, Cretaceous, Spain, 237
Loydell, D. K. Isolated graptolites from the Llan-
dovery of Kallholen, Sweden, 671
Lucas, S. G. See Hunt, A. P. and Lucas, S. G.

Lucas, S. G. See Hunt, A. P. and Lucas, S. G.

M

Machaeridians : Silurian, USSR, 637
Malagasy: Permian dicynodont, 837
Maletz, J. See Beckly, A. J. and Maletz, J.

Mazin, J. M. and King, G. M. The first dicynodont
from the late Permian of Malagasy, 837
Metaclimacograptus hughesi, 675
Milner, A. R. and Evans, S. E. The Upper Jurassic
diapsid Lisboasaurus estesi - a maniraptoran thero-
pod, 503

Modiolopsisl elegantulus , 123
Monilipsolus mirabilis sp. nov., 66
Monoclimacis sp., 679

Monograptus communis , 683 ; denticulatus , 684 ; mille-
peda , 686

980

INDEX

Mosasaurs: Cretaceous, Niger, 653
cf. Mosasaurus hoffmanni , 665
Munchoaspis denisoni, 422
Munsterellid gen. et sp. nov., 173
Murornate tetrads, 619

Mutvei, H. See Dalingwater, J. E., Hutchinson, S. J.,
Mutvei, H. and Siveter, D. J.

Muyzer, G. See Collins, M., Curry, G. B., Muyzer, G.,
Quinn, R., Shanjin Xu, Westbroek, P. and
Ewing, S.

Myoplusia bilunata , 120

N

Nannolytoceras polyhelictum , 866
Neopolygrapta leallensis sp. nov., 526
Niger: Cretaceous mosasaurs, 653
Norway: Ordovician graptolites, 283, 887

O

Ordovician: bivalves, Spain, 109; brachiopod,

Sweden, 195; bryozoans, Wales, 77; coral, Scot-
land, 179; graptolites, Canada, 1; graptolites,
Scandinavia and Britain, 887 ; graptolites, southern
Scandinavia, 283; trilobites, Portugal, 329
‘ Orthograpius ' cyperoides , 678; insectiformis, 678
Otischalkia elderae sp. nov., 929
Oudenodon sakamenensis sp. nov., 841

P

Paleorhinus bransoni , 491
Paracarinachites leshanensis, 385; spinus , 384
Paracarinachitids : Cambrian, China, 357
Parade lograptus elongatus sp. nov., 317; tenuis sp.
nov., 319

Paratenmograptus isolatus sp. nov., 19
Permian: dicynodont, Malagasy, 837; fishes and
amphibians, Brazil, 561 ; mammal-like reptile.
South Africa, 547
Phialaspis symondsi , 403
Phytosaur: Triassic, USA, 487
Pickard, N. A. H. See Simmons, M. D., Emery, D.

and Pickard, N. A. H.

Pinnatoporella carinata, 98
Plant: Devonian, Canada, 149
IPlatecarpus sp., 665
Plioplatecarpus sp., 663

Portugal: Jurassic theropod, 503; Ordovician trilo-
bites, 329

Praenucula costae , 114; sharpei sp. nov., 115
Predation: role in the evolution of cementation in
bivalves, 455

Pribylograptus leptotheca , 682

Prionocheilus costai , 347 ; mendax , 343 ; cf. pulcher,
346

Prionosuchus plummeri, 565
Pristiograptus concinnus, 680
Problematica : Cambrian, China, 357
Pseudodyadospora cf. laevigata , 587
Pseudograpta jonesi sp. nov., 540; morrisi sp. nov.,
542; murchisonae, 538; orbita , 540
Pseudoscorpion: Cretaceous, Canada, 971

Q

Quinn, R. See Collin, M., Curry, G. B., Muyzer, G.,
Quinn, R., Shanjin Xu, Westbroek, P. and
Ewing, S.

R

Rastrites peregrinus , 687
Rechnisaurus cristarhynchus, 768
Redonia deshay esi , 129

Reptiles: dicynodont, Permian, Malagasy, 837;
dicynodont, Triassic, East Africa, 767; lizard egg-
shells, Cretaceous, Spain, 237; mosasaurs, Cre-
taceous, Niger, 653; phytosaur, Triassic, USA,
487; rhynchosaur, Triassic, USA, 927; therapsid,
Permian, southern Africa, 547; theropod, Jurassic,
Portugal, 503
Rhabdinopora sp., 43
Rhynchosaur: Triassic, Texas, 927
Richardson, J. B. See Burgess, N. D. and Richardson,
J. B.

Rigby, S. Feeding strategies in graptoloids, 797
Rimasventeraspis angusta , 425
Rimosotetras problematica sp. nov., 586
Romano, M. Trilobites from the Ordovician of
Portugal, 329

Rubidge, B. S. A new primitive dinocephalian
mammal-like reptile from the Permian of southern
Africa, 547

Rugosphaera cf. R. ? cerebra , 593

S

Salterocoryphe salteri, 342

Schawaller, W. The first Mesozoic pseudoscorpion,
from the Cretaceous Canadian amber, 971
Scotland: Jurassic conchostracans, 515; Ordovician
coral, 179; Silurian brachiopods, 439
Scrutton, C. T. and Clarkson, E. N. K. A new
scleractinian-like coral from the Ordovician of the
Southern Uplands, Scotland, 179
Segestrespora laevigata sp. nov., 589; sp. A, 592;
( Dyadospora ) membranifera , 588; ( Pseudodyado-
spora) rugosa , 589

Selden, P. A., Shear, W. A. and Bonamo, P. M. A
spider and other arachnids from the Devonian of
New York, and reinterpretations of Devonian
Araneae, 241

Sennicaulis hippocrepiformis, 754

INDEX

981

Shanjin Xu. See Collins, M., Curry, G. B..
Muyzer, G., Quinn, R., Shanjin Xu, Westbroek, P.
and Ewing, S.

Shear, W. A. See Selden, P. A., Shear, W. A. and
Bonamo, P. M.

Silurian: bivalve feeding strategies, Sweden, 219;
brachiopods, Ireland and Scotland, 439; grap-
tolites, Sweden, 671; machaeridians, USSR, 637;
spores, England, 601; spores, Wales, 575
Simmons, M. D., Emery, D. and Pickard, N. A. H.
Hensonella dinarica , an originally calcitic Early
Cretaceous dasycladacean alga, 955
Siveter, D. J. See Dalingwater, J. E., Hutchinson, S. J.,
Mutvei, H. and Siveter, D. J.

Skyestheria elliptica sp. nov., 532; intermedia sp. nov.,
532

Smith, A. B. and Gallemi, J. Middle Triassic holo-
thurians from northern Spain, 49
South Africa: Permian mammal-like reptile, 547
Spain: Cretaceous lizard eggshells, 237; Ordovician
bivalves, 109; Triassic holothurians, 49
Spongiophyton minutissimum , 1 54
Spores: Silurian, England, 601; Silurian, Wales, 575
Stemmatoceras frechi , 880

Stephanoceras (Normannites) orbignyi , 878; sp. aff.
fords, 879

Stephanoceras ( Stephanoceras ) leoniae , 876; mutabile ,
877; scalare, 872; sturanii , 878; triplex, 870; sp.,
874

Stevens, R. K. See Williams, S. H. and Stevens, R. K.
Strobilothyone rogenti sp. nov., 54
Strophomorpha ovata, 593

Sweden: Cambrian trilobite ultrastructure, 205;
Ordovician brachiopod, 195; Ordovician grapto-
lites, 283, 887; Silurian bivalve feeding strategies,
219; Silurian graptolites, 671
Synorisporites cf. S.'l libycus , 617

T

Tapinocaninus pamelae sp. nov., 548
Tarrant, P. R. The ostracoderm Phialaspis from the
Lower Devonian of the Welsh Borderland and
South Wales, 399

Tetragraptus krapperupensis sp. nov., 310; longus sp.
nov., 308

Tetrahedraletes medinensis , 580, 604
Teuthids: Jurassic, Antarctica, 169
Therapsid: Permian, southern Africa, 547
Theropod: Jurassic, Portugal, 503
Tilasia rugosa sp. nov., 197

Toombsaspis pococki, 420; sabrinae, 422
Trachyteuthis cf. hastiformis, 172
Traquairaspis campbelli, 423

Triassic: dicynodont. East Africa, 767; holothurians,
Spain, 49; phytosaur reptile, USA, 487; rhyn-
chosaur, USA, 927
Trilete miospore type I, 618

Trilobites: Ordovician, Portugal, 329; patterns of
diversification, 461 ; ultrastructure, Cambrian,
Sweden, 205

Turrilepas modzalevskae sp. nov., 648; wrightiana ,

USA: Devonian arachnids, 241; Triassic phytosaur
reptile, 487; Triassic rhynchosaur, 927

USSR: Silurian machaeridians, 637

V

Valongia wattisoni, 351

Van Iten, H. Anatomy, patterns of occurrence, and
nature of the conulariid schott, 939

Velatitetras laevigata sp. nov., 583 ; reticulata sp. nov.,
585; rugulata sp. nov., 585; sp. A, 586.

W

Wales: Devonian fish, 399; Devonian plant, 751;
Jurassic bivalve, 963; Ordovician graptolites, 887;
Silurian spores, 575

Walton, D. and Curry, G. B. Amino acids from
fossils, facies and fingers, 851

Westbroek,P.SeeCollins,M., Curry, G. B.,Muyzer,G.,
Quinn, R., Shanjin Xu, Westbroek, P. and
Ewing, S.

Williams, S. H. and Stevens, R. K. Late Tremadoc
graptolites from western Newfoundland, 1

Y

Young, J. R. and Bown, P. R. An ontogenetic
sequence of coccoliths from the Late Jurassic
Kimmeridge Clay of England, 843

Z

Zhijinites longistriatus, 366

Zinsmeister, W. J. See Blake, D. B. and Zinsmeister,
W. J.

VOLUME 34

Palaeontology

1991

PUBLISHED BY THE
PALAEONTOLOGICAL ASSOCIATION
LONDON

Dates of Publication of Parts of Volume 34

Part 1, pp. 1-240
Part 2, pp. 241-501
Part 3, pp. 503-749
Part 4, pp. 751-981

12 March 1991
14 June 1991
19 September 1991
29 November 1991

THIS VOLUME EDITED BY M. J. BENTON, J. E. DALINGWATER, D. EDWARDS,
P. D. LANE, P. A. SELDEN AND P. D. TAYLOR

Date of Publication of Special Paper in Palaeontology
Special Paper No. 44, 17 July 1991

© The Palaeontological Association , 1991

Printed in Great Britain
by Cambridge University Press

CONTENTS

Part

Adrain, J. M., Chatterton, B. D. E. and Cocks, L. R. M. A new species of machaeridian
from the Silurian of Podolia, USSR, with a review of the Turrilepadidae 3

Babin, C. and Gutierrez-Marco, J.-C. Middle Ordovician bivalves from Spain and their
phyletic and palaeogeographic significance 1

Baker, P. G. Morphology and shell microstructure of Cretaceous thecideidine brachiopods and
their bearing on thecideidine phylogeny 4

Beckly, A. J. and Maletz, J. The Ordovician graptolites Azygograptus and Jishougraptus in
Scandinavia and Britain 4

Blake, D. B. and Zinsmeister, W. J. A new marsupiate cidaroid echinoid from the Maastrichtian
of Antarctica 3

Bonamo, P. M. See Selden, P. A., Shear, W. A. and Bonamo, P. M.

Bown, P. R. See Young, J. R. and Bown, P. R.

Burgess, N. D. Silurian cryptospores and miospores from the type Llandovery area, south-west
Wales 3

Burgess, N. D. and Richardson, J. B. Silurian cryptospores and miospores from the type
Wenlock area, Shropshire, England 3

Buttler, C. J. A new upper Ordovician bryozoan fauna from the Slade and Redhill Beds, South
Wales 1

Chaloner, W. G. See Gensel. P. G., Chaloner, W. G. and Forbes, W. H.

Chatterton, B. D. E. See Adrain, J. M., Chatterton, B. D. E. and Cocks, L. R. M.
Christensen, W. K. Belemnites from the Coniacian to lower Campanian chalks of Norfolk and
southern England 3

Clarkson, E. N. K. See Scrutton, C. T. and Clarkson, E. N. K.

Cocks, L. R. M. See Adrain, J. M., Chatterton, B. D. E. and Cocks, L. R. M.

Conway Morris, S. and Chen Menge Cambroclaves and paracarinachitids, early skeletal
problematica from the Lower Cambrian of South China
Chen Menge See Conway Morris, S. and Chen Menge

Chen Pei-Ji and Hudson, J. D. The conchostracan fauna of the Great Estuarine Group, Middle

Jurassic, Scotland 3

Collins, M., Curry, G. B., Muyzer, G., Quinn, R., Shanjin Xu, Westbroek, P. and Ewing, S.

Immunological investigations of relationships within the terebratulid brachiopods 4

Cox, C. B. The Pangaea dicynodont Rechnisaurus and the comparative biostratigraphy of
Triassic dicynodont faunas 4

Cox, C. B. and Hutchinson, P. Fishes and amphibians from the Late Permian Pedra de Fogo
Formation of northern Brazil 3

Curry, G. B. See Collins, M., Curry, G. B., Muyzer, G., Quinn, R., Shanjin Xu,
Westbroek, P. and Ewing, S.

Curry, G. B. See Walton, D. and Curry, G. B.

Dales, R. C. See Kendrick, P., Edwards, D. and Dales, R. C.

Dalingwater, J. E., Hutchinson, S. J., Mutvei, H. and Siveter, D. J. Cuticular ultrastructure
of the trilobite Ellipsocephalus polytomus from the Middle Cambrian of Oland, Sweden 1

Doyle, E. N., Hoey, A. N. and Harper, D. A. T. The rhynchonellide brachiopod Eocoelia from
the Upper Llandovery of Ireland and Scotland 2

Doyle, P. Teuthid cephalopods from the Upper Jurassic of Antarctica 1

Edwards, D. See Kendrick, P„ Edwards, D. and Dales, R. C.

Emery, D. See Simmons, M. D., Emery, D. and Pickard, N. A. H.

Evans, S. E. See Milner, A. R. and Evans, S. E.

Ewing, S. See Collins, M., Curry, G. B., Muyzer, G., Quinn, R., Shanjin Xu, Westbroek, P.
and Ewing, S.

Foote, M. Morphologic patterns of diversification : examples from trilobites 2

Page

637

109

815

887

629

575

601

77

695

357

515

785

767

561

205

439

169

461

IV

CONTENTS

Forbes, W. H. See Gensel, P. G., Chaloner, W. G. and Forbes, W. H.

Galacz, A. Bajocian stephanoceratid ammonites from the Bakony Mountains, Hungary 4

Gallemi, J. See Smith, A. B. and Gallemi, I.

Gensel, P. G., Chaloner, W. G. and Forbes, W. H. Spongiophyton from the late Lower
Devonian of New Brunswick and Quebec, Canada 1

Gutierrez-Marco, J.-C. See Babin, C. and Gutierrez-Marco, J.-C.

Harper, D. A. T. See Doyle, E. N., Hoey, A, N. and Harper, D. A. T.

Harper, E. M. The role of predation in the evolution of cementation in bivalves 2

Hodges, P. The relationship of the Mesozoic bivalve Atreta to the Dimyidae 4

Hoey, A. N. See Doyle, E. N., Hoey, A. N. and Harper, D. A. T.

Holmer, L. E. The taxonomy and shell characteristics of a new elkaniid brachiopod from the
Ashgill of Sweden 1

Hudson, J. D. See Chen Pei-Ji and Hudson, J. D.

Hunt, A. P. and Lucas, S. G. The Paleorhinus biochron and the correlation of the non-marine
Upper Triassic of Pangaea 2

Hunt, A. P. and Lucas, S. G. A new rhynchosaur from the Upper Triassic of West Texas, and
the biochronology of Late Triassic rhynchosaurs 4

Hutchinson, P. See Cox, C. B. and Hutchinson, P.

Hutchinson, S. J. See Dalingwater, J. E., Hutchinson, S. J., Mutvei, H. and Siveter, D. J.
Kenrick, P., Edwards, D. and Dales, R. C. Novel ultrastructure in water-conducting cells of
the Lower Devonian plant Sennicaulis hippocrepiformis 4

King, G. M. See Mazin, J. M. and King, G. M.

Kohring, R. Lizard egg shells from the Lower Cretaceous of Cuenca Province, Spain 1

Liljedahl, L. Contrasting feeding strategies in bivalves from the Silurian of Gotland 1

Lindholm, K. Ordovician graptolites from the early Hunneberg of southern Scandinavia 2

Lingham-Soliar, T. Mosasaurs from the Upper Cretaceous of Niger 3

Loydell, D. K. Isolated graptolites from the Llandovery of Kallholen, Sweden 3

Lucas, S. G. See Hunt, A. P. and Lucas, S. G.

Lucas, S. G. See Hunt, A. P. and Lucas, S. G.

Maletz, J. See Beckly, A, J. and Maletz, J.

Mazin, J. M. and King, G. M. The first dicynodont from the Late Permian of Malagasy 4

Milner, A. R. and Evans, S. E. The Upper Jurassic diapsid Lisboasaurus estesi - a maniraptoran
theropod 3

Mutvei, H. See Dalingwater, J. E., Hutchinson, S. J., Mutvei, H. and Siveter, D. J.
Muyzer, G. See Collins, M., Curry, G. B., Muyzer, G., Quinn, R., Shanjin Xu, Westbroek, P.
and Ewing, S.

Pickard, N. A. H. See Simmons, M. D., Emery, D. and Pickard, N. A. H.

Richardson, J. B. See Burgess, N. D. and Richardson, J. B.

Rigby, S. Feeding strategies in graptoloids 4

Romano, M. Trilobites from the Ordovician of Portugal 2

Rubidge, B, S. A new primitive dinocephalian mammal-like reptile from the Permian of
southern Africa 3

Schawaller, W. The first Mesozoic pseudoscorpion, from the Cretaceous Canadian amber 4
Scrutton, C. T. and Clarkson, E. N. K, A new scleractinian-like coral from the Ordovician of
the Southern Uplands, Scotland 1

Selden, P. A., Shear, W. A. and Bonamo, P. M. A spider and other arachnids from the
Devonian of New York, and reinterpretations of Devonian Araneae 2

Shanjin Xu See Collins, M„ Curry, G. B„ Muyzer, G., Quinn, R., Shanjin Xu, Westbroek, P.
and Ewing, S.

Shear, W. A. See Selden, P. A., Shear, W. A. and Bonamo, P. M.

Simmons, M. D., Emery, D. and Pickard, N. A. H. Hensonella dinarica , an originally calcitic
Early Cretaceous dasycladacean alga 4

Siveter, D. J. See Dalingwater, J. E., Hutchinson, S. J., Mutvei, H. and Siveter, D. J.

Smith, A. B. and Gallemi, J. Middle Triassic holothurians from northern Spain 1

Stevens, R. K. See Williams, S. H. and Stevens, R. K.

Tarrant, P. R. The ostracoderm Phialaspis from the Lower Devonian of the Welsh Borderland
and Wales 2

859

149

455

963

195

487

927

751

237

219

283

653

671

837

503

797

329

547

971

179

241

955

49

399

CONTENTS

Van Iten, H. Anatomy, patterns of occurrence, and nature of the conulariid schott 4 939

Walton, D. and Curry, G. B. Amino acids from fossils, facies and fingers 4 851

Westbroek, P. See Collins, M., Curry, G. B., Muyzer, G. Quinn, R., Shanjin Xu,

Westbroek, P. and Ewing, S.

Williams, S. H. and Stevens, R. K. Late Tremadoc graptolites from western Newfoundland 1 1

Young, J. R. and Bown, P. R. An ontogenetic sequence of coccoliths from the late Jurassic

Kimmeridge Clay of England 4 843

Zinsmeister, W. J. See Blake, D. B. and Zinsmeister, W. J.

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform
in style to those already published in this journal, and should be sent to Dr Dianne Edwards, Department of Geology,
University of Wales College of Cardiff CF1 3YE, who will supply detailed instructions for authors on request (these are
published in Palaeontology 1990, 33, pp. 993-1000).

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 1991 is £45 00 (U.S. $80) for Institutional
Members, and £20 00 (U.S. $36) for Ordinary and Student Members. A single copy of each Special Paper is available on a
non-subscription basis to Ordinary and Student Members only , for their personal use, at a discount of 25 % below the listed
prices: contact the Marketing Manager. Non-members may obtain Nos 35-43 (at cover price) from Basil Blackwell Ltd.,
Journal Subscription Department, Marston Book Services, P.O. Box 87, Oxford OX2 0DT, England, and older issues from
the Marketing Manager. For all orders of Special Papers through the Marketing Manager, please add £1-50 (U.S. $3) per
item for postage and packing.

PALAEONTOLOGICAL ASSOCIATION PUBLICATIONS

Special Papers in Palaeontology

For full catalogue and price list, send a self-addressed, stamped A4 envelope to the Marketing Manager. Numbers 2-32 are
still in print and are available together with those listed below :

33. (for 1985): Evolutionary Case Histories from the Fossil Record. Edited by j. c. w. cope and p. w. skelton. 202 pp., 80
text-figs., 4 plates. Price £30 (U.S. $60).

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. $20).

35. (for 1986): Studies in palaeobotany and palynology in honour of N. F. Hughes. Edited by d. j. batten and d. e. g.
briggs. 178 pp., 29 plates. Price £30 (U.S. $60).

36. (for 1986): Campanian and Maastrichtian ammonites from northern Aquitaine, France, by w. J. Kennedy. 145 pp., 43
text-figs., 23 plates. Price £20 (U.S. $40).

37. (for 1987): Biology and revised systematics of some late Mesozoic stromatoporoids, by rachel wood. 89 pp., 31 text-
figs., 1 plates. Price £20 (U.S. $40).

38. (for 1987): Taxonomy, evolution, and biostratigraphy of late Triassic-early Jurassic calcareous nannofossils, by p. r.
bown. 118 pp.. 19 text-figs., 15 plates. Price £30 (U.S. $60).

39. (for 1988): Late Cenomanian and Turonian ammonite faunas from north-east and central Texas, by w. j. Kennedy,
131 pp., 39 text-figs., 24 plates. Price £30 (U.S. $60).

40. (for 1988): The use and conservation of palaeontological sites. Edited by p. r. crowther and w. A. Wimbledon. 200 pp.,
31 text-figs. Price £30 (U.S. $60).

41. (for 1989): Late Jurassic-early Cretaceous cephalopods of eastern Alexander Island, Antarctica, by p. j. howlett.
72 pp., 9 text-figs., 10 plates. Price £20 (U.S. $40).

42. (for 1989): The Palaeocene flora of the Isle of Mull, by m. c. boulter and z. kvacek. 149 pp., 23 text-figs., 23 plates.
Price £40 (U.S. $80).

43. (for 1990): Benthic palaeoecology of the Late Jurassic Kimmeridge Clay of England, by p. r. wignall, 74 pp., 50 text-
figs. Price £30 (U.S. $60).

Field Guides to Fossils and Other Publications

These are available only from the Marketing Manager. Please add £100 (U.S. $2) per book for postage and packing plus
£ 1 -50 (U.S. $3) for airmail. Payments should be in Sterling or in U.S. dollars, with all exchange charges prepaid. Cheques
should be made payable to the Palaeontological Association.

1. (1983): Fossil Plants of the London Clay, by m. e. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $16) (Members £6

or U.S. $12).

2. (1987): Fossils of the Chalk, compiled by E. owen; edited by A. b. smith. 306 pp., 59 plates. Price £11 -50 (U.S. $23)

(Members £9-90 or U.S. $20).

3. (1988): Zechstein Reef fossils and their palaeoecology, by N. hollingworth and T. Pettigrew, iv + 75 pp. Price £4-95

(U.S. $10) (Members £3.75 or U.S. $7-50).

4. (1991): Fossils of the Oxford Clay, edited by d. m. martill and j. d. Hudson. 286 pp., 44 plates. Price £15 (U.S. $30)

(Members £12 or U.S. $24).

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $40).

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, 1991

Palaeontology

VOLUME 34 • PART 4

CONTENTS

, Novel ultrastructure in water-conducting cells of the Lower
Devonian plant Sennicaulis hippocrepiformis

P. KENRICK, D. EDWARDS and R. C. DALES 751

The Pangaea dicynodont Rechnisaurus and the comparative

biostratigraphy of Triassic dicynodont faunas

c. b. cox 767

Immunological investigations of relationships within the terebratulid
brachiopods

M. COLLINS, G. B. CURRY, G. MUYZER, R. QUINN, SHANJIN XU,

P. WESTBROEK and S. EWING 785

Feeding strategies in graptoloids

S. RIGBY 797

Morphology and shell microstructure of Cretaceous thecideidine

brachiopods and their bearing on thecideidine phylogeny

p. G. baker 815

The first dicynodont from the Late Permian of Malagasy

J. M. mazin cmd G. M. king 837

An ontogenetic sequence ofcoccoliths from the Late Jurassic Kimmeridge
Clay of England

j. r. young and p. r. bown 843

Amino acids from fossils, facies and fingers

D. WALTON and G. B. CURRY 851

Bajocian stephanoceratid ammonites from the Bakony Mountains,
Hungary

a. galacz 859

The Ordovician graptolites Azygograptus and Jishougraplus in
Scandinavia and Britain

a. J. beckly and j. maletZ 887

A new rhynchosaur from the Upper Triassic of West Texas, and the
biochronology of Late Triassic rhynchosaurs

a. p. hunt and s. G. lucas 927

Anatomy, patterns of occurrence, and nature of the conulariid schott

H. VAN ITEN 939

Hensonella dinarica , an originally calcitic early Cretaceous dasycladacean
alga

M. D. SIMMONS, D. EMERY and N. A. H. PICKARD 955

The relationship of the Mesozoic bivalve Atreta to the Dimyidae
p. HODGES 963

The first Mesozoic pseudoscorpion, from Cretaceous Canadian amber
W. SCHAWALLER / 971

• ■' " •

Printed in Great Britain at the University Press , Cambridge

ISSN 0031-0239

vj|,

m ~ m • ' x; -yKA^'j^ m X^ostuX m

HlllNS S3 iyvy9ll — LIBRARIES SMITHSONIAN~INSTITUTIONL NOlinillSNI ~ NVIN0SH1IWS S3 I B V B 8 I
oo 2 , w z , to 2:

DNIAN INSTITUTION NOlifllllSNI _ NVINOSHil INS SBIBVaan LIBRARIES SMITHSONIAN INSTITUTE

^ 2 \ ^ - -t. tO

uJ /T*Istit77^\ //v 'W UJ iii

Q 'N^VOSV^X' 2 O Q X^VOSH^ ~ o

Z _J 2 -j 2 j z

■uiins saiavaan libraries Smithsonian institution NoiinmsNi nvinoshiiws saiavaa

Z •“ Z r- Z 1“ Z

o Z O . “ o xoawEtx ~ o xsswt':

bs

m Xfo p-<Lx EJ m x^uoajx m

oo „ c/> ±: y)

DNIAN INSTITUTION NOlinillSNI NVIN0SH1IINS S3IBVBan LIBRARIES SMITHSONIAN INST1TUTK

z ,v </> z ^ w z w 2: ,v.

< ^ 2 .< S . < Xx\ > S < a£s..

z
o
00

£

5 1 >

^ k. 2 </5 Z 00 *- z - W

hihns ssiavaan libraries Smithsonian institution NouniiiSNi nvinoshiiins saiavaa

CO ” 05 5 ^ 00 . — .... . 05

^ Xhso^x UJ ^ >rw»so^x w

[i\ —

'is oc

<
tr

O \“\oC/ “ O X!ft ocx ^ q

OMAN”* INSTITUTION 2 NOlinillSNI ^NVINOSHil INS S3 IBVaail^LIBRAR I ES^SMITHSONIAN^INSTITUTK
r- z «“ „ z ~

m oi*ss :« xjvA$v^ rrj >" v®?- \|wsx m XL'vosniX 'ii m __

to S oo ' r; o> ± co

hihns saiHvaan libraries Smithsonian institution NoiiniiiSNi nvinqshiiins satavaa

m . Z „ t/> y CO Z </>

JX § S < \<v 2 2

> |

«/> '" Z oo y* z in —

0NIAN INSTITUTION NOlinillSNI NVINOSHil INS SBIBVMail LIBRARIES SMITHSONIAN INSTITUTE

CO .. 7 \ 00 “ <o “

z

HiiiNS S3 tava an libraries Smithsonian institution NoiiniiiSNi nvinoshhiais saiavaa

z r- z *“ Z r- ^ z

Iv.

m x; '^8^ m x; x*ua«x m

C/5 — C/5 ™ CO _

IONIAN INSTITUTION NOlinillSNI NVINOSHilWS S3IHVaan LIBRARIES SMITHSONIAN INSTITUTK

Z .... to Z w Z -V W z

co fr co ± co _ u.

RARIES SMITHSONIAN INSTITUTION NOIiniliSNI NVINQSH1IINS S3ldVaan LIBRARIES SMITHSC

CO z . CO Z CO 2 , CO

5 t § 4 P ^ !Z £ Nft?

^ 5 2 vw > 5 >• s 'v# > X*«s

A' 2 CO ' ■ 2 CO 2 CO A' z

millSN! _ NV1N0SH1I WS S31HVHail LIBRARIES SMITHSONIAN INSTITUTION NOlifllllSNI NVINOSI
<0 = co — co 5

w *■'

CO

CO

o

2

*va5«SX O O XOvos*£X — O

2 «J 2 Z

RARIES SMITHSONIAN INSTITUTION NOlifllllSNI NVINOSHillMS S3ISVa9l1 LIBRARIES SMITHS!
2 s" 2 r , z r- Z

© . ™ o /ra?v4T\ . o /SS\ xfv'so^>. o

,oV - — m w V^^X m " X 5 XSS^ m

— co _ to \ 5 CO

miiiSNi nvinoshiiws SBiavaan libraries Smithsonian institution NouniiiSNi nvinosi

2 _ co 2 .... CO Z . CO 2

~ . . .^vc i < -^v- S .<

S 5 ^ s 'x 5 g X | 'W I ^

CO 2 CO V 2 rn V z Jo

RARIES SMITHSONIAN INSTITUTION NOIlDlllSNI NVINOSHHWS S3iavaail LIBRARIES SMITHSC

00 2 ^ ~ .... co 5 co

«</> w .v^c “ - ,w

fc***«\x ,mw

f|S

O “ '^nfnv^ 5 X “ 5 Xo£“T*X CQ

fUllSNI NVIN0SH1IWS S3 I U VH 8 11 “"l I B RAR I ES^SMITHSONIAN^INSTITUTION ^NOlinillSNI^NVINOSL
n 2 r~ z Z r~

m *' .VS?' X?vas>'_>^ fr( XOvosv'^2 'LL FI 'X'jco Vtvs 'LL XJyAst''>/' m

<0 ~ co 5 co ? co

TAR I ES SMITHSONIAN INSTITUTION NOIiniliSNI NVIN0SH1IWS S3iavaail LIBRARIES SMITHSC
55 _ 2 « co z co 2 co

? < * 2 s < V 2

X

co AMP'/faZ. CO |£jjg- ^5) CO

| '^iP? t » i.

> S )CX > s XLoosKjX > s

niliSNI_NVIN0SHimS^S3 S HVd 0 nZLI B RAR I ES^SMITHSONIAN^INSTITUTION ^NOIlfllliSNsXviNOSI-

x ^ ___ m — rn v

TARIES SR
2

O “ ■ 0 V!2 ^

, -* 2 j 2 j

N INSTITUTION NOIiniliSNI NVIN0SH1IWS S3!UVdan LIBRARIES SMITHSC
^ . 2 <r z r* 2

.ccvwovJJx. o O ^cTwivAJX XiiJTPx. O

nillSN! NVINOSHillMS S3 1 avaan LIBRARIES SMITHSONIAN INSTITUTION NOIiniliSNI NVINOSI

2 CC? 2 .... CO y ... ^