STITUTION °° NOlinillSNI^NVINOSHlIWS^SH I B VH 8 n\l BRAR I ES^SMITHSONIAN^INSTITUTION ** N<

^ Z \ ^ C/) GO

UJ XXLI?cX co Ak UJ m x"5T7 UJ £ XoTI^x UJ

o

2:

/^/r H _
o xfoosnj^

_j 2 _j z j 2 j

navy an libraries Smithsonian institution NoiinuxsNi nvinoshiiws saiavasn l

m X-vdcx y, m xw/ ^ m V, 'il m

to £ to £ co \ E co

STITUTION NOlinXIlSNI NVINOSHIIWS SBIHVaan LIBRARIES SMITHSONIAN INSTITUTION N
to 2 _ to 2 .... . in 2

<

, a/ Z//, * 2

tn

> w I I 1 ’VS vw =

2 CO 2 CO * 2 CO

1 LIBRARIES SMITHSONIAN INSTITUTION NOliniiJLSNI NVINOSHIIWS SBIHVaail L

— v (/) CO CO

^ . w ^ X^L^ w a4\\ u to

07

O p.c/ ~ O ''Z? DV-^ __ X")jus^x o " _ '<^Un^y q

STITUTION^ NOlinillSNI^NVINOSHHWS^S 3 IdVaail^LlBRARI ES*2 SMITHSONIAN-* INSTITUTION ^ N
2 n > 2 z z

5<T*oX P m V 2 Z X'ToT^X O m 0

03
>

- (W f~

m >V§r' ^ XhTsXX m X^o ££X ^ m

_ co ‘ ' ' £ to £ to

3iavaan libraries Smithsonian institution Noiiniiism nvinoshiiws saiavaan l

2 CO ^ 2 OO 2 CO Z

Mfo,- 2 \\ .„v *h 2 =5 wM///*- z /^esw^\ h

~ ■ ■ o x fmjyWi. 0

00 'Ol®' co co

5 •' OW > ^ v 'W > ' g Xilvo^X >

to •» 2 •> g ^ 2

STITUTION NOIlDlllSNI NVINOSHIIWS S3iyvy0I1 LIBRARIES SMITHSONIAN INSTITUTION N
to 5 \ to 5 <0 to

u w X u to xfooSoX W M UJ

•■ c

£Q XSTHKKP’ £ TmW* cq VJfJSSffij/ » m-*' ' /’ 5 63

k<y ^ Yy€Ak§r&

’tKoiWSf Z XxasXX o '' LW Z XjvAst^X q _

— j 2 ' j 2 — j 2 «j

aiavaan libraries Smithsonian institution NoiininsNi nvinoshiiws saiavaan l

r~ 2 r* _ z r* . 2

(TJ X*D<X 7 ■ rn X2C0&X Xl2±!i2x ^ N

07 - CO £ ’ to v £ »„

ISTITUTION NOlIfUlISNI NVINOSHIIWS S3IHVHan LIBRARIES SMITHSONIAN INSTITUTION, N
11 w 2 to 2 v- to z

> ’W" S > W 2

2 ' CO 2 OT ® 2 40

aiHvaan libraries Smithsonian institution noiioiiisni nvinoshiiws saiavaan l

«=» irn, °*s gn —=’ (fA

CO

nr

2 >• 5 xtgsgg/ > 5 :

NVINOSHIIWS^SB I dvaa lAl BRAR I ES^SMITHSONIAN INSTITUTION NOIifUIlSNI NVINOSHillNs'

CO 5 CO — CO 5 \ c

h = r. 3 - /£i^\ "* V

.,?> • c
/ v '

o

2

SMITHSONIAN INSTITUTION N0I1IUI1SNI NVINOSH1IINS SBIHVHaH LIBRARIES SMITHSONIAN
- . 2 2 r; ^ r

m Vo\dc^/ co m \ co XyTTC^ m Xo~c^/ - r

O) £ CO \ 5 CO £ c

NViNOSHiiiNS saiavaan libraries Smithsonian institution NouniiiSNi nvinoshiiins

"**’’"* Z C/5 2

CO

S > _ _

"^SMITHSONIAN __ INSTITUTION NOliniliSNI^NVINOSHIlWS^Sa I BVB 9 II^LI BRAR I ES^SMITHSONIAN
■■n -7 co — . co “ co

jJ m cy .» yj £ ✓fHsoi lx uj

30 5 5 V “ 5 n%fc;

"VlVINOSHlMS S3 I y VH an LI B RAR I ES^SMITHSONIAN^lNSTITUTION^NOIinillSNI^NVlNOSHlilNS'
v Z _ r- 2 r- z _ v :

— jam>A 03 /£*,___» .

Kj, - ^ ,

SMITHSONIAN INSTITUTION NOlifUllSNI _ NVINOSHIIINS" S3 I H V9 9 11 ”LI B RAR I ES SMITHSONIAN

Si fo -zr fn -y CO

I

>' 2 Xovos^ >' 5 \i£p' >

_NVIN0SHimSS°S3 I H VH a n2LI B R AR I ES^SMITHSONIAN^INSTITUTION ^NOlinilJLSNI nvinoshirms
2 \ ^ 5 co _ — co “

1 W 2 5 _ _ _ _ _ _

2 -* 2 _j 2 .J 2

SMITHSONIAN INSTITUTION NOlinillSNS NVIN0SH1IWS S3IUVHSn LIBRARIES SMITHSONIAN

/ / y5 2 5 1 5 S y

'•/ XW “3i A- * T rA.'tX r- Av^frJ^vX i— /^}JMlr9kz\ /^^irJkvX h- . /

m co p, ' 10 m Kr''%Ak9ra/ -

NV!N0SH1IINS^S3 I H ViJ 3 n~LI B RAR I ES ^SMITHSONIAN "INSTITUTION ^ NOlinillSNI^NVINOSHlI WS

^ CO

CO

2 £

2 > •-sir’' 2 ^ i ^ >

■S) 2 co * 2 — OT v 2 co :

SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3iaVUan LIBRARIES SMITHSONIAN
w =5 — co ~ co — co :

uj

Z

wso A/7N

"TLkvjiu

Palaeontology

VOLUME 32 • PART 1 FEBRUARY 1989

Published by

The Palaeontological Association ■ London

Price £25-50

THE PALAEONTOLOGICAL ASSOCIATION

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

COUNCIL 1988-1989

President : Dr J. D. Hudson, Department of Geology, University of Leicester, Leicester LEI 7RH
Vice-Presidents : Dr L. B. Halstead, Department of Geology, University of Reading, Reading RG6 2AB
Dr P. W. Skelton, Department of Earth Sciences, The Open University, Milton Keynes MK7 6AA
Treasurer: Dr M. E. Collinson, Department of Biology, King’s College, London W8 7 AH
Membership Treasurer: Dr H. A. Armstrong, Department of Geology, University of Newcastle,

Newcastle upon Tyne NE1 7RU

Institutional Membership Treasurer: Dr A. W. Owen, Department of Geology, University of Dundee, Dundee DD1 4HN
Secretary: Dr P. Wallace, The Croft Barn, Church Street, East Hendred, Oxon 0X12 8LA
Circular Reporter: Dr D. Palmer, Department of Geology, Trinity College, Dublin 2
Marketing Manager: Dr C. R. Hill, Department of Palaeontology, British Museum (Natural History), London SW7 5BD
Public Relations Officer: Dr M. J. Benton, Department of Geology, The Queen’s University of Belfast, Belfast BT5 6FB

Editors

Dr M. J. Benton, Department of Geology, The Queen’s University of Belfast, Belfast BT5 6FB
Dr J. E. Dalingwater, Department of Environmental Biology, University of Manchester, Manchester M13 9PL
Dr D. Edwards, Department of Geology, University of Wales College of Cardiff CF1 3YE
Dr C. R. C. Paul, Department of Earth Sciences, University of Liverpool, Liverpool L69 3BX
Dr P. A. Selden, Department of Extra-Mural Studies, University of Manchester, Manchester M13 9PL
Dr P. D. Taylor, Department of Palaeontology, British Museum (Natural History), London SW7 5BD

Other Members

Dr J. A. Crame, Cambridge Dr E. A. Jarzembowski, Brighton

Dr G. B. Curry, Glasgow Dr R. A. Spicer, London

Dr V. P. Wright, Reading

Overseas Representatives

Argentina: Dr Miguel O. Mancenido, Division Paleozoologia Invertebrados, Facultad de Ciencias Naturales y Museo,
Universitad Nacional de la Plata, Paleo del Bosque, 1900 La Plata, Argentina
Australia: Dr K. J. McNamara, Department of Palaeontology, Western Australian Museum, Francis Street, Perth, Western
Australia 6000

Canada: Professor S. Henry Williams, Department of Earth Sciences, Memorial University of Newfoundland, St John’s,
Newfoundland, Canada, A1B 3X5

New Zealand: Dr R. A. Cooper, New Zealand Geological Survey, PO Box 30368, Lower Hutt, New Zealand
U.S. A.: Professor M. Wilson, Geology Department, College of Wooster, Ohio 44691, U.S.A.

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1989 are:

Institutional membership . . . £50 00 (U.S. $90)

Ordinary Membership .... £21-00 (U.S. $38)

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

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

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr A. W.
Owen, Department of Geology, The University, Dundee DD1 4HN. Student members are persons receiving full-time
instruction at educational institutions recognized by the Council. On first applying for membership, an application form
should be obtained from the Membership Treasurer: Dr H. A. Armstrong, Department of Geology, University of Newcastle,
Newcastle upon Tyne NE1 7RU. Subscriptions cover one calendar year and are due each January; they should be sent to
the Membership Treasurer. All members who join for 1989 will receive Palaeontology , Volume 32, Parts 1-4. Enquiries
concerning back numbers should be directed to the appropriate Membership Treasurer.

Non-members may subscribe, and also obtain back issues up to 3 years old, at cover price through Basil Blackwell,
Journals Department, 108 Cowley Road, Oxford OX4 1 JF, England. For issues older than 3 years contact the Marketing
Manager.

Cover: Ammonite (Grossouvria) from the Oxford Clay (Jurassic) of Woodham, Bucks., containing geopetal pyrite stalac-
tites. Direct print from a thin section; pyrite white. Diameter of specimen approximately 10 mm. (See Sediment ology, 29,

639-667, 1982.)

AMURAL ARACHNOPHYLLID CORALS FROM THE
SILURIAN OF THE NORTH ATLANTIC AREA

by COLIN T. SCRUTTON

Abstract. Species previously assigned to Arachnophyllum and related genera from the Silurian of north-west
Europe and North America are revised and their phylogeny and biogeography are discussed. Intraspecific
variation, particularly in the English sample of Arachnophyllum murchisoni , is described in detail. The genera
Arachnophyllum and Prodarwinia are redefined on the basis of redescriptions of their type species, A.
murchisoni and P. speciosa. In addition the species A. sinemurum, A. separatum , A. pentagonum, A. striatum ,
A. pygmaeum , P. striata , P. mamillaris, and P. distans are accepted as senior synonyms and revised. A new
species of Prodarwinia is described and comments are made on species of doubtful status. Species removed
from this group are assigned to Iowaphyllum , IZenophila , Radiastraea , and IMazaphyllum.

Arachnophyllids are the only common amural rugose corals in the Llandovery to Ludlow of
north-west Europe and eastern North America, amural Rugosa being defined as massive colonies
in which no epitheca is present between constituent corallites (Scrutton 1988). They play an
important role in the development of bioherms in some parts of this North Atlantic area, although
much subordinate to the stromatoporoids, with which they can be confused on broken longitudinal
surfaces, and tabulate corals. Many species have been erected, often under the generic name
Strombodes, and the local use of different species names obscures the true distribution of lineages
in space and time. The only substantial revision this century was by Lang and Smith (1927) for
the English material. Thus, as McLean (1975, p. 54) remarked, revision of Arachnophyllum is long
overdue.

My original intention was to revise north-west European species of the genus, including the type
species, to clarify the septal structure in A. murchisoni , and to document the remarkable variability
of the English material. This has led to re-establishing the genus Darwinia (as Prodarwinia ), long
regarded as a junior synonym of Arachnophyllum. The necessity for comparisons with North
American species indicated that these would also require at least preliminary revision. I therefore
comment on all species known to me in the North Atlantic area based on a re-examination of
their type material as far as possible (Table 1). This has allowed initial clarification of the
biostratigraphy and biogeography of those species I regard as valid (text-figs. 1 and 2).

The section on Systematic Palaeontology at the end of the paper provides the basis for the
sections on Biostratigraphy, Biogeography, and Phylogeny, and Intraspecific Variation and Species
Discrimination. Much of the work is based on museum material, but includes extensive new
collections from the Much Wenlock Limestone of England (deposited in the British Museum
(Natural History)), together with small amounts of material from Gotland, Estonia, and Canada.

BIOSTRATIGRAPHY, BIOGEOGRAPHY, AND PHYLOGENY OF
NORTH ATLANTIC AMURAL ARACHNOPHYLLIDS

The earliest member of the Arachnophyllum- Prodarwinia group is of late Aeronian age (text-
fig. 1). P. speciosa from the Rumba Formation of the Adavere Stage in Estonia is certainly of this
age but may be little if any older than P. distans , from the Waco Member of eastern Kentucky.
The Waco Member is referred to the Noland Formation of late Llandovery age and is itself
regarded as C^2, possibly as young as C3 in age (Rexroad and KlelTner, in press; Rexroad, pers.
comm.). Further afield in North America, P. mamillaris and members of the A. pentagonum

IPalaeontology, Vol. 32, Part 1, 1988, pp. 1-53, pis. 1-8.|

© The Palaeontological Association

2

PALAEONTOLOGY, VOLUME 32

table 1 . Revision of North Atlantic species formerly or currently assigned to Arachnophyllum or

related genera.

Species and author

Original genus

Assignment herein

alpenensis Rominger

Strombodes

Iowaphyllum alpenensis

approximatus Parks

Strombodes

IMazaphyllum approximation

diffluens Edwards and Haime

Strombodes

Arachnophyllum murchisoni

eximius Billings

Strombodes

? Prodarwinia gigas

gigas Owen

Astrea?

IP. gigas

gracile Nicholson and Hinde

Astraeophyllum

A. pygmaeum

granulosum Foerste

Arach. ( Strombodes )

IP. granulosa

incertus Davis

Strombodes

A. murchisoni

infundibularia Owen

Lamellopora

indet.

kayi Merriam

A rachnophyllum

IZenophila kayi

mamillaris Owen

Astrea

P. mamillaris

mamillare-distans Foerste

Arachnophyllum

P. distans

mamillare-wilmingtonensis Foerste

Arachnophyllum

P. striata

murchisoni Edwards and Haime

Strombodes

A. murchisoni

pentagonus Goldfuss

Strombodes

A. pentagonum

phillipsii d’Orbigny

Actinocyathus

A. murchisoni

pygmaeus Rominger

Strombodes

A. pygmaeum

quadrangularis Davis

Strombodes

A. sinemurum

richardsoni Salter

Arachnophyllum

Radiastraea richardsoni

separatus Ulrich

Strombodes

A. separatum

sinemurus Davis

Strombodes

A. sinemurum

speciosa Dybowski

Darwinia

P. speciosa

striata d’Orbigny

Favastraea

A. striatum

striata James

Lye Ilia

P. striata

typus M‘Coy

Arachnophyllum

A. murchisoni

unicus Davis

Strombodes

A. separatum

verneuili Edwards and Haime

Phillipsastrea

‘IP. verneuili

group appear in the later Telychian, e.g. in the Fossil Hill Formation of Manitoulin Island, Ontario
in which the coral biostromes are regarded as of C3-C4 age (Copper 1978). P. striata appears
slightly later in the Dayton Formation of southern Ohio where it is no older than C5.6 (Kletfner
via Rexroad, pers. comm.).

P. speciosa, P. distans, P. mamillaris , and P. striata could have shared a common ancestor or
have had a more direct relationship. A possible phylogeny for this lineage is given in text-fig. 1.
P. speciosa appears to have migrated from Europe to North America, where it appeared at least
by mid-Telychian times (text-fig. 2). Current knowledge of species distributions suggests that P.
distans is confined to North America (Kentucky and Iowa) and its appearance may predate this
event. Even so, its evolution from P. speciosa seems most likely. P. mamillaris is also known only
from North America and it may have evolved from either P. speciosa , to which it is similar in
dimensions and calical form, or less likely from P. distans. P. striata is confined to more or less
coeval levels in the late Telychian of both North America (Ohio) and Gotland. It is very similar
to P. speciosa , from which it is distinguished by little more than dimensions and calical form, and
undoubtedly arose from that species. Some Ontario material of P. speciosa shows transition to P.
striata and the reverse is true in Ohio. If P. striata evolved in North America, then the species
appears to have migrated back to Europe. Thus trans-Iapetus migration in both directions may
have been possible in the Telychian.

The origin of the A. pentagonum group is more cryptic. The most likely evolutionary sequence
is from smaller forms of P. speciosa through A. pygmaeum to A. pentagonum and A. striatum. All

SCRUTTON: SILURIAN ARACHNOPH YLLI D CORALS

3

0 A. separatum
|; ' | Land

A A. pygmaeum

8 P. mamillaris
[2 P. di starts

text-fig. 1 . Distribution of species of Arachnophyllum and Prodarwinia in the North Atlantic area. Continental
distributions in the early Silurian largely based on Cocks and Fortey (1982) and McKerrow (1988).

text-fig. 2. Stratigraphical distri-
bution and tentative phylogeny of
species of Prodarwinia and Arachno-
phyllum. In vertical ranges of species,
dots indicate gaps in record and
dashes indicate range within which
records fall when precise age is un-
certain.

are approximately of the same age. The A. pentagonum group again has a relatively local distribu-
tion in the Telychian of Michigan, Ontario, and Quebec and did not migrate across to Europe.
However, A. murchisoni, which almost certainly evolved from A. pentagonum , also occurs in
Quebec, where the two species coexisted in the late Llandovery (although carinae in the specimen
of A. murchisoni available is slightly abnormal; text-fig. 3). A. murchisoni spread widely, reach-
ing England in the latest Llandovery on the basis of a new record in the Petalocrinus Limestone.

4

PALAEONTOLOGY, VOLUME 32

text-fig. 3. a, b, Arachnophyllwn murchisoni ; Silurian, late Telychian, La Vieille Formation; Little Port
Daniel River, Gaspe. Quebec (GSC 91554). a, cross-section; b, longitudinal section, c, d, Arachnophyllwn
pentagonum; same horizon; railway cut above Anse-a-la-Vieille, east of Port Daniel, Gaspe, Quebec (GSC
91555). c, cross-section; b, longitudinal section. All figs. x2-5.

This seems to be a clear example of easterly trans-Iapetus migration (in terms of present geography).
It is also recorded in the mid- and particularly late Wenlock of England and Gotland and the late
Wenlock (?early Ludlow) Louisville Limestone (Rexroad et al. 1978; Shaver et al. 1985) of the
Kentucky area. There is a single record apparently from the English early Wenlock (Woolhope
Limestone) in an old collection (SM A5700) but this specimen is almost certainly wrongly labelled.
I also have some doubt about a specimen of A. sp. cf. A. sinemurum recorded from Iowa where it
would most likely be from the La Porte City Lormation of C5 to earliest Wenlock age (Witzke
1983 and pers. comm.); its preservation is identical to that of material from the Louisville
Limestone. Otherwise, A. sinemurum and A. separatum are later descendants of A. murchisoni in
the Louisville Limestone of the Kentucky area, whilst the rare specimens of A. sinemurum recorded
from the English Wenlock may be independently evolved from A. murchisoni rather than the result
of migration. The American and English material of A. sinemurum is indistinguishable and
essentially coeval but the species is of questionable validity (see below). P. mamillaris also survived
in the Kentucky area until the late Wenlock (?early Ludlow). A. murchisoni persisted into the mid-
Ludlow Brownsport Formation in Tennessee (Shaver et al. 1985), which is the youngest record of
Arachnophyllwn currently known.

The species of Prodarwinia and A. murchisoni not only clearly demonstrate the possibility of
trans-Iapetus migration in the Telychian, but suggest a reversal of direction during that time. Non-
migration is more difficult to deal with. For the rarer, local species such as P. distans , they may
have become extinct before a suitable distribution of environments could facilitate their wider

SCRUTTON: SILURIAN ARACHNOPHYLLID CORALS

5

distribution. In the cases of P. mamillaris and A. pentagonum, one possibility is that these species,
most common in the North American dolomite suite (Berry and Boucot 1970), were facies restricted
by water temperature and/or salinity. Cocks and Fortey (1982, p. 473 et seq.) suggest a latitudinal
difference between England and the American continental interior in the late Llandovery and there
is also a significant facies contrast between the American dolomitized pure carbonates and the
marls or muddy limestones of England and Gotland. However, P. mamillaris coexists with A.
murchisoni in the Louisville Limestone of late Wenlock (?early Ludlow) times, when it appears
that little latitudinal or facies differences existed. Oliver (1977, pp. 95, 96) has shown that by
Ludlow times there was 34-36% generic endemism in eastern North America, with the mid-west
isolated from the northern Appalachians and Europe by an area of dolomite and clastic deposition.
He also considered that migration was from the European area towards eastern North America
in the late Silurian. The absence of P. mamillaris from the Anglo-Baltic area may reflect the early
expression of either or both of these inhibitors to migration towards Europe along the shelf. There
is no evidence of a land barrier in this area in the late Silurian.

Finally, the origin of the Arachnophyllum-Prodarwinia lineage as a whole is obscure. Iwanowskii
(1965, p. 50, fig. 6) proposed descent from Paleophyilum (Stauriina) via Entelophyllum , whilst Hill
(1981, p. 44) suggested derivation from the Streptelasmatina. Another possibility seems to be descent
from the Upper Ordovician to Silurian kyphophyllid genus Donacophyllum (? = Strombodes ), in
which lonsdaleoid dissepiments had already evolved. This genus is unknown from North America
but present in the Upper Ordovician of Estonia, and acceptance of this suggestion would accord
with the origin of the lineage in Europe through P. speciosa. Of the Ketophyllina ( sensu Hill 1981)
the Kyphophyllidae and Endophyllidae appear to be quite closely related to the Arachnophyllidae
and should be merged into the same suborder.

VARIATION AND SPECIES DISCRIMINATION IN
ARACHNOPHYLLUM AND PRODARWINIA SPECIES

Of the species described here, by far the largest single sample is that of A. murchisoni from
biohermal facies of the late Wenlock Much Wenlock Limestone from the Welsh Borderland. Its
intraspecific variation is described together with comparative comments on other species.

Variation in A. murchisoni.

A. murchisoni , is a highly variable species, reflected to some extent by the fact that five different
species names have been erected for its subsets. However, fundamental internal structure, such as
the typical septal carination, the form of dissepiments and tabulae, microstructure and increase
are all relatively uniform. In addition, the principal variable characters, viz. growth form of the
calicular surface, corallite size, and the degree of septal development, all show more or less
continuous variation within the population samples available, and support the recognition of a
single species.

In the British Wenlock as a whole, colony mean tabularium diameter (dt) ranges from 1-85 to
4-63 mm, corallite area (Ac) from 0-70 to 6-33 cm2, and major septa (n) from 15-3 to 22. Both Ac
and n are loosely correlated with dt and the sample plots as a single but rather dispersed group
(text-fig. 5). However, both dt and n are difficult to measure accurately, tabularium definition often
being vague in cross-section and septa sometimes irregularly developed, with breakdown of a clear
major and minor distinction, and more rarely very weak and only partial development (e.g. PI. 3,
fig. 3). The data are regarded as imprecise although the likely degree of error is not greater than
10%. The range of variation is similar for material from a single locality, the biohermal facies of
the Farley Dingle road cutting, which has yielded the largest single collection in the present study
(text-fig. 6).

Early species discrimination within A. murchisoni as defined here largely depended on external
features (PI. 1). The principal calicular surface form end members can conveniently be termed the

6

PALAEONTOLOGY, VOLUME 32

murchisoni morph and the diffluens morph. The murchisoni morph has distinct intercorallite ridges
(PI. 1, figs. 1, 2, 4) whereas they are completely suppressed in the diffluens morph (PI. 1, figs. 3
and 7). The latter has more or less prominent peritabularial swellings (mamelons bearing the
tabularia). Variation between these extremes is expressed by the recognition of a sub-diffluens
morph in which the swellings coexist with intercorallite walls that vary from very weak to absent
across the calicular surface (PI. 1, figs. 5 and 6). Both in the total sample and in the Farley Dingle
sample all gradations exist from colony to colony between the two extremes. In both cases, the
diffluens plus sub-diffluens morphs account for approximately 25% of the samples. These morphs
tend to have smaller tabularium and corallite size and septal number ranges ( 1 -85—2-96 mm, 1T-
2-55 cm2, and 15-8-19, respectively), than the total sample, although it is covariate with the lower
end of the murchisoni morph range (text-fig. 5). In terms of the cyclicity of dissepiment size in the
dissepimentarium (dense bands; see below), the morphs are essentially covariate in the ranges of
cycle thickness and number. Furthermore, there is no known disjunct distribution data for the
morphs in the English Wenlock that suggest a sharply defined ecological significance for the
differences in growth form. In the biohermal facies of Farley Dingle, all variants coexist, suggesting
that any ecological control operated on a microenvironmental scale. The rare, isolated colonies in
the bedded interbiohermal sediments of the Much Wenlock Limestone tend to be tabular masses
of rather weak walled murchisoni morph type. Elsewhere, colonies from the argillaceous M ulde
Marl on Gotland are broadly conical with domed calicular surfaces and subdiffluens morph
corallites (PI. 1, fig. 6); North American material is exclusively of murchisoni morph type (PI. 3,
figs. 7 and 8). Both cases may reflect regional genetic variation in local populations, with the sub-
diffluens and diffluens morphs arising in the European area after migration of the species from
North America.

In longitudinal section, there is a tendency for the intercorallite ridges to show signs of sharper
definition in association with zones of smaller dissepiments which bear the septal crests. In some
cases, the ridge is enhanced at these levels by short vertical series of very small, globose dissepiments
(PI. 3, figs. 2, 6, 8). These periodic levels of denser tissue in the skeleton are similar to the ‘dense
bands’ described in Silurian favositid corals from the same beds by Scrutton and Powell (1981),
and in other corals by many previous authors, and are interpreted in the same way as annual
growth increments. Colonies with ragged margins show the death of peripheral tissue with sediment
cover to occur always immediately above dense bands (e.g. PI. 2, fig. 7). They are thus interpreted
as autumn and/or winter growth with notching due to sediment movement by winter storms. The
implication is thus that stronger intercorallite ridge development is associated with winter growth
and probably with increased turbidity and local sediment settling rates. Ridges may be an
adaptation to inhibit sediment encroachment across the colony surface, or may be associated with
increased polypal activity to shift smothering sediment (see below). Furthermore, where diffluens
morph colonies do show a tendency to intercorallite wall development, it occurs at the margins of
the colony, sometimes with evidence of sediment interference with growth. The predominance of
the murchisoni morph may in part reflect the winter period as the most common time of colony
death due to sediment burial. However, there is also evidence of diffluens morph colonies that have
suffered at least minor peripheral sediment encroachment but show no significant signs of
intercorallite ridge formation. Ultimately, the full range of variation may have a genetic com-
ponent that is not apparently subject to strong selection pressure. A parallel may be drawn
with the pattern of ecomorph development in scleractinian coral species (Veron and Pichon
1976).

Considering size variation, correlation between mean corallite area and mean tabularium
diameter is relatively weak (text-fig. 5b). Corallite area varies widely within some colonies and little
in others, whilst tabularium diameter shows relatively little intracolonial variation. Corallite area
of apparently mature corallites in the sample as a whole varies by an order of magnitude, as does
the volume of corallite skeletal material added annually (from 200 mm3 to 2021 mm3). The
relationship between colony longevity, rate of growth, and corallite area is also very weak (text-
fig. 4). There is considerable scatter in the data but some tendency for enhanced survivorship to

SCRUTTON: SILURIAN A R ACHNOPH YLLI D CORALS

7

40

30

to

s_

>.

+-«

>

a)

U)

c

o

20

10

o

o

_l

100

o

i i

200

Ac/cycle length (mm)

text-fig. 4. Longevity (in years) plotted against mean corallite area (Ac) divided by mean length of annual
growth cycle (result in mm) for colonies of Arachnophyllum murchisoni from the Much Wenlock Limestone
Formation. Farley Dingle sample indicated by open circles.

be associated with larger corallite areas and slower growth rates. There is no evidence of differential
survivorship between the calicular surface morphs.

A priori , a correlation between larger corallite area (and strong corallite walls) and the ability
to survive periodic burial by a thin layer of sediment might be expected. Such a relationship has
been demonstrated for some living scleractinians (Marshall and Orr 1931) and confirmed by
Hubbard and Pocock (1972), who also conclude that more active polyps have greater calical relief
and V-shaped calical floors (as here in the murchisoni morph). The weak correlation found here
may reflect the importance of microenvironmental variation within these bioherms (essentially
isolated patches of dense colonization with minor relief on the sea-floor which are persistent in
time) and the survivorship value of settlement site. The inverse correlation between survivorship
and growth rate may be no more than a reflection of a link between growth rate and sedimentation,
periodic (annual) higher rates of sediment accumulation stimulating faster vertical rather than
lateral growth, and also increasing the likelihood of death by burial. On the other hand, persistent
turbidity resulting from repeated sediment resuspension correlates with decreased growth rates in
living scleractinian corals (Dodge et al. 1974), possibly due to the deflection of metabolic effort
into constant cleaning activity and the effect of decreased light intensity.

Variation in other species of Arachnophyllum

Other species in this group and members of the A. pentagonum group all possess very similar
internal structure. In particular, they all develop the characteristic reticulate septal structure of A.
murchisoni, identical in form and dimensions (text-fig. 7a, b, e). The features used to discriminate
these species here appear relatively minor when compared with the variation described in the
English sample of A. murchisoni. In each case, however, variation in these features appears to be dis-
continuous between species on the basis of previous descriptions and the material available here.

A. sinemurum and A. separatum are both discriminated on growth form and calicular size alone

PALAEONTOLOGY, VOLUME 32

(see systematic descriptions). Their retention as distinct species may be questionable and they are
undoubtedly closely related to A. murchisoni. They occur mainly in argillaceous facies of the
Louisville Limestone and if growth form alone was involved, an ecophenotypic explanation of
their distinct form might be acceptable. Nevertheless, they are also characterized by large corallite
area, the two species appear to coexist without intermediaries and specimens of typical A. murchisoni
are known from similar facies, e.g. the Mulde Marl. However, some of the latter specimens show
a tendency for occasional peripheral corallites to vary towards the form found in A. sinemurum.
The presence of roughly contemporary sinemurum morphs in England and North America could
be due to migration, but might also represent independent evolutionary events (if specific distinction
is accepted) or simply extreme variants in large samples (if this material is assigned to A. murchisoni).
In any event, there is no morphological basis for separating the English and North American
sinemurum.

Members of the A. pentagonum group appear to be more securely distinguished from the A.
murchisoni group on their significantly coarser dissepiments. However, the former is almost
restricted in distribution to the North American dolomite belt of Berry and Boucot (1970), whereas
the latter is similarly characteristic of limestone and marly facies. The two groups overlap in
distribution only in Quebec, where A. murchisoni is considered to have evolved from A. pentagonum.
The possibility has to be considered that the distinct dissepimentaria may be a reflection of
contrasting environmental and latitudinal distribution, but their sequential ranges and individually
consistent internal structure suggests that this is a case of punctuated phyletic speciation. Indirect
evidence from the distribution of P. mamillaris supports this conclusion. Within the A. pentagonum
group, all three species display only murchisoni- morph calicular surfaces, least well developed in
A. pygmaeum. This latter species is more clearly distinct from the other two, whereas A. pentagonum
itself and A. striatum appear to differ only in the much greater mean calicular area of the latter.
Both species may show enhanced intercorallite ridge height at dense bands (even though these are
less clearly developed), often involving short supplementary colums of small globular dissepiments,
as in A. murchisoni (PI. 5, figs. 2, 4, 8). Although only a small sample is quoted here, the disjunct
corallite size ranges of A. pentagonum and A. striatum are supported by large samples from the
Fossil Hill Formation of Fossil Hill, Manitoulin Island (Dr Paul Copper, pers. comm.). The three
species appear to show low intraspecific variability, although variation in A. pygmaeum is
insufficiently known.

Variation in species of Prodarwinia

Species of Prodarwinia never develop intercorallite walls and show little variation in the development
but more in the spacing of mamelons bearing the tabularia. This is most pronounced in P. speciosa,
less in P. striata and P. distans, and least in P. mamillaris , which always has crowded, prominent,
rounded mamelons. Mamelon form is particularly distinctive in P. distans.

P. mamillaris and P. distans appear to show relatively limited intraspecific variation and differ
in several, if individually minor, details. This is despite the fact P. mamillaris bridges the distributions
of A. murchisoni and A. pentagonum in time and space in North America, maintaining consistency
of internal structure, particularly characteristics of the dissepimentarium, over a range of carbonate
facies. This suggests that the contrast in dissepimentaria between A. murchisoni and A. pentagonum
is a valid specific criterion.

Septal crests are only very weakly developed in P. mamillaris and P. distans (PI. 6, fig. 10;
PI. 8, figs. 2, 5, 7). In P. speciosa and P. striata , the degree of development of solid septal crests
shows considerable variation, particularly in the former where in some colonies they may be no
better developed than in P. mamillaris (PI. 6, figs. 3, 5, 7, 10, 12; text-fig. 11C). Throughout this
genus, however, their structural characteristics, particularly the peculiar pseudorhabdacanthine
appearance of trabeculae in the dissepimentarium (text-fig. 7c, d), appear to be consistent.

SCRUTTON: SILURIAN ARACHNOPH YLLID CORALS

9

SYSTEMATIC PALAEONTOLOGY

Terminology follows the standard usage of Hill (1981) and others. In addition, to save repetition, the
distinctive structure produced by the fusion of carinae from adjacent septa in A. murchisoni and other species
is referred to as murchisoni-type carination or murchisoni- type reticulate septal structure. It is fully described
under A. murchisoni. The characteristic trabecular structure of Prodarwinia, described under P. speciosa, is
described here as pseudorhabdacanthine.

Measurements are also standard except that corallite area has been measured for a group of representative
corallites and averaged for each colony using the method described by Scrutton (1981). In addition, the
coarseness of dissepiment sections in cross-section had been measured as sections intersected along a line
between adjacent corallite centres 10 mm long, or scaled up to 10 mm.

Horizon and locality for specimens from old collections have been given as registered with the specimen
with clarification in parentheses. Interpretation of this information will be clear from the range quoted for
the species. Ranges are based only on specimens studied here or sufficiently well illustrated for their identity
to be certain. Information from species lists is suspect because of the confusion over the identity of North
American species.

For convenience, closely related species of Arachnophyllum are placed in two species groups, the A.
murchisoni and A. pentagonum groups. Brief notes are given for species probably referable to either
Arachnophyllum or Prodarwinia but not suitable for revision in detail for some reason. Finally, species
removed from these two genera are considered briefly and redescribed if appropriate.

Abbreviations. BMNH — British Museum (Natural History), London; BGS — British Geological Survey,
Keyworth; SM— Sedgwick Museum, Cambridge; BMAG— City Museum and Art Gallery, Bristol; UAGC—
Geological Collections, University of Aberdeen; MNHN IP— Institut de Paleontologie, Museum National
d’Histoire Naturelle, Paris; UCBL— Universite Claude Bernard, Lyon; RM — Riksmuseet, Stockholm; SGU —
Swedish Geological Survey, Uppsala; IGT— Institute of Geology, A.N. Est. SSR, Tallinn; USNM— National
Museum of Natural History, Washington; AMNH — American Museum of Natural History, New York;
MCZ— Museum of Comparative Zoology, Harvard University, Cambridge, Mass.; UMMP— Museum of
Paleontology, University of Michigan, Ann Arbor; GSC— Geological Survey of Canada, Ottawa.

Subclass rugosa Edwards and Haime, 1850
Order stauriida Verrill, 1865
Suborder arachnophyllina Zhavoronkova, 1972

e.p. 1981 Arachnophyllina; Hill, p. 206.
e.p. 1981 Ketophyllina; Hill, p. 217.

Diagnosis. Mainly colonial stauriids in which septa usually become multitrabecular and discontinu-
ous in the dissepimentarium, either lonsdaleoid or as septal crests. Tabularia well defined, with or
without peripherally dished tabellae which may rise to meet septal sections in the axis, or with an
axial series of arched or mesa-shaped tabellae, or with simple flat to gently depressed more or less
complete tabulae. Dissepiments variable in form but usually elongate, poorly inflated.

Discussion. A full revision of this suborder is beyond the scope of this paper. In general terms,
this revision of Arachnophyllum suggests that the Arachnophyllidae have more in common with
the Kyphophyllidae and Endophyllidae of Hill (1981), which constitute the Arachnophyllina as
here defined, than with the Entelophyllidae, which I suggest should be removed to the Columnariina.

Family arachnophyllidae Dybowski, 1873

1873
? 1887
1949
1956
e.p. 1965
e.p. 1965
e.p. 1971
1981
non 1977

Arachnophyllidae Dybowski, p. 339.
Chonophyllidae Holmes, p. 25.
Chonophyllidae; Stumm, p. 48.
Arachnophyllidae; Hill, p. 274.
Arachnophyllidae; Iwanowskii, p. 114.
Arachnophyllidae; Stumm, p. 30.
Chonophyllidae; Oliver and Galle, p. 67.
Arachnophyllidae; Hill, p. 214.
Arachnophyllinae; Pedder, p. 173.

10

PALAEONTOLOGY, VOLUME 32

Discussion. Considerable divergence of opinion on the interpretation of the Arachnophyllidae
exists. Pedder (1976, p. 133; 1977, p. 175) and McLean (1976a, p. 185) regard this group of corals
as very close to the Cyathophyllidae, to the extent that Pedder made the Arachnophyllinae a
subfamily of the Cyathophyllidae. The predominant characteristic of Arachnophyllum is the
development of septa in the dissepimentarium only as crests on discrete dissepimental surfaces. I
agree with Hill (1981, p. 214) that affinities are probably closer to the chonophyllids. In addition
to the genera included in the family by Hill. I add Iowaphyllum.

1839
? 1844
1846

1850

1851
1855
1874
1876
1927
1940
1956

e.p. 1965
e.p. 1975
1981

non 1819
non 1819

Genus arachnophyllum Dana, 1846

Acervularicr, Lonsdale, p. 691.

Lamellopora Owen, p. 70.

Arachnophyllum Dana, p. 186.

Strombodes; Edwards and Haime, p. lxx.

Strombodes ; Edwards and Haime, pp. 172, 426.

Strombodes ; Edwards and Haime, p. 293
Astraeophyllum Nicholson and Hinde, p. 152.

Strombodes ; Rominger, p. 130.

Arachnophyllum ; Lang and Smith, p. 452.

Arachniophyllum\ Lang, Smith and Thomas, p. 19.
Arachnophyllum', Hill, p. 274.

Arachnophyllum', Iwanowskii, p. 114.

Arachnophyllum', McLean, p. 54.

Arachnophyllum', Hill, p. 215.

Acervularia Schweigger, tab. 6.

Strombodes Schweigger, tab. 6.

Diagnosis. Astraeoid, thamnasterioid colonial corals. Septa in tabularium thin, usually solid blades
but rarely may be almost completely suppressed. Major and minor septa usually distinguished,
often irregularly developed. In dissepimentarium, septa reduced to crests on discrete dissepimental
surfaces, weakly to strongly developed, carinate, more rarely naotic. Dissepiments small, poorly
inflated to coarse, irregular, moderately inflated, usually more or less clearly zoned by size, with
septal crests on surfaces of smaller dissepiments. Intercorallite wall, if present, formed by arching
of dissepimental surface, with or without supplementary vertical series of more globose dissepiments.
Tabulae complete or incomplete, with or without dished tabellae in narrow peripheral zone, usually
with axial arched or mesa-shaped tabellae. Increase peripheral, non-parricidal. Septal microstructure
of monacanthine trabeculae.

Type species (by subsequent designation of Lang and Smith 1927, p. 452). Acervularia baltica Schweigger;
Lonsdale 1839, pi. 16, fig. 8 b-e ( non fig. 8, 8a; non Schweigger 1819, tab. 6) = Strombodes murchisoni Edwards
and Haime 1851, p. 428; 1855, p. 293, pi. 70, figs. I and la-d.

Discussion. Contrary to Laub (1979, p. 193), whilst Lonsdale’s A. baltica was deemed to contain
two different species, Lang and Smith’s (1927, p. 452) selection of one of them, A. murchisoni , as
type species was necessary and valid. In fact I show the second species, A. diffluens , to be a junior
subjective synonym of A. murchisoni and thus all of Lonsdale’s material is regarded as conspecific.

Edwards and Haime (1850, p. lxx) wrongly quoted Strombodes pentagonus Goldfuss as type
species of Strombodes and listed Arachnophyllum as a junior synonym of that genus . This usage
was widely followed until the early part of this century. Strombodes , type species Madrepora
stellaris Linnaeus (designated by M‘Coy 1849, p. 10), is a phaceloid genus quite distinct from
Arachnophyllum although possibly ancestral to the arachnophyllids. Goldfuss’ species is congeneric
with A. murchisoni and is described below.

Lamellopora Owen is based on L. inf undibul aria Owen, 1844 which is shown below to be

SCRUTTON: SILURIAN ARACHNOPH YLLI D CORALS

11

unrecognizable. It could be an Arachnophyllum but equally likely it could be a stromatoporoid.
The genus should be set aside.

Astraeophyllum Nicholson and Hinde is based on A. gracile Nicholson and Hinde, 1874
which is shown below to be a synonym of A. pygmaeum. This species possesses typical murchisoni-
type septal structure and Astraeophyllum is therefore established as a junior synonym of Arachno-
phyllum.

Arachnophyllum is distinguished from Prodarwinia principally on the basis of its septal structure.
Septa in Arachnophyllum are only thickened to contiguity in a very thin veneer on dissepimental
surfaces, if at all, and possess distinctive murchisoni- type carination. In Prodarwinia , on the other
hand, septa are weakly and irregularly carinate, or non-carinate, and are present in the
dissepimentarium of some species as solid septal crests, which may be quite thick. Septa may be
multitrabecular in both but in Arachnophyllum trabeculae are monacanths, whilst in Prodarwinia
they appear to be pseudorhabdacanths, at least in the outer dissepimentarium where the trabeculae
are set in lamellar sclerenchyme.

Zenophila , tentatively placed in synonymy with Arachnophyllum by McLean (1975, p. 55), has
unthickened, more continuous septa composed of very long trabeculae. I regard it as a distinct
genus, although provisionally retaining it in the Arachnophyllidae.

Range. Silurian, Llandovery (?Aeronian, Telychian) — Ludlow; England, Gotland, USA, Canada, 7USSR,
?Australia.

1839

1841

1850

1850

1850

1851
1851
1851
1851
1851
1855
1855

1855

1855

1887

1887

1887

1887

1927

e.p. 1927
1927
1949
1965
1965
non 1819

Arachnophyllum murchisoni group

Arachnophyllum murchisoni (Edwards and Haime, 1851)

Plate 1, figs. 1-7; Plate 2, figs. I ll; Plate 3, figs. 1-8;
text-figs. 3a, b, 4-6, 7a, b, e

Acervularia baltica Schweigger; Lonsdale, p. 689, pi. 16, fig. 8, 8a-e.

Acervularia baltica Phillips, p. 13, pi. 7, fig. 18e, a, b , c.

Actinocyathus balticus d’Orbigny, p. 48.

Actinocyathus phillipsii d’Orbigny, p. 107.

Arachnophyllum typus M’Coy, p. 278.

Arachnophyllum typus M‘Coy; M‘Coy, p. 38, pi. 1b, fig. 27 and 21a.

Strombodes labechii Edwards and Elaime, p. 427.

Strombodes murchisoni Edwards and Elaime, p. 428.

Strombodes phillipsi (d’Orbigny) Edwards and Haime, p. 429.

Strombodes diffluens Edwards and Haime, p. 431.

Strombodes typus (M‘Coy) Edwards and Haime, p. 293, pi. 71, fig. 1, la, b.

Strombodes murchisoni Edwards and Haime; Edwards and Haime, p. 293, pi. 70, fig. 1 and
1 a-d.

Strombodes phillipsi (d’Orbigny); Edwards and Haime, p. 294, pi. 70, fig. 2 and 2a.
Strombodes diffluens Edwards and Haime; Edwards and Haime, p. 294, pi. 71, fig. 2 and 2a.
Strombodes striatus (d’Orbigny); Davis, pi. 121, fig. 1; pi. 122, figs. 1 and 2.

Strombodes pentagonus Goldfuss; Davis, pi. 121, figs. 2 and 3.

Strombodes pygmaeus Rominger; Davis, pi. 123, fig. 1.

Strombodes incertus Davis, pi. 123, fig. 2.

Arachnophyllum murchisoni (Edwards and Haime); Lang and Smith, p. 467, pi. 34, figs. 2
and 3.

Arachnophyllum diffluens (Edwards and Haime); Lang and Smith, p. 468, pi. 34, fig. 4.
Arachnophyllum typus M'Coy; Lang and Smith, p. 469.

Arachnophyllum pentagonum (Goldfuss); Amsden, p. 104, pi. 26, figs. 1-6.

Arachnophyllum pentagonum (Goldfuss); Stumm, p. 30, pi. 20, figs. 1-3; pi. 21, fig. I.
Arachnophyllum striatum (d’Orbigny) Stumm, p. 30, pi. 20, figs. 4-6.

Astrea baltica Schweigger, tab 6.

12

PALAEONTOLOGY, VOLUME 32

Diagnosis. Arachnophyllum of variable external form comprising tabular, discoidal, or low domal
colonies with or without more or less sharply defined intercorallite ridges on the calicular surface.
Surface flat, or may funnel gently to tabularial pit, or tabularia centred on broad, low mamelons.
Tabularia more or less well defined, 1 -8-4-7 mm in diameter with fifteen to twenty-two thin major
septa usually with weak bilateral arrangement and reaching the axis where they may form a weak
columella. In the dissepimentarium, major and minor septa reduced to crests on discrete
dissepimental surfaces, occasionally naotic peritabularially and always distinctly carinate. Carinae
of adjacent septa normally fuse to form a characteristic box-work structure with dissepimental
plates. Dissepiments variable in size but characteristically small on average. Dissepimentarium
surface matches colony surface but with no structural modifications at intercorallite boundaries
except for intermittent columns of small, globose dissepiments in some specimens. Tabulae usually
flat-topped domes. Increase marginarial, non-parricidal.

Type material. Lectotype of Strombodes murchisoni Edwards and Haime (chosen by Lang and Smith (1927,
p. 467)): BGS GSC6577 (Much) Wenlock Limestone (Lormation), Shropshire (= Acervularia baltica
Schweigger; Lonsdale (1839, pi. 16, fig. 8<?)). Paralectotypes: BGS GSC6579 and BGS GSC6578, same horizon
and locality ( = A. baltica Schweigger; Lonsdale (1839, pi. 16, fig. 86, c, d respectively)).

Two other species placed in synonymy with A. murchisoni here are also based on one or other of the
specimens illustrated as A. baltica Schweigger by Lonsdale (1839, pi. 16, figs. 8 and 8 a-e). S. diffluens Edwards
and Haime was said by them to be the species illustrated by Lonsdale (1839) in pi. 16, figs. 8 and 8a. Edwards
and Haime (1855, pi. 71, figs. 2 and 2a) figured a specimen of their own under this name, which can be
construed as a syntype with Lonsdale's material. Lang and Smith (1927, p. 468) reported Lonsdale’s specimens
as missing and therefore chose the original of Edwards and Haime (1855, pi. 71, figs. 2 and 2a), which is
MNHN IP SI 1664, from Wenlock, as the lectotype. Arachnophyllum typus M'Coy was said to be Acervularia
baltica Lonsdale (1839, pi. 16, fig. 8a) by M'Coy (1850). However, the figure quoted may have been in error
as in the same paper, he noted fig. 8a as probably the same as his new species Strombodes wenlockensis
(M'Coy 1950, p. 274), whilst later (M'Coy 1851, p. 38) he quoted Lonsdale’s fig. 8 b with reference to A.
typus. As noted above, the original of Lonsdale’s fig. 8a, a paralectotype of S. diffluens, is missing and the
original of fig. 8 b (BGS GSC6579) is a paralectotype of A. murchisoni. However, M'Coy in 1850 clearly had
in front of him at that time the specimen he figured for the species in 1851 as he refers to it and quotes its
locality (M'Coy 1850, p. 278). This figured specimen (M'Coy 1851, pi. IB, figs. 27, 27a) is SM A4828, from
the Wenlock Limestone of Aymestry, and is a syntype with Lonsdale’s specimen. Lang and Smith (1927,
p. 469) quoted SM A4828 as the holotype of the species in error and to avoid any confusion, I hereby
designate it as lectotype of A. typus.

The holotype by monotypy of A. phillipsii is the original of Phillips (1841, pi. 7, fig. 18e a, b , c) from an
unknown locality in the (Much) Wenlock Limestone (Lormation). This appears to be the same specimen as
that illustrated by Edwards and Haime (1855, pi. 70, figs. 2 and 2a), said to be from 'Wenlock’. MNHN IP
SI 1668 is labelled as probably the original of Edwards and Haime’s figures, but I regard this specimen as
not convincingly similar to either Phillips’ or Edwards and Haime’s illustrations. The holotype, therefore,
appears to be missing.

An original of Strombodes labecheii is MNHN IP SI 1666 and SI 1665 may be another syntype (Dr Pierre
Semenoff-Tian-Chansky, pers. comm.). I hereby select MNHN IP SI 1666 as lectotype on his advice. Both
specimens are labelled as from Wenlock.

EXPLANATION OF PLATE 1

Pigs. 1 7. Arachnophyllum murchisoni, calicular surfaces of type and other material, all x 1. 1, lectotype of
A. murchisoni (BGS GSC6577); Wenlock Limestone; Shropshire. 2, lectotype of A. typus (SM A4828);
Wenlock Limestone; Aymestry. 3, lectotype of Strombodes diffluens (MNHN IP SI 1664); Wenlock
Limestone; Wenlock. 4, lectotype of Strombodes labecheii (MNHN IP SI 1666); Wenlock Limestone;
Wenlock. 5, BMNH R46415; Wenlock Limestone; ?Dormington, Hereford and Worcester. 6, RM
Cn67213; Mulde Marl; Eksta, Djupvik, Gotland. 7, BMNH R53513; Much Wenlock Limestone Lormation;
Lea Quarry, north side of B4371, \\ km south-west of Much Wenlock, Shropshire (SO 5972 9861).

PLATE 1

SCRUTTON, Arachnophyllum

14

PALAEONTOLOGY, VOLUME 32

Material. English material: BM R536 16-536 17 (?same colony), Silurian, Llandovery, Upper Haugh Wood
beds (Petalocrinus Limestone); temporary trench in Haugh Wood, 2-3 km west of Woolhope, Hereford and
Worcester (SO 58853558) [SM A5700, Woolhope Limestone; Wenlock ridge (label almost certainly a
transpositional error for Wenlock Limestone, Woolhope, there being no Woolhope Limestone on the ridge
of Wenlock Edge)]. SM A46667, Silurian, Wenlock, Wenlock Shale; Blaisdon Hall, May Hill, Glos. BGS
37964, Silurian, Wenlock, (Much) Wenlock Limestone (Lormation); May Hill Longhope, Glos. MNHN IP
SI 1669, Silurian; Dudley (figured Edwards and Haime 1855, pi. 70, fig. lc). IP SI 1667, no details recorded
but almost certainly Much Wenlock Limestone Lormation; England. BMNH R52622, Much Wenlock
Limestone Lormation; old quarry on Lincoln Hill, Ironbridge, Shropshire (SJ 67050390). R52623-52627,
Much Wenlock Limestone Lormation; roadside exposure on B4378 (Earley Dingle), 2-5 km north of Much
Wenlock, Shropshire (SJ 633023). R52628 -52630, Much Wenlock Limestone Lormation; Parley Quarry,
Gleedon Hill, 1-7 km north of Much Wenlock, Shropshire (SJ 629017). R52631, Much Wenlock Limestone
Lormation; Coates Quarry, north side of B4371, 2 km south-west of Much Wenlock, Shropshire (SO
60419935). R53513, Much Wenlock Limestone Lormation; Lea Quarry, north side of B4371, 2-75 km south-
west of Much Wenlock, Shropshire (SO 59809875).

Gotland material: RM Cn20146, 66040 66045, 67213-67215, 67075, Silurian, Wenlock, Mulde Marl; Eksta,
Djupvik. Cn20155, 55254, Silurian; Gotland. SGU 61 10, Silurian, Wenlock, Klinteberg Beds; 4-75 m south-
south-west of point 35,2 on railway line, Projel Parish, Hemse. RM Cn 12366, Silurian, Wenlock, Halle Beds;
Horsna Canal.

North American material: GSC 91554, Silurian, Telychian, La Vieille Lormation; Little Port Daniel River,
Gaspe, Quebec. GSC 90726, Silurian, ?upper Wenlock, Sayabec Lormation; 1 m east-north-east of La
Redemption, Quebec. MCZ 8071, Niagara ‘red clay’ (Silurian, late Wenlock (?early Ludlow), Louisville
Limestone); first quarry on Beargrass Creek, East Louisville, Kentucky. UMMP 5307, 34212, USNM 422818
422819, Louisville Limestone; Louisville, Kentucky. USNM 422820 422821, Lobelville Limestone (Silurian,
Ludlow, Brownsport Lormation); Sewell’s Spring, Tennessee.

Range. Silurian, late Llandovery, La Vieille Lormation, Quebec; latest Llandovery, Petalocrinus Limestone;
Wenlock (?Woolhope Limestone), Wenlock Shale, Much Wenlock Limestone Lormation, England; Wenlock,
Halle Beds, Mulde Marl, Klinteberg Beds, Gotland; late Wenlock (?early Ludlow), Louisville Limestone,
Kentucky; mid Ludlow, Brownsport Lormation, Tennessee; Wenlock — early Ludlow (probably later Wenlock
part), Sayabec Lormation, Quebec.

Description. Astraeoid to thamnasterioid colonies < 300 x 200 x 1 50 mm. Colony form thin tabular sheets to
thick tabular or domal masses with or without ragged margins, irregular to subcircular in plan. A majority
have moderate to well-developed intercorallite walls <2-5-3 mm, exceptionally 8 mm high, formed by
uparching of the dissepimental surface between adjacent corallites (PI. 1). In these, corallites are polygonal
with straight to slightly curved walls, equidimensional to elongate, with the longer axis <l-6x the shorter.
Corallite size highly variable, 0-70-6-33 cm2 in area. All gradations exist to colonies in which an intercorallite
ridge is completely absent and adjacent tabularia are separated by a flat or gently depressed common
dissepimentarium. The peritabularial area may be a broadly inflated mamelon, or flat or gently depressed,
curving down to the edge of the more steeply declined wall of the tabularial pit, normally 1 2 mm deep with
a small but often prominent axial boss.

In transverse section, tabularia prominent but rather imprecisely defined, 1-8 4-7 mm diameter and
5-33 mm apart centre to centre (PI. 2, figs. 1, 2, 4, 6, 8, 10; PI. 3, figs. 1, 3, 5, 7). Corallite margins either un-
defined (PI. 2, figs. 2 and 6) or weakly to strongly defined by a more or less distinct row of subcircular dis-

EXPLANATION OF PLATE 2

Pigs. 1-11. Arachnophyllum murchisoni, internal structure of type and Gotland material, all x 2-5. 1, cross-
section of lectotype of A. murchisoni (BGS PP4533); Wenlock Limestone; Shropshire. 2, cross-section and
3, longitudinal section of lectotype of Strombodes diffluens (MNHN IP SI 1664); Wenlock Limestone;
Wenlock. 4, cross-section and 5, longitudinal section of lectotype of A. typus (SM A4828c, d)\ Wenlock
Limestone; Aymestry. 6, cross-section and 7, longitudinal section of lectotype of Strombodes labecheii
(MNHN IP SI 1666); Wenlock Limestone; Wenlock. 8, cross-section and 9, longitudinal section (RM
Cn67214); Mulde Marl; Eksta, Djupvik, Gotland. 10, cross-section and II, longitudinal section (SGU
6110); Klinteberg Beds; Projel Parish, Hemse, Gotland.

PLATE 2

SCRUTTON, A rachnophyllum

16

PALAEONTOLOGY, VOLUME 32

24r

22

20

16

14 -

/

mA

o •

• • •

18

©

o °

••• ° •• :

• • •

•

dA°

• •

•

t a* •

o • iA • o

• □

dt (mm)

Ac

(cm2)

•

murchisoni morph

mA

A. murchisoni lectotype

o

sub -diffluens morph

tA

A. typus lectotype

o

diffluens morph

dA

A. diffluens lectotype

m

Gotland material

□

N. American material

text-fig. 5. Variation in a, mean number of major septa (n) and b, mean corallite area
(Ac in cm2), both plotted against mean tabularium diameter (dt in mm) for Arachnophyllum
murchisoni. For discussion see text.

SCRUTTON: SILURIAN ARACHNOPH YLLID CORALS

17

24 r

22

20

n is

16

14 -

o'
' o

• •

dt (mm)

8r

6-

Ag

(cm2)4

2-

0

i

i

2

3

4

J

5

dt (mm) B

text-fig. 6. Variation in a, mean number of major septa (n) and b, mean corailite
area (Ac in cm2), both plotted against mean tabularium diameter (dt in mm) for the
Farley Dingle sample of Arachnophyllum murchisoni. Dotted line outlines held of variation
for the total sample of A. murchisoni ; legend otherwise as for text-fig. 5. For discussion

see text.

sepimental sections reflecting a strong wall on the calicular surface (PI. 2, fig. 1; PI. 3, Figs. 1, 3, 7). Septa
generally well developed in the tabularium as usually continuous, slightly and uniformly thickened vertical
plates; occasionally they are weakly developed, or even rarely locally absent (PI. 3, fig. 3). Division into
major and minor septa may be rather irregular but usually more or less clearly distinguished. Minor septa

18

PALAEONTOLOGY, VOLUME 32

may or may not just penetrate the tabularium. Major septa extend into the axial area, sometimes with free
axial ends which may be slightly rhopaloid, or more commonly fused with corresponding septa of the adjacent
quadrant to form a more or less strongly bilaterally symmetrical pattern. Septa best developed, thickest, and
sometimes coated with sclerenchyme on tabularial surfaces. Axial intersections of arched tabulae may thus
suggest substellate or aulate axial structures. In the dissepimentarium, septa reduced to crests on discrete
dissepimental surfaces. Intracolonial development of crests slightly variable, but intercolonial variation from
scarcely developed to very dense. At the tabularium boundary, corresponding with the peritabularial zone
of steeply dipping dissepiments, septa usually carinate, with yard-arm or zigzag carinae, and may be
discontinuous or naotic in structure. As the dissepimental surface curves over to a sub-horizontal attitude
towards the corallite periphery, the septa are either present or absent depending on their degree of development
and the level of section. Where present, further septa intercalate between those radiating from the tabularium
to maintain a constant lateral spacing of 0-20-3 mm. Carinae become regularly and well developed, four to
seven per mm, crossbar to zigzag in form, and fused with those of adjacent septa to form a distinctive
reticulate pattern in cross-section (text-fig. 7a, e). Septa may be multi-trabecular and contiguous immediately
above the dissepimental surface on which they are based, bounding the reticulate pattern by a more or less
narrow band of dense tissue where this surface is intersected, creating the impression of one or more partial
false walls around the corallite (PI. 2, figs. 6 and 10; PI. 3, figs. 1 and 7). In some specimens, carinae clearly
extend higher than the septal crest so that isolated crossbars may occur in otherwise aseptate parts of the
dissepimentarium. Dissepimental sections vary considerably in size and shape, from irregularly arcuate to
subcircular, twelve to eighteen in 5 mm.

In longitudinal section, tabularia well defined, more or less vertical, and generally of uniform width. Axial
tabcllae arched and dense, occasionally flat-topped domes and less dense, about three-quarters of the
tabularium wide, fifteen to forty in 5 mm. They may be disrupted by septal sections. A peripheral series of
flat to dished tabellae of variable development, ten to thirty in 5 mm, interlock with axial tabellae and
dissepiments. Dissepiments dominantly weakly inflated, rather elongate, regular to irregular vesicles. Highly
inflated dissepiments rare but may occasionally occur in zones. Dissepimental size highly variable overall,
<3x3-5 mm. Layers of coarser dissepiments grade distally into finer dissepiments capped by septal crests,
these units strongly and pervasively developed, 0- 1 3-5 mm thick (PI. 2, figs. 6 and 11; PI. 3, figs. 2 and 8).
Septal crests appear as successive surfaces bearing thin, parallel and regularly spaced, more or less vertical
sections through carinae, varying from barely discernible thorns to continuous structures penetrating <6
layers of dissepiments and 3 mm high (PI. 2, figs. 5 and 11; PI. 3, figs. 2 and 8; text-fig. 7b). Usually, they
are about 0-5 mm high and penetrate only one or possibly two layers. Their orientation is approximately
normal to dissepimental surfaces of low to moderate relief but intermediate between normal and the vertical
when surfaces are of high relief. Carinae may be associated with one or more thickened dissepimental surfaces
and may merge into solid plates when the septal plane is sectioned. When well developed, successive septal
crests may become locally vertically continuous but normally, intervening zones of uninterrupted dissepiments
are wider than septate zones, although the former may sometimes show scattered, isolated, and low carinal
sections. Dissepimental surface steeply inclined downwards to the tabularium boundary, gently arched or
flat peritabularially and flat to sharply arched between tabularia. A distinctive vertical column of particularly
globose dissepiments, usually discontinuous vertically, or of limited vertical extent may correspond to well-
developed intercorallite ridges on the calicular surface (PI. 3, figs. 6 and 8). Uparching of the dissepimentarium
profile maximized at calicular surfaces based on zones of small dissepiments.

Increase marginarial non-parricidal, more common in peripheral areas of colonies. Rare offsets in central
areas occur more or less equidistant from surrounding established tabularia.

Septa and carinae constructed of equidimensional, smooth to tufted monocanths generally 0 06-0- 10 mm
diameter (text-fig. 7a, b). All or nearly all of any thickening of these elements appears to be trabecular;
lamellar sclerenchyme is rarely developed. Dissepiments and tabulae constructed of fibronormal tissue.

EXPLANATION OF PLATE 3

Figs. 1-8. Arachnophyllum murchisoni , variation in internal structure in Farley Dingle sample and North
American material, all x2-5. 1, cross-section and 2, longitudinal section (BMNH R52626). 3, cross-

section and 4, longitudinal section (BMNH R52627). 5, cross-section and 6, longitudinal section (BMNH
R52623). All Much Wenlock Limestone Formation; roadside exposure on B4378 (‘Farley Dingle’),
2-5 km north of Much Wenlock, Shropshire (SJ 633023). 7, cross-section and 8, longitudinal section

(UMMP 34212); Louisville Limestone; Louisville, Kentucky.

PLATE 3

SC R L) TT ON, A rachnophyllum

20

PALAEONTOLOGY, VOLUME 32

Discussion. Five species of Arachnophyllum have been described from the Much Wenlock Lime-
stone; Actinocyathus phillipsii d’Orbigny, Arachnophyllum typus MLCoy, Strombodes labecheii
Edwards and Haime, S. murchisoni Edwards and Haime, and S. diffluens Edwards and Haime.
D’Orbigny’s species name has been unused as a senior synonym since Edwards and Haime (1855,
p. 294), who even then remarked on the close similarity of phillipsii, murchisoni, and typus. In
the same work, they placed their own species S. labecheii , without illustration, in synonymy with
M’Coy’s species.

Lang and Smith (1927, p. 467 et seq.) placed phillipsii and murchisoni in synonymy but, in
apparent ignorance of d’Orbigny’s priority, described the species as Arachnophyllum murchisoni.
The paliform lobes attributed to Strombodes phillipsii by Edwards and Haime (1855, p. 294) they
dismissed as unmodified intrathecal portions of the septa. However, the appearance, rather
exaggerated in Edwards and Haime’s illustration, is real and a consequence of a prominent and
well-preserved axial boss in the calice. Lang and Smith (1927, p. 469) maintained A. typus distinct
from A. murchisoni, which it very closely resembles in surface features, by virtue of its smaller
corallite size. The type material of A. diffluens, the third species they recognized, lacks the
intercorallite ridges of the previous two species. Whilst maintaining these three separate species,
Lang and Smith (1927, p. 465) clearly recognized that they represented morphological varieties of
what ‘would probably more correctly be considered as ... a single species’.

Iwanowskii (1965, p. 115) placed A. diffluens in synonymy with A. murchisoni, although the
specimen he figures (p. 28, fig. 3) is certainly not conspecific with this species. McLean (1975,
p. 54 et seep), on the other hand, considered that there might be value in distinguishing two species
groups, one with and the other without intercorallite ridges. However, although other species, such
as A. pentagonum and Prodarwinia mamillare, consistently possess or lack ridges, there is no doubt
in the British material of A. murchisoni that this feature is variable to some extent within individual
colonies, and completely so in the sample as a whole.

In the large sample of English Wenlock material described here, considerable variation is
recorded not only in the development of intercorallite ridges but also in other exterior features
such as the presence of an annular swelling surrounding the tabularium, corallite area, and the
form of the colony as a whole. On the other hand, in sections, although septa may vary in the
strength of their development, their structure otherwise, together with dissepimental and tabularial
characteristics, is broadly consistent throughout the sample. Thus only one species of Arachnophyl-
lum is recognized in the English Wenlock. In the relatively limited recent discussion of the English
material, only the species names murchisoni and diffluens have been considered, both of which have
extant lectotypes, and of which murchisoni is clearly the senior synonym. In addition, A. murchisoni
is type species for the genus. Thus, although the names phillipsi, typus, and labecheii have priority,
I make a case for the use of the name murchisoni for this species in the interests of stability.

McLean (1975, p. 55) disputed Lang and Smith’s (1927, p. 466) description of the septal structure
in this species. However, whilst in certain details the description of septal structure given here also
differs from that of Lang and Smith, the reticulate structure produced in cross-section by septal
crests and their carinae is a constant and distinctive feature of this species. Lang and Smith’s
description of ‘vertical and horizontal rows of spines’ appears to arise from the appearance of the
carinae in cross and longitudinal section. They describe these spines as forming a solid boxwork
with secondary tissue, whereas in fact the septal crests, which are usually solid blades, and the
carinae appear to be more or less wholly trabecular. The floors and ceilings of the boxwork are
provided by intersected dissepiments. In fact, Lang and Smith's (1927, pp. 452-453, figs. 5-7) line
drawings are very accurate despite their description.

A. murchisoni also occurs in North America (PI. 3, figs. 7 and 8), where it has usually been
recorded as A. pentagonum or more rarely A. striatum, and in Gotland (PI. 2, figs. 8-11). A.
pentagonum, A. pygmaeum , and A. striatum all differ in possessing dissepimentaria in which the
vesicles, although variable in size as in this species, are consistently and significantly on average
larger. The only other species possessing similar reticulate septal structure in cross-section are A.
sinemurum and A. separatum, whose relationships are discussed below.

SCRUTTON: SILURIAN AR ACHNOPHYLLID CORALS

21

text-fig. 7. a, b, Araclmophyllum murchisoni ; Silurian, Wenlock, Much Wenlock Limestone Formation;
roadside exposure on B4378 (Farley Dingle), 2-5 km north of Much Wenlock, Shropshire (SJ 633023) (BMNH
R 52625), x 40. a, cross-section, edge of tabularium at upper left; b. Longitudinal section in dissepimentarium.
C, D, Prodarwinia speciosa; Silurian, upper Aeronian, Rumba Formation; Pari, 1 km north-west of Tallinn
Virtsu road, Estonia (BMNH R53511), x40. c, cross-section, edge of tabularium at left; d, longitudinal
section in dissepimentarium. e, Araclmophyllum murchisoni; Silurian, Wenlock, Mulde Marl; Eksta, Djupvik,
Gotland (RM Cn67214), x 8; detail of septal carination on weathered surface, edge of tabularium at top left.

22

PALAEONTOLOGY, VOLUME 32

A. murchisoni is recorded here for the first time from the English latest Llandovery as the result
of recent trenching across the outcrop of the Petalocrinus Limestone in the Woolhope Inlier.

Arachnophyllum sinemurum (Davis, 1887)

Plate 4, figs. 1-6; text-fig. 8

e.p. 1887 Strombodes sinemurus Davis, pi. 121, fig. 4, pi. 122, figs. 4 and 6; non pi. 123, fig. 3.

1887 Strombodes quadrangularis Davis, pi. 122, fig. 3.

1965 Arachnophyllum sinemurum (Davis) Stumm, p. 31, pi. 21, fig. 2.

1965 Arachnophyllum quadr angular e (Davis) Stumm, p. 31, pi. 21, figs. 7-10.

Diagnosis. Broadly based, low colonies of variable form, from a single large corallite with one or
a few small peripheral offsets to a small number of more equidimensional large corallites. Corallite
walls broad to sharp ridges, fading at periphery of colony. Internal structure of murchisoni type.
Mean tabularium diameter ranging from 3-4-4-75 mm with sixteen to twenty-two major septa and
corallite areas typically 8-16 cm2.

Type material. Lectotype of A. sinemurum (chosen by Stumm 1965, p. 31): MCZ 8072 (original of Davis
1887, pi. 121, fig- 4), Silurian, Upper Niagara ‘white clay’ (late Wenlock (?early Ludlow), Louisville Limestone);
fourth quarry on Beargrass Creek, East Louisville, Jefferson Co., Kentucky. Lectoparatypes: MCZ 8070
(original of Davis 1887, pi. 122, fig. 4), Silurian, Upper Niagara ‘white clay’; quarry on Poplar Level Road,
4-8 km east of Louisville, Jefferson Co., Kentucky. MCZ 8073 (original of Davis 1887, pi. 122, fig. 6), Silurian,
Upper Niagara ‘red clay’ (late Wenlock (?early Ludlow), Louisville Limestone); Workhouse Quarry, East
Louisville, Jefferson Co., Kentucky.

Elolotype of S', quadrangularis: MCZ 8059, Silurian, Upper Niagara ‘red clay’; Daly Farm, 3-2 km south-
east of Charlestown, Clark Co., Indiana.

Other material. UMMP 21400, USNM 422822, Silurian, late Wenlock (?early Ludlow), Louisville Limestone;
Louisville, Kentucky. 7USNM 37376, Niagara Group (?late Telychian or early Wenlock, La Porte City
Formation); Masonville, Iowa. BMAG Ccl259-1260, Much Wenlock Limestone Formation; Dudley, West
Midlands.

Range. Silurian, ?late Telychian-early Wenlock, La Porte City Formation; Iowa. Late Wenlock, Much
Wenlock Limestone Formation; West Midlands. Late Wenlock (?early Ludlow), Louisville Limestone;
Indiana and Kentucky.

Description. Colonies either small, <60 x 65 x 23 mm, with a flat to broadly conical base (PI. 4, figs. 1, 3, 5),
or tabular <950 x 850 x 30 mm with a flat base and a gently to strongly arched upper surface (PI. 4, fig. 6).
Small colonies, circular in plan, consisting of one or two large mature corallites with one to four immature
peripheral offsets, delimited by sharply ridged walls towards the colony centre which fade and disappear
towards the edge of the colony. Larger colonies with several large, more equidimensional corallites separated
by well-defined to more diffuse ridges, and occasionally with peripheral offsets. Maximum corallite area may

EXPLANATION OF PLATE 4

Figs. 1 -6. Arachnophyllum sinemurum. 1-4, lectotype (MCZ 8072). 1, calicular surface, x 1. 2, cross-section,
x2-5. 3 and 4, longitudinal section; 3, x 1, 4, x2-5. Silurian, upper Niagara ‘white clay’ (Louisville

Limestone); fourth quarry on Beargrass Creek, East Louisville, Kentucky. 5, calicular surface (BMAG
Ccl259); Wenlock Limestone; Dudley, West Midlands, x 1. 6, calicular surface of holotype of Strombodes
quadrangularis (MCZ 8059); Silurian, upper Niagara ‘red clay’ (Louisville Limestone); Daly Farm, 2 m
south-east of Charlestown, Indiana, x 1.

Figs 7-9. Arachnophyllum separatum , all x 1. 7, 8, lectotype of A. separatum (AMNH FI 36584); 7, calicular
view; 8, side view; Silurian, Niagara Group (Louisville Limestone); Louisville, Kentucky. 9, holotype of
Strombodes unicus (MCZ 8076); Silurian, upper Niagara ‘red clay’ (Louisville Limestone); Workhouse
Quarry, East Louisville, Kentucky.

PLATE 4

S( RUTTON, Arachnophyllum

24

PALAEONTOLOGY, VOLUME 32

reach 22-5 cm2 (text-fig. 8b). All corallites flat-floored to saucer-shaped with deep tabularial pits <3 mm
deep containing a prominent axial boss.

In cross-section, tabularia circular, well defined, 3-4-4-75 mm mean diameter with sixteen to twenty-two
major septa (text-fig. 8a). Major septa thin to slightly thickened in tabularia, some reaching and fusing in
axis. Minor septa irregularly developed, just penetrating tabularium. Septa may be weakly naotic at the
dissepimentarium tabularium boundary. In dissepimentarium, septal traces sporadic as thin crests on discrete
dissepimental surfaces. Small, suboval dissepiment sections bearing septal crests more or less distinctly radially
orientated. Characteristic murchisoni carination well developed, usually three to four carinae per mm but
sometimes rather irregularly spaced. New septa are intercalated in the dissepimentarium to maintain a lateral
spacing of 0-25-0-33 mm. Septa may be thickened almost to contiguity immediately above dissepimental
surface, apparently composed of thickened multiple trabeculae. Dissepiment sections arcuate to subcircular,
crowded, about twelve to fourteen in 10 mm. In larger colonies, positions of walls on colony surface marked
by rows of elliptical dissepimental sections flanked by arcuate sections sometimes bearing septal crests.

In longitudinal section, tabularia defined by very steep, small dissepiments. Tabellae flat to dished peri-
pherally with a very wide axial zone of gently arched tabellae. Spacing of tabellae fifteen to twenty in
5 mm. Dissepiments very low to moderately inflated, averaging about 1 -5x0-3 mm with a maximum of
4x1-5 mm. Dissepimentarium prominently zoned with septal crests, averaging 2-5 mm apart, developed on
successive surfaces of small, crowded dissepiments. Sections of carinae <2 mm high, constructed of
monacanthine trabeculae.

Offsets peripheral, non-parricidal.

Discussion. Although typical sinemurum and quadrangulare colonial morphologies appear distinct,
they share corallites with identical structure and similar dimensions. I conclude that sinemurum-
morphs represent colonies aborted at an early stage of development, probably by unfavourable
environmental conditions, and that the quadrangulare-morph represents colonies that survived to
form several generations of offsets.

A. sinemurum is closely related to A. murchisoni , with which it shares almost identical internal
structure. The former is distinguished by the very large size of the corallites (text-fig. 8b).

Arachnophyllum separatum (Ulrich, 1886)

Plate 4, figs. 7-9; text-fig. 8

1 886 Strombodes separatus Ulrich, p. 32, figs. 1 and 2.

1887 Strombodes unicus Davis, pi. 122, fig. 5.

1965 Arachnophyllum separatum (Ulrich) Stumm, p. 31, pi. 21, figs. 3-6.

Diagnosis. Colonies or pseudocolonies of up to three trochoid to turbinate corallites. Internal
structure apparently of murchisoni type. Tabularial diameters approximately 40-4-3 mm with
about eighteen major septa and corallite areas in the range 4-8 cm2.

Type material. Lectotype (here designated): AMNH FI 36584 (Ulrich 1886, p. 32, fig. 1). Paralectotype: FI
36585 (Ulrich 1886, p. 32, fig. 2). Both Niagara Group (Silurian, late Wenlock (?early Ludlow), Louisville
Limestone); Louisville, Kentucky.

Holotype of S. unicus : MCZ 8076, Upper Niagara ‘red clay’ (Louisville Limestone); Workhouse Quarry,
East Louisville, Jefferson Co., Kentucky.

Other material. BMNH R5186, R52632-52635, Niagara Group (Louisville Limestone); Louisville, Kentucky.
Range. Silurian, late Wenlock (?early Ludlow), Louisville Limestone; Kentucky.

Description. Unfortunately all of this material is variably but badly beekitized and is unsuitable for sectioning
(PI. 4, figs. 7-9). From external appearance, although some groups of corallites are clearly associated through
offsetting, others may be the result of fusion of solitary coralla. Ulrich (1886, p. 32) describes the corallum
as ‘. . . consisting of one or more individuals, . . .’ implying that it may be solitary. All are trochoid to
narrowly turbinate, 40 mm high and 20 mm (35 mm according to Ulrich) diameter.

Corallites have flat to slightly dished surfaces with shallow to deep tabularial pits in which a low, axial
boss is sometimes preserved. Mean tabularium diameters approximately 4 0-4-3 mm with eighteen major

SCRUTTON: SILURIAN A R ACH NOPH Y LLI D CORALS

25

22r

20 -

..H •••.

n 18-

16 -

14

3

_L

4

dt (mm)

A

20 -

A

Ac

(cm2)

10-

A

° o

o

■A-

dt (mm)

□

□

B

field of A. murchisoni (total sample)

o A. sinemurum A lectotype □ English material

▼ A. separatum A lectotype A holotype A. quadrangulare

text-fig. 8. Variation in a, mean number of major septa (n) and b, maximum corallite area (Ac in cm2),
both plotted against mean tabularium diameter (dt in mm) for Arachnophyllum sinemurum and A. separatum.

For discussion see text.

septa and corallite areas up to 8 cm2 (text-fig. 8a, b). In some cases, septal structure is clearly visible on the
corallite surface, showing the typical murchisoni- type carinate nreshwork with carinae spaced four per mm.
Tabulae cannot be seen but dissepiments appear to be small and weakly inflated.

Discussion. I agree with Stumm (1965, p. 31) that S. unions (PI. 4, fig. 9) is a junior synonym of
A. separatum. From what can be seen, A. separatum is closely related to A. murchisoni , having the
same internal structure. The former is distinguished through its growth form, pseudocolonial or
weakly colonial structure and rather large corallite area. Colonial structure and growth form also
distinguish A. separatum from A. sinemurum , both species occurring together specifically at the
type locality of S. unicus and generally in the Louisville Limestone.

Arachnophyllum pentagonum group

Arachnophyllum pentagonum (Goldfuss, 1826)
Plate 5, figs. I 4; text-figs. 3c, D and 9a, 10

1826
1851
e.p. 1876
1901
1966
1981
non 1887
non 1949
non 1965

Strombodes pentagonus Goldfuss, p. 62, pi. 21, fig 3a, b.

Strombodes pentagonus Goldfuss; Edwards and Haime, p. 430.

Strombodes pentagonus Goldfuss; Rominger, p. 131, pi. 48, fig. 2 only.
Arachnophyllum pentagonum (Goldfuss) Lambe, p. 181, pi. 15, figs. 3 and 3 a.
Arachnophyllum pentagonum (Goldfuss); Bolton, pi. 9, fig. 4.

Arachnophyllum mamillare (Owen); Bolton, pi. 4, fig. 7.

Strombodes pentagonus Goldfuss; Davis, pi. 121, figs. 2 and 3.

Arachnophyllum pentagonum (Goldfuss); Amsden, p. 104, pi. 26, figs. 1-6.
Arachnophyllum pentagonum (Goldfuss); Stumm, p. 30, pi. 20, figs. I 3; pi. 21, fig. I

26

PALAEONTOLOGY, VOLUME 32

Diagnosis. Tabular colonies of pentagonal corallites, mean area 1 -2—4- 1 cm2, separated by more
or less sharply defined straight or more rarely curved ridges. Well-defined tabularial pit with axial
boss. Major septa, fourteen to nineteen, well developed in tabularia, mean diameter 2-4-3-4 mm.
Minor septa, variably developed, may penetrate tabularium or are limited to dissepimentarium
where both orders are reduced to low crests on dissepimental surfaces and develop typical
murchisoni- type carination. Tabularium with flat to slightly dished tabellae surrounding a narrow
axial zone of arched tabellae or merging directly with axial septal traces. Dissepiments typically
coarse. Increase peripheral, non-parricidal.

Type material. Neotype (here selected): GSC 90730, Silurian, Llandovery, Telychian, Fossil Hill Formation;
Fossil Hill, Manitoulin Island, Ontario.

Other material. GSC 20550, Fossil Hill Formation; Ridge in Long Bay area, south-west bay of Lake Manitou,
Manitoulin Island (Bolton 1966, pi. 9, fig. 4). GSC 30534, Silurian, upper Llandovery (?lower Wenlock),
Thornloe Formation; roadside exposure, lot 12, concessions V VI, Dymond Township, Ontario (Bolton and
Copeland 1972, pi. 12, fig. 1 1). BMNH R36122, R36124, Silurian, Niagara Group; Drummond Island, Lake
Huron, Michigan, USA. R32154, Silurian, Lockport Formation (Fossil Hill Formation); Fossil Hill,
Manitoulin Island. SM A6298, Silurian, Niagaran Formation (Fossil Hill Formation); Manitouwakning,
Manitoulin Island. BMNH R20864 Silurian, Niagaran (Fossil Hill Formation); Manitoulin Island. GSC
66806, Silurian, Upper Llandovery, Telychian, La Vieille Formation (lower member); west end of railroad
cut, east of highway 132(6) crossing east of Gascons, Gaspe, Quebec (Bolton 1981, pi. 4, fig. 7). GSC 90725,
91555, La Vieille Formation (lower member); railroad cut above Anse-a-la-Vieille, east of Port Daniel,
Quebec. GSC 91556, La Vieille Formation; Point La Roche, Charlo area, east New Brunswick.

Range. Silurian, Llandovery, Telychian (?Lower Wenlock); Michigan and Ontario and Quebec.

Description. Thin to thick discoidal to tabular colonies <150x140x30 mm (present sample). Calicular
surface flat to gently arched. Corallites regularly to irregularly pentagonal, separated by well defined, generally
sharply crested, straight or less commonly curved walls (text-fig. 9a). Tabularial pit, with or without a weak
peritabularial arch, about 1 mm deep with a more or less strong axial boss.

In cross-section, tabularia clearly but sometimes crudely defined by arcuate sections of steeply inclined
dissepiments. Major septa, uniformly thin, generally well but sometimes irregularly developed, usually reach
or almost reach the axis where they may fuse, are sometimes slightly thickened and rarely coated with minor
amounts of stereome. Rarely a weak vortex is developed, or bilateral symmetry may be evident. Minor septa
may penetrate tabularia up to half the radius, but are usually poorly developed and limited to the
dissepimentarium. Septal development may be so irregular that major and minor septa are not clearly
distinguished. In the dissepimentarium, further septa are intercalated to maintain a regular spacing of about
0-25 mm. Here, all septa are restricted to low crests on discrete dissepimental surfaces and appear in arcuate
zones and patches where these surfaces are intersected by the plane of section. Septal traces may be sparse
(text-fig. 3c). Septa are thickened almost to contiguity at dissepimental surfaces, becoming successively
strongly carinate with typical murchisoni- type reticulate structure, more weakly carinate and non-reticulate,
and non-carinate at higher levels in the septal crest (PI. 5, figs. 1 and 3). There are four to five carinae per
mm. Dissepimentarium with sparse, irregular dissepimental traces, usually about five to seven sections in
10 mm. Chains of smaller, elliptical dissepimental sections mark the position of the intercorallite ridge.

In longitudinal section, tabularia well defined by steeply inclined, usually smaller dissepiments. Tabellae
flat to dished peripherally, rising to a slight peak in the axis where they are disrupted by septal traces, or
merging with a narrow axial zone of arched to flat-topped tabellae about a third the tabularium in width.
There are fifteen to twenty-five tabellae in 5 mm. Dissepiments coarse, generally rather irregular, usually flat
and elongated, poorly to moderately inflated, <11x2 mm, rarely more highly inflated <5x3-5 mm. They
are organized into crude zones by size 4-8 mm thick, with septal crests developed on surfaces of smaller
dissepiments (PI. 5, figs. 2 and 4). Sections of carinae < 1-5 mm high, normal to the dissepimental surface or
between normal and vertical when the surface is highly inclined. Dissepimental surface generally flat,
funnelling into tabularium and strongly arched to a vertical zone of slightly smaller dissepiments to define
the intercorallite ridge.

Mean tabularium diameters 2-42-3-44 mm with 14-6 19 major septa. Mean corallite areas 1-24-4-14 cm2.

Increase peripheral, non-parricidal.

SCRUTTON: SILURIAN ARACHNOPHYLLID CORALS

27

text-fig. 9. a, Arachnophyllum pentagonum (neotype); Silurian, Telychian, Fossil Hill Formation; Fossil Hill,
Manitoulin Island, Ontario (GSC 90730). b, A. striatum (neotype); same horizon and locality (GSC 90731).
c, A. pygmaeum (syntype); Silurian, ‘Niagaran’ (?Telychian); Point Detour, Michigan (UMMP (Rominger
Collection) 5310). A-c, all x 1. d-f, A. pygmaeum (?syntype of Astraeophyllum gracile); Silurian, ‘Niagaran’
(?Telychian); Owen Sound, Ontario (UAGC 518). d, cross-section, x2-5; e, detail of cross-section showing
murchisoni- type carination, x 5; F, longitudinal section, x 2-5.

Discussion. The original of Goldfuss (1826, pi. 21, fig. 3a, b ), from Drummond Island in Lake
Huron, appears to be lost (Professor Dr Winifred Haas, pers. comm.). There can be little doubt
as to the identity of the species both from Goldfuss’ illustrations and the relatively consistent
contemporary and subsequent interpretation of the species from the type area of Michigan and
Ontario. I select a neotype from Manitoulin Island, adjacent to Drummond Island, from where in
situ material is available. A. pentagonum is abundant at the neotype locality, colonies reaching
500 mm across and 100-200 mm thick (Dr Paul Copper, pers. comm.).

In most respects, A. pentagonum is structurally similar to A. murchisoni , both sharing in particular
the distinctive reticulate septal structure in the dissepimentarium. A. pentagonum is distinguished
through its consistently and considerably coarser dissepiments, a feature noted by Rominger (1876,
p. 132) in contrast to Louisville Limestone material, but apparently subsequently overlooked.
North American material of A. murchisoni has usually been referred to A. pentagonum. In addition

28

PALAEONTOLOGY, VOLUME 32

in this species, axial arched to flat-topped tabellae tend to be narrower or absent in comparison
to the generally broadly arched tabellae of A. murchisoni. Neither does A. pentagonum appear to
show the wide variety of surface morphology found in A. murchisoni.

A. striatum and A. pygmaeum both have similar internal structure to A. pentagonum, including
the characteristically coarse dissepiments. These two species are distinguished principally on size,
A. striatum having much larger corallites and A. pygmaeum having smaller coralites than A.
pentagonum (text-fig. 10).

Arachnophyllum striatum (d’Orbigny, 1850)

Plate 5, figs. 7-8; text-figs. 9b and 10

1850

1851

e.p.

1876

1966

non

1887

non

1965

Favastraea striata d’Orbigny, p. 48.

Strombodes striatus (d'Orbigny) Edwards and Haime, p. 430.

Strombodes pentagonus Goldfuss; Rominger, p. 131, pi. 48, fig. 1 only.
Arachnophyllum striatum (d’Orbigny); Bolton, pi. 9, fig. 1.

Strombodes striatus (d’Orbigny); Davis, pi. 121, fig. 1; pi. 122, figs. 1 and 2.
Arachnophyllum striatum (d’Orbigny) Stumm, p. 30, pi. 20, figs. 4-6.

Diagnosis. Tabular colonies with large, regularly pentagonal corallites, about 8 cm2 mean area,
separated by sharply defined ridges. Major septa, twenty-two in tabularium diameter 5 0 mm,
reach axis in well-defined tabularia. Minor septa just penetrate tabularium. In dissepimentarium,
septa developed only on discrete dissepimental surfaces as low crests with typical murchisoni- type
carination. Tabellae gently arched in axial third of the tabularium. Dissepiments coarse.

Type material. Neotype (here selected): GSC 90731, Silurian, Llandovery, Telychian, Fossil Hill Formation;
Fossil Hill, Manitoulin Island, Ontario.

Other material. GSC 20547, Fossil Hill Formation; 0-48 km south of Manitowaning-South Baymouth Road,
lot 9, concession I, Assiginack township, Manitoulin Island (Bolton 1966, pi. 9, fig. 1). BM R32153, Silurian,
Lockport Formation (Fossil Hill Formation); Fossil Hill, Manitoulin Island.

Range. Silurian, Landovery, Telychian; Michigan and Ontario.

Description. Colonies tabular, <100x80x40 mm. Corallites separated by straight, sharply crested walls
<4 mm high, curving down into saucer-shaped dissepimental surface, with a central tabularial pit <3 mm
deep containing a prominent axial boss (text-fig. 9b).

Tabularia well defined by smaller, steeply inclined dissepiments, usually with a thickened surface. Major
septa well developed, tapering to uniformly attenuate, most just reaching and a few fusing in the axis (PI. 5,
fig. 7). Minor septa just penetrate tabularium. In the dissepimentarium, septa present only as low crests on
dissepimental surfaces and develop typical murchisoni- type reticulate structure, with carinae four to five per
mm and septa about 0-25 mm apart. Sections of dissepiments wide spaced and irregular, about six in 10 mm.
Arcuate traces of septal crests symmetrically distributed about the positions of intercorallite walls in peripheral
parts of corallites.

In longitudinal section, tabularia fairly well defined (PI. 5, fig. 8). Peripheral flat to dished tabellae interlock
with gently arched plates in the axial third of the tabularium, where they are disrupted by septal traces.
Tabellae spaced fifteen in 5 mm. Peritabularial dissepiments relatively small and steeply inclined adaxially,
becoming coarse, moderately inflated and shallowly inclined adaxially away from tabularium. Intercorallite
wall intermittently strongly defined by short columns of very small dissepiments. Dissepiments <10x3 mm.
Septal crests 3-4 mm apart on surfaces of slightly smaller dissepiments with characteristic sections of carinae
up to 1 mm high.

Tabularium diameter < 5 mm with twenty-two major septa and mean corallite area of 8 cm2.

Discussion. D’Orbigny’s specimen cannot be found (Dr Pierre Semenoff-Tian-Chansky, pers.
comm.). According to Edwards and Haime (1851, p. 430), d’Orbigny gave the location of his
specimen as the Falls of the Ohio in error; it should have been Drummond Island, Lake Huron.
Although neither d’Orbigny nor Edwards and Haime figured this species, there seems to be little

SCRUTTON: SILURIAN A R ACH NOPH Y LLI D CORALS

29

field of A. murchisoni (total sample)

• A. pentagonum A neotype o A. pygmaeum

t A. striatum A neotype

text-fig. 10. Variation in a, mean number of major septa (n) and b, mean corallite
area (Ac in cm2), both plotted against mean tabularium diameter (dt in mm) for
Arachnophyllum pentagonum, A. striatum, and A. pygmaeum. For discussion see text.

30

PALAEONTOLOGY, VOLUME 32

doubt from their descriptions of its identity as a large corallite equivalent of A. pentagonum (text-
fig. 10). Rominger (1876, p. 131) considered the two species to grade into one another, but Dr
Paul Copper (pers. comm.), who kindly provided the neotype specimen, states that in the Fossil
Hill Formation the range of size variation is discontinuous between A. pentagonum and A. striatum.
I therefore retain the species here on that basis.

This species name has occasionally been applied to North American material of A. murchisoni.

Arachnophyllum pygmaeum (Rominger, 1876)

Plate 5, figs. 6 and 7; text-figs. 9c-f and 10

1874 Astraeophyllum gracile Nicholson and Hinde, p. 138, fig. 4a, b.

1876 Strombodes pygmaeus Rominger, p. 132, pi. 48, fig. 3.
non 1887 Strombodes pygmaeus Rominger; Davis, pi. 123, fig. 1

Diagnosis. Laminar to tabular colonies of small, equidimensional polygonal corallites, mean area
0-5-09 cm2, separated by straight, low rounded walls. Major septa extend to axis in tabularia,
twelve to seventeen in mean diameter 1 -8-2-7 mm, forming a boss in the tabularial pit. Septa in
dissepimentarium very sparsely developed as crests on dissepiments, showing typical murchisoni-
type reticulate structure. Tabularium with or without narrow axial series of flat-topped tabellae.
Dissepiments coarse, very variably inflated.

Type material. Syntypes UMMP 8604 and 5410, Silurian, Niagaran; Drummond Island and Point Detour,
Michigan respectively.

Syntypes? of A. gracile: UAGC 518, 12101 12103 (Of these four specimens, one, UAGC 12103, is probably
a specimen of Prodarwinia speciosa), Niagaran; Owen Sound, Ontario.

Other material. GSC 2639, Middle Silurian, Lockport; Derby, lot 13, concession 2, near Owen Sound,
Ontario. BMNH R36121, R36123, Silurian, Niagaran; Drummond Island, Lake Huron.

Range. Silurian, upper Llandovery, Telychian (?Wenlock); Michigan and Ontario.

Description. Thin laminar to thick tabular colonies <100x80x60 mm composed of small, equidimen-
sional polygonal corallites (text-fig. 9c). Calices saucer-shaped with or without slight peritabularial ridge and
tabularial pit <2 mm deep containing prominent axial boss. Calices separated by low ridges, highest at wall
junctions.

In cross-section, tabularia subcircular, outlined by arcuate sections of small steeply inclined dissepiments.
Major septa, slender, extend to axis and fuse to form a boss of variable size and definition. Minor septa just
penetrate tabularium. Dissepimentarium consisting of dissepiment sections of very variable size and shape,
but typically four to five sections in 10 mm. Traces of septa rare as very low unthickened crests on discrete
dissepimental surfaces. Typical murchisoni- type reticulate structure developed but seldom observed in cross-
section (PI. 5, fig. 5; text-fig. 9e). No distinct arrangement of dissepiments in area of corallite wall.

In longitudinal section, tabularia well defined (PI. 5, fig. 6). Peripheral flat to slightly dished tabellae curve
upwards to meet axial septal traces, or to narrow axial series, about 0-25 tabularium wide, of flat-topped
tabellae where septal traces are poorly developed. There are twelve to twenty tabellae in 5 mm. Dissepiments

EXPLANATION OF PLATE 5

Figs. 1-4. Arachnophyllum pentagonum. 1, cross-section and 2, longitudinal section of neotype (GSC 90730);
Silurian, Telychian, Fossil Hill Formation; Fossil Hill, Manitoulin Island, Ontario. 3, cross-section and
4, longitudinal section (BMNH R20864); Silurian, Niagaran; Manitoulin Island, Ontario.

Figs. 5 and 6. A. pygmaeum (BMNH R36121). 5, cross-section and 6, longitudinal section; Silurian, Niagaran;
Drummond Island, Lake Huron, Michigan. Figs. 7 and 8. Neotype of Arachnophyllum striatum (GSC
90731). 7, cross-section and 8, longitudinal section; Silurian, Telychian, Fossil Hill Formation; Fossil Hill,
Manitoulin Island, Ontario.

All figs, x 2-5.

PLATE 5

SC RUTTON, A rachnophyllum

32

PALAEONTOLOGY, VOLUME 32

very variable, poorly to strongly inflated, coarse, <6x2 mm. Smaller dissepiments occur in ill-defined zones.
Septal crests thin and sparse with sections of carinae <0-75 mm high.

Tabularium diameters 1 -8-2-7 mm with 12-3 17 septa and mean corallite areas 0-49-0-89 cm2.

Discussion. Benton (1979, p. 31) has recognized UAGC 518, 12101-12103 as Nicholson material,
and possibly therefore the syntype set of Astraeophyllum gracile, among other Nicholson type
specimens in the University of Aberdeen. Unfortunately, the four badly preserved fragments cannot
be recognized from Nicholson and Hinde’s (1874, p. 138, fig. 4 a, b) very stylized figures although
they could have been the basis for them. The material is so poor that sections of the only specimen
suitable for preparation, UAGC 518, reveal little more than a trace of typical murchisoni- type
septal reticulation, proving beyond doubt that the specimen belongs to Arachnophyllum , and the
general dimensions and sparse septal development of A. pygmaeum (text-fig. 9d-f). UAGC 12103,
of doubtful identity, is unsuitable as a basis for firmly establishing A. gracile as a junior synonym
of Prodarwinia speciosa. Thus, it would simply be better to set gracile aside in favour of the better
established species.

Unfortunately, the original of Rominger’s (1876, pi. 48, fig. 3) figured specimen, UMMP 8604,
the most suitable lectotype for A. pygmaeum , is mislaid but probably not lost (Dr J. A. Kitchell,
pers. comm.). 1 therefore refrain from making a type selection for this species at the moment. The
identity of the species appears to be quite clear from Rominger’s illustration and account.

A. pygmaeum is most closely related to A. pentagonum , from which it differs principally is its
small size (text-fig. 10), particularly its small corallite area, its relatively weak intercorallite walls,
and the very weak development of septa in the dissepimentarium.

Unrecognizable species of ? Arachnophyllum

Lamellopora infundibularia Owen, 1844

1844 Lamellopora infundibularia Owen, p. 70, pi. 14, fig. 1.

1851 Strombodes infundibularius (Owen) Edwards and Haime, p. 432.

Type material. Owen’s figured specimen is lost. It was recorded from the ‘coralline beds of the Upper
Magnesian Cliff Limestone of Iowa and Wisconsin’, likely to represent either the Hopkinton or Kankakee
Dolomite according to Laub (1979, p. 194).

Discussion. Edwards and Haime (1851) thought Owen’s species to be either a small S. pentagonum
or perhaps S. mamillaris. From Owen’s figure, which is a side view of the colony, little can be
deduced except that it could be an amural arachnophyllid or even a stromatoporoid (Lang et al.
1940, p. 74; Hill 1981, p. 669). The species is unrecognizable.

Genus prodarwinia Cotton, 1973

1873 Darwinia Dybowski, pp. 336, 404.

1940 Darwinia; Lang, Smith and Thomas, p. 49.

1973 Prodarwinia Cotton, p. 161.

Diagnosis. Astraeoid to thamnasterioid coralla lacking intercorallite ridges. Tabularia usually set
on more or less prominent mamelons. Septa monacanthine thin continuous plates in tabularium, but
multitrabecular, becoming pseudorhabdacanthine in the dissepimentarium, weakly and irregularly
carinate or non-carinate. Solid septal laminae of variable thickness may develop on successive
discrete dissepimental surfaces by trabecular thickening and/or embedding in lamellar sclerenchyme.
Tabellae flat to dished peripherally, merging axially into septal sections or with axial series of flat
topped tabellae. Dissepiments variable in size and weakly inflated.

Type species. Darwinia speciosa Dybowski, 1873.

Discussion. The generic name Darwinia Dybowski, 1873 is preoccupied by Darwinia Bate, 1857 for
a crustacean (Lang et al. 1940, p. 49). The genus was therefore renamed Prodarwinia by Cotton
(1973, p. 161).

SCRUTTON: SILURIAN ARACHNOPHYLL1D CORALS

33

Prodarwinia is almost identical in structure to the Devonian genus Iowaphyllum (see Oliver and
Galle 1971; Oliver 1978). The only distinctions are of uncertain significance, such as the lack of
corallite walls on the external calicular surface and the pseudorhabdacanthine trabeculae of
Prodarwinia. Species of Iowaphyllum usually have external corallite walls which are sometimes
reflected in the internal structure by other than the form of the dissepimentarium surface alone.
The septa in Iowaphyllum may be unitrabecular coarse simple monacanths or multitrabecular.
Septal structure in Prodarwinia is distinctive (text-fig. 7c, d). Septa are multitrabecular and
in the dissepimentarium when solid septal crests are developed, trabeculae appear to be thin,
<c. 01 mm diameter, and loosely branching (pseudorhabdacanthine), embedded in lamellar
sclerenchyme. The possibility that this appearance is a preservational artefact has been considered,
but the standard of preservation, its appearance in material from several different localities, and
its relationship with the surface topography of the septal crests suggest that it is an original
structure. Although species of Iowaphyllum appear to have little additional sclerenchyme in their
septa, traces of a similar microstructure to that in Prodarwinia are illustrated in I. nisbeti from the
Upper Devonian (Frasnian) of Arizona by Oliver (1978, fig. 3 b, d,f).

The principal reason for maintaining two distinct genera is the long stratigraphical gap between
the Silurian and Devonian species groups. No species referable to either genus are known from
the Ludlow (or at least later Ludlow), Pridoli, or Lochkov. Coral faunas are sufficiently well
known from these levels to suggest that this gap is real and not a preservational or descriptive
artefact. I conclude that the two genera represent an extreme case of homoeomorphy.

Range. Llandovery (Aeronian)-late Wenlock (?early Ludlow); Estonia, Gotland, USA, southern Canada.

Prodarwinia speciosa (Dybowski, 1873)

Plate 6, fig. 1; Plate 7, figs. 1 -7; text-figs. 7c, D, 11, 12

1858 Strombodes diffluens Edwards and Haime; Schmidt, p. 232.

1873 Darwinia speciosa Dybowski, p. 404, pi. 2, fig. 8 and 8 a.

? 1901 Arachnophyllum diffluens (Edwards and Haime); Lambe, p. 183, pi. 14, fig. 12.
e.p. 1927 Arachnophyllum diffluens (Edwards and Haime); Lang and Smith, p. 468 (syn. only).

1973 Prodarwinia speciosa (Dybowski) Cotton, p. 161.

Diagnosis. Flat, tabular astraeoid to thamnasterioid colonies. Tabularial pits set in broad, low,
rounded mamelons; no intercorallite walls. Tabularia fairly well defined. Mean tabularium diameter
1 -82-3-23 mm with fifteen to twenty-one major septa and mean corallite area 0-63-2-41 cm2. Major
septa usually extend to axis and fuse to form columella. In dissepimentarium, septa may extend a
short distance from tabularium as discontinuous or continuous attenuate blades. More frequently,
septa are multitrabecular and contiguous, forming arcuate belts and patches of solid tissue with
unthickened, coarse, irregular dissepimental sections in between. Tabularia clearly defined by
steeply inclined dissepimental sections in longitudinal section. Tabellae periaxial, dished, merging
with septal sections in axis. Dissepiments very variable, poorly to strongly inflated, crudely zoned.
Solid septal crests usually developed on specific dissepimental surfaces, but may be reduced to
discrete trabeculae in some specimens.

Type material. Lectotype (chosen by Lang et al. 1940, p. 49): the original of Dybowski's (1873) pi. 2, fig. 8
of Darwinia speciosa var. major , which is 1GT Col 334; Silurian, Adavere Stage, Rumba Formation (late
Aeronian), Pari (ex Kattentack), Estonia.

Other material. IGT Col348, BMNH R53510 53512, horizon and locality as for lectotype. Col349, Silurian,
Adavere Stage, Rumba Formation; Tammikaare, Estonia. Co 1350, Rumba Formation; Vaike Rohda,
Estonia. GSC 2639 a, Lockport; Owen Sound, Ontario (?figured Lambe 1901. pi. 14, fig, 12). GSC 90732,
Fossil Hill Formation; Fossil Hill, Manitoulin Island, Ontario. 7UAGC 12103, Niagaran; Owen Sound,
Ontario. UCBL EM 15171 (de Verneuil Collection), (?Silurian, Hopkinton Formation); Wisconsin.

Range. Silurian, Llandovery, late Aeronian Telychian; Estonia, USA, Canada.

34

PALAEONTOLOGY, VOLUME 32

Description. Thin, tabular astraeoid thamnasterioid colonies <c. 240 mm diameter and 35 mm thick.
Calicular surface undulose to flat. Tabularial pits with strong axial boss set in broad, low, rounded mamelons
(PI. 6, fig. I; text-fig. 1 1a,d). Septa more or less sharp granular ridges, more often continuous than abutting
between corallites.

In cross-section, tabularia fairly well defined by thickened sections of one to three concentric series of
steeply inclined dissepiments, or set in solid septal material (PI. 7, figs. 1, 2, 4, 6; text-fig. 11b). Major and
minor septa clearly distinguished. Majors usually extend to axis, attenuate to slightly thickened <01 mm,
where they fuse together, or in groups, sometimes forming a stellate axial structure with periaxial tabellae
and occasional minor sclerenchyme. Minor septa just penetrate tabularium, but individual septa rarely may
be longer. Outside tabularium, septa may extend a short distance as unitrabecular attenuate blades, either
continuous or as alligned series of septal crests, or are absent. In specimens with well-developed solid septal
crests tabularia more usually set in annular, arcuate, or irregular patches of solid material in which individual
multitrabecular septa are difficult to distinguish. Margins of solid areas dentate where septa rapidly taper
and disappear. Elsewhere, unthickened dissepiment sections are coarse and irregular, four to eight in 10 mm.

In longitudinal section, tabularia well defined by steeply inclined inner dissepiment faces. Tabellae
peripherally dished, close-spaced twelve to twenty-two in 5 mm, rising to axial septal sections. Where major
septa do not reach the axis, there may be a narrow axial series of flat tabellae. Dissepiments very variable
in form, poorly to strongly inflated, <8xL5 mm, crudely zoned. Solid septal crests of varying thickness
<4 mm thick, rarely 7 mm where several coalesce laterally, usually developed on discrete dissepimental
surfaces (PI. 7, figs. 5 and 7), but may be reduced to discrete trabeculae in some specimens (PI. 7, fig. 3).
There may be four to seven crests in 5 mm vertical growth. In some specimens, layers of thin septal crests
may be capped or sandwiched between zones of more or less solid septal material <15 mm thick (text-fig.
11c). In dissepimentarium, crests with dentate upper surfaces formed of thin, <c. 0T mm diameter, crudely
branching trabeculae (pseudorhabdacanths) set in lamellar sclerenchyme (text-fig. 7d), but becoming solidly
trabecular at tabularium.

Mean tabularium diameters vary 1 -82-3-23 mm with 15-20-5 major septa and mean corallite area 0-63-
2-4 1 cm2 (text-fig. 12).

Discussion. Dybowski (1873, pp. 404-406) distinguished major and minor varieties of D. speciosa ,
of which the original of var. minor is now missing. He recorded both varieties from Piiri and the
small collection of topotypic material described here suggests continuous variation in the sample,
although it may not include the extreme limit represented by var. minor.

P. speciosa is very close to P. striata. The latter has smaller tabularium diameters and septal
number but more overlap in corallite area (text-fig. 12). In addition, the weak mamelons on the
calicular surface of P. striata are confined to the immediate peritabularial area in contrast to the
equally weak but broader swellings bearing the tabularia in P. speciosa (compare PI. 6, fig. 1 and
text-fig. 11a, D with PI. 6, fig. 2).

EXPLANATION OF PLATE 6

Fig. 1. Prodarwinia speciosa, calicular surface (BMNH R53512); Silurian, Aeronian, Rumba Formation; Piiri,
Estonia.

Fig. 2. P. striata , calicular surface (RM Cn20154); Silurian, Telychian, Rode Mergel; Norderstrand, Visby,
Gotland.

Figs. 3 and 4. P. distans, calicular surfaces. 3, lectotype (USNM 113639); Silurian, Llandovery, Noland
Formation. Waco Member; Irvine, Kentucky; 4, BMNH R50215; Silurian, Telychian, Hopkinton
Formation; Manchester, Iowa.

Figs. 5 and 6. P. mamillaris , calicular surfaces. 5, GSC 20534; Silurian, Telychian, Fossil Hill Formation;
Fossil Hill, Manitoulin Island, Ontario; 6, USNM 422823; Silurian, Louisville Limestone; Louisville,
Kentucky.

Figs. 7 and 8. P. striata (USNM 422827). 7, cross-section and 8, longitudinal section; Silurian, Telychian,
Dayton Formation; Todd’s Fork, 3-2 km north of Wilmington, Ohio.

Figs. 9 and 10. P. distans (USNM 422825). 9, cross-section and 10, longitudinal section; Silurian, Llandovery,
Noland Formation, Waco Member; Indian Fields, Kentucky.

Figs. 1-6, x 1. Figs. 7-10, x2-5.

PLATE 6

SCRUTTON, Prodarwinia

36

PALAEONTOLOGY, VOLUME 32

text-fig. 1 1 . Prodarwinia speciosa , North American material, a, calicular surface; Silurian (?Hopkinton
Formation); Wisconsin (UCBL (de Verneuil Collection) EM 15171), xl.B, cross-section and c, longitudinal
section; Silurian, Telychian, Fossil Hill Formation; Fossil Hill, Manitoulin Island, Ontario (GSC 90732),
both x 2-5. d, calicular surface; Silurian, Lockport; Owen Sound, Ontario (GSC 2639 a), x 1.

Specimens of P. striata consistently develop solid septal crests, but some specimens of P. speciosa
do not. These approach P. distans and P. mamillaris in internal structure but lack the strong calical
surface features of these species.

P. speciosa has most frequently been referred to as Arachnophyllum diffluens in the literature.
There is some similarity between the calicular surfaces of the diffluens- morph of A. murchisoni and
P. speciosa but considerable contrasts in their internal structures, particularly with the lack of
murchisoni-type carination in the latter (contrast PI. 2, figs. 2 and 3 of the lectotype of Strombodes
diffluens with PI. 7, figs. 1-7).

Prodarwinia striata (James, 1878)

Plate 6, figs. 2, 7, 8; Plate 7, figs. 8-12; text-fig. 12
1878 Lyellia striata James, p. 10.

? 1888 Strombodes pygmaeus Rominger; Foerste, p. 120, pi. 13, fig. 18.
1906 Arachnophyllum mamillare-wilmingtonensis Foerste, p. 320.
e.p. 1979 Arachnophyllum mamillare (Owen); Laub, p. 194, pi. 7, fig. 3 only.

SCRUTTON: SILURIAN ARACHNOPH YLLID CORALS

37

Diagnosis. Tabular astreoid to thamnasterioid colonies with tabularial pits surrounded by low
peritabularial swellings; no intercorallite walls. Tabularia strongly defined with major septa extend-
ing to axis and fusing to form prominent columella. Mean tabularium diameter 1-52-1 -93 mm
with 11-16-5 major septa; mean corallite area 0-78-1-52 cm2. In dissepimentarium, major and
minor septa present as contiguous, multitrabecular laminae on dissepimental crests. Tabularium
with peripheral series of dished tabellae merging with septal traces in axis. Dissepiments variable
in size, weakly inflated.

Type material. Neotype (here selected): USNM 422826, Silurian, late Telychian, Dayton Limestone; Todds
Fork, 3-2 km north of Wilmington, Ohio.

USNM 422826 is labelled ‘ Lyellia striata, James’ and was collected by James and donated to the National
Museum of Natural History in 1883. It agrees closely with James’ description of his species, based on one
of two specimens, but cannot certainly be identified as that syntype; it does not agree in size with the other
syntype. In this case, stability seems best served by designating the available specimen as neotype for the
species.

The original of Foerste’s Strombodes pygmaeus (Foerste 1888, pi. 13, fig. 18) from the Dayton Limestone
of Ohio, for which he later erected the subspecies wilmingtonensis , has not been located. However, Foerste
(1906, p. 320) also stated that A. mamillare-wilmingtonensis ‘. . . was described as Lyellia striata, by U. P.
James . . .’ and could therefore be considered technically as an objective synonym of James’ species. USNM
422826 has a second label bearing Foerste’s subspecies name and the specimen could have been seen by
Foerste at that time.

Material. USNM 422827, Dayton Limestone; Todd’s Fork, 3-2 km north of Wilmington, Ohio. RM
Cn20114, Cn 20135-6, Cn20154, Cn67085, Silurian, Upper Llandovery, Rode Mergel (Arachnophyllum Bed);
Norderstrand, Visby, Gotland.

Range. Silurian, Llandovery (Telychian); Ohio and Gotland.

Description. Tabular colonies <150x80x40 mm. Calicular surface flat but for peritabularial ridge about
1 mm high surrounding tabularial pit 1-2 mm deep crossed by thin radiating septal plates to prominent axial
boss (PI. 6, fig. 2). Outside tabularium, septa low granular ribs tending to parallel perferred orientations,
predominantly confluent with those of adjacent corallites.

In cross-section, tabularia clearly defined with thin to thickened, 0-025 0- 1 5 mm thick, major septa,
sometimes slightly tapering, extending to the axis where they fuse, sometimes coated with a little sclerenchyme,
to form a prominent columella. Minor septa may or may not just penetrate tabularium. Tabularium
surrounded by thick to thin wall of steeply inclined dissepiments bearing septal crests, or set in an area of
solid trabecular material in which individual septa are not usually clearly distinguished (PI. 6, fig. 7; PI. 7,
figs. 8, 9, 11). In places, small trabeculae 0-05-0- 1 mm diameter are set in sclerenchyme; elsewhere thickened
trabeculae may be contiguous. Laterally in the dissepimentarium, individual septa consists of three to four
trabeculae abreast and taper more or less abruptly to a point or short, unitrabecular blade. Between septal
material, dissepiment sections coarse and irregular, five to ten in 10 mm.

In longitudinal section, tabularia sharply defined with periaxial, close-spaced, trough-shaped tabellae
merging into an irregular axial septal columella (PI. 6, fig. 8; PI. 7, figs. 10 and 12). Tabellae eighteen to thirty
in 5 mm. Dissepimentarium, arched adjacent to tabularium but flat elsewhere, consists of thick solid septal
crests based on successive series of wide, poorly inflated dissepiments, <8 mm x 2 mm. Septal crests up to
5 mm thick but usually 0-5-1 mm thick. Three to eight discrete crests separated by dissepiments may be
superposed in 5 mm and crests may coalesce laterally to form thicker solid laminae. Trabeculae about
0 13 mm wide in plane of septa, either contiguous or set in lamellar sclerenchyme when they may appear
pseudorhabdacanthine in form.

Mean tabularium diameter range from 1-52-1-93 mm with 1 1 -46-16-5 major septa. Mean corallite areas
0-78-1-52 cm2.

Increase intercalicinal.

Discussion. The neotype from Ohio is almost indistinguishable from material from the Arachno-
phyllum Bed at the very base of the succession on Gotland.

P. striata is closest to P. speciosa with which it is likely to have a direct evolutionary relationship
as discussed earlier. The two species are compared under the discussion of the latter.

38

PALAEONTOLOGY, VOLUME 32

Prodarwinia mamillaris (Owen, 1844)

Plate 6, figs. 5 and 6; Plate 8, figs. 1-5; text-fig. 12

1844 Astrea mamillaris Owen, p. 70, pi. 14, fig. 3.

1876 Strombodes mamillatus (Owen) Rominger, p. 133, pi. 48, fig. 4.

1887 Strombodes mamillaris (Owen); Davis, pi. 123, fig. 4.

1965 Arachnophyllum mammillare (Owen); Stumm, p. 31, pi. 20, figs. 7-9.

1966 Arachnophyllum mammillare (Owen); Bolton, pi. 6, fig. 15.

e.p. 1979 Arachnophyllum mamillare (Owen); Laub, p. 194, non pi. 7, figs. 2 and 3; pi. 25, figs. 1 and 3.

Inon 1901 Arachnophyllum mamillare (Owen) Lambe, p. 182, pi. 15, fig. 4.
non 1981c Arachnophyllum mammillare (Owen); Bolton, pi. 4, fig. 7.

Diagnosis. Tabular colonies of close-spaced corallites, mean area 1 -3-2-4 cm2. Deep tabularial pit
with axial boss in centre of prominent, broadly rounded mamelons. No intercorallite walls. Major
and minor septa usually not distinguished, twenty-four to thirty-six total in mean tabularium
diameter 2-6-34 mm. All septa thin, extending more or less to axis of tabularium. Dissepiments
very variable, coarse, with weakly developed septal crests. Crests may be solid, multitrabecular on
dissepimental surface, irregularly carinate or non-carinate slender blades above. Tabellae variable,
with or without broad axial series of flat-topped domes.

Type material. Owen’s material appears to be lost. It was stated as coming from the ‘coralline beds of the
Upper Magnesian Cliff Limestone of Iowa and Wisconsin’. Laub (1979, p. 194) considered this to represent
either the Hopkinton or Kankakee Dolomite. As no material of this species is available from these horizons,
I refrain from selecting a neotype (but see discussion).

Other material. GSC 20534, Silurian, Landovery, Telychian, Fossil Hill Formation; Manitoulin Island,
Ontario. BMNH 90001, Silurian, Niagaran; Lake Huron. USNM 422823, UMMP 34296, Silurian, upper
Wenlock (?lower Ludlow),/ Louisville Limestone; Louisville, Kentucky.

Range. Silurian, Llandovery (Telychian) to late Wenlock (?early Ludlow); northern USA and southern
Canada.

Description. Tabular colonies < 1 10 x 80 x 40 mm with close-spaced, prominent, broadly rounded mamelons
on flat to gently undulating calicular surface (PI. 6, figs. 5 and 6). Mamelons <5 mm high, with central
tabularial pit <2 mm deep containing an axial boss. Septa thin blades in tabularium, expressed as broad,
low arched to ridged, radiating granular bands separated by furrows on dissepimental surface. Septa of
adjacent corallites astraeoid to thamnasterioid; no corallite walls.

In cross-section, tabularia clearly but often crudely outlined by arcuate sections of steeply inclined
dissepiments with slightly thickened surfaces (PI. 8, figs. 1, 3, 4). Septa usually not clearly divisible into major
and minor series, all extending as thin, rarely thickened blades to or almost to axis, where some may fuse.
Septa may fail between tabularial surfaces which may be cut in cross-section to form an aulos. In dis-
sepimentarium, dissepiment sections very variable but generally coarse, four to thirteen sections in 10 mm,
irregularly circular to elliptical except adjacent to tabularium. Septa developed as very thin crests on discrete
dissepiment surfaces. Septal traces rare in cross-section, either contiguous and solid (PI. 8, fig. 1), with granular

EXPLANATION OF PLATE 7

Figs. 1-7. Prodarwinia speciosa. 1 and 2, cross-sections and 3, longitudinal section of lectotype (IGT Col 334).
4, cross-section and 5, longitudinal section (BMNH R53511). 6, cross-section (BMNH R53510).

7, longitudinal section (IGT Co 1348). All Silurian, Aeronian, Rumba Formation; Pari, Estonia.

Figs. 8-12. P. striata. 8 and 9, cross-sections and 10, longitudinal section of neotype (USNM 422826);
Silurian, Telychian, Dayton Formation; Todd’s Fork, 3-2 km north of Wilmington, Ohio. 11, cross-
section and 12, longitudinal section (RM Cn20154); Silurian, Telychian, Rode Mergel; Norderstrand,
Visby, Gotland.

All figs. x2-5.

PLATE 7

SCRUTTON, Prodarwinia

40

PALAEONTOLOGY, VOLUME 32

o P. speciosa A lectotype • P. mamillaris

v P. striata A neotype ▼ P. distans A holotype

text-fig. 12. Variation in a, mean number of major septa (n) and b, mean corallite area
(Ac in cm2), both plotted against mean tabularium diameter (dt in mm) for Prodarwinia
speciosa, P. striata, P. mamillaris, and P. distans. For discussion see text.

SCRUTTON: SILURIAN A R ACHNOPH Y LLI D CORALS

41

multitrabecular structure, or irregularly carinate or non-carinate and present in narrow arcuate zones where
crested dissepiment surfaces are cut (PL 8, fig. 3). Reticulate structure of murchisoni type not developed.

In longitudinal section, tabularium with peripheral series of flat to dished, horizontal to abaxially declined
tabellae and usually an axial series, <0-6 tabularium wide, of flat to mesa-shaped tabellae (PI. 8, figs. 2 and
5). There are fifteen to thirty tabellae in 5 mm. Dissepiments very variable in size and inflation, 0-3-7-5 mm
across, height width ratio varying from 1:11-5 to 1:1 but mostly 1:2-3. Surface of dissepimentarium
strongly arched 1-2 mm outside tabularium. Dissepimentarium zoned with coarse dissepiments grading
distally to finer dissepiments; zones about 2-5 mm thick. Very low septal crests, <05 mm high, on surface
of smaller vesicles.

Mean tabularium diameter 2-67-3-4 with 23-8-36 total septa and mean corallite area I 34 2-37 cm2.

Discussion. Despite the fact that Owen’s material is missing, the species seems to be readily
identified from his figure (Owen 1844, pi. 14, fig. 3), and subsequently has been generally consistently
interpreted. There is a possibility that Owen’s specimen could be an example of P. speciosa , which
is known from the type area (text-fig. 1 1a). This cannot be resolved until the type is located and/or
substantial new collections are available from Iowa and Wisconsin but in any case, stability is best
served by retaining the traditional interpretation of the species.

P. mamillaris is a very distinctive species, the main point of dispute being its relationship to P.
distans and P. striata , both of which have been considered subspecies of P. mamillaris but which
are treated as distinct species here. P. mamillaris can be readily distinguished from P. striata
through the prominent mamelons on the calicular surface, the larger tabularium diameter (text-
fig. 12), lack of septal differentiation, and lack of well-developed solid septal crests in the former.
It is compared with P. distans in the discussion under that species.

Prodarwinia distans (Foerste, 1906)

Plate 6, figs. 3, 4, 9, 10; plate 8, figs. 6 and 7; text-fig. 12

1906 Arachnophyllum mamillare-distans Foerste, p. 319, pi. 3.
e.p. 1979 Arachnophyllum mamillare (Owen); Laub, p. 194, pi. 7, fig. 2; pi. 25, figs. 1 and 3 only.

Diagnosis. Large, flat, thin tabular colonies with wide-spaced conical mamelons bearing shallow
tabularial pits. Mean corallite area 2 7-3-9 cm2, with fifteen to nineteen major septa in mean
tabularium diameter 34-4-5 mm. Major septa cross tabularium to axis. Minor septa may reach
<0-5 tabularium diameter but both orders reduced to sparse low crests on dissepiments in
dissepimentarium. Tabularia with peripheral series of close-spaced dished tabellae rising to meet
septal traces in axis. Dissepiments characteristically low, elongate vesicles.

Type material. Holotype: USNM 1 13639, Silurian, Niagaran, Clinton-Waco (Plate Aeronian-early Telychian,
Waco Member); Irvine, Kentucky.

Other material. USNM 422825, Clinton Waco; Indian Fields, Kentucky. USNM 422124, Silurian, Telychian,
Hopkinton Formation; Quarry SW \ sec. 30', T.88N., R.3E., Dubuque South Quadrangle, Iowa. BMNH
R50215, Silurian, Telychian, Hopkinton Formation; Manchester, Iowa.

Range. Silurian, Llandovery, ?Aeronian, Telychian; Kentucky and Iowa.

Description. Large, flat, thin tabular colonies <210 x 140 x 50 mm. Calicular surface flat to gently undulating,
bearing wide-spaced conical mamelons <6 mm high with central, shallow tabularial moat crossed by septal
blades to a prominent axial boss (PI. 6, figs. 3 and 4). Septa continue outside tabularium as very low, broad
ribs with 0-1 mm scale granular ornament, separated by narrow grooves. Septa of adjacent corallites mainly
abutting.

In cross-section, tabularia well defined, with thin, tapering major septa extending more or less to axis,
where some may fuse. Septa may show weak bilateral distribution. Minor septa variably developed, sometimes
reaching <0-5 distance to axis but in other cases only just penetrating tabularium boundary. Outside
tabularium, traces of septa mainly seen as thin discontinuous blades cresting dissepiments in immediate
peritabularial area (PI. 6, fig. 9; PI. 8, fig. 6). Rare crossbar carinae may occur. Septa almost absent elsewhere

42

PALAEONTOLOGY, VOLUME 32

but rarely solid multi-trabecular septal deposits are seen cresting dissepiments. Dissepimentarium with variable
but generally large and irregular to subcircular dissepiment sections, five to eight in 10 mm.

In longitudinal section, tabularia well defined with close-spaced peripheral series of flat to dished tabellae
rising to the axis where they are cut by septal traces (PI. 6, fig. 10; PI. 8, fig. 7). There may be some steeply
inclined tabellae in the axial area. Peripheral tabellae spaced eleven to twenty-five in 5 mm. Septal trabeculae
inclined axially and upwards at about 20° at edge of tabularium, increasing in inclination to about 70° near
axis. Trabeculae about 01 mm diameter, either tufted monacanths or possibly pseudorhabdacanths.
Dissepiments characteristically low and elongate, <8-3 mm wide and 1 : 3-5-12 height: width ratio. Crest of
dissepimentarium 0-7-2-5 mm outside tabularium, funnelling steeply into tabularium but sloping gently
abaxially to generally flat or undulating peripheral calicular surface. Dissepiments may be weakly zoned by
size. Septal crests not evident.

Discussion. Although P. distans was originally described as a subspecies of P. mamillaris, I consider
there to be sufficient difference between them to justify full separation at the species level. Externally,
P. distans is characterized by wide but variably spaced conical mamelons and very flat septal ribs
in contrast to the close-spaced, broadly rounded mamelons and more arched to ridged septal ribs
of P. mamillaris. P. distans consequently has a higher mean corallite area, as well as a larger mean
tabularium diameter, than P. mamillaris (text-fig. 12). Internally, P. distans has clearly distinguished
minor septa, rather more uniform, less inflated dissepiments and lacks an axial series of arched or
mesa-shaped tabellae in contrast to typical P. mamillaris. Furthermore, although both have only
weakly developed septa in the dissepimentarium, they are notably more weakly developed in
P. distans.

P. distans is readily distinguished from both P. speciosa and P. striata by its distinctive calicular
surface morphology and lack of solid septal crests. However, occasional specimens of P. speciosa
also lack well-developed septal crests (see discussion under that species). In addition, P. striata has
much smaller tabularium diameters and mean corallite areas (text-fig. 12). These parameters are
larger in P. speciosa but still slightly smaller than in P. distans.

Prodarwinia sp. nov.

Text-fig. 13a-c

Material. USNM 79840, Silurian, Clinton, Dayton Limestone (late Telychian); Todds Fork, 3-2 km north
of Wilmington, Ohio.

Description. Small ovoid hand specimen, 80 x 60 x 35 mm with domed upper surface bearing two or three
very large tabularial pits 30 40 mm apart on low conical to rounded elevations (text-fig. 13a). Septa gently
arched ribs, astreoid to thamnasterioid.

In cross-section, tabularia well defined by slightly thickened dissepiment sections, 5-6-67 mm diameter
with seventeen to twenty major septa (text-fig. 13b). Major septa thin, tapering in tabularium to axis, or as
short septal crests on tabulae. Minor septa weakly developed, just penetrating tabularium, and both orders
discontinuous or rarely continuous in dissepimentarium within 3-4 mm of tabularium, where they may
appear faintly beaded or develop very weak crossbar carinae. Here also septa may be multitrabecular, three
or more trabeculae across, and thickened to contiguity in a very thin zone on discrete dissepimental surfaces.
Dissepimental sections somewhat irregular and moderately coarse.

EXPLANATION OF PLATE 8

Figs. I 5. Prodarwinia mamillaris. 1, cross-section and 2, longitudinal section (UMMP 34296); 3, cross-
section (USNM 422823); all Silurian, Louisville Limestone; Louisville, Kentucky. 4, cross-section and 5,
longitudinal ^section (BMNH 90001); Silurian, Niagaran; Lake Huron, Michigan.

Figs. 6 and 7. P. distans lectotype (USNM 113639). 6, cross-section and 7, longitudinal section; Silurian,
Llandovery, Noland Formation, Waco Member; Irvine, Kentucky.

All figs. x2-5.

PLATE 8

SCRUTTON, Prodarwinia

44

PALAEONTOLOGY, VOLUME 32

text-fig. 13. a-c, Prodarwinia sp. nov.; Silurian, Telychian, Dayton Formation; Todd’s Fork, 3-2 km north
of Wilmington, Ohio (USNM 79840). a, calicular surface, x 1; b, cross-section, x2-5; c longitudinal section,
x2-5. d, ? Prodarwinia gigas (holotype of Strombodes eximius); Silurian, ‘Lockport’ (?Telychian); Manitoulin

Island, Ontario (GSC 2633), x 1.

In longitudinal section, tabularium somewhat irregularly defined by steep, vertical, inner faces of
peritabularial series of dissepiments, some of which may extend into and across the tabularium as flat to
dished plates (text-fig. 13c). Tabulae complete or incomplete, flat or low, wide flat-topped domes, fifteen
to sixteen in 5 mm. Dissepiments generally scarcely to poorly inflated, except the peritabularial series
when the dissepimentarium surface is arched there. Dissepiments <7xT5 mm, not clearly zoned and
bearing scattered septal spines which tend to be more common on discrete surfaces, with thin dense
septal crests developed adjacent to the tabularium. Crests are <0-25 mm thick and individual spines
<0-75 mm long.

Discussion. This specimen is unique. It is like no well-established species of Prodarwinia and only
resembles IP. gigas (see below) in the size of its corallites. Septal structure is similar to that in P.
mamillaris but tabularium and corallite size, as well as the character of the dissepiments set this
specimen apart. It is inappropriate to name a new species on a single specimen. Further material
is necessary to establish its range of variation and to confirm its status.

SCRUTTON: SILURIAN ARACHNOPHYLLID CORALS

45

Unrecognizable species of 1 Prodarwinia

1 Prodarwinia verneuili (Edwards and Haime, 1851)

1851 Phi/lipsastrea verneuili Edwards and Haime, p. 447, pi. 10, fig. 5.

non 1859 Phillipsastraea verneuili Edwards and Haime; Billings, p. 127, fig. 24.
non 1876 Phillipsastraea verneuilli Edwards and Haime; Rominger, p. 128, pi. 38, fig. 2.
non 1917 Phillipsastraea ( Billingsastraea ) verneuili (Edwards and Haime) Grabau, p. 957.
non Billingsastraea verneuili of authors ( = Asterohil/ingsa magdisa Oliver, 1974).

Type material. Edwards and Haime’s figured specimen is unlocated and possibly missing. It was recorded as
from Wisconsin.

Discussion. Phillipsastraea verneuili Edwards and Haime was designated as type species of
Billingsastraea by Grabau (1917, p. 957) who quoted the species name as Phillipsastraea
( Billingsastraea ) verneuili (M.E. & H.) in a species list. The species was thought to be of
Devonian age and Billingsastraea became widely used for a North American group of astraeoid,
thamnasterioid, and aphroid Devonian corals lacking horseshoe dissepiments and characterized
by undilated septa. Billings’s material, which probably seeded the misconception to judge from
Grabau’s choice of name, together with Rominger’s illustrated specimen from the drift of Michigan,
belong to this group. Oliver (1974, p. 167; 1976, p. 88) finally unravelled the confusion, renaming
this group of Devonian corals as Asterohil/ingsa. On the basis of another specimen in the de
Verneuil Collection very close in character to Edwards and Haime’s illustration, he suggested that
P. verneuili was probably an Arachnophyllum and likely to have come from the Hopkinton Dolomite
of north-east Iowa and south-west Wisconsin. That specimen, UCBL EM 15171, has been re-
examined and is illustrated here (text-fig. 11a); it is assigned to Prodarwinia speciosa.

This species is probably therefore a Prodarwinia , and possibly a senior synonym of P. speciosa ,
although this cannot be confirmed until the illustrated specimen is found.

? Prodarwinia gigas (Owen, 1844)

Text-fig. 13d

1844 Astreal gigas Owen, p. 70, pi. 14, fig. 7.

? 1866 Strombodes eximius Billings, p. 93.

1893 Strombodes gigas (Owen) Calvin, p. Ill, pi. 5, fig. 5.

? 1901 Arachnophyllum eximium (Billings) Lambe, p. 184, pi. 16, figs. 3, 3 a, 4.

non 1859 Phillipsastrea gigas (Owen) Billings, p. 128.

non 1876 Phillipsastraea gigasO (Owen); Rominger, p. 129, pi. 37, fig. 3.

non 1887 Phillipsastrea gigas (Owen); Davis, pi. 118, figs. 1 and 2.

Type material. Owen’s original material, said to be from the 'coralline beds of the Upper Magnesian Cliff
Limestone of Iowa and Wisconsin’ interpreted by Laub (1979, p. 194) as either the Hopkinton or Kankakee
Dolomite, is lost.

Holotype of Strombodes eximius: GSC 2633, Silurian, Lockport (Telychian); Manitoulin Island, Ontario.

Discussion. Owen’s species was widely interpreted subsequently as a Devonian coral. Calvin (1893,
p. 109) considered that ‘The specimen collected and studied by Owen is a Strombodes'' and renamed
Billings’s (1859) misidentified Phillipsastrea gigas (Owen) as Phillipsastrea billingsi. Oliver (1976,
pp. 91, 129, 132) subsequently showed that Phillipsastrea billingsi and Rominger’s material are
species of Heliophyllum , and Davis’s specimen probably also belongs to the same Devonian genus.

Calvin’s (1893, pi. 5, fig. 5) illustration of Strombodes gigas does look to be conspecific with
Owen’s illustration. Furthermore, both appear likely to be conspecific with S. eximius Billings, of
which the holotype, GSC 2633, is in existence. Unfortunately it is a poorly preserved, silicified
specimen unsuitable for sectioning. Its external features suggest that it is a unique species of

46

PALAEONTOLOGY, VOLUME 32

Prodarwinia with mamillaris- type septal structure and tabularia about 5 mm diameter with eighteen
major septa. No other material referable to this species is known although Prodarwinia sp. nov.
shows some similarity and when more material is available to show variation in these corals, could
possibly prove to be conspecific.

? Prodarwinia granulosa (Foerste, 1906)

1906 Arachnophyllam ( Strombodes ) granulosum Foerste, p. 318, pi. 3, fig. 1.

1979 Arachnophyllum granulosum Foerste; Laub, p. 199.

Type material. Foerste’s material appears to be lost. It was described as from the Silurian, Waco Limestone
(?late Aeronian-early Telychian) from near Waco, Kentucky.

Discussion. Foerste compared his species with Strombodes alpenensis Rominger, which was shown
more recently to be an lowaphyllum from the Devonian, Hamilton Group of Michigan (Stumm
1953; see Oliver 1978). Laub (1979, p. 199) compared Foerste’s species with A. pentagonum.

Foerste’s illustration certainly looks more like an iowaphyllid than an Arachnophyllum. There-
fore, his species is most probably a Prodarwinia , close to if not conspecific with P. striata , but
fundamentally unrecognizable from Foerste’s description and illustration alone.

Species excluded from Arachnophyllum and Prodarwinia

Genus iowaphyllum Stumm, 1949

1949 Iowaphyllum Stumm, p. 50.

1978 Iowaphyllum; Oliver, p. 797 .

Discussion. Oliver (1978) lists species he considers properly assigned to Iowaphyllum and briefly
discusses the relationships of the genus. I regard Iowaphyllum as a member of the Arachnophyllidae
(see discussion above).

Iowaphyllum alpenensis (Rominger, 1876)

1876 Strombodes alpenensis Rominger, p. 133, pi. 38, fig. 1.

Discussion. Stumm (1953) has shown that Rominger’s species, which is from the Devonian Hamilton
Group of Michigan, is an Iowaphyllum (see Oliver 1978, p. 798).

Genus zenophila Hill, 1940

1940 Zenophila Hill. p. 414.

1981 Zenophila; Hill, p. 217.

Discussion. I regard Zenophila as likely to belong to the Arachnophyllidae, as tentatively indicated
by Hill (1981).

1 Zenophila kayi (Merriam, 1974)

1974 Arachnophyllum kayi Merriam, p. 43, pi. 5, figs. 7 and 8.

Remarks. Merriam’s species, from the Llandovery of Nevada, does not appear to agree with the
concept of Arachnophyllum as revised here, appearing to have much more continuous septa in the
dissepimentarium. A full revision of this species is outside the scope of this paper, but I suggest
that it is more closely allied to Zenophila. Pickett (1975, p. 153) had reached the same conclusion.
If correct, this would significantly extend the range of this genus, previously recorded from the
Upper Silurian of Australia.

SCRUTTON: SILURIAN ARACHNOPH YLLID CORALS

47

Suborder columnariina Soshkina, 1941
Family disphyllidae Hill, 1939
Subfamily paradisphyllinae Jell, 1969
Genus radiastraea Stumm, 1937

1937 Radiastraea Stumm, p. 439.

1976 Radiastraea ; Pedder and McLean, p. 135.

1982 Radiastraea ; Pedder, p. 77.

Diagnosis. Astraeoid or thamnasterioid, more rarely aphroid or pseudocerioid colonies. Septa
uniformly attenuate or slightly fusiform, smooth but more commonly carinate. Major septa may
more or less reach axis usually with a counter-clockwise vortex. Dissepiments small, moderately
inflated, uniformly developed; dissepimentarium surface may be broadly arched periaxially with a
broad fan of monocanthine trabeculae. Tabulae complete or incomplete, usually with broad flat-
topped tabellae in axial area.

Type species. R. arachne Stumm, 1937. Devonian, Zlichovian; Lone Mountain, Nevada.

Discussion. Pedder (1982, p. 77) has revised Radiastraea and removed the genus from the
Arachnophyllinae where it was placed by Pedder and McLean (1976). In both papers, the possibility
that Radiastraea evolved from Arachnophyllum (and in turn gave rise to some species at least of
Phillipsastrea ) was advanced. It seems more likely to me that Radiastraea evolved from an ancestor
with thin, solid septal blades, and more uniform small dissepiments, such as an entelophyllid, with
the acquisition of an astraeoid growth form.

Radiastraea richardsoni (Salter, 1852)

Text-fig. 14a-c

1852 Arachnophyllum richardsoni Salter, p. 232, pi. 6, figs. 10 and 1 ()a.

1878 Arachnophyllum richardsoni Salter; Etheridge, p. 585.

Diagnosis. Thin tabular astraeoid to thamnasterioid colonies. Well-defined circular tabularia 2-3-3 0 mm
diameter with twelve to fifteen major septa extending just short of or to the axis with a pronounced counter-
clockwise vortex. Septa uniformly attenuate or weakly fusiform, minors just penetrating tabularium. Tabellae
flat to dished periaxially, rising to septal sections in axis. Dissepiments small, globose; dissepimental surface
gently arched peritabularially with broad trabecular fan.

Type material. The original of Salter (1852, pi. 6, figs. 10 and 10a) appears to be lost. The type locality was
given as Niagaran, Pt. Eden, South side of Baring Bay, Arctic America. Neotype: BMNH R4959, Upper
Silurian; Cape Riley, North Devon, Arctic America.

Other material. 7BMNH R4934, Silurian, Wenlock; Cape Hilgard, Dobbin Bay, west of Kane Sea, Arctic
America (described as Arachnophyllum richardsoni by Etheridge 1878, p. 585).

Range. Silurian, ?Ludlovian; Arctic Canada.

Description. Tabular colonies <120x 100 x 10-22 mm with very broadly conical lower-surface and flat to
undulose calicular surface (text-fig. 13a). Tabularial pits, with or without slight peritabularial swelling, near
vertically sided, < 1-5 mm deep, with low, weak axial boss.

In cross-section, tabularia circular, well defined by innermost two to three series of close-spaced, adaxially
concave sections of steeply inclined dissepiments (text-fig. 13b). Major septa extend two-thirds to full radius
to axis, strongly curved in pronounced counter-clockwise vortex, with some septa fused at their axial ends.
All septa thin, uniformly attenuate to weakly fusiform peritabularially. Minor septa just penetrate tabularium.
In the dissepimentarium, septa straight to weakly zigzag with scattered weak, thorn-like carinae, continuous
with or less frequently abutting septa from adjacent corallites. Dissepimental sections regular, uniserial
between septa.

In longitudinal section, tabellae flat to dished periaxially, rising towards axis where, with scattered subsidiary
steeply abaxially inclined tabellae, they mingle with septal sections (text-fig. 13c). There is a narrow axial

48

PALAEONTOLOGY, VOLUME 32

text-fig. 14. A-C, Radiastraea richardsoni (neotype); Silurian, ?Ludlow; Cape Riley, North Devon, Arctic
Canada (BMNH R4959). a, calicular surface, x 1; b, cross-section, x2-5; c, longitudinal section, x 2-5.
d, e, IMazaphyllum approximation (holotype); Silurian, late Telychian, La Vieille Formation; Anse-a-la-
Vieille, Gascons, Quebec (GSC 9149). a, cross section; b, longitudinal section; both x2-5.

zone 0-3 diameter wide of flat tabellae. There are twenty-three tabellae in 5 mm. Dissepiments small, globose,
< L5 x 1 mm but generally 1 x 0-5 mm. Dissepimentarium surface gently arched peritabularially with broad
trabecular fan and steep dissepiment flanks facing and defining tabularium.

Trabeculae 0 04 0 08 mm thick, monacanths or tufted monacanths. Mean tabularium diameter 2-48
2-75 mm, with 1 3-7 1 5 major septa and corallite areas 0-3 1-0-59 cm2 (neotype first in each case).

SCRUTTON: SILURIAN A R ACH NOPH YLLI D CORALS

49

Discussion. Salter’s original would be expected to be housed in the British Museum (Natural
History) with the rest of the material collected during the series of Arctic voyages searching for
traces of the Franklin expedition. Although it cannot be traced, two other specimens collected
from the same general area as the figured specimen during these expeditions and identified as
Salter’s species are available. Of these, R4959 is closest to the original description and is selected
as neotype. The species is clearly a Radiastraea.

Although Radiastraea is predominantly a Lower and Middle Devonian genus, Pedder and
McLean (1976) have described new unnamed species from the Upper Silurian, mainly of Pridoli
age but with two specimens, from the Douro Formation of the Grinnell Peninsula, Devon Island,
Arctic Canada, possibly of Upper Ludlow age. These latter specimens (Pedder and McLean, 1976,
figs. 24.15-24.18) are almost certainly conspecific with R. richardsoni , the neotype cross-section
matching fig. 24.15, and the longitudinal section that of the other specimen, fig. 24.18.

Order cystiphyllida Nicholson, 1889
Family cystiphyllidae Edwards and Haime, 1850
Subfamily cystiphyllinae M’Coy, 1951

19766 Cystiphyllinae; McLean, p. 52.

1981 Holmophyllidae; Hill, p. 105.

Discussion. McLean (1976/?, p. 52) took a broad view of the composition of the Cystiphyllidae
given the great variability in internal structure of this group of corals. Within this large group, the
species described below agrees most closely with members of Hill’s Holmophyllidae, although
lacking the relatively well-defined tabularium usually found in that group in contrast to most
cystiphyllids.

Genus mazaphyllum Crook, 1955

1955 Mazaphyllum Crook, p. 1052.

1976 Mazaphyllum; Pedder, p. 287.

1981 Mazaphyllum; Hill, p. 107.

Diagnosis. Thamnasterioid or aphroid colonies. Septa present as discrete spines piercing several
dissepimental layers, randomly distributed but usually showing radial alignment in cross-section.
Tabularium more or less well defined, surface profile flat to shallowly sagging and composed of
vesicular tabulae. Dissepiments small, weakly inflated. Trabeculae rhabdacanthine and holacanthine.

Type species. Mazaphyllum cortisjonesi Crook, 1955. Silurian, Wenlock; Palmer’s Oaky District, New South
Wales, Australia.

Discussion. The genus is known mainly from the Wenlock and Ludlow of eastern Australia,
although the new, unnamed, species described by Pedder (1976, p. 287) comes from the Pridoli of
Arctic Canada.

1 Mazaphyllum approximation (Parks, 1933)

Text-fig. 14d, e

1933 Strombodes approximate Parks, p. 38, pi. 8, fig. 5.

1939 Arachnophyllum approximation (Parks) Northrop, p. 145.

Diagnosis. Tabular, aphroid colony. Tabularia vaguely defined in cross-section. Acanthine septa
densely developed in peritabularial area and less densely in tabularium; sparse elsewhere. Tabulae
vesicular in funnel-shaped tabularium, merging with surrounding dissepiments. Dissepimentarium
zoned, with septal spines short and sparse on crests of smaller dissepiments, becoming dense,
thicker and longer around tabularia. Tabularium diameter about 1-7 mm with twenty-six septa.
Mean corallite area 0-44 cm2.

50 PALAEONTOLOGY, VOLUME 32

Type material. Holotype: GSC 9149, Silurian, Telychian, La Vieille Formation; Anse-a-la-Vieille, Gascons,
Quebec.

Description. Tabular colony 120 x 100x40 mm with no clear surface features.

In cross-section, tabularia vaguely defined by annular zones of arcuate dissepiment sections steeply inclined
(text-fig. 14d). Septa acanthine, not separated into two orders, well developed in area of tabularia, sparse
elsewhere. Colony aphroid, with no trace of intercorallite walls. Septal spines developed in zones on
dissepimental and tabularial crests, densest immediately pcritabularially where spine bases are thickened
almost to contiguity and spines taper adaxially penetrating one or two vesicular layers. Spines may extend
to the axis. Spines only rarely seen elsewhere in the dissepimentarium. Dissepiment sections irregularly
elliptical, fifteen in 10 mm.

In longitudinal section, tabularia rather better defined, about 1-7 mm diameter (text-fig. 14e). Dissepi-
ments moderately inflated, <3-5x1 mm. Surface broadly arched in dissepimentarium, with slightly smaller,
less inflated and steeply inclined dissepiments adjacent to tabularium, merging with small, slightly inflated
vesicular to flat labellae in tabularium. Tabellae twenty to twenty-two in 5 mm. Dissepimentarium zoned
with layers of slightly smaller dissepiments, 3 4 mm apart, bearing sporadic very short septal spines. Spines
well developed in area of tabularia, 0-75 mm long. Microstructurc obscure, either tufted monacanths or pos-
sibly rhabdacanths.

There are about twenty-six septa at a diameter of 1-7 mm in cross-section. Mean corallite area 0-44 cm2.

Discussion The holotype of Strombodes approximatus had not previously been sectioned. The
species is clearly not an Arachnophyllum. The vesicular tabellae and acanthine septa suggest
affinities with the Cystiphyllida, possibly as a colonial holmophyllid. The only existing genus with
which it can be compared is Mazaphyllum, although the latter has a relatively well-defined
tabularium and very long septal spines, pervasively developed in the dissepimentarium but more
or less excluded from the tabularium. This species probably represents a new genus but as at the
moment it is known only from this single specimen, the establishment of a new generic name is
not felt to be justified.

Acknowledgements. I thank Paul Copper (Laurentian University, Sudbury), Dimitri Kaljo (Institute of
Geology, Estonian Academy of Sciences, Tallinn), Bjorn Neuman (University of Bergen), Bill Oliver, Jr.
(United States Geological Survey, Washington DC), and Pierre Semenoff-Tian-Chansky (Institut de
Paleontologie, Museum National d’Histoire Naturelle, Paris) for valuable discussion on various aspects of
the paper. All made material available, as did Roger Batten (American Museum of Natural History, New
York), Tom Bolton (Geological Survey of Canada, Ottawa), Peter Crowther (City Museum and Art Gallery,
Bristol), Ronald Eng (Museum of Comparative Zoology, Harvard), Valdar Jaanusson (Riksmuseet,
Stockholm), Jennifer Kitchell (University of Michigan, Ann Arbor), Peter Osborne (University of Birm-
ingham), David Price (Sedgwick Museum, University of Cambridge), Brian Rosen and Sue Nelson (British
Museum (Natural History), London), Nigel Trewin (University of Aberdeen), and Dennis White (British
Geological Survey, Keyworth). Joan Bennett (Newcastle upon Tyne) allowed me free access to her extensive
English Wenlock material. Carl Rexroad (Indiana Geological Survey, Bloomington) and Brian Witzke (Iowa
Geological Survey, Iowa City) gave valuable stratigraphical advice and Ross McLean (Amoco Canada
Petroleum Co. Ltd., Calgary) also contributed useful discussion. Brian Richardson and Christine Jeans
(University of Newcastle upon Tyne), made thin sections and acetate peels, and drafted the text-figures
respectively. To all of these I am most grateful. I acknowledge the support of NATO Research Grant no.
439/84 for this project, together with contributions from the University of Newcastle upon Tyne Research
Fund and the Royal Society towards some aspects of the work.

REFERENCES

amsden, T. w. 1949. Stratigraphy and paleontology of the Brownsport Formation (Silurian) of Western
Tennessee. Bull. Peabody Mus. nat. Hist. 5, i viii + 1 138, 34 pis.
benton, m. j. 1979. H. A. Nicholson (1844 1899), invertebrate palaeontologist: bibliography and catalogue
of his type and figured material. R. Scott. Mus. Inf Ser., Geology , 7, viii + I 94.
berry, w. B. N. and boucot, a. j. 1970. Correlation of the North American Silurian rocks. Spec. Pap. geol.
Soc. Am. 102, i xii+ I 289.

SCRUTTON: SILURIAN AR ACHNOPH YLLI D CORALS

51

billings, E. 1859. On the fossil corals of the Devonian rocks of Canada West. Can. J. Ind. Sci. & Art , 4, 97
140, figs. 1 -29.

— 1866. Catalogues of the Silurian fossils of the island of Anticosti, with descriptions of some new genera
and species , 1 93, 28 figs. Geological Survey of Canada, Montreal.

bolton, T. E. 1966. Illustrations of Canadian fossils — Silurian faunas of Ontario. Geol. Surv. Pap. Can.
66-5, 1 46, 19 pis.

— 1981. Early Silurian Anthozoa of Chaleurs Group, Port Daniel-Black Cape region, Gaspe Peninsula,
Quebec. In lesperance, p. j. (ed.). Field meeting, Anticosti- Gaspe, Quebec, 1981, vol. 2: Stratigraphy and
Paleontology, 299 314, 6 pis. University of Montreal.

— and copeland, m. j. 1972. Paleozoic formations and Silurian biostratigraphy, Lake Timiskaming region,
Ontario and Quebec. Geol. Surv. Pap. Can. 72-15, 49 pp., 13 pis.

Calvin, s. 1893. Notes on some of the fossil corals described by David Dale Owen in his report of work
done in the autumn of 1859, with observations on the Devonian species, Phillipsastrea gigas of later
authors. Am. Geol. 12, 108 112, pis. 5 and 6.

cocks, l. r. m. and fortey, r. a. 1982. Faunal evidence for oceanic separations in the Palaeozoic of Britain.
J. geol. Soc. Lond. 139, 465-478, 6 figs.

copper, p. 1978. Paleoenvironmcnts and paleocommunities in the Ordovician Silurian sequence of Manitoulin
Island. Sp. Pap. Mich. Basin Geol. Soc. 3, 47-61, 14 pis.

COTTON, G. 1973. The rugose coral genera, 358 pp. Elsevier, Amsterdam.

crook, k. a. w. 1955. Mazaphyllum, a new cystiphyllid coral from the Silurian of New South Wales. J.
Paleont. 29, 1052 1056, 3 figs.

dana, j. d. 1846. Genera of fossil corals of the family Cyathophyllidae. Am. J. Sci. Arts (2), 1, 178-189,
5 figs.

davis, w. J. 1887. Kentucky fossils corals, a monograph of the fossil corals of the Silurian and Devonian of
Kentucky , Pt. 2, xiii pp. 139 pis. Kentucky Geological Survey, Frankfort, Kentucky.
dodge, r. e., aller, r. c. and Thomson, j. 1974. Coral growth related to resuspension of bottom sediments.
Nature, Lond. 247, 574-577, 4 figs.

dybowski, w. N. 1873-1874. Monographic der Zoantharia sclerodermata rugosa aus der Silur Formation
Estlands, Nord-Livlands und der Insel Gotland. Arch. Naturk. Liv-, Ehst- Kurlands (/), 5 (3), 257 414, pis.

1 and 2 (1873); 415-532, pis. 3-5 (1874).

edwards, h. m. and haime, j. 1850. A monograph of the British fossil corals. Part 1. Introduction, etc.
Palaeontogr. Soc. [ Monogr .], i lxxxv+1 71, pis. 1 11.

— 1851. Monographie des polypiers fossiles des terrains palaeozoiques. Archs Mus. Hist. nat. Paris,
5, 1-502, pis. I 20.

— 1855. A monograph of the British fossil corals. Part V. Corals from the Silurian Formation.
Palaeontogr. Soc. [Monogr. \, 245 299, pi. 57 72. London.

etheridge, r. 1878. Palaeontology of the coasts of the Arctic lands visited by the late British expedition
under Capt. Sir George Nares, r.n., k.c.b., f.r.s. Q. Jl geol. Soc. Lond. 34, 568 -636, pis. 25-29.
foerste, f. a. 1888. Notes on Palaeozoic fossils. Bull. Denison Univ., Sci. Lab. 3, 117-136, pi. 13.

— 1906. The Silurian, Devonian and Irvine Formations of east-central Kentucky, etc. Bull. Ky geol. Surv.
7, 1-369, 8 pis.

goldfuss, g. a. 1826 1833. Petrefacta Germaniae, I, 1 76, pi. 1 25 (1826); 77-164, pis. 26-50 (1829); 165
240, pi. 51-71 (1831); 241 -252 ( 1833). Arnz & Co., Dusseldorf.
grabau, a. w. 1917. Stratigraphic relationships of the Tully Limestone and the Genessee Shale in eastern
North America. Bull. geol. Soc. Am. 28, 945-958.

hill, D. 1940. The Silurian Rugosa of the Yass-Browning district, N.S.W. Proc. Linn. Soc. NS W, 65, 388
420, pis. 1113.

— 1956. Rugosa. In moore, r. c. (ed.). Treatise on invertebrate paleontology. Part F, 233-324, figs. 165
219. Geological Society of America and the University of Kansas Press, Boulder, Colorado and Lawrence,
Kansas.

— 1981. Rugosa and Tabulata. In teichert, c. (ed.). Treatise on invertebrate paleontology. Part F.
Coelenterata, suppl. 1, xl+ 1-762, figs. 1 462. Geological Society of America and University of Kansas
Press, Boulder, Colorado and Lawrence, Kansas.

holmes, m. e. 1887. The morphology of the carinae upon the septa of rugose corals, 31 pp, 16 pis. Bradlee
Whidden, Boston.

hubbard, j. a. e. b. and pocock, y. p. 1972. Sediment rejection by recent scleractmian corals: a key to palaeo-
environmental reconstruction. Geol. Rdsch. 61, 598-626, figs. 1-10.

52

PALAEONTOLOGY, VOLUME 32

iwanowskii, a. b. 1965. Drevneishie rugozy , 152 pp., 39 pis. Nauka, Moskva. [In Russian.]
james, u. p. 1878. Descriptions of newly discovered species of fossils and remarks on others, from the Lower
and Upper Silurian rocks of Ohio. Paleontologist , 1, 2, 9-13.
lambe, l. m. 1 899-190 1 . A revision of the genera and species of Canadian Palaeozoic corals. The Madreporaria
Aporosa and the Madreporaria Rugosa. Contr. Can. Palaeont. 4 (2), 97 197, pis. 1-18.
lang, w. d. and smith, s. 1927. A critical revision of the rugose corals described by W. Lonsdale in
Murchison’s 'Silurian System’. Q. Jl geol. Soc. Loncl. 83, 448-491, pis. 34-37.

— and thomas, h. d. 1940. Index of Palaeozoic coral genera , 231 pp. British Museum (Natural
History), London.

laub, r. s. 1979. The corals of the Brassfield Formation (mid-Llandovery; Lower Silurian) in the Cincinnatti
Arch region. Bull. Ain. Paleont. 75, 305, 1-457, 42 pis.
lonsdale, w. 1839. Corals. In Murchison, r. i. The Silurian System. Part 2, 675-699, pis. 15, 15 bis, 16, 16
bis, 17. John Murray, London.

m'coy, f. 1849. On some new genera and species of Palaeozoic corals and Foraminifera. Ann. Mag. nat. Hist.
2, 3, 1-136.

— 1850. On some new genera and species of Silurian Radiata in the collection of the University of
Cambridge. Ibid. 2, 6, 270-290.

1851 1855. In sedgwick, a. A synopsis of the classification of the British Palaeozoic rocks, by the Rev.
Adam Sedgwick, . . . with a systematic description of the British Palaeozoic fossils in the Geological Museum
of the University of Cambridge by Frederick M' Coy, i iv, I -184 (1851); i— [x], 185-406 (1852); Introduction,
i-xcviii, 407-661 (1855); pis. iA-iL, ii a iiD, iii a iiiD. J. W. Parker and Son, and Cambridge University Press,
London and Cambridge.

mckerrow, w. s. 1988. The development of the Iapetus Ocean from the Arenig to the Wenlock. In Harris,
a. l. and fettes, d. j. (eds.). The Caledonian- Appalachian Orogen, 405 412, 3 figs. Special Publication of
the Geological Society, London.

mclean. r. a. 1975. Lower Silurian rugose corals from central New South Wales. J. Proc. R. Soc. NSW, 108,
54 69.

— 1976a. Aspects of the Silurian rugose coral fauna of the Yass region. New South Wales. Proc. Linn.
Soc. NSW, 100, 179-194, pis. 18 21 T

1976 b Genera and stratigraphic distribution of the Silurian and Devonian rugose coral family
Cystiphyllidae Edwards and Haime. Geol. Surv. Pap. Can. 76- IB, 295-301, fig. 52.1.
marshall, s. m. and orr, a. p. 1931. Sedimentation on Low Isles Reef and its relation to coral growth. Sci.

Rept. Gt. Barrier Reef Exped. 1, 93-133, pis. 1-3. British Museum (Nat. Hist.), London.
merriam, c. w. 1974. Silurian rugose corals of the central and southwest Great Basin. Prof. Pap. US geol.
Surv. Ill, iv + 1-66, 16 pis.

nicholson, h. a. and hinde, G. J. 1874. Notes on the fossils of the Clinton, Niagara and Guelph formations
of Ontario, with descriptions of new species. Can J. Sci. Lit. and Hist. 14, 137 152, 6 figs.

Northrop, s. a. 1939. Paleontology and stratigraphy of the Silurian rocks of the Port Daniel- Black Cape
region, Gaspe. Spec. Pap. geol. Soc. Am. 21, 301 pp., 28 pis.

Oliver, w. a. jr. 1974. Classification and new genera of noncystimorph colonial rugose corals from the
Onesquethaw Stage in New York and adjacent areas. J. res. US Geol. Surv. 2, 2, 165-174, 5 figs.

1976. Noncystimorph colonial rugose corals of the Onesquethaw and Lower Cazenovia Stages (Lower
and Middle Devonian) in New York and adjacent areas. Prof. Pap. US Geol. Surv. 869, vi + 156 pp.,
108 pis.

1977. Biogeography of late Silurian and Devonian rugose corals. Palaeogeogr. Palaeoclimat. Palaeoecol.
22, 85-135, 13 figs.

— 1978. Iowaphyllum (rugose coral) from the Upper Devonian of Arizona. J. res. US Geol. Surv. 6, 6,
797-805, 4 figs. ’

and galle, a. 1971. ‘ Calceola ’ (= Rhizophyllum) and ' Billingsastraea' (= Iowaphyllum) in Bohemia.
Vest, itstfed. Ust. geol. 46, 209-216, pis. 1-4.

orbigny, a. d’. 1850. Prodrome de pa/eontologie . . ., 1. pp. lx+ 1-394. Victor Masson, Paris.

owen, d. d. 1844. Report of a geological exploration of part of Iowa, Wisconsin, and Illinois ... in 1839;

etc. US 28tli Congr. Senate Exec. Doc. 407, 191 pp., 24 pis.
parks, w. a. 1933. New species of stromatoporoids, sponges and corals from the Silurian strata of Baie des
Chaleurs. Univ. Toronto Stud. geol. Ser. 33, 1-40, 8 pis.
pedder, a. e. h. 1976. First records of five rugose coral genera from Upper Silurian rocks of the Canadian
Arctic Islands. Geol. Surv. Pap. Can. 76-1 B, 287-293, figs. 51.1-20.

SCRUTTON: SILURIAN A R ACH NOPH Y LLI D CORALS

53

— 1977. Systematics and biostratigraphic importance of the Lower Devonian rugose coral genus Exilifrons.
Ibid. 77-1 B, 173-180, figs. 34.1-19

1982. New Zlichovian (early Devonian) rugose corals from the Blue Fiord Formation of Ellesmere
Island. Ibid. 82-1C, 71-82, pis. 9.1 a-c.

— and mclean, r. a. 1976. New records and range extensions of seven rugose coral genera in Silurian
strata of northwestern and arctic Canada. Ibid. 76-1C, 131-141, figs. 24.1-24.

Phillips, J. 1841. Figures and descriptions of the Palaeozoic fossils of Cornwall, Devon and West Somerset ,
xii+ 1-231, pis. 1 60. Geological Survey of Great Britain and Ireland, London.
pickett, j. 1975. Continental reconstructions and the distribution of coral faunas during the Silurian. J.
Proc. R. Soc. NSW , 108, 147-156, 5 figs.

rexroad, c. b. and kleffner, m. a. In press. The Silurian stratigraphy of east-central Kentucky and adjacent
Ohio. In rast, N. and hay, h. (eds.). Field trip guides for G.S.A. annual meeting , southeastern and north-
centra! sections, Lexington , Kentucky, April 4 -6 , 1984, 44 63.

— noland, a. v. and pollock, c. a. 1978. Conodonts from the Louisville Limestone and the Wabash
Formation (Silurian) in Clark County, Indiana and Jefferson County, Kentucky. Sp. Rep. Geol. Surv.
Indiana, 16, [iv]+15 pp., 1 pi.

rominger, c. 1876. Palaeontology. Fossil corals. Geol. Surv. Michigan, 3, 2, I 161, 55 pis.

Salter, J. w. 1852. Geology. In Sutherland, p. c. Journal of a voyage in Baffin s Bay and Barrow Straits, in
the years 1850-1851, etc. 2, Appendix, 217 233, pis. 5 and 6. John Murray, London.
schweigger, A. F. 1819. Beobachtungen auf Naturhistorischen Reisen, xii+ 1-127, pis. I 8. Reirner, Berlin.
schmidt, f. 1858. Untersuchungen fiber die Silurische Formation von Ehstland, Nord-Livland und Oesel
Arch. Naturk. Liv- Ehst-Kurlands, ser. i, 2, 1, 1-248.
scrutton, c. t. 1981. The measurement of corallite size in corals. J. Paleont. 55, 687 688, 3 figs.

— 1988. Patterns of extinction and survival in Palaeozoic corals. In larwood, g. p. (ed . ). Extinction and
survived in the fossil record, 65-88, 6 figs. (Systematics Association Special Volume 34.) Clarendon Press,
Oxford.

— and powell, j. h. 1981. Periodic development of dimetrism in some favositid corals. Acta Palaeont. pol.
25, 477 491, 8 figs.

shaver, R. H. et at. 1985. Correlation of stratigraphic units of North America (COSUN A project): Midwestern
basin and arches region. A.A.P.G., Tulsa, Oklahoma.

Stumm, E. C. 1937. The Lower Middle Devonian tetracorals of the Nevada Limestone. J. Paleont. II, 423
443, pis. 53-55.

— 1949. Revision of the families and genera of the Devonian tetracorals. Mem. geol. soc. Am. 40, vii +
1-92, pis. 1-25.

— 1953. Tetracoralla. In Type invertebrate fossils of North America (Devonian), Division 1, Unit IF, pt. B,
cards 89-127. Wagner Free Inst. Sci., Philadelphia.

— 1965. Silurian and Devonian corals of the Falls of the Ohio. Mem geol. Soc. Am. 93, x+ 1 184, pis.
1-80.

ulrich, e. o. 1886. Descriptions of new Silurian and Devonian fossils. Contr. Am. Paleont. 1, 3-35,
pis. 1-3.

veron, J. E. n. and pichon, m. 1976. Scleractinia of eastern Australia, Part I. Families Thamnasteriidae,
Astrocoeniidae, Pocilloporidae. Monogr. Austr. Inst. Mar. Sci. 1, v + I 86, 166 figs.
witzke, b. j. 1983. Silurian benthic invertebrate associations of eastern Iowa and their paleoenvironmental
significance. Proc. Wise. Acad. Sci. Arts Lett. 71, 21 47.

Typescript received 26 January 1988
Revised typescript received 3 May 1988

COLIN T. SCRUTTON

Department of Geology
The University

Newcastle upon Tyne NE1 7RU

EVALUATION OF A THECIDEIDINE BRACHSOPOD
FROM THE MIDDLE JURASSIC OF THE
COTSWOLDS, ENGLAND

by P. G. BAKER

Abstract. The discovery of sources of material of Aalenian (Middle Jurassic) age has enabled restudy of
Moorellina dundriensis (Rollier). Although in general morphology the species resembles other monoseptate
moorellinins, the absence of brachial lobes, together with the characteristically thickened shell and distinctive
skeletal shell fabric, clearly separates it from all other contemporaneous representatives of Moorellina. On
this basis it is assigned to a new genus Pachymoorellina. The shell microstructure, with its suppressed
secondary fibrous mosaic, resembles that of the polyseptate Mimikonstantia sculpta Baker and Elston with
which it is associated. The microstructure of the pedicle valve indicates a phylogenetic link with the Lower
Cretaceous Thecidiopsis tetragona (Roemer). Furthermore, correlation of the microstructure of certain
tubercles in the brachial valve of Pachymoorellina dundriensis with almost identical structures in the Recent
Thecidel/ina barretli (Davidson) suggests a more rectilinear evolutionary pattern for the Thecidellina group
than has previously been supposed.

The phylogenetic importance of the suppressed (Williams 1973) secondary fibrous shell mosaic
found in a recently described (Baker and Elston 1984) upper Aalenian micromorphic polyseptate
genus prompted the search for possible ancestral stock. Study of the ontogeny (Baker and Elston
1984, p. 784) of Mimikonstantia sculpta Baker and Elston clearly showed the development of the
polyseptate condition from an early juvenile monoseptal phase, indicating derivation from a
monoseptate ancestor. Further, the absence of brachial lobes suggested the possibility of descent
via a bouchardi or dundriensis- type moorellinid ancestral line. A normal ( sensu Williams 1973,
p. 465) secondary shell layer has been identified in the Liassic Moorellina bouchardi (Davidson)
and, although this does not completely rule out bouchardi stock, the simultaneous acquisition of
a suppressed secondary mosaic and polyseptate condition in Mimikonstantia sculpta is thought to
be unlikely. Serial sectioning of micromorphic dundriensis- type specimens associated with M.
sculpta in loosely consolidated marly horizons overlying hardgrounds in the Pea Grit at Crickley
Hill (N.G.R. SO 928 163), near Cheltenham, yielded startling results. They were found to have a
strongly suppressed secondary fibrous mosaic totally unlike that of any associated Moorellina
species and a pedicle valve skeletal fabric which, in part, parallels that exhibited by certain Lower
Cretaceous thecideidines. The Crickley specimens are morphologically identical with the Bajocian
specimens assigned to Moorellina dundriensis (Rollier) from the Inferior Oolite of Dundry Hill,
near Bristol. The Dundry material also shows the same suppressed secondary fibrous mosaic. M.
dundriensis , with its empty brachial cavities and enormously thickened pedicle valve, has always
been noticeably different from the associated Dundry species M. granulosa (Moore) and M.
duplicata (Moore) with their characteristically developed brachial lobes. The differences in
morphology, together with a fundamentally different shell secretory regime, clearly separate M.
dundriensis from all other known moorellinin genera. In view of this and the phylogenetic
implication discussed later, it is proposed to assign M. dundriensis to a new genus and to include
the Crickley specimens therein. The confusion surrounding the identity, description, and location
of Moore’s original type specimens and the selection of hypotypes from the Crickley material
necessitates emendation of the previous diagnoses.

| Palaeontology, Vol. 32, Part 1, 1989, pp. 55-68, pis. 9-1 1.|

© The Palaeontological Association

56

PALAEONTOLOGY, VOLUME 32

Registration of material. The paratypes (M2848) and the hypotypes figured in this paper (PB3250 3256) are
housed in Bath Geology Museum. Other hypotypes are located in the Institut de Paleontologie (Natural
History), Paris (MNHN IP B44600 and B44601). The specimen of Thecidiopsis tetragona is housed in the
Musee Geologique, Lausanne (42533).

Preparation of material. The techniques used for the recovery and preparation of the specimens were fully
described by Baker and Elston (1984, p. 777) and require no further elaboration here.

SYSTEMATIC PALAEONTOLOGY

Order spiriferida Waagen, 1883
Suborder thecideidina Elliott, 1958
Superfamily thecideacea (Gray, 1840) H. and G. Termier, 1949
Family thecideidae Gray, 1840
Subfamily moorellininae Pajaud, 1966
Genus pachymoorellina gen. nov.

Etymology. From the Greek pakhus (thick) after the greatly thickened free ventral wall of the pedicle valve.

Diagnosis. Endopunctate, monoseptate moorellinin having a strongly suppressed secondary fibrous
shell mosaic, brachial valve with a fragile, very wide peripheral flange and a high median septum
separating brachial cavities devoid of brachial lobes, and pedicle valve with a large area of
attachment and characteristically thickened free ventral wall.

Type species. Thecidea ( Davidsonella ) Dundriensis Rollier, 1915 nom. subst. [ pro Thecideum Bouchardii Moore,
1854 (non Davidson, 1851)].

Pachymoorellina dundriensis (Rollier, 1915)

Plates 9 and 10; Plate 11, figs. 1-6; text-figs. 1 3

1854 Thecideum Bouchardii [non Davidson] Moore, p. 116, pi. 1, figs. 11 13.

1854 Thecideum Bouchardi Davidson, p. 79.

1876 Thecideum Bouchardii Davidson, p. 106, pi. 13, figs. 1-3.

1915 Thecidea ( Davidsonella ) Dundriensis Rollier, p. 53.

1963 Moorellina dundriensis (Rollier); Pajaud, p. 996.

1966 M. ( Moorellina ) dundriensis (Rollier); Pajaud, p. 633, figs. 4 and 6.

1970 Moorellina (Moorellina) dundriensis (Rollier); Pajaud, p. 164, pi. 3, fig. 4; pi. 8, figs. 1 and 2;
text-figs. 66Dc and 67.

explanation of plate 9

Figs. 1 8. Pachymoorellina dundriensis (Rollier). Stereoscan photomicrographs of specimens from the
Aalenian (Middle Jurassic) Pea Grit of Crickley Hill, near Cheltenham, Gloucestershire; all specimens were
coated with evaporated gold before photography. 1 and 2, PB3250, brachial and posterior views of
hypotype brachial valve in which the delicate endopunctate marginal flange is almost intact, x 22. 3 and
4, PB3251, brachial and posterior views of a hypotype to show the morphology typical of separated
brachial valves in which the marginal flange is broken away and the median septum usually damaged,
x20. 5, three-quarters profile view of same specimen (rotation, 40°; tilt angle, 45°), showing the form of
the brachial cavities and the remains of the ridges uniting the posterior of the median septum with the
bridge abutments, x 20. 6, PB3254, three-quarters profile view (rotation, 45°; tilt angle, 35°) of an

immature hypotype pedicle valve showing the large area of attachment, the characteristic marginal crest,
and the precursor of the enormously thickened free ventral wall, x 17. 7, PB3250, enlarged portion of
the lateral margin of a brachial valve showing the terminal faces of a small patch of secondary fibres (lower
centre) at the point where the marginal flange unites with the sub-peripheral rim, x 1200. 8, PB3250,

enlarged view of part of the outer margin of the sub-peripheral rim of a brachial valve showing the tubercles
composed of granular calcite, x 1000.

PLATE 9

BAKER, Pachymoorellina

58 PALAEONTOLOGY, VOLUME 32

Tvpe specimens. No holotype designated. Four paratypes on M2848; hypotypes PB3250-3256 and MNHN
IP B44600- 44601.

Original diagnosis (Moore 1854). Shell inequivalve, flattened, subcircular; attached by the principal portion
of the ventral valve; deltidium large, elevated, triangular; area large and extended showing lines of growth;
hinge line depressed in centre, leaving a small flat area under the deltidium; dorsal valve much smaller than
the ventral. The interior of the ventral valve shows a slight middle septum, on either side of which are two
large scars, due to the attachment of the cardinal muscle, on the outer edge of which are two small depressions,
which received the adductor muscles; interior rugosely striated; the cavity of the valve in adult shells
surrounded by a broad margin, having a wavy appearance, due to lines of growth. Interior of the dorsal
valve has a broad granulated margin, within which is a very high central septum, nearly reaching the surface
of the opposite valve from whence proceeds a granulated ridge, united by a bridge over the visceral cavity;
within this ridge is a smooth slightly concave space, between which and the granulated interior is a small
granulated ridge.

Emendation of original diagnosis (Pajaud 1970). Coquille de taille moyen (L = 6-5 mm; 1 = 8 mm), legerement
transverse, tres inequivalve et a surface de fixation relativement peu etendue. Grand valve subcirculaire, a
cavite distincte, peu developee, avec pseudodeltidium bombe; bord cardinal long. Petite valve subcirculaire,
megathyride; processus cardinal saillant, aires articulaires etendues; limbc marginal granulcux. Septe median
long, haut et mince; formation en V tres visible. Aires lophophoriennes lisses ou legerment granuleuses.

Emended diagnosis. Moderate sized Pachymoorellina , up to about 5-5 mm in length, 6-5 mm in
width, and 3-5 mm in thickness. Outline transversely elliptical; weakly dorsiconvex up to the point
at which the free ventral wall of the pedicle valve begins to develop, after which the lateral profile
becomes flattened triangular. Pseudodeltidium large, transversely convex, and elevated above the
surface of the ventral interarea, with a flattened exposed anterior terminal face. Hinge line long,
only slightly shorter than the maximum width of the shell. The undamaged brachial valve is
characterized by a wide, very thin, peripheral flange which, in complete shells, conceals the
disproportionately thickened rim of the pedicle valve.

Description. A moderate sized moorellinin with a large attachment scar and a variably developed free ventral
wall, giving a range from dorsiconvex to flattened triangular lateral profile. The ventral interarea is clearly
developed with a large, well-defined, transversely convex pseudodeltidium. There is never more than a trace
of an interarea in the rather weakly convex brachial valve. In articulated shells the delicate peripheral flange
of the brachial valve is almost invariably broken away, producing a marked inequivalve appearance with the
characteristically thickened free ventral wall of the pedicle valve exposed.

Distribution. Geographical distribution unknown but probably more widespread than its recorded occurrence.
Although the precise details remain unknown, the locality from which Moore collected his original material
is given (Moore 1854, p. 110) as a small quarry in the lower part of the Dundry Freestone situated above
the village of Bishport, near Dundry (Avon). As far as can be ascertained this would approximate
stratigraphically to a parkinsoni Zone, truelli Subzone position. The new material was obtained from near
the top of the Pea Grit (Aalenian, murchisonae Zone, murchisonae Subzone) at Crickley Hill (Grid Ref. SO
928 163), near Cheltenham, Gloucestershire, indicating a range of at least Aalenian to Bajocian for the
species.

MORPHOLOGY, GROWTH AND SHELL MICROSTRUCTURE
Valve characters

Pedicle valve (PI. 9, fig. 6; text-fig. 1e, f). Hemispondylium sessile with the elongate diductor muscle scars
deeply impressed into the floor of the valve and united posteriorly to form a characteristic inverted V; the
median myophragm is prominent, bifurcating, and grooved anteriorly to accommodate the median adductor
muscle scars. The lateral adductor muscle scars are large and widely separated with a narrow groove along
their dorsal margin. The teeth are weakly to moderately developed.

Brachial valve (PL 9, figs. 1-5; text-fig. 1a-d). The cardinal process is relatively small and obscurely bilobed.
The lateral adductor muscle scars are large, rounded rectangular, and erect (almost perpendicular to the
commissural plane), occupying most of the posterior border. The inner socket ridges are extended laterally

BAKER: JURASSIC THECIDEIDINE BRACHIOPOD

59

text-fig. 1. ‘Wild’ stereomicroscope traces (except a) to show the essential morphology of Pachymoorellina
dundriensis (Rollier). a, PB3252, reconstruction of the juvenile brachial valve, based on three specimens, to
show the form of the ‘horn-like’ connections between the posterior of the median septum and the bridge
abutments, b, PB3253, immature brachial valve showing fusion of the connections between the median septum
and the bridge abutments with the floor of the valve; the peripheral llange is missing in this specimen, c,
M2848, interior view of a paratype brachial valve showing the morphology typical of mature brachial valves
in which the peripheral flange is missing, d, PB3250, interior view of a hypotype brachial valve showing the
almost intact peripheral flange. E, PB3254, immature pedicle valve showing the large area of attachment
(entire ventral surface of the valve as figured) and the characteristic marginal crest with its very wide externally
inclined border, f, M2848, interior view of a paratype pedicle valve showing the morphology typical of
mature pedicle valves; note the similarity in the appearance of the tuberculate margin, in dorsal view, both
before (e) and after the development of the free ventral wall. Abbreviations: a.g., articulatory groove; a.r.,
articulatory ridge; b ., bridge; b.c., brachial cavity; c.p., cardinal process; d.s., diductor muscle scar; e.b.,
externally sloping finely tuberculate border; h.c., horn-like connection; h.t., hinge tooth; /., ventral interarea;
l.a.s., lateral adductor muscle scar; m., median myophragm; m.a.s., median adductor muscle scar; m.c.,
marginal crest; m.s., median septum; p., pseudodeltidium; p.f., peripheral flange; s., dental socket; s.p.r ., sub-
peripheral rim; t., tubercle; v.c., visceral cavity; and v.w., free ventral wall. Scale bar represents 100 mm.

60

PALAEONTOLOGY, VOLUME 32

along the dorsal margin of the lateral adductor muscle scars to form low, narrow ridges which articulate
with the complementary grooves in the pedicle valve. The median septum is thin and blade-like, extending
over the body cavity. Posteriorly, two ridges diverge from the base of the median septum. These ridges may
extend across the floor of the valve for only a short distance or, in some cases, form low ridges extending
round the perimeter of the visceral cavity to unite with the bridge abutments. The brachial cavities are smooth
or weakly granulose. The outer surface of the sub-peripheral rim is ornamented with small tubercles. A wide
but very thin and fragile endopunctate peripheral flange (PI. 9, figs. 1 and 2) is present. This invariably shows
some damage and in separated valves is usually broken away altogether (PI. 9, figs. 3 and 4; text-fig. lc).

Ontogeny

Pedicle valve. All the characters of the pedicle valve are present in the smallest valves available. Ontogenetic
development involves the development of the hinge line, slight changes in the relative proportion of the
pseudodeltidium, the muscle scars, the hinge teeth, and, most noticeably, the development of the free ventral
wall. Up to the point at which the valve is about 5-0 mm long, attachment is effected by the whole surface
(PI. 9, fig. 6). The precursor of the free ventral wall is present, however, even during this phase of development.
It is represented by a clearly defined marginal crest demarcating the limit of the internal cavity and falling
away laterally and anteriorly (PI. 9, fig. 6; text-fig. 1 e) to form a very wide, finely tuberculate, externally
sloping border. As development proceeds, the outer edge of the valve begins to grow away from the substrate
but shell material continues to be accreted to the whole area external to the marginal crest so that, in relative
terms, an enormously thickened free ventral wall is developed (text-figs. If and 3g).

Brachial valve. The juvenile valve, apart from being slightly more circular, with less erect lateral adductor
muscle scars, differs from the adult valve only in the presence of two posteriorly directed horn-like processes
(text-fig. 1a) extending from near the base of the posterior edge of the median septum. These structures often
extend backwards to unite with the bridge abutments. Initially they remain separated from the valve floor
but as growth proceeds, accretion of material along their dorsal edges eventually unites them with the floor
of the valve (text-fig. 1b). Evidently, in the adult valve shell resorption activity outstrips accretion as the
structures are eliminated altogether or reduced to low ridges bounding the visceral cavity (PI. 9, fig. 5).

Microstructure

In contrast with all other described monoseptate moorellinins, the shell of P. dundriensis shows strong
suppression of the secondary fibrous layer. Significant differences in the microstructure of the brachial and
the pedicle valve necessitate description of the two valves separately.

Pedicle valve. A thin, granular primary layer is present, composed of approximately equant granules about
1 ym across. Orthodoxly stacked secondary fibres (sensu Williams 1973, p. 454) are restricted to the hinge
teeth, tooth ridges, and occurrence as occasional strands (PI. I 1, fig. 1) along the posterior and posterolateral
margins of the valve. The lateral, anterolateral, and anterior sectors of the valve are characterized by tracts

EXPLANATION OF PLATE 10

Figs. 1 6. Pachymoorellina dundriensis (Rollier). Stereoscan photomicrographs of specimens and cellulose
acetate peels of sectioned specimens; all specimens were coated with evaporated gold before photography.
I, PB3251, enlargement of a tubercle on the floor of the right brachial cavity showing its component
secondary fibres arranged into a low cone, x 1500. 2, PB3254, enlargement of a pedicle valve tubercle
showing the terminal surface of an acicular crystallite tract emerging from a matrix of finely granular
calcite to form a tubercle; seen end-on, the acicular crystallites are virtually indistinguishable from granules,
x 1000. 3, PB3254, floor of pedicle valve adjacent to the marginal crest, showing detail of the much

coarser granular calcite which is deposited as the valve thickens in association with the developing free
ventral wall, x 3000. 4, PB3255, acetate peel section (3255/4) through a pedicle valve showing the general
appearance of the acicular crystallite tracts which form the tubercle cores (section orientation— parallel
with the area of attachment; section location— anterolateral sector), x 750. 5, enlargement of the area

outlined in fig. 4 to show the splayed relationship of the terminal crystallites, x 1250. 6, detail of acicular
crystallite tracts in longitudinal section showing the characteristic splayed arrangement of the component
crystallites along their entire length (section orientation and location as in fig. 4), x 1250.

PLATE 10

BAKER, Pachymoorellina

62

PALAEONTOLOGY, VOLUME 32

text-fig. 2. ‘Wild’ stereomicroscope traces of cellulose acetate peels of horizontal sections through
the pedicle valve of Pachymoorellina dundriensis (Rollier). a, PB3255/6, posterolateral to anterolateral
sector of specimen showing the arrangement and distribution of the acicular crystallite tracts in
longitudinal section, relative to the granular calcite shell. B, PB3255/14, part of the anterior sector
of the specimen showing the arrangement and distribution of acicular crystallite tracts close to the
rim of the valve; note the relative thickening of the granular shell layer in the free ventral wall.

Scale bar represents 100 /im.

(PI. 10, fig. 4) of acicular crystallites, about 1 /an wide and about 10 /mi long, aggregated into series that
exhibit a typically splayed (PI. 10, fig. 5) relationship. The characteristic grouping and arrangement of the
crystallites creates an overlapping series of large apparent megafibres up to 100 /mi wide and up to 0-8 mm
long (PI. 10, fig. 6; text-fig. 2). In many cases the aggregates appear to be almost contiguous (PI. 11, fig. 2)
or separated only by narrow strips of granular calcite. The tracts of acicular crystallites form a thick wedge
almost totally comprising the marginal crest (text-fig. 3f) of the immature pedicle valve. Because of the
peculiarity of the ontogenetic development of the immature pedicle valve, sections orientated parallel with
the plane of the attachment scar (text-fig. 2a) enable the path of these crystallite aggregates to be traced
through to the external surface of the sloping border. Their tips are seen to emerge as small tubercles (PI.
10, fig 2; text-fig. 3f). In reality, therefore, the apparent megafibres are revealed as a series of closely packed
tubercle cores. As the growth pattern of the valve changes and the shell begins to grow away from the
substrate to form a free ventral wall, granular primary shell forms the outer layer. As the free ventral wall
develops and increases in height, it is thickened by the deposition of granular calcite on its inner surface
(text-figs. 2b and 3g). This inner layer may be up to 0-3 mm thick and the component granules are much
coarser (PI. 10, fig. 3) than the granules of the primary layer. The shell substance is apparently impunctate
as no endopunctae have been observed in any of the material studied.

Brachial valve. A thin granular primary layer is present. Fibrous secondary shell is well represented in the
cardinal process, inner socket ridges, and posterior border. Elsewhere it occurs as thin sheets, only a few
fibres thick, providing almost continuous cover in posterolateral sectors (PI. 11, fig. 5; text-fig. 3d) and
appearing as intermittent patches in lateral sectors (PI. 9, fig. 7; text-fig. 3e) of the valve. Additionally,
sporadic secretion of small aggregates of about fifty secondary fibres (PI. 10, fig. 1) forms low conical mounds

BAKER: JURASSIC TIIECIDEIDINE BRACHIOPOD

63

OQ(
O G~

fine-grained calcile shell.

coarse-grained calcile shell.

acicular crystallite shell.

secondary fibrous shell.

mm? m

L 1

text-fig. 3. Diagrammatic reconstruction of the shell microstructure of Pachymoorellina
dundriensis (Rollier). a-c, locational diagrams. D G, block diagrams showing the varying
microstructure of brachial and pedicle valves, d, posterolateral and e, anterolateral sectors of
the brachial valve, f, anterolateral sector of the immature pedicle valve. G, anterior sector of
the mature pedicle valve. Abbreviations: a.r., articulatory ridge; l.a.s., lateral adductor muscle
scar; m.c., marginal crest; /?./., primary layer; s.p.r., sub-peripheral rim; and t., tubercle.

64

PALAEONTOLOGY, VOLUME 32

on the floor of the brachial cavities. With these exceptions, the remainder of the shell is composed of granular
calcite. This occurs in two principal layers: an outer, fine-grained layer including the luberculate sub-peripheral
rim (PI. 9, fig. 8; PI. 11, fig. 3); and an inner, much coarser layer with granules up to 5 /an across flooring
the brachial cavities (PI. 1 1, fig. 4; text-fig. 3d, e). Clearly, as in the case of the pedicle valve, the accretion
of the coarse granular material is associated with the thickening of the shell underlying the brachial cavities
as they increase in size through resorptive activity around the inner margin of the sub-peripheral rim (Baker
1970) as the valve increases in size. In contrast to the pedicle valve, the brachial valve is clearly endopunctate,
with the peripheral flange in particular being penetrated by numerous endopunctae.

DISCUSSION

Relationship with substrate

In almost all cases the peripheral flange of the brachial valve is damaged. Even in complete shells
the brachial valve is nearly always slightly impressed into the interior of the pedicle valve, so that
the flange is split away and lost, leaving the erroneous impression of a markedly inequivalve shell.
As noted, the pedicle valve margin slopes from the crest to the exterior (PI. 9, fig. 6). This inclination
is paralleled by the ventral deflection of the peripheral flange (PI. 9, fig. 2). The few pedicle valves
which have been found in situ are attached to shells cemented into hardground surfaces. In P.
dundriensis the large area of attachment, the thickened shell, the flattened dorsoventral profile, and
the very well-developed lateral adductor muscle scars, at first sight all seem to be indicative of life
in a high energy environment. However, this explanation is incompatible with the very delicate
peripheral flange which undoubtedly could not have survived long in high energy conditions. On
the other hand, the ridge in the brachial valve and complementing groove in the pedicle valve
suggests that precise articulation of the valves was essential. The organo-detrital debris associated
with the hardground may offer a clue. Together with broken shells, the matrix encloses gaping but

EXPLANATION OF PLATE 1 1

Figs. 1 -6. Pachymoorellina dundriensis ( Rollier). Stereoscan photomicrographs of gold-coated cellulose acetate
peels (except fig. 6) of sectioned specimens. 1, PB3255, section through the posterolateral margin of pedicle
valve showing an acicular crystallite tract in association with a strand of secondary fibres (left) (section
orientation— parallel with the area of attachment), x 1000. 2, high oblique section through acicular

crystallite tracts showing detail of their contiguous relationship (section orientation— horizontal, almost
perpendicular to the shell surface; section location— anterior sector, pedicle valve, close to the base of the
free ventral wall), x 1400. 3, PB3256, low oblique section through brachial valve showing the distribution
of coarse granular calcite in the brachial cavity (left), fine granular calcite in the sub-peripheral rim (centre),
and contact with the thin layer (right) of secondary fibrous shell which, in places, separates it from the
primary granular layer (section orientation— parallel with the plane of the commissure; section location -
posterolateral sector), x 600. 4, section showing detail of the junction between the fine granular calcite
of the sub-peripheral rim and the coarse granular calcite flooring the brachial cavities (section orientation
and location as in fig. 3), x 1600. 5, section showing detail of the fine granular calcite of the sub-peripheral
rim and the junction with the almost continuous layer of secondary fibrous shell in the posterolateral
sectors of the brachial valve (section orientation and location as in fig. 3), x 1300. 6, PB3255, transmitted
light photomicrograph of a cellulose acetate peel (3255/14) through a pedicle valve showing the distribution
and gross mosaic of the acicular crystallite tracts in high oblique section through the free ventral wall;
because of their splayed arrangement, small areas of component crystallites will be sectioned transversely,
appearing as darker patches (arrowed) (section orientation as in fig. 2; section location as in fig. 2, but
approximately 0-5 mm below the lip of the valve), x 75.

Fig. 7. Thecidiopsis tetragona (Roenrer). Musee Geologique, Lausanne, 42533. Transmitted light photomicro-
graph of a cellulose acetate peel (42533/12) of a section through the free ventral wall of a pedicle valve
showing the gross mosaic of the acicular crystallite tracts in high oblique section and its close resemblance
to the gross mosaic of P. dundriensis (section orientation— parallel with the commissural plane; section
location —anterior sector, close to the attachment scar), x 70.

PLATE 11

BAKER, Pachymoorellina, Thecidiopsis

66

PALAEONTOLOGY, VOLUME 32

still articulated bivalves and unbroken juvenile echinoid coronas. The assemblage indicates rapid
accumulation with limited transport and little sorting, possibly as a result of storm activity,
followed by quiet conditions. It seems probable that P. dundriensis subsequently colonized suitable
niches. The growth profile and the inclination of the peripheral flange relative to the commissural
plane is such that any ‘snapping’ action of the brachial valve would ensure the expulsion of a
water current downwards towards the substrate. In this way the extraordinary development of the
adductor muscle field can be reconciled with a functional requirement. It is envisaged that P.
dundriensis , once established in fortuitous niches, employed a snapping valve action to waft away
fine sediment or to discourage the unwelcome spatfall of competitors in the vicinity of the shell.
The anterior wall of mature pedicle valves often shows exposed segments of border (text-fig. 3g)
indicating mantle retraction. This is followed by accretion of granular primary shell on to the
exposed tuberculate surface until the valve profile has been re-established. Such discontinuities
may reflect minor damage to the edge of the brachial valve’s peripheral flange.

Relationship with other species

Jurassic. As is the case with so many other micromorphic species described in the mid-nineteenth
century, confusion exists regarding type specimens. Careful comparison of Bath Geology Museum
material with the figures in Moore’s (1854) paper leaves no room for doubt that Moore based his
original description on four paratypes comprising a complete shell, a pedicle valve, and two
brachial valves. Although Moore identified most of the essential morphological characters of the
species, he unfortunately confused his specimens with the Liassic Thecidea Bouchardi Davidson,
1851. The error was later noticed (Rollier 1915) and the species T. (D.) Dundriensis Rollier, 1915
was erected to separate the Bajocian specimens. The distinctive ‘lame circumpalleale papilleuse’
(Rollier 1915) may now be identified as the granulose sub-peripheral rim, clearly seen on some
Crickley specimens. Subsequently (Pajaud 1963) the species was assigned to Moorellina dundriensis
(Rollier). In his comprehensive monograph, Pajaud (1970) overlooked the existence of the Moore
paratypes in Bath. He concluded correctly that a holotype had never been designated and selected
a brachial valve from the Paris Museum collections to serve as a neotype. Unfortunately, since
the type material was never lost and Moore was in error only over the name, the neotype
designation must be regarded as invalid and the registered specimen (MNHN IP B44600) relegated
to hypotype status.

In comparison with contemporaneous Middle Jurassic monoseptate forms, P. dundriensis , with
its large empty brachial cavities (without brachial lobes) and grossly thickened pedicle valve, has
always rested uncomfortably alongside the more conventional morphologies of M. granulosa
(Moore), M. duplicata (Moore), and M. dubia (D’Orbigny). These species, with their well-developed
brachial lobes, reduced areas of attachment, and relatively thin-walled pedicle valves, also show a
marked distinction at the level of shell microstructure. Investigation (Baker 1970) of the shell
microstructure of M. granulosa revealed the presence of a secondary layer of normally fashioned
fibres. Although the secondary shell fabric was modified by the presence of tubercle cores and the
effects of shell resorption, it was found (Baker 1970, text-fig. 6, p. 91) to form a complete lining
in both valves. A similar, fully developed secondary fibrous layer was subsequently identified
(Williams 1973) in the Liassic M. bouchardi , M. deslongchampsi (Davidson), Eudesella mayensis
(Eudes-Deslongchamps), and the Oxfordian Rioultina ornata (Moore). Recently (author’s unpub-
lished work), fully developed secondary fibrous shell layers have been found in the Aalenian M.
duplicata and M. dubia , and also in the Bathonian R. triangularis (D’Orbigny). Nothing remotely
resembling the microstructure of P. dundriensis has been discovered in any of the above species.
The fibrous tubercle cores (Baker 1970, pi. 21, fig. 5) seen in the free ventral wall of the pedicle
valve of M. granulosa are composed of conically arranged aggregates of normal fibres. The
microstructure of M. prima Elliott has not been investigated owing to the unavailability of material
for sectioning. In addition to the moorellinin species, the monoseptate Lower Jurassic davidsonellin
Davidsonella must be considered. As Williams’s comprehensive (1973) survey has shown, David-
sonella sinuata (Eudes-Deslongchamps), in addition to its concavo-convex adult profile and

BAKER: JURASSIC THECIDEIDINE BRACHIOPOD

67

well-developed brachial lobes, also has a fully developed secondary fibrous shell layer in both
valves.

On morphological and microstructural evidence, therefore, a genetic link between P. dundriensis
and any described monoseptate species is not easily established. As mentioned earlier, the detailed
investigation of P. dundriensis was prompted by the discovery that the polyseptate Mimikonstantia
sculpta developed from an early juvenile monoseptal phase and, by implication, was descended
from a monoseptate ancestor. However, although both species are without brachial lobes and
possess an inner granular shell layer, the suppression of the secondary fibrous shell is much more
pronounced in P. dundriensis where, as described, secondary fibrous shell is restricted in its
occurrence in the brachial valve and even more localized in its distribution in the pedicle valve. In
M. sculpta , the secondary fibrous layer, although reduced in thickness and penetrated by granular
calcite tubercle cores (Baker and Elston 1984, pi. 71; text-fig. 4) is retained as a continuous layer
in both valves. Nothing resembling the acicular crystallite tracts of P. dundriensis has been seen in
M. sculpta , although a capability apparently enjoyed by both species was thickening of the shell
by rapid deposition of granular calcite, around the anterior of the juvenile median septum in M.
sculpta and near the base of the free ventral wall in P. dundriensis. Clearly, although close to the
same line of descent, the contemporaneous occurrence of the two species and the higher degree of
suppression of secondary fibrous shell precludes the possibility of P. dundriensis being ancestral to
M. sculpta. However, the view that they share a common plexus of descent is strengthened by the
evidence obtained from certain Lower Cretaceous species.

Lower Cretaceous. The relationship between the Aalenian M. sculpta and the Lower Cretaceous
Thecidiopsis tetragona (Roemer) was considered by Baker and Elston ( 1984, p. 789). A phylogenetic
link was postulated on the basis of morphology, ontogeny, and shell microstructure. The elucidation
of the shell microstructure of P. dundriensis enables a remarkable similarity in the microstructure
of the pedicle valve of P. dundriensis and T. tetragona to be demonstrated. In both species the
anterior and anterolateral sectors of the pedicle valve are characterized by acicular crystallite tracts
(PI. 11, figs. 6 and 7) which are barely distinguishable from each other. The only discernible
difference is that in T. tetragona the tracts, or so-called ‘Fir-tree structures’ of Smirnova (1979),
often have a central axis containing much larger crystallites. It is not proposed to offer a detailed
comparison here as a critical revision of the microstructure of the shell of T. tetragona is to be
published separately. Additionally, in P. dundriensis the distal splaying of the acicular crystallites
(PI. 10, fig. 5) is almost identical (Baker and Laurie 1978, pi. 62, cf. fig. 8) with the pattern seen
in the Aptian Bifolium faringdonense (Davidson).

Recent. The very detailed investigation (Williams 1973) of the monoseptate Thecidellina barretti
(Davidson) revealed the presence of tubercles in the brachial valve which showed (Williams 1973,
pi. 43, fig. 25) the characteristic development of ten to twelve secondary fibres. Larger but otherwise
almost identical structures (PI. 10, fig. 1 ) composed of thirty to fifty fibres are found in the brachial
cavities of P. dundriensis.

Phytogeny

In considering thecideidine phylogeny, Pajaud (1970) noted a rectilinear evolutionary pattern with
offshoots in the Lacazella clan and a discontinuous zigzag progression in the Thecidellina clan.
Although the database needs further extension, a more rectilinear pattern of evolution in the
Thecidellina group sensu Williams (1973, fig. 100, p. 468) is beginning to emerge. Essentially, it
seems that two plexi of descent are involved: one, including Moorellina, Eude sella, and Rioultina
exhibits retention of a fully developed secondary fibrous layer in both valves; the other, with its
early representatives as yet little known but including P achy moorellina. Bifolium , Mimikonstantia ,
Thecidiopsis, and Thecidellina , exhibits a secondary fibrous layer suppressed to a greater or lesser
extent and a continuity of organization of microstructural features (such as acicular crystallite
tracts and characteristic secondary fibre segregations) traceable over long periods of time.

68

PALAEONTOLOGY, VOLUME 32

CONCLUSIONS

The present study further illustrates the recognized (Nekvasilova 1967, p. 128) inherent weakness
of thecideidine taxonomy based on the morphology of separated valves in which delicate structures
such as brachial lobes and peripheral flanges are so easily damaged, particularly as convergent
evolution (Baker 1984) appears to have played such an important role in the emergence of
representatives of the suborder.

It is remarkable to find the shell microstructure of an Aalenian monoseptate species so closely
paralleled in a polyseptate species of early Cretaceous age. Also, the discovery of a suppressed
secondary fibrous shell mosaic in M. sculpta , its correlation with T. tetragona , and, via ontogeny,
the probability of its descent from a monoseptate ancestor now assumes greater significance. It
appears that the Thecidellina group had its origin in a Pachymoorellina- type ancestor in the Lias.
The search for this must continue, however, since all the described contenders have a fully developed
secondary fibrous shell layer. It is concluded that the Moorellina plexus, already diverged from the
main line of descent by the Lias, did not survive the Jurassic. Also, the new evidence substantiates
the view that the main line of descent of the Thecidellina plexus, via Pachymoorellina-type stock,
probably followed a rectilinear trend similar to that exhibited by the Lacazella group.

Acknowledgement . I thank Mr S. A. Taylor (Derbyshire College of Higher Education) for help with the
preparation of the plate figures.

REFERENCES

baker, p. g. 1970. The growth and shell microstructure of the thecideacean brachiopod Moorellina granulosa
(Moore) from the Middle Jurassic of England. Palaeontology , 13, 76-99.

— 1984. New evidence of a spiriferide ancestor for the Thecideidina (Brachiopoda). Ibid. 27, 857-866.

— and elston, d. g. 1984. A new polyseptate thecideacean brachiopod from the Middle Jurassic of the
Cotswolds, England. Ibid. 777-791.

and laurie, k. h. 1978. Revision of the Aptian thecideidine brachiopods of the Faringdon Sponge

Gravels. Ibid. 21, 555-570.

moore, c. 1854. On new Brachiopoda from the Inferior Oolite of Dundry. Proc. Somerset archaeol. nat. Hist.
Soc. 5, 107 128.

nekvasilova, o. 1967. Thecidiopsis (Thecidiopsis) bohemica imperfecta n. subsp. (Brachiopoda) from the
Upper Cretaceous of Bohemia. Sbor. geol. Ved. Praha, 9, 1 15 136.
pajaud, d. 1963. Note sur les Thecideidae (Brachiopodes) jurassiques. Bull. Soc. geol. Fr. (7) 5, 995 1000.

— 1966. Problemes relatifs a la determination des especes chez les Moorellininae (Thecideidae, Brachio-
podes). Ibid. 8, 630-637.

— — 1970. Monographic des Thecidees (Brachiopodes). Mem. Soc. geol. Fr. 112, 1-349.
rollier, l. 1915. Synopsis des Spirobranches (Brachiopodes) jurassiques celto-souabes. Mem. Soc. pal. suisse,
41,45-56.

Smirnova, t. n. 1979. Shell microstructure in Early Cretaceous thecidean brachiopods. Paleont. J. 13, 339-
344.

williams, a. 1973. The secretion and structural evolution of the shell of thecideidine brachiopods. Phil.
Trans. R. Soc. B264, 439-478.

Typescript received 27 January 1988
Revised typescript received 6 May 1988

p. G. baker

Derbyshire College of Higher Education
Kedleston Road
Derby DE3 1GB

A NEW MITRATE FROM THE UPPER ORDOVICIAN
OF NORWAY, AND A NEW APPROACH TO
SUBDIVIDING A PLESION

by a. j. cr ask e and R. p. s. Jefferies

Abstract. This paper reconstructs, describes, and places systematically the mitrate Barrandeocarpus norvegicus
sp. nov. from the Upper Ordovician (Hirnantian stage, Ashgill series) of Rambergoya in Oslo Fjord, Norway.
This new species is the first mitrate described from Norway and a stem-group craniate in the plesion of
Mitrocystella. To locate the species within its plesion in an objective manner, without using traditional
categorial ranks, a number of new terms are proposed: a scion is a monophyletic group comprising a crown
group and an adjacent crownward part of a stem group. A scion ought to be named after its basal plesion.
Within a plesion, a first order apical group is a small monophylum (ideally a pair of sister species) further
removed from the stem lineage than are any other species of the plesion, i.e. separated from the stem lineage
by a greater number of phylogenetic segments (= species). The first order apical lineage is the direct line of
descent leading from the stem lineage to the apical group. A first order paraplesion comprises all those
members of a plesion which are equally closely related to the first order apical group. A first order parascion
is a monophylum containing the first order apical group and one or more adjacent paraplesions, and should
be named after its basal paraplesion. When a first order paraplesion contains several known species, it should
be possible to recognize a second order apical group, a second order apical lineage, etc. and so on with still
higher orders as the cladogram becomes more complex.

In these terms, Barrandeocarpus norvegicus is placed in the plesion of Mitrocystella in its own first order
paraplesion. This is less apical than the parascion of Ateleocystites guttenbergensis (i.e. the Anomalocystilida
in conventional terms) and more apical than the paraplesion of Barrandeocarpus jaekeli Ubaghs.

The locomotory cycle of B. norvegicus , crawling rearwards through the mud pulled by its tail, is
reconstructed. The internal features of the head of B. norvegicus are similar to those of Placocystites forbesianus
de Koninck in most respects, but show indications never seen before in mitrates, of the ventral surface of
the hypophysis.

The aims of this paper are to reconstruct, describe, and place systematically the mitrate
Barrandeocarpus norvegicus sp. nov. from the latest Ordovician of the Oslo region, Norway, and
to consider how a plesion should be subdivided.

The material on which this species is based was collected in July 1978 by Dr L. R. M. Cocks,
of the Department of Palaeontology, British Museum (Natural History) and all of it is preserved
at the BMNH. It was reconstructed, under the supervision of R. P. S. Jefferies, by A. J. Craske
when working as a vacation student at the BMNH in the summer of 1985. The discussion of
phylogenetic methodology in this paper, with particular reference to the placing of fossils within
a plesion, is the work of R. P. S. Jefferies as also is the discussion of the systematic position of
the species and of its locomotion. The account of the stratigraphy is by Dr L. R. M. Cocks.

PHYLOGENETIC METHODOLOGY

The cladistic approach to fossils is a matter of dispute. In this section we shall first discuss the
cladistic treatment of extant organisms and then argue that fossils, though they should be inserted
into the same system as recent organisms, ought to be treated differently from these. Concerning
recent organisms, we shall argue that paraphyletic groups are always regrettable, discuss why this
is so, and advocate the abandonment of traditional categorial ranks (families, orders, etc.).

IPalaeontology, Vol. 32, Part 1, 1989, pp. 69-99, pis. 12-13-1

© The Palaeontological Association

70

PALAEONTOLOGY, VOLUME 32

Concerning fossils, we assert that fossil groups must be defined, as well as recognized, on the basis
of their features and that paraphyletic groups are necessary. We also distinguish between stem
groups and stem lineages, discuss the division of stem groups into plesions, propose a standardized
method for naming monophyletic groups that comprise a crown group and a crownward part of
a stem group, and discuss how plesions may themselves be subdivided objectively.

Our systematic viewpoint is fundamentally that of Hennig (1965, 1966, 1969, 1987) but has been
influenced by Ax (1984, 1985, 1987), by Patterson and Rosen (1977), and by the penetrating
philosophical essay of Griffiths (1976). We have also been affected by Fordham (1986) and
Willmann (1985) who have discussed, in particular, the nature of fossil species and how such
species should be treated when the fossil record is excellent.

Fossils differ from recent organisms, cladistically speaking, in two basic ways: 1 , recent organisms
are far better known than fossils, both as to percentage of species known and as to what is known,
or knowable, about each species; and 2, recent organisms are accurately contemporaneous, whereas
fossils differ in age. It is these two contrasts, but particularly the second, which imply that fossils
cannot be systematized in exactly the same way as recent organisms. The fact that recent organisms
are far better known than fossils makes them the starting point for systematic endeavours, though
groupings based initially on recent organisms may be confirmed, modified, or sometimes even
confuted, by information from fossils.

The exact contemporaneity of recent organisms, whereas fossils differ in age, has subtler
implications. For, according to the biological species concept (Mayr 1963), a species is a population
of interbreeding individuals separated from other such populations by mechanisms which prevent
breeding. This implies, as Willmann (1985) has pointed out, that species owe their separate existence
to barriers with respect to contemporaneous species, not with respect to species which existed in
the past. This again implies, since an interspecific barrier to breeding normally works in two
directions, that at a speciation event two new species arise, not one. In other words, a species is
an undivided segment of the phylogenetic tree which originates at one speciation event and ceases,
either by dissolution at the next speciation event which affects it, or else at extinction. If the present
time can be taken as infinitesimal in duration, it follows that no extant species can be ancestral to
any other extant species. This is the fundamental reason why a complete system of extant species
will include nothing but monophyletic groups (= monophyla in Ax’s convenient usage; 1984,
p. 15) and single species. Nothing, however, prevents past species from being ancestral to other
past species or to recent species. Indeed, every species must be descended from earlier species, and
the fact that ancestor-descendant relationships are difficult to recognize does not lessen the certainty
that they existed. It follows that any complete system containing extinct species must accommodate
ancestors and must therefore include groups or species ancestral to non-members. Such groups
are by definition paraphyletic (paraphyla of Ax 1984, p. 32). Some authors, such as Ax himself
(1984, p. 209), have sought to systematize fossils without using paraphyla but all such attempts
are doomed— the paraphyla are covertly there though not acknowledged.

Why ought the system of recent species to be constructed entirely of species and of monophyla?
To illustrate the reason, text-fig. 1 shows the phylogenetic tree of a group of five extant species (a
to e) and records the evolutionary origin of morphological features 1 to 5. It also shows two
different ways, among many, of systematizing the five species— in text-fig. 1 a the monophylum
[c + d + e], characterized by feature 3, is distinguished from the paraphylum [a + b], characterized
by feature 1 and by the primary lack of feature 3. In text-fig. 1 b, on the other hand, the group
[a + b + c + d + e], itself a monophylum characterized by autapomorphy 1, is recognized to contain
four smaller monophyla, characterized respectively by features 2, 3, 4, and 5. Assuming that the
phylogenetic relationships of the species in text-fig. 1 are completely known, then the system shown
in text-fig. I a implies an arbitrary decision. It implies, namely, that the origin of feature 3 is more
important than that of 2, 4, or 5— important enough to have a group based upon it. The system
shown in text-fig. 16, on the other hand, is in no way arbitrary. It merely involves historical
reconstruction, i.e. the placing of features I to 5 in their correct sequence of origin, on the basis
of their observed distribution among recent organisms. In other words, features are commensurable

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

71

text-fig. I . The non-objectivity of para-
phyla in the systematization of extant
organisms, a, the extant species a-e are
placed in a monophylum [c e] and a para-
phylum [a b], on the grounds that novelty
3 is ‘more important than' novelties 2 or 4.
b , an objective systematization using mono-
phyla only, c, an objective systematization
based on ignorance of the relationships of
species a and b.

paraphylum monophylum

not in themselves, but only because they arose in the shared dimension of time. A system of
monophyla for recent groups is based on reconstructing the origin of successive features in time
and, as such, is an objective hypothesis which may be either right or wrong. Arbitrary and
subjective decisions about the relative importance of features are simply not required for such a
system.

Ignorance brings its own form of objectivity. If, in text-fig. la, the distributions of features 2,
4, and 5 were not known among recent organisms, it would then be legitimate, as a provisional
measure, to divide the group [a + b + c + d + e] into the monophylum [c + d + e] and the paraphylum
[a + b]. However, under these circumstances, [a + b] would not have been recognized as a
paraphylum, for the best available cladogram would be as shown in text-fig. lc. Gauthier (1986,
p. 8) has referred to such groupings, doubtful whether monophyletic or paraphyletic, as metataxa.

The system of extant organisms is necessarily formed from inter-nested groups— of older more
inclusive monophyla containing newer less inclusive monophyla. Each monophyletic group with

72

PALAEONTOLOGY, VOLUME 32

text-fig. 2. The extant sister groups I
and 2 with their respective latest common
stem species z and y, their shared latest
common stem species x and autaponror-
phies (1-7) acquired in the respective
stem lineages.

other extant species, more closely
related to it than is any other among extant organisms. These were referred to by Hennig (1966,
p. 139) as the sister group or sister species of the first-mentioned group, while Ax (1985, p. 16)
has proposed the term adelphotaxon to cover both sister group and sister species. Every
monophylum with extant members will have had in the past a latest common stem species. Every
pair of adelphotaxa will itself form a monophylum with its own latest common stem species. A
monophylum is recognized as such by possessing advanced features of its own — these are the
autapomorphies of the monophylum and, as such, were not present in the latest common ancestor
shared with the adelphotaxon. The line of descent exclusive to an extant monophyletic group and
in which the autapomorphies of this group were acquired can be called the stem lineage (Stammlinie
of Ax 1984, p. 6, though Ax’s use of this term is somewhat ambiguous when applied to fossils, as
discussed below). At a different level of analysis, the autapomorphies of a monophylum, by
definition acquired in the stem lineage of this monophylum, are synapomorphies of the two largest
adelphotaxa within the monophylum.

Text-fig. 2 makes this usage of the term ‘stem lineage’ more explicit. 1 and 2 are sister groups
with extant members. All extant members of 2 share a latest common stem species y which is not
shared with any other extant organism. Similarly, all extant members of 1 share exclusive descent
from their latest common stem species z. And the two adelphotaxa together combine to form a
larger monophyletic group, all members of which share an exclusive descent from their latest
common stem species x. The stem lineage of 2 will run from x to y (though in our usage it excludes
these two species populations). Features 2, 3, and 4, acquired in the stem lineage of group 2, will,
if retained in y, be autapomorphies of group 2. Likewise features 5, 6, and 7, acquired in the stem
lineage of group 1, between x and z, will be autapomorphies of group 1. And feature 1, acquired
in the stem lineage of group [1 +2], will be an autapomorphy of group [1 +2] and a synapomorphy
of groups 1 and 2.

Traditional categorial ranks cannot objectively be assigned to the groups in a phylogenetic
system. The fundamental reason for this is that adelphotaxa will nearly always differ in the number
of hierarchical levels which exist within them. Thus text-fig. 3 shows the phylogenetic trees of sister
groups 1 and 2. These contain respectively two and seven species, while one hierarchical level exists
in 1, as against four such levels in 2. If we attempt to apply to such a non-uniform ‘truncated’
hierarchy (Griffiths 1976) a traditional series of categorial ranks (subgenus, genus, subfamily,
family, superfamily), should the group comprising [1 +2] be regarded as a genus, since it contains
two hierarchical levels with respect to the species in 1? Or should it be regarded as a superfamily,
since it contains five hierarchical levels with respect to two of the species in 2? (There are similar,
though smaller, discrepancies with respect to each nodal point within 2.) Any system which aims
at assigning categorial rank to all the monophyla which exist will produce endless paradoxes of
this sort. The only solution, explicitly advocated by Ax (1984, 1987), and adopted also, for example.

1 2

extant members will have one other extant monophylum, or

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

73

text-fig. 3. The arbitrariness of categorial rank (see text).

by Gauthier (1986, p. 8), is to abandon any formal system of categorial rank. This proposal departs
uncomfortably from tradition, but also stops a well-spring of pointless quarrelling.

As concerns fossils, we have discussed the treatment of these elsewhere (Jefferies 1986, chapter
1; Jefferies et al. 1987) and we elaborate these earlier proposals here.

A fundamental difference between any group known only from extant organisms, on the one
hand, and one known to include fossils, on the other, can also be illustrated from text-fig. 3. It
concerns the distinction between definition and recognition. Thus group 2, for example, as to its
recent members, can be defined exactly in terms of phylogeny— it comprises, namely, all extant
descendants of the latest common stem species y. On the other hand, group 2 can be recognized
as a monophyletic group because all its extant members have the autapomorphies 2, 3, and 4, the
distribution of which among recent organisms is coextensive with group 2 (apart from secondary
loss). Where fossils are concerned, however, the situation is more complicated. For features 2, 3,
and 4 did not arise simultaneously but originated in that sequence in the stem lineage of group 2.
There were forms more closely related to the extant members of 2 than to those of 1 (i.e. descended
from x but not belonging to group 1) which primitively lacked features 2, 3, and 4; there were
others which had 2 but primitively lacked 3 and 4; and others which had 2 and 3 but primitively
lacked 4. (These forms would not all have been members of the stem lineage of 2, since some of
them would have belonged to the side branches from that lineage.) It follows that groups containing
fossils, unlike purely recent groups, must be defined, as well as recognized, in terms of their
features, since the distribution of features reflects the phylogeny only inexactly.

As to the distinction between crown group and stem group, text-fig. 4 shows the phylogenetic
trees for the sister groups 1 and 2 with extant representatives. Following Hennig (1969, 1983),
there are two obvious delimitations of group 2, one narrow and one broad. The narrower delimits
the group as the latest common stem species of 2, plus all descendants of that stem species. This
delimitation defines the *group of Hennig, which we call the crown group. The wider delimitation
contains all those forms which are more closely related to the extant members of 2 (i.e. share a
more recent common ancestor with them) than to the extant members of 1. This wider delimitation
defines the total group of Hennig. By subtracting the crown group of 2 from the total group of 2,
a paraphyletic assemblage remains called the stem group of 2. Through the stem group of 2 runs
the stem lineage of 2 in our sense. This is the direct lineage of ancestors and descendants leading
from x (the latest common ancestor of [1+2]) to y (the latest common ancestor of the extant
members of 2), but excluding both x and y. In this lineage, the autapomorphies of 2 would have
been acquired, and, in this respect, it corresponds to the stem lineage in the sense used by Ax

74

PALAEONTOLOGY, VOLUME 32

text-fig. 4. Stem group, crown group, nodal group, and plesion.

(1984, 1987) when discussing extant groups with no fossils. We should point out, however, that
Ax, when discussing fossils, also uses the term ‘stem lineage’ or ‘ Stammlinie' in a broader sense
exactly equivalent to Hennig’s ‘stem group’ (Ax 1984, 1987, chapter J). Here we explicitly exclude
this broader sense of the term ‘stem lineage’.

A plesion, for us, comprises all those members of a stem group which, so far as can be discerned,
are equally closely related to the crown group. It therefore includes all those forms which possess,
or have secondarily lost, one novelty acquired in the stem lineage, but primitively lack the next
recognized novelty acquired in the stem lineage. For us, therefore, a plesion is, in principle, a
paraphyletic grouping since it contains part of the stem lineage and this will have been ancestral
to non-members of the plesion. This differs from the concept proposed by Patterson and Rosen
(1977) and espoused by Ax (1984; 1987, chapter J) who all imply that plesions are, in principle,
monophyletic and who thus have no place in their system for true members of a stem lineage.

Plesions, although in principle paraphyletic, can be defined and recognized objectively because
they are made as small as possible. Thus, as soon as it can be shown that some members of a
former plesion are more closely related to the crown group than others are, by demonstrating a
previously unnoticed novelty in the contained part of the stem lineage, then the former plesion
splits into two plesions, one more crownward than the other. In other words, the recognition of
plesions is objective since it is based on reconstructing, so far as possible, the complete sequence
of evolutionary novelties, as they actually happened in time, in the stem lineage.

A plesion can be delimited in terms of two monophyletic groups — the larger containing the
smaller (text-fig. 5). Both groups would have the same extension among recent organisms, and
would include the same crown group. Within the relevant stem group, however, they would differ
in extension, in that the plesion in question would be included in the larger group but excluded
from the smaller one. We propose that these partly fossil and partly extant monophyla should be
called ‘scions’. The name refers to the horticultural operation of grafting, in which a scion comprises
all parts of a branch distal to the point cut through by the gardener. We propose that each scion
should be named after the plesion included at its base. The reason for this proposal is that each
monophylum has a unique latest common stem species, and could, ideally, be named after this
stem species, if such could ever be identified (which it probably never can be). Naming a scion
after its basal plesion is the best practicable approach to naming it after its latest common stem
species. This procedure would standardize and simplify the naming of partly fossil monophyla.

The desirability of such standardization is exemplified by the recent work of Gauthier who has
published an extremely thorough cladistic analysis of the origin of birds (1986; text-fig. 6 herein).
In doing so, he implies a succession of nine plesions in the avian stem group. However, he does
not describe the situation in terms of a stem group with plesions. Instead, he restricts the term

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY 75

‘Aves’ to the crown group of the birds and sets up eight additional monophyla, each co-extensive
with the Aves in the recent fauna, but differing in extent in the fossil fauna. To each of these eight
additional monophyla he gives a name, not etymologically related to any particular member of
the group and often new. For example, Gauthier’s group Avialae differs from his group Ornithurae
only by including the plesion of Archaeopteryx. We suggest that the Avialae could conveniently
be referred to as the scion of Archaeopteryx , whereas the Ornithurae could perhaps be called the
scion of Hesperornis. This would greatly reduce the burden on the reader’s memory. The ‘scion of
Archaeopteryx' would not be the same as the ‘monophylum of Archaeopteryx' since the latter
expression would most naturally refer to a monophyletic group containing Archaeopteryx alone.

In naming a plesion, we ought not to use the name of any arbitrary member, nor even the name
first given to the plesion in the literature. Rather it should be named after the known form most
closely related to the stem lineage, and thus most closely related to the latest common stem species
of the relevant scion. The name of a plesion, or of a scion, would therefore not be fixed by priority
but would change as new forms, more closely related to the latest common stem species of the
scion, were discovered. (Perhaps the choice of ‘ Hesperornis ' as the scion name for Gauthier's
Ornithurae is for this reason unsuitable.)

We have proposed the term ‘nodal group’ (Jefferies et al. 1987, p. 432) for all those fossil
members of an extant monophylum which possess all the autapomorphies of the crown group,
but primarily lack any of the autapomorphies of any of the extant subgroups of the crown group.
In this sense, the nodal group includes the latest common ancestor of the extant members of a
group. We now extend the use of this term by speaking of the nodal group of a plesion to mean
all those members of the plesion which, so far as known features are concerned, do not differ from
the stem lineage. Whenever possible, a plesion should be named after a member of its nodal group.

Systematization within a plesion is the next problem to be considered and is especially relevant
to the fossil described in this paper. We shall not use traditional categorial ranks since, as already
pointed out, they cannot be assigned objectively. The problem, therefore, is to recognize and name
monophyla and objective paraphyla within a plesion, without using traditional ranks and without
creating names unnecessarily, i.e. the naming procedure for each included group should be related,
in a standard way, to the contents of the group.

The real phylogenetic tree of a plesion will consist of part of the stem lineage of an extant
monophyletic group, plus all side branches from that part of the stem lineage. If this real tree is

76

PALAEONTOLOGY, VOLUME 32

text-fig. 6. Monophyletic groups recognized in the evolution of birds.
Modified after Gauthier (1986, fig. 8).

big and complicated, there will be a pair of sister species within it further removed from the stem
lineage than any other such pair, i.e. separated from the stem lineage by a greater number of
phylogenetic segments (= species). We propose to call this pair of species the real first order apical
group of the plesion (first apex for short). The lineage leading from the stem lineage to the first
apex can be called the first order apical lineage. All monophyla which include the first apex (text-
fig. 7) can be called first order parascions (this name should suggest a scion cut off a side branch).
There will have been successive evolutionary novelties in this apical lineage and, on the basis of
these, the plesion can be divided into first order paraplesions (text-fig. 8), all members of any such
paraplesion being, so far as discernible, equally closely related to the first apex. A first order
paraplesion can equally be defined as comprising all those members of a plesion which are contained
in one first order parascion, but not contained in the included next smaller recognizable first order
parascion. So long as paraplesions are made as small as possible they are objective. This means
that, if it can be shown that some members of a former paraplesion are more closely related to
the relevant apex than other members, then that former paraplesion must be split into two
paraplesions, one more apical than the other. If this is done, a paraplesion, though paraphyletic,
is not arbitrary, depending, as it does, on the most complete possible reconstruction of evolutionary
history.

A first order paraplesion may itself contain several species and, if the branching is complicated,
a pair of these species will be further removed from the first order apical lineage than is any other
species of the paraplesion. This pair will constitute a second order apical group. Within the first
order paraplesion it will therefore be possible to recognize second order paraplesions arranged in
order of relationship to the second order apex. With any plesion of complicated and abundant
branching there will be only one first order apex, but several second order apices. Similarly, a
second order paraplesion may contain a third order apex and be divisible into third order
paraplesions, and so on.

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

77

1st order apical group,

text-fig. 7. a, first order apical group and parascions in the real tree
of a plesion. The curved arrows point to recognized evolutionary
novelties, b , subdivision of a first order paraplesion. Compare text-
figs. 8 and 9.

We suggest that parascions ought to be named after their least apical paraplesion, and that this
ought to be named after that known included species which is most closely related to the relevant
apical lineage. (As to what 'relevant’ means here, in dividing a plesion into first order parascions
and first order paraplesions, the apical lineage leading to the first order apical group is the relevant
lineage.) This is analogous to the procedure suggested above for naming scions and is based on
the same consideration— that such a procedure is the best practicable approximation to naming a
monophylum after its latest common stem species.

In palaeontological practice, we do not know the real phylogenetic tree of a plesion. However,
if more than two species are known within a plesion, then it is likely that they can be placed on
a cladogram, which we see as a reconstructed phylogeny formalized by placing at the ends of

78

PALAEONTOLOGY, VOLUME 32

1st order apical group

a

text-fig. 8. a , first order apical group, paraplesions, and nodal group
in the real tree of a plesion. b , subdivision of a first order paraplesion.

branches all the known forms considered. In this cladogram (text-fig. 9), a pair or a small number
of species (probably sister species but possibly ancestors and descendants of each other) will be
further from the stem lineage than the other known species are, i.e. separated from the stem lineage
by a greater number of cladogram segments than any other known species is. This pair or small
number of species will be the provisional first order apex of the plesion and the other species can
be arranged in provisional paraplesions, each one of which will contain all species which, so far
as discernible, are equally closely related to this first order apex. Thus, at any one stage of study,
the known members of a plesion can objectively be arranged in a best possible approximation to
the real phylogenetic tree. The best possible cladogram will change as new species are discovered
and studied. It may be that the new species will lie distal to the first apex as originally set up, the
members of which would thus become a first order paraplesion. Or it may be that numerous new
members of what was originally a single paraplesion will be discovered, so that the recognized first
apex shifts to that former paraplesion. Always, however, if a cladogram can be constructed, it will
be possible to arrange the known members of a plesion in terms of apices and paraplesions, of

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

79

text-fig. 9. Relation between: a, a partly known plesion (in
which the species a-e have been discovered); and b, its
cladogram. The thick line in a represents the parts of the real
tree which are represented in the cladogram.

first, second, or nth orders, as suggested here. As study proceeds, the arrangement will become
steadily more complex and should converge on the topology of the real tree.

The naming of the constituent parascions and paraplesions of a plesion will change as the
cladogram changes and will follow the rule already suggested— that a monophylum will be named
after that known constituent species which is most closely related to the latest common stem
species. Thus the names will change as the cladogram does but should, like the cladogram, converge
on a stable condition corresponding to the true tree.

In describing and naming the parts of a river system, geographers use a procedure analogous
to the one suggested here for systematization within a plesion. They consider, namely, the longest
watercourse as being one river (first order apical lineage) and regard other watercourses flowing
into it as different rivers (second order apical lineages). Traditional hierarchical ranks would
suggest that the phylogenetic system is uniformly hierarchical like an army, with equal numbers
of hierarchical levels in all its constituent groups of the same rank. In fact it is more like a drainage
pattern, in which different rivers have different numbers of tributaries. Any sound approach to

80

PALAEONTOLOGY, VOLUME 32

describing the phylogenetic system must reflect the orderliness, and disorderliness, which evolution
has produced in Nature. We hope that what we suggest here will contribute to that aim.

Thus the cladistic approach to fossil groups will, in one sense, be the same as to recent groups —
whenever a monophylum can be recognized, it should be named and incorporated in the system.
However, unlike recent groups, usable fossil groups must be defined, as well as recognized, by
their features. Also, fossil studies strive towards a different end point from recent studies. For,
with purely recent organisms, a system comprising nothing but monophyla and species is
theoretically attainable, but with fossils, simply because some of them will have been ancestors, it
is not.

SYSTEMATIC PALAEONTOLOGY
Superphylum: Deuterostomia Grobben 1908
Subsuperphylum: Dexiothetica Jefferies 1979
Phylum: Chordata Bateson 1886
Subphylum: Craniata Goodrich 1909
(Stem group of the Craniata)

Plesion of Mitrocystella Jefferies 1979
Paraplesion of Barrandeocarpus norvegicus herein
Genus Barrandeocarpus Ubaghs 1979
Species Barrandeocarpus norvegicus sp. nov.

Systematic position

This statement of phylogenetic position gives conventional categorial ranks (superphylum, etc.) for the higher
ranking groups, but, as already discussed, such ranks have no objective meaning and will probably come to
be abandoned (Ax 1984, 1987). The subphylum Craniata, as given here, is coextensive with the subphylum
Vertebrata as used in Jefferies (1986). We now adopt Janvier’s usage (1981) by which the term Vertebrata is
restricted to the group [Petromyzonida + Gnathostomata] and Craniata is used for the wider group of
[Myxinoida + Petromyzonida + Gnathostomata],

Material , horizon , and locality

All of the specimens described in this paper come from a single bedding plane 71 m above the top of the
‘Brown Sandstone’, which is a I -5 m thick, heavily bioturbated sandstone marking the local base of the
Langoyene Sandstone Formation. They were collected by Dr L. R. M. Cocks in July 1978. The locality is a
natural beach outcrop at the south-western end of Rambergoya in the Oslo Fjord, Norway. The locality
(text-fig. 10) is at grid reference NM 962394 and is 3-7 km south-south-east of Oslo Town Hall.

The bedding plane is not rich in fossils but the assemblage is dominated by the brachiopod Eostropheodonta
hirnantensis (M’Coy), with Barrandeocarpus norvegicus the only other fossil collected in more than single
specimens. The other fossils recorded are the brachiopod Hirnantial sp., Tentaculites sp., a colonial bryozoan,
and a crinoid ossicle.

Within the Brown Sandstone, however, and in the beds 2-3 m above it, there is a rich and diverse Hirnantia
fauna of Hirnantian (latest Ashgill) age (Brenchley and Cocks 1982, p. 796, locality 197) and other Hirnantian
faunas occur above the bedding plane with Barrandeocarpus norvegicus.

As to geographical conditions, in the late Ordovician the Oslo area was on the west of the palaeocontinent
of Baltica and lay at a latitude of about 30° S. (Cocks and Fortey, 1982). During the Hirnantian stage there
is much evidence of glacial deposits in and around the large Gondwana palaeocontinent which lay south of
Baltica, but there were contemporary warm-water limestones in the equatorial North American palaeoconti-
nent. It is likely, therefore, that Barrandeocarpus norvegicus lived and died in warm to temperate waters.

The material is all preserved in the British Museum (Natural History) and comprises remains of about
fifty individuals. Some of it is part and counterpart (signified by a and b at the end of the registration
number). Some of the slabs show more than one individual, in which case each individual was given a letter
after the registration number. In the following list, an asterisk indicates a morphologically informative
specimen.

The registration numbers are as follows: E29381a/n*, b *, c*, d* (holotype), e*,f, /*, m*\ E293816/g*, h*,
i,j, k*, n*\ E29383a; E29384/a, 6, c, d, e,f*, h, /*; E28385/a*; E63163/a; E63164o/a*; E631646/a*; E63165/«*,

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

81

text-fig. 10. Map to show the locality for Barrandeocarpus norveg-
icus sp. nov. Traced from the Norwegian 1:50 000 sheets for Asker
and Oslo. The numbers refer to the Norwegian national grid. The
crosses, taken from the grid, are 1 km apart.

b, c; E63166/a*; E63167/a*; E63168/6; E63170/a, b\ E63171/a, b, c, d*, e; E63172/a; E63173/a, 6; E63174/a*;
E63175/6; E63 1766/a*; E63 177/a*; E63 178/a.

The holotype (E29381 a/d) shows the most complete dorsal surface available.

Most specimens were preserved as calcite plates embedded in the rock, though the calcite has been removed
by acid preparation. There is some crushing and dislocation, but the good articulation of many of the
specimens suggests burial at, or immediately after, death.

Anatomical description

In this description, individual plates are labelled according to the system proposed in Jefferies and
Prokop (1972) and expanded in Jefferies and Lewis (1978, figs. 29 and 30). Plates given the same
letter or number in different species are thought to be homologous with each other. Four different
notations are used simultaneously:

1. Some anterior and posterior marginal plates are given lower-case letters (b, c, g, h, i, j, n,
and p) implying homology with plates in primitive mitrates and, in some cases, with the crownward
cornute Reticulocarpos hanusi.

2. Some marginal plates and all centrodorsal plates are given arabic numbers (1 to 12) implying
homology with mitrates closely related to Barrandeocarpus norvegicus.

3. The ventral plates are, so far as possible, given lower case roman numerals, implying
homology, in the first instance, with the specialized mitrate Placocystites forbesianus.

4. Four of the most anterior ventral plates are given Greek letters a to 3. These plates are
constant in B. norvegicus but cannot certainly be homologized with plates in any other mitrate.
These different notations are complicated and difficult to remember but text-figs, \ \a-e should
clarify their meaning.

In general form, Barrandeocarpus norvegicus has a box-like head, with convex and probably

82

PALAEONTOLOGY, VOLUME 32

text-fig. 1 1 . Barr ancle o car pus norvegicus sp. nov. Reconstruction of external features: a, dorsal aspect;
b , ventral; c, left lateral; d , anterior (the oral plates are purely reconstructional); e, posterior (tail omitted).

rigid dorsal and ventral surfaces. There is a projecting, but rounded, peripheral keel running along
the lateral margins of the dorsal surface.

The fore tail is anteriorly almost as wide as the posterior part of the head, being almost circular
at its proximal end. In dorsal aspect it narrows to where the styloid is situated. The hind tail is
usually somewhat longer than the head, although in one specimen (PI. 13, fig. 7) it is almost twice
as long as the head.

Dorsal surface of the head (text-figs. 1 1 a and 12). This is made up of fourteen plates, being three
centrodorsals and eleven marginals.

The anterior margin of the head is framed by plates b, n, and c. These make a fairly straight
edge, and constitute the upper lip of the mouth. There are three marginal plates along each side
of the head, plates 1, 2, and (3 or 4) on the left and 8, 7, and (5 or 6) on the right. These six plates
also form the lateral faces of the head.

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

83

text-fig. 12. Barrandeocarpus norvegicus sp. nov. Explanatory diagram
of Plate 12, fig. 2, showing plates in anterior part of dorsal skeleton.

The dorsal surface makes an angle of about 90°, or slightly less, with the lateral faces. Where
they meet, the dorsal surface slightly overhangs the lateral faces, giving a keel along the edge (PI.
12, figs. 1 and 5). Where the lateral and ventral faces meet there is no such keel, for the ventral
plates abut against the marginals.

Plates i and h adjoin the posterior edge of the head. They do not help to form the posterior
face.

There are three centrodorsal plates, one large one in a median position (plate 12) and two
smaller ones anterior to and left of this (plates 1 1 and 10, respectively). Plate 1 1 can be seen plainly
in several specimens (PI. 12, fig. 2) but plate 10 is harder to trace. Its existence is shown by a triple
junction on the median edge of plate 2 in one specimen (PI. 12, fig. 2). The exact size and shape
of plate 10 is uncertain. Plates 11 and 12 are almost bilaterally symmetrical but plate 10 lies to
the left, being the only external sign of asymmetry in the dorsal skeleton.

In number and general position of plates the dorsal surface is identical both to Ateleocystites
guttenbergensis as described by Kolata and Jollie (1982) and Barrandeocarpus jaekeli as described
by Ubaghs (1979). Differences are matters only of shape and proportion.

Ventral surface of the head (text-fig. 11/?). This surface is again convex, but less so than the dorsal
surface. It consists of about twenty-three plates which seem to have formed a rigid armour.

The plates of the posterior part of the ventral surface, forward to the transverse level of plate
xi, are standardized between specimens, as also are the ventral plates adjacent to the mouth.
Between these two regions, however, the plates vary considerably in number and position. All the
plates of the ventral surface seem to be tessellate, giving a flush surface (except plate xi which
protrudes somewhat). It is therefore likely that the entire ventral surface of the head was rigid
(except, presumably, that it could warp slightly posteriorly, so that the atrial openings could gape).
In having the anterior part of the ventral surface completely tessellate and completely rigid,
Barrandeocarpus norvegicus differs from B. jaekeli , where each plate of the anterior part of the
ventral surface overlaps the posterior edges of its anterior neighbours.

The two largest plates in the ventral surface of B. norvegicus are plates g and j, which together
form the posterior edge of the head in ventral aspect. Like plates i and h on the dorsal surface,
they are more convex than the rest of the ventral surface, so as to accommodate the fore tail.
There is a transverse ridge near the posterior edge of plates g and j (pi. 13, fig. 6). This, like a
similar ridge in Mitrocystites and Mitrocy Stella, defines the anterior margin of the posterior surface.
In B. norvegicus this surface is very small.

The other plates which appear to be standardized in the posterior part of the ventral surface
are p, i (roman numeral), vii, viii, x, and xi.

Plate xi (the ‘placocystid plate’ of Caster 1952) is very peculiar. It is rounded rather than angular
in outline (PI. 12, fig. 11), and slightly more convex ventrally than the surrounding surface. Also,
all other ventral plates have flat edges which meet at sutures perpendicular to the ventral surface,
but the edges of plate xi slope dorsalwards toward the centre of the plate, making it a truncated
cone with the internal surface smaller in area than the external surface. The edges of the adjoining
plates (viii, x, and three unnamed, irregular more anterior plates) are sloped to receive plate xi.

84

PALAEONTOLOGY, VOLUME 32

In related species plate xi is likewise markedly different from the other ventral plates (Jefferies
1984). In addition to the peculiarities already mentioned in B. norvegicus (accentuated ventral
convexity, non-angular outline, and truncated conical shape), there are two other distinctions, as
follows:

1. It lacks ornament when its neighbours are ribbed, as seen in Ateleocystites guttenbergensis
and B. jaekeli', this distinction seldom holds in B. norvegicus for, though plate xi is always ribless,
the neighbouring plates are usually ribless also.

2. It is very variable in size and does not enlarge as the head grows (as shown by an ontogenetic
series in Placocystites forbesianus (Jefferies 1984))— these distinctions may or may not apply to B.
norvegicus , since the known specimens do not vary enough in size to give a plausible ontogenetic
series.

The anterior edge of the ventral skeleton of B. norvegicus is, in all observed cases (PI. 12, fig.
8), made up of five plates arranged in bilateral symmetry. In text-fig. 1 1/? these have been labelled
a, /i, iv, y and S. These plates have a fairly prominent anterior edge while the parts nearest the
mouth are bent upward to form the lowest portion of the anterior surface of the head. By
comparison with A. guttenbergensis (which, as shown below, was more apical in the plesion) and
with B. jaekeli (which was less apical in the plesion), there were probably spike-like oral plates in
the lower lip of B. norvegicus , but no traces of these were observed. Text-fig. 11 d shows a
reconstruction of the anterior face of the head, including such oral plates.

The five plates which form the anterior edge of the ventral skeleton carry a series of external
projections on this edge. These projections are particularly prominent on the leftmost and rightmost
plates.

The ventral surface of B. norvegicus is in most ways similar in its constituent plates to those of
B. jaekeli and A. guttenbergensis. However, there is a morphological series in which the plates
become increasingly standardized and the surface increasingly rigid. This series passes from B.
jaekeli (flexible anteriorly and the least standardized in its plating) to B. norvegicus (rigid throughout,
standardized posteriorly and at the anterior margin) to A. guttenbergensis (rigid throughout and
standardized throughout).

A few cuesta-shaped ribs exist on the head plates of B. norvegicus. On the ventral surface they
are always found on plates g, p, and j and sometimes on plates immediately anterior to these
except for plate xi which, as in related forms, is never ornamented (PI. 12, figs. 6 and 1 1). On the
dorsal surface ribs are found on plates 1, i, h, and 8 (PI. 12, fig. 5) though there is always a medial

EXPLANATION OF PLATE 12

Figs. I 11. Barrandeocarpus norvegicus. Latex casts of specimens. 1, E29381a/d; x 5. Dorsal surface of
holotype. Most of the plates of the dorsal surface can be seen except 10. Note peripheral keel, cuesta-
shaped ribs, and the form of i and h near the tail insertion. 2, E29381a/c, d\ x 5. Anterior part of dorsal
surface. Note peripheral keel on plates 7 and 8. Plates 11 and 12 are visible and the evidence for the
position of plate 10 shows as an obtuse point (triple junction) on plate 2 (arrowed). Compare text-fig. 12.
3, E63162/6; x 5. Latex. Left side of head with peripheral keel, fore-tail plates and cerebral basin. 4,
E29381r///»; x 5. Dorsal ossicles and ventral plates of hind tail in right lateral aspect. 5, E29381o/c, d\
x 5. Specimen d (the holotype) overlies c, the posterior dorsal surface of which is seen at the top of the
figure. Specimen d exposes the dorsal surface and clearly shows i, h, 1, 8, and 12. The peripheral keel can
be seen on 8, together with cuesta-shaped ribs on 1 and 8. Anteriorly the dorsal surface is missing, allowing
the interior surfaces of a, b, iv, c, and d to be seen. 6, E2938 1 A; x 5. Ventral surface of head and ventral
aspect of styloid and hind tail. Note cuesta-shaped ribs on g, p, and j. 7, E29381fi/g; x 5. Fore, mid, and
hind tail. 8, E29384/i; x 10. Anterior part of the ventral surface of the head. Note the most anterior five
plates a, /l, iv, y, and <5 and the irregular plates behind them. 9, E29385/a; x2-5. Complete animal in left
aspect to show common observed posture. 10, E63 161 //?; x 7-5. Dorsal aspect of styloid. Note sharp ridge
connecting the two transversely elongate blades. 11, E63161//;; x 10. Latex of posterior part of ventral
surface of head. Note plate xi and cuesta-shaped ribs on plates viii, x, g, p, and j.

PLATE 12

CRASKE and JEFFERIES, Barrandeocarpus

86

PALAEONTOLOGY, VOLUME 32

ribless strip. Plates 1 and 8 also carry ribs on their lateral faces. Thus, cuesta-shaped ribs are
confined to the posterior third of the head but are there found ventrally, dorsally (except for the
medial strip), and laterally.

Comparing the ribbing with other species the dorsal distribution of ribs is similar to that of B.
jaekeli. However, on the ventral surface B. norvegicus has far fewer ribs than B. jaekeli , in which
the posterior two-thirds is ribbed. A. guttenbergensis has more ribs than B. norvegicus on both
surfaces, but fewer than B. jaekeli. The relative sparsity of ribs on B. norvegicus may be due, at
least in part, to the small size of the animal. For the head measures about 10 mm in length on
average (e.g. specimen E63 176/6 in PI. 13, fig. 6), whereas A. guttenbergensis is somewhat larger
(some individuals reach a head length of 25 mm according to Kolata and Jollie 1982) and B.
jaekeli is much larger (about 30 mm head length). This can be related to the situation in the
ontogenetic series of P. forbesianus where Jefferies (1984) noted that large specimens have many
ribs while small ones (presumed juveniles) have far fewer. In these juveniles the ribs were located
as in B. norvegicus— on the posterior right and left areas of the dorsal surface and the posterior
part of the ventral and lateral surfaces.

The tail As in all known mitrates, the tail has three distinct regions^ the fore, mid, and hind tail.

The skeleton of the fore tail is made up of six rings (PI. 13, fig. 8), each consisting of four plates
(right and left dorso- and ventrolaterals) sutured together to make the ring inflexible. Each ring
overlaps the ring posterior to it. The sutures in each ring are in the dorsal and ventral mid-lines
and left and right. Plate 12, fig. 7 shows the plates separated at these sutures.

The fore tail is almost circular in section where it joins the head and so makes up almost the
whole posterior aspect of the head. It narrows rearwards to about two-fifths of its anterior width
and the depth reduces only to about two-thirds of the anterior depth. Thus the distal end of the
fore tail is elliptical and laterally compressed in section.

The skeleton of the mid tail consists of the styloid and associated plates. As in all known mitrates
except the stem-group acraniate Lagynocystis (Jefferies 1973, 1986), the styloid seems to represent
two ossicles of the hind tail fused together and otherwise modified. It articulates in B. norvegicus
with two pairs of ventral plates, again as in all known mitrates except Lagynocystis.

The styloid (text-fig. 13) is complex in structure, with two transversely elongate blades dorsally,
connected by a median ridge (PI. 12, fig. 10). The anterior blade is the taller of the two and is
slightly wider than the proximal part of the hind tail. Anterior to this blade, the styloid decreases
in height and is overlapped by the posterior part of the dorsal region of the fore tail. The blade
itself is concave on its posterior face which is divided into two by the median ridge. This ridge is

EXPLANATION OF PLATE 13

Figs. 1-9. Barrandeocarpus norvegicus sp. nov. Latex casts and natural moulds. 1, E63166; x 10. Ventral
aspect of posterior part of natural internal mould. Note the posterior coelom, bounded anteriorly by a
transverse fissure in the rock, and the likely position of the hypophysis, showing as a swelling posteriorly
in the mid-line. 2, E63161u; x 10. Ventral aspect of natural internal mould. Note twin buttons on plate
xi and the two large buttons flanking this plate. Compare text-fig. 17. 3, E63177; x 10. Natural mould
of cerebral basin, imitating the brain. Compare text-fig. 18. 4, E63165; x 10. Ventral aspect of a contorted
natural internal mould of the head (same specimen as fig. 5). The transverse groove about one-quarter of
the length forward from the posterior end is the anterior boundary of the posterior coelom. 5, E63165;
x 10. Dorsal aspect of contorted internal mould (same specimen as fig. 4). Compare text-fig. 16.
6, E631766/«; x 10. Plates of the ventral surface preserved as calcite. Note plate xi, rounded in outline
and slightly protruding from the surrounding plates. Cuesta-shaped ribs on g, p, and j. Styloid shows in
ventral aspect. 7, E29381a/6; x 5. Latex showing twenty-seven segments of hind tail, lit from beneath.
8, E293816/«; x 5. Fairly complete left lateral aspect of head and part of tail, lit from beneath. Note fore
tail and styloid and plate d, showing its junction with plates b and (3 or 4) and with the rest of the
ventral surface. The upside-down individual is specimen k. 9, E63161/e; x 5. Posterior part of natural
internal mould of head, in dorsal aspect. Compare text-fig. 15.

PLATE 13

CRASKE and JEFFERIES, Barrandeocarpus

PALAEONTOLOGY, VOLUME 32

text-fig. 13. Barrandeocarpus norvegicus sp. nov. The styloid: a , left lateral aspect; b , posterior; c, dorsal;

d, left posterodorsal.

sharp-edged and continues rearward into the anterior face of the posterior blade of the styloid
which is also concave rearward.

The successive segments of the hind tail do not much decrease distally in width but lessen in
height (PI. 12, fig. 7). As in all other mitrates, the skeleton of each segment of the hind tail consists
of three calcite elements— two ventral plates and one dorsal ossicle. The ventral plates are joined
ventrally in the mid-line, while the dorsal ossicle overlaps slightly the dorsal edges of these plates.
The posterior part of each segment of the hind tail overlaps the anterior edge of the segment
behind it.

Each pair of ventral hind-tail plates has a flattened ventral surface, from which the left and right
sides rise almost vertically. In the larger specimens there is a distinct bulge at the bottom of these
vertical sides. There also seems to be a slight ridge or lip on the anterior edge of the ventral plates,
but normally this is hidden by the plate next anterior overlapping this area (PI. 13, fig. 7).

Most specimens that retain the tail have up to twenty segments in the hind tail. No specialized
terminal segment has been found. As with all other described mitrates, the tails seem to be broken
off at their distal ends. The structure of the tail would allow a little flexion from side to side in
the fore tail. Considerable dorsoventral flexion would be possible throughout the length of the tail.

The fore and hind tail are similar to those in B. jaekeli and A. guttenbergensis. The styloid
differs, however, in that neither of the other two species seems to have the transversely elongate
blades of B. norvegicus and, in particular, the posterior blades in B. jaekeli and A. guttenbergensis
are longitudinally, not transversely, elongate. In these respects, the styloids of B. jaekeli and A.
guttenbergensis are probably primitive, since they are more like the hind-tail ossicles of these same
species and also more like the styloids of Mitrocystites, Mitrocystella and Plaeocystites.

Introduction to internal anatomy of the head. A detailed description of the internal anatomy of the
mitrate head is given in Jefferies and Lewis (1978) and Jefferies (1986). In these works it is argued
that some mitrates are stem-group craniates (there called stem-group vertebrates) and that all
mitrates fed by means of a mucous filter inside the pharynx, as seen today in tunicates, acraniates,
and the ammocoete larvae of lampreys. Evidence is also given that the brain and cranial nervous
system were fundamentally similar to those of vertebrates today. Inside the tail, a dorsal nerve
cord, spinal ganglia, and a notochord are deduced to have existed.

There are two complementary ways of deducing internal features — examination of: 1, the internal
surfaces of plates (or of latex casts of natural moulds of plates); and 2, of natural moulds. The
natural moulds are helpful in reconstruction because they represent positives in rock of the soft
parts. Unfortunately, the natural moulds available for B. norvegicus are imperfect.

Dorsal features of the natural mould (text-fig. 14 a). Incomplete dorsal surfaces of the natural mould
are seen in two specimens (PI. 13, figs. 5 and 9), while much of the internal surface of the calcite
plates can be seen in others. There are also other specimens showing smaller areas. Text-fig. 14#
is a reconstruction of an idealized natural mould in dorsal aspect.

CRASKE AND JEFFERIES: NEW ORDOVICIAN M1TRATE FROM NORWAY 89

probable position of buccal cavity elongate ant. buttons

text-fig. 14. Barr andeocar pus norvegicus sp. nov. Reconstructed internal mould of the head, representing

the soft parts, a, dorsal aspect, b , ventral aspect.

The most obvious feature of the dorsal surface of the mould is the oblique groove which runs
across the head from anterior right to posterior left (PI. 13, fig. 9; text-fig. 15). As argued by
Jefferies (1986, chapter 8), this groove separated the fields of the right and left anterior coeloms.
A second, weaker groove runs from approximately the centre of the oblique groove towards the
right posterior corner of the head. This weaker groove divides the right field into two parts and
represents the left boundary of the right pharynx, i.e. the limit between the right pharynx and the
cavity of the right anterior coelom. No signs of the posterior boundary of the buccal cavity could
be seen, although such a cavity presumably existed. Peripheral ridges (PI. 13, fig. 5; text-fig. 16)
run along the lateral edges of the natural mould and represent nerves n2 (in vertebrate terms the
maxillary branches of the trigeminal nerve). The brain is discussed below. As in P. forbesianus and
A. guttenbergensis , flat raised areas are visible on the dorsal surface of the natural mould over the
right and left pharynges. Since these areas run without break across plate junctions, they probably
represent resorbtion of the internal layers of the skeleton during life, and replacement of the calcite
by soft connective tissue.

Ventral features of the natural mould (text-fig. 146). The ventral surface of the natural mould is
more complex than the dorsal surface but, fortunately, is shown by more specimens.

text-fig. 15. Barrandeocarpus norvegicus sp. nov. Explanatory dia-
gram of Plate 13, fig. 9. lbrp = left boundary of right pharynx;
og = oblique groove; pb = plate boundary; rc = resorbtion cliff in
region of left pharynx.

90

PALAEONTOLOGY, VOLUME 32

Most plates in the anterior part of the ventral surface of the head, as far rearward as plates i,
vii, and xi, carry a protrusion on the internal surface evident on the internal mould as a pit (PI.
13, figs. 2 and 4). These ‘buttons’ were simple disc-shaped columns of calcite, usually at the centre
of the plate and concave-sided and concave- or convex-topped. They represent the inner layer of
calcite for this region of the skeleton.

Some plates, however, have buttons differing from the usual structure. Thus, the buttons of the
plates at the anterior edge of the ventral skeleton (a, /?, vi, y, <5) have ridges running forward from
the plate centres, evident as grooves in the natural mould. Also, plate xi seems to be unique in
having two normal-sized buttons, instead of one (PI. 13, fig. 2). These buttons are paired and lie
forward of the centre of the plate. Possibly the endostyle ran between these two buttons.

There are two other areas where inner-layer calcite has formed button-like depressions in the
natural mould, these being on the plates anterior to plates viii and x, adjoining plate xi. Each of
these plates has two buttons, that nearer the median plane of the animal being much larger than
the other and located near the boundary with plate xi. The smaller button is central on these plates
as is normal (PI. 13, fig. 2; text-fig. 17). Sometimes there is a faint ridge of calcite running rearward
from the larger buttons on to plates viii and x, approximately along the edge of plate xi.

text-fig. 16. Barrandeocarpus norvegicas sp. nov. Explanatory diagram of
Plate 13, fig. 5. Dorsal aspect of distorted internal mould of head, lbrp = left
boundary of right pharynx; og = oblique groove; n2 = right nerve n2.

The posterior part of the ventral surface of the natural mould has a fairly complex structure,
and is best described by reference to text-figs. 146 and 17 (the latter being a diagram of PI. 13, fig.
2). The posterior coelom shows itself on the natural mould as an approximately semicircular area
at the posterior end of the head. In the natural mould this region is delimited by a transverse
fissure anterior to it and by a pair of fissures anterolateral to it on left and right. As in Placocystites,
the nerves nx would have emerged left and right of the transverse anterior fissure (Jefferies and
Lewis 1978, fig. 25; Jefferies 1986, fig. 8.25). The anterolateral fissures would represent walls of
calcite in the skeleton— by comparison with Placocystites , these walls would perhaps have supported
a suboesophageal process on the right and a subrectal process on the left (Jefferies 1986, chapter
8). To right and left of them the palmar nerves would have passed out of the posterior coelom.
In the mid-line of the posterior coelom, at the posterior end of the head, there is a slight
protuberance in the natural mould. This protuberance underlies the mid-line of the brain, being
situated just beneath the place where left and right hypocerebral skeletal processes end beneath
the brain without meeting. This position suggests that the protuberance may represent the
position of the hypophysis (in vertebrate terms) or of the neural gland (in tunicate terms).

Channels for nerves n0 are clearly visible as grooves in the natural mould leading from the
posterior coelom towards plate xi. (Jefferies and Lewis (1978) and Jefferies (1986) argued that the
nerves n0 supplied the endostyle.) In the specimen shown in PI. 13, fig. 2, these grooves reach plate
xi. However, the positions of the plates of this specimen, together with its size, suggest that it is a
juvenile. In a mature specimen, there would be a larger gap between the posterior coelom and
plate xi, as indicated in text-fig. 146.

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

91

text-fig. 17. Barrandeocarpus norvegicus sp. nov. Explana-
tory diagram of Plate 13, fig. 2; ventral aspect of natural
internal mould of head, b = button (natural cast of inner
layer calcite at centre of plate); du = deuterencephalon;
hy = hypophysis; lb = large button just lateral to plate xi;
n0 = nerve n0; nx = nerve nx; pc = posterior coelom;
pr = prosencephalon; sp = suboesophageal process; xi =
plate xi, with paired buttons.

P du hy pr

The cerebral basin , brain, and tail. Four plates of the head are in contact with the tail, g, j, i, and
h. Anterior to the tail, plates h and i form a basin which is deduced to have contained the brain.
The front wall of this basin would have separated the brain from the general cavity of the head.
The basin is divided into a central, more anterior portion, in life containing the prosencephalon,
and a more peripheral shallower portion corresponding to the deuterencephalon. Anteroventrally
(text-fig. 18; PI. 13, fig. 3), the prosencephalar portion of the basin is penetrated by the transversely
elongate optic foramen, beneath which are the paired hypocerebral processes which almost meet
each other in the mid-line. The cerebral basin, and therefore the brain, is very similar to that seen
in other mitrates such as Mitrocystella and Placocystites (Jefferies 1986, chapter 8).

No internal details of the tail were observable, except that no dorsal longitudinal canal existed
in the styloid or in the hind tail. In this respect, B. norvegicus differed from Mitrocystella and
Mitrocystites but resembled Placocystites and perhaps all other members of the Anomalocystitida.

LOCOMOTION IN BARRANDEOCARPUS NORVEGICUS

Text-fig. 19 reconstructs the locomotory cycle of B. norvegicus , as it crawled rearwards through
the top layer of mud in the sea bed. The reconstruction is based on an adjustable model in which

text-fig. 18. Barrandeocarpus norvegicus sp. nov. Explanatory
diagram of Plate 13, fig. 3. Natural mould of the cerebral cavity
representing the brain in anteroventral aspect. 1, i, h, 8 = dorsal
plates of the head; du = deuterencephalon; hp = hypocerebral
processes; of = optic foramen; pi = plate of fore tail pushed in
to form a pit in the rock; pr = prosencephalon.

92

PALAEONTOLOGY, VOLUME 32

a flexible rule was used to imitate the notochord as described for some other mitrates in Jefferies
(1984) and for the cornute Protocystites menevensis in Jefferies et al. (1987). The use of a flexible
rule is appropriate since, like a notochord, it can bend but not shorten or lengthen. The pictures
ignore any yaw in locomotion. Unlike some other mitrates, B. norvegicus had an elongate head,
so yaw was, indeed, probably small. In this respect, therefore, text-fig. 19 is probably not misleading.
Obviously the graphical reconstruction can be no more than suggestive, but it brings the difficulties
and mechanical requirements into focus.

A rearwards direction of movement in B. norvegicus , probably like that of all other calcichordates,
is indicated by the cuesta-shaped ribs on the head, which, as in all other mitrates that possess
them, consistently have the steeper slope anterior. They can be compared with the cuesta-shaped
ribs of recent crabs and bivalves (Jefferies 1986, p. 248) which, by gripping sandy or silty sediment,
serve to reduce movement in the wrong direction during the return stroke of the locomotory cycle.
The belief that the tail moved mainly ventral to the head is based partly on the observed position
of the tail in specimens of B. norvegicus (PI. 12, figs. 4, 6, 7, 10) and partly on the fact that the
same posture is deduced to hold for all other mitrates (Jefferies 1984). This deduction is particularly
well founded for P/acocystites (Jefferies 1986, fig. 8.18), in which plated, overlapping, dorsal folds
in the fore tail would stretch and disappear when the tail flexed ventrally in locomotion. The view
that B. norvegicus moved just beneath the surface of the sea bed is based on the fact that the right
and left parts of the dorsal surface are ribbed but the median part ribless. This suggests that
sediment was gripped by the ribs as it passed, in some thickness, at right and left over the dorsal
surface but medially, where the head stood higher than the right and left portions, the dorsal
surface was either unburied, or too thinly buried for the mud to be grippable. Ribs are similarly
present at right and left of the dorsal surface, but absent medially, in all members of the parascion
of A. guttenbergensis (the traditional group Anomalocystitida) and in B. jaekeli , and the same
functional explanation probably holds for this pattern in all these forms. On the ventral surface
in B. norvegicus, B. jaekeli, Mitrocystella incipiens, and Anomalocystitida with elongate heads, the
transverse ribbing is not interrupted near the mid-line — presumably because the ventral surface
was everywhere in contact with grippable sediment beneath.

As against this, there are some mitrates which have ventral ribs at right and left but a central
ribless area ventrally. This is true, for example, of Placocystites and of Mitrocystites (Jefferies
1984). The dorsal and ventral patterns of such forms are thus, at first glance, similar (though in
Placocystites the ribbing on the dorsal surface ends abruptly in a median direction, whereas the
ribbing on the ventral surface breaks up and fades out gradually towards the centre). The functional
reasons are probably not the same, however. For the forms with the central, ventral ribless patch
have relatively broad heads and this suggests that they yawed strongly in locomotion, alternately
pivoting about right and left sets of ventral ribs. If so, they would have no use for ventral ribs
placed centrally, for these could never be used as a pivot. Forms like B. norvegicus, in which the
head is elongate, most likely did not yaw, and have medianly placed ventral ribs in consequence
(PI. 12, figs. 6 and 11; PI. 13, fig. 6).

Arrows are shown in text-fig. 19 attached to three landmarks on the tail. The heads of the
arrows show where, in space, the same landmark would be in the next diagram. The arrows
therefore suggest, approximately, the direction of travel of the three landmarks in question, and
the lengths of the arrows are proportionate, in any one diagram, to the velocities of the three
landmarks. On the other hand, text-fig. 19 does not imply that successive diagrams are separated
from each other by equal intervals of time. The proportionate lengths of arrows in different
diagrams therefore mean nothing.

text-fig. 19. Barrandeocarpus norvegicus sp. nov. Reconstructed locomotory cycle. The arrows record the
movements of three landmarks on the tail as between one figure and the next. The time intervals between
figures are not supposed to be equal. The horizontal lines represent the sea bed. The vertical line is an

arbitrary fixed mark. Further explanation in text.

94

PALAEONTOLOGY, VOLUME 32

Forces provoked in the mud or water will be opposite in direction to the arrows of movement
and equal to the forces produced by this movement. The actual size, in newtons, of such forces,
will depend on many factors such as the velocity, the shape of the moving part (e.g. whether a
bearing surface or a cutting edge), and whether the part is moving in undisturbed mud, in mud
weakened by disturbance, or in water (here assumed to produce no resistance). In other words,
the provoked force can only be guessed, not measured. The forces provoked in the mud by
movements of the tail will be transmitted to the rear of the head and cause the head to move—
whether by translation (forwards or rearwards) or rotation (yawing, pitching, and rolling).
Rotational movements are assumed to have been too small to be shown in the diagrams.

Ribs are restricted to the rear part of the head in B. jaekeli and similar members of the scion of
A. guttenbergensis. This is probably for two reasons: 1, the mud in contact with the anterior part
of the head will have been disturbed and weakened by the previous motion of the head and tail
across it and so, perhaps, would not have been strong enough for the ribs to grip; and 2, rotational
forces exerted by the tail on the rear part of the head during the return stroke (at which time the
ribs would function) would be stronger than those exerted indirectly on the front part of the head.

Text-fig. 19/1 shows the beginning of the locomotory cycle, i.e. the start of the return stroke.
The hind tail was moving downwards and the styloid rearwards. Resistance to this movement
would be decreased by the fact that the dorsal mid-line of the hind tail was a series of knife edges,
serving to cut downwards through the mud. Nevertheless, the movement of the tail would provoke
an upwards and forwards force in the mud which, transmitted to the head, would tend to lift the
rear of the head and push the head as a whole forwards (in the unwanted direction). This force
would be resisted by the cuesta-shaped ribs, particularly of the dorsal surface of the head.

In text-figs. 19/2 to 19/12, the hind tail moved rearwards through the mud, but resistance would
be reduced by the fact that the hind tail was moving mainly along its own length. The force acting
on the styloid, however, would be considerable, and would be directed downwards and forwards.
It would be transmitted to the head and would tend to move the rear of the head downwards and
the whole of the head forwards. The movement of the styloid, as shown in text-fig. 19/3, 4, and
5, would be partly perpendicular to the length of the tail and dorsalward. Thus the transverse
and longitudinal cutting edges on the dorsal surface of the styloid would cut through the mud
and decrease the forward force exerted by the mud on the styloid. The shape of the styloid is
therefore a compromise between two irreconcilable needs: that of exerting (a) the greatest possible
forward force on the mud during the power stroke (text-fig. 19/20); and ( b ) the least possible force
on the mud during the return stroke (text-fig. 19/2-6). At some stage between text-fig. 19/1 and
5, the force exerted on the tail, and therefore on the head, by the mud, would change in direction
from forwards and upwards to forwards and downwards. As this change proceeded, the dorsal
cuesta-shaped ribs would cause less resistance to forward translation, and the ventral cuesta-shaped
ribs more. Forward translation of the head, in the unwanted direction, would be worst in the early
stages of the return stroke (text-fig. 19/1-3), when the styloid was moving rearwards. Later in the
return stroke there would have been little forward translation. For the styloid would have been
moving mainly upwards, while the hind tail would have exerted little force on the mud, because
it was sliding rearwards along its own length, and because of its dorsal cutting edges.

In the latest parts of the return stroke (text-fig. 19/11-13) the styloid and hind tail would, to an
increasing extent, be surrounded by water, not mud, and would therefore exert almost no force
on the head.

The power stroke began at text-fig. 19/14. The almost fully extended hind tail began to move
forwards and downwards. As soon as the hind tail touched the mud the flattened ventral bearing
surfaces of the hind-tail plates would push downwards and forwards against the mud, producing
a rearward and upward force which, transmitted to the head, would translate it rearwards. As the
power stroke continued, ventral bearing surfaces situated more and more proximally would push
forward against the mud and, at the same time, the distal bearing surfaces of the dorsal ossicles
of the hind tail would likewise push forward against the mud, particularly in the later stages of
the power stroke (text-fig. 19/18-20). Finally the styloid itself, with its transversely expanded cusps,

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

95

would push forward (text-fig. 19/19-21) until the end of the power stroke at text-fig. 19/22. At this
time the cycle would be ready to begin again, the head having moved considerably rearward.

To make a comparison among the forms whose locomotion was reconstructed in Jefferies (1984),
the most similar to B. norvegicus, in most respects, is A. guttenbergensis Kolata and Jollie. B.
norvegicus differs from this species, however, in some aspects of the tail morphology. In particular,
the anterior cusp of the styloid is transversely elongate in B. norvegicus and there are well-developed
bearing services on the ventral plates of the hind tail in this species. The reconstruction of the
locomotory cycle in B. norvegicus explains these differences by proposing that, at the beginning of
the power stroke, the ventral bearing surfaces would push forward more effectively than in A.
guttenbergensis and that, towards the end of the power stroke, the styloid would push forward
more effectively.

Naturally, the reconstructed locomotion has a large element of guesswork, especially in its quanti-
tative aspects such as how far the head would move rearwards in the power stroke or forwards in
the return stroke. Qualitatively, however, it is likely to be right, since it explains many observed
details. Those who think, with Philip (1981 ), that mitrates lived with the dorsal surface downwards,
but none the less pulled themselves rearwards by their tails, should try to produce a reconstruction
of the locomotory cycle as coherent as ours. We believe they will never succeed.

SYSTEMATIC POSITION

In Jefferies (1986, chapters 8, 9) it is argued that all known mitrates can be regarded as stem-group
acraniates, as stem-group tunicates, or as stem-group craniates (there called stem-group vertebrates).
B. norvegicus , along with its closest relatives, is a stem-group craniate.

The stem-group craniate mitrates can, at present, be divided into three plesions, as discussed in
Jefferies (1986). In crownward order these are the plesions of: 1, Chinianocarpos thoralv, 2, Mitro-
cystites mitra; and 3, Mitrocy Stella. B. norvegicus belongs to the last and most crownward of these
plesions, on the basis of many detailed resemblances with other members.

The Mitrocystella plesion is a diverse one comprising the following genera: Allanicytidium Caster
and Gill (in Ubaghs 1967); Anomalocystites Hall 1859; Ateleocystites Billings 1858;
Australocyst is Caster 1954; Barrandeocarpus Ubaghs 1979; Basslerocystis Caster 1952 ; Diamphidio-
cystis Kolata and Guensburg 1979; Enoploura Wetherby 1 879; Mitrocystella Jaekel 1 900; Notocarpos
Philip 1981; Placocy Stella Rennie 1936; Placocystites de Koninck 1869; Rhenocystis Dehm 1933;
Tasmanicytidium Caster 1983; Victoriaecystis Gill and Caster 1960; Willmanocystis Kolata and
Jollie 1982.

It would be premature to attempt a complete cladogram for these forms since many of them
are inadequately known. The nodal group of the plesion is the genus Mitrocystella since this has
not lost its lateral line and nor is it otherwise disqualified, in terms of features, from belonging to
the chordate stem lineage. Both of the two described species, however— M. barrandei Jaekel and
M. incipiens Barrande— are stratigraphically too late to be ancestral to crown-group craniates. For
they are known respectively from the Llanvirn and the Llandeilo whereas the earliest known fossil
fish, with a phosphatic skeleton, is Anatolepis cf. heintzi from the Upper Cambrian (Repetski
1978).

Within the plesion of Mitrocystella , however, there is a well-marked monophyletic group
traditionally called the Anomalocystitida Caster 1952 and ranked by Ubaghs (1967) as a suborder.
This group is characterized by a pair of oral spines articulated to the head right and left of the
mouth. The spines are probably homologous with the rightmost and leftmost spike-shaped oral
plates in such a form as Mitrocystella. As spines, they are very distinctive and probably evolved
only once. The Anomalocystitida comprise all the genera listed above as belonging to the
Mitrocystella plesion, except for the spineless genera Mitrocystella and Barrandeocarpus. Within
the Mitrocystella plesion, the first order apical group, whatever it may be, would have had oral
spines (or conceivably would have secondarily lost them), and so lay somewhere within the
Anomalocystitida. The latter group was therefore a first order parascion, as defined above. One

96

PALAEONTOLOGY, VOLUME 32

of the least apical members of this parascion is the species A. guttenbergensis, which has been
thoroughly described by Kolata and Jollie (1982). It differs from B. norvegicus mainly by having
oral spines. We shall therefore refer to the Anomalocystitida, henceforth, as the parascion of A.
guttenbergensis. This follows the procedure proposed above, that a monophyletic group containing
fossils, whether a scion or a parascion, should be named after its most primitive known member,

parascion of

A. guttenbergensis I

I

text-fig. 20. The systematic position of Barrandeocarpus norvegicus sp. nov. and some
other members of the plesion of Mitrocystella incipiens. The basal part of the apical
lineage of the plesion is shown adnate to the craniate-stem lineage because it may, or
may not, have been part of that lineage. The arrows represent evolutionary novelties,
as follows: 1, cuesta-shaped ribs appear on ventral surface, loss of plate 3 or 4 (from
anterior left of head); 2, major plates of fore tail sutured to form rings, reduction in
area of posterior surface of head, loss of lateral line, loss of plate 5 or 6 (from anterior
right of head), cuesta-shaped ribs on dorsal surface, individualization of plate xi,
reduction of number of ventral plates with individualization of ventral plates i, viii, x,
vii, projection of plate p rearwards so as almost to separate g from j; 3, individualization
and regularization of plates a, /?, iv, y, and 3 to form rigid anterior frame of ventral
skeleton, tessellation (rather than imbrication) of all plates in anterior third of ventral
skeleton so making it rigid; 4, individualization and symmetrization of ventral plates
between the anterior edge of the ventral skeleton and plate xi; modification of the
leftmost and rightmost oral plates to form oral spines articulated to plate (3 or 4) on
the left and plate (5 or 6) on the right -the changes at 4 delimit the parascion of
Ateleocystites guttenbergensis (Anomalocystitida) which is not further analysed here.

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

97

i.e. after the known species most closely related to the latest common stem species of the
monophylum.

Within the Mitrocystella plesion, B. norvegicus represents a paraplesion less apical than the
parascion of A. guttenbergensis but more apical than the paraplesion of B.jaekeli. It is more apical
than B. jciekeli in that the most anterior ventral plates are standardized within the species and
individualized as plates a, /(, vi, y, <5, and that the ventral skeleton immediately behind these five
plates is rigid and consists of fewer plates than in B. jaekeli. It is less apical than A. guttenbergensis
in lacking oral spines and in the fact that the ventral plates of the region in front of the transverse
level of plate xi, but behind plates a, /?, vi, y, and <5, are irregular and not bilaterally symmetrical.
Text-fig. 20 summarizes the likely evolutionary relationships.

The Mitrocystella plesion is the most crownward known among the mitrate craniates— just
crownward of it the skeleton was lost in the craniate stem lineage. We cannot tell therefore where
the more crownward part of this stem lineage separated from the plesion. Of the two described
species of Mitrocystella , M. incipiens is more closely related to the A. guttenbergensis parascion
than is M. barrandei. This is shown by two synapomorphies— -the presence of ventral cuesta-shaped
ribs and the loss of plate 3 or 4 (it is impossible to say which) on the left side of the head. However,
as already explained, the two described species of Mitrocystella belong to the nodal group of the
plesion (i.e. they are not disqualified by their features from belonging to the craniate stem lineage)
and are put in the same plesion precisely because we cannot tell which, if either, is more crownward.
The synapomorphies shared by M. incipiens with the A. guttenbergensis parascion but not with M.
barrandei could have evolved in the craniate stem lineage, or in the apical lineage of the plesion.
We express this uncertainty in text-fig. 15 by making one of the segments linking the two species
run parallel (adnate) to the craniate stem lineage as drawn. This signifies ignorance— we do not
know whether or not these parts of the respective lineages were one and the same.

CONCLUSIONS

This paper reconstructs and describes the first mitrate known from Norway. It is very similar to
other known forms, though it undoubtedly deserves to be placed in a new species. It is a stem-
group craniate and thus contributes to our knowledge of the earliest evolution of a group to which
we ourselves belong.

Its placement in cladistic terms, within the plesion of Mitrocystella , requires discussion of the
plesion concept and an attempt to refine how fossils can be systematized within a plesion. Using
the concepts discussed above under ‘Phylogenetic methodology’, we treat the group traditionally
called the Anomalocystitida as a first order parascion within the plesion of Mitrocystella and we
refer to it as the parascion of A. guttenbergensis , after its most primitive well-described member.
B. norvegicus represents a paraplesion less apical than this parascion, but more apical than the
paraplesion of B. jaekeli. The least apical paraplesion within the plesion of Mitrocystella is that
of Mitrocystella itself. This represents the nodal group of the plesion.

We believe that, unlike recent organisms, fossils require paraphyletic groups for their complete
systematization. Consideration of B. norvegicus helps to answer the basic question of how these
groups can be made objective.

Acknowledgements. We should like to thank Dr Robin Cocks, Keeper of the Department of Palaeontology
in the British Museum (Natural History), for finding the fossils and handing them to us for description and
for writing the account of the stratigraphy in this paper. Mr Angus Johnson, Map Librarian at the British
Museum (Natural History), made some interesting comments on how geographers describe drainage patterns.

REFERENCES

ax, p. 1984. Das Phylogenetische System , 349 pp. Fischer, Stuttgart.

— 1985. Stem species and the stem lineage concept. Cladistics 1, 279-287.

98

PALAEONTOLOGY, VOLUME 32

ax, p. 1987. The phylogenetic system , 340 pp. Wiley, Chichester.

bateson, w. 1886. The ancestry of the Chordata. Q. Jl microsc. Sci. 26, 535-571.

brenchley, p. j. and cocks, l. r. m. 1982. Ecological associations in a regressive sequence: the latest
Ordovician of the Oslo-Asker District, Norway. Palaeontology, 25, 283-315.
caster, k. e. 1952. Concerning Enoploura of the Upper Ordovician and its relation to other carpoid
Echinodemrata. Bull. Am. Paleont. 34 (141), 1 47.

— 1954. A new Silurian carpoid echinoderm from the Parana Devonian. Anais Acad. bras. Cienc. 26, 123-
147.

— 1983. A new Silurian carpoid echinoderm from the Silurian of Tasmania and a revision of the
Alanicytiidae. Alcheringa , 7, 321-336.

cocks, l. r. m. and fortey, r. a. 1982. Faunal evidence for oceanic separations in the Palaeozoic of Britain.
J. geol. Soc. Lond. 139, 465 478.

dehm, r. 1933. Cystoideen aus dem rheinischen Unterdevon. Neues Jb. Miner. Geol. Palaont. BeilBd , (Abt.
B), 69, 63-93.

fordham, b. G. 1986. Miocene- Pleistocene planktic foraminifers from the D.S.D.P. sites 208 and 77, and
phylogeny and classification of Cenozoic species. Evol. Monogr. 6, 1-200.
gauthier, j. 1986. Saurischian monophyly and the origin of birds, 1-55. In padian, k. (ed.). The origin of
birds and the evolution of flight. Mem. Calif. Acad. Sci. 8, 1-98.

Goodrich, e. s. 1909. Vertebrata craniata (first fascicule: cyclostomes and fishes). In lankester, e. r. (ed.).

A treatise on Zoology, 529 pp. Oxford University Press, London.

Griffiths, G. c. d. 1976. The future of Linnaean nomenclature. Syst. Zool. 25, 168-173.

grobben, k. 1908. Die systematische Einteilung des Tierreiches. Verh. zool.-bot. Ges. Wien, 58, 491 -511.

hennig, w. 1965. Die Acalyptratae des baltischen Bernsteins. Stuttg. Beitr. Naturk. 145, 1-215.

1966. Phylogenetic Systematics, 263 pp. University of Illinois, Urbana.

— 1969. Die Stammesgeschichte der Insekten, 436 pp. Kramer, Frankfurt-am-Main.

1983. (trans. pont, a. c.). Insect phylogeny, 514 pp. Wiley, Chichester.

jaekel, o. 1900. Uber Carpoideen, eine neue Klasse von Pelmatozoen. Z. dt. geol. Ges. 52, 661-677.
janvier, p. 1981. The phylogeny of the Craniata, with particular reference to the significance of the fossil
'agnathansk J. vert. Paleont. 1, 121-159.

jefferies, R. p. s. 1973. The Ordovician fossil Lagynocystis pyramidalis (Barrande) and the ancestry of
amphioxus. Phil. Trans. R. Soc. Lond. B265, 409-469.

— 1979. The origin of chordates— a methodological essay, 443-477. In house, m. r. (ed.). The origin of
major invertebrate groups. Systematics Association Special Volume , 12, 1-515.

— 1984. Locomotion, shape, ornament and external ontogeny in some mitrate calcichordates. J. vert.
Paleont. 4, 292-319.

— 1986. The ancestry of the vertebrates, 376 pp. British Museum (Natural History), London.

— and lewis, d. n. 1978. The English Silurian fossil Placocystites forbesianus and the ancestry of the
vertebrates. Phil. Trans. R. Soc. Lond. B282, 205-323.

— lewis, m. and donovan, s. k. 1987. Protocystites menevensis — a stem-group chordate (Cornuta) from
the Middle Cambrian of south Wales. Palaeontology, 30, 429-484.

— and prokop, r. j. 1972. A new calcichordate from the Ordovician of Bohemia and its anatomy,
adaptations and relationships. Biol. J. Linn. Soc. 4, 69- 1 15.

kolata, d. r. and guensburg, t. e. 1979. Diamphidiocystis, a new mitrate carpoid from the Cincinnatian
(Upper Ordovician) Maquoketa group in southern Illinois. J. Paleont. 53, 1 121 1135.

and jollie, m. 1982. Anomalocystitid mitrates (Stylophora, Echinodermata) from the Champlainian
(Middle Ordovician) Guttenberg Formation of the Upper Mississippi Valley Region. Ibid. 56,
531-565.

koninck, m. l. de. 1869. Sur quelques echinodermes remarquables des terrains paleozoiques. Bull. Acad. r.
Belg. Cl. Sci. (2) 28, 544-552.

mayr, e. 1963. Animal species and evolution, xiv + 797 pp. Harvard, Cambridge (Mass.).
patterson, c. and rosen, d. e. 1977. Review of the ichthyodectiform and other Mesozoic teleost fishes and
the theory and practice of classifying fossils. Bull. Am. Mas. nat. Hist. 158, 85 172.
philip, G. m. 1981. Notocarpos garratti gen. et sp. nov., a new Silurian mitrate carpoid from Victoria.
Alcheringa, 5, 29-38.

rennie, J. v. l. 1936. On Placocy Stella, a new genus of cystids from the Lower Devonian of South Africa.
Ann. S. Afr. Mus. 31, 269-275.

repetski, j. e. 1978. A fish from the Upper Cambrian of North America. Science, NY, 200, 529-531.

CRASKE AND JEFFERIES: NEW ORDOVICIAN MITRATE FROM NORWAY

99

ubaghs, G. 1967. Stylophora, S495-S565. In MOORE, r. c. (ed.). Treatise on invertebrate paleontology. Part S.
Echinodermata. /, 2 vols, 650 pp. Geological Society of America and University of Kansas Press, Lawrence,
Kansas.

— 1979. Trois Mitrata (Echinodermata: Stylophora) nouveaux de FOrdovicien de Tchecoslovaquie. Pa/dont.
Z. 53, 98 119.

wetherby, a. G. 1879. Description of a new family and genus of Lower Silurian Crustacea. J. Cincinn. Soc.
nat. Hist. 1, 162 166.

wiley, e. o. 1981. Phylogenetics— the theory and practice of phylogenetic systematics , 439 pp. Wiley, New
York.

willmann, r. 1985. Die Art in Raum und Zeit , 207 pp. Parey, Berlin and Hamburg.

A. J. CRASKE

18 Fairview Road
Chigwell, Essex IG7 6HN

Typescript received 15 December 1987
Revised typescript received 25 April 1988

R. P. S. JEFFERIES

Department of Palaeontology
British Museum (Natural History)
London SW7 5BD

A NEW CAMERATE CRINOID FROM THE
ARENIG OF SOUTH WALES

by STEPHEN K. DONOVAN and JOHN C. W. COPE

Abstract. Celtocrinus ubaghsi gen. et sp. nov., from the Middle Arenig of Dyfed, South Wales, is only the
second lower Ordovician camerate known. It is of approximately the same antiquity as the other Arenig
camerate, the diplobathrid Proexenocrinus inyoensis Strimple and McGinnis, but the Welsh species is a
monobathrid.

Although the first Ordovician crinoid from the British Isles was described over 140 years ago,
fewer than fifty species have been recognized during the intervening period. Two species of lower
Ordovician crinoids, out of a world fauna of only about twenty species (Donovan 1988), have
been described from the UK. Ramseyocrinus cambriensis (Hicks) is well known from the Arenig
of Dyfed, South Wales (Bates 1968; Donovan 1984; Cope 1988). Aethocrinus murchisoni Donovan
is based on a pluricolumnal and a dissociated brachial from the Mytton Flags of Shropshire
(Donovan 1986). A third Arenig species from the UK is described herein and is exceptional in
being only the second camerate of undoubted lower Ordovician age to be recognized. The unique
specimen formed part of the collections of the Department of Geology, University College of
Swansea, until its importance was recognized by J.C.W.C.

SYSTEMATIC PALAEONTOLOGY

Class CRiNOiDEA J. S. Miller, 1821
Subclass camerata Wachsmuth and Springer, 1881

Remarks. Now that Kelly (1986) has demonstrated that the primitive crinoids Reteocrinus
E. Billings, Colpodecrinus Sprinkle and Kolata, and Cleiocrinus E. Billings are not camerates, and
has produced a convincing cladistic analysis of the Class Crinoidea (in prep.), the Subclass
Camerata has been redefined based on three advanced characters: pinnulation; rigid thecae having
both fixed brachials and fixed interbrachials; a radial series which bifurcates at the second
primibrachial and the second secundibrachial. In addition, all camerates have a holomeric stem,
whereas merism seems to be a primitive condition in the inadunates (Donovan 1988). All of these
camerate features are shown by the new species from Triffleton.

Order monobathrida Moore and Laudon, 1943a

Remarks. Plates below the basal circlet, if present, are apparently hidden by the top of the column
(PI. 14, figs. 2 and 3; text-fig. 1). The diameter of the stem is only about 4-5 mm in this region. If
infrabasals were present, they would have to be vanishingly minute, unless the base of the cup was
strongly concave. It is therefore suggested that this species was most probably monocyclic, that
is, a monobathrid. However, in any deduction regarding cup cyclicity we are cautious. The only
other camerate crinoid of comparative antiquity, Proexenocrinus inyoensis Strimple and McGinnis,
1972, was originally considered to be a monobathrid, but has now been shown to be dicyclic, that
is, a diplobathrid (Ausich 1986).

I Palaeontology, Vol. 32, Part 1, 1989, pp. 101-107, pi. 14.|

© The Palaeontological Association

102

PALAEONTOLOGY, VOLUME 32

Incertae faniiliae

Remarks. The tall, conical crown of the Triffleton species most closely resembles the calyx of two
groups of primitive camerates, the monobathrid xenocrinaceans and the diplobathrid archaeocrinids
(Frest et al. 1976, fig. 2). The Triffleton species is most probably a monobathrid (above) and is
easily compared with the four known genera of xenocrinacean (Ubaghs 1978, pp. T440-T443, figs.
249 and 250; Ausich 1986 has shown that Proexenocrinus is a diplobathrid rhodocrinitid). Although
superficially similar, the arm branching pattern of Xenocrinus S. A. Miller is simpler than that of
the Welsh species. X. multiramus Ramsbottom and X. breviformis Brower (Brower 1974) both
branch at lBr2 and 2Br2; X. pencillus S. A. Miller branches only at lBr2 (Ubaghs 1978, fig. 249.1);
the arms of the Welsh species branch at least four times. The Triffleton species also has a stem
with a circular, rather than a square, section and has a continuous circlet of radial plates which
are not separated by interradial ossicles. The interbrachial plates are small and depressed in
Xenocrinus , whereas in the Triffleton species and all other xenocrinaceans these ossicles are large
and prominent. The crown architecture of Abacocrinus Angelin (Ubaghs 1978, fig. 249.2a) differs
considerably from the new taxon and no detailed analysis is worthwhile. Canistrocrinus has five
or six fixed secundibrachs and Compsocrinus has two fixed secundibrachs, branching at 2Br2, or
does not branch further after lBr2 (Ubaghs 1978, pp. T440-T441). Neither is closely similar to
the Triffleton species.

text-fig. f. Celtocrinus ubaghsi gen. et sp. nov., holotype, NMW
87.44G.lc. Camera lucida drawing oflatex cast. BB = basal plates,
RR = radial plates.

DONOVAN AND COPE: ORDOVICIAN CAMERATE CRINOID

103

text-fig. 2. Celtocrinus ubaghsi gen. et sp. nov., paratypes. a, NMW 87.44G.la. b, NMW 87.44G.16. Camera

lucida drawings of latex casts.

Apart from these differences, there are two other points of variance between these genera and
the new taxon. Excepting Silurian Abacocrinus , all of the xenocrinaceans are of late Ordovician
age, so that they are considerably younger than the Welsh species. The Triffleton species also
differs from the known xenocrinaceans in that it apparently lacks an anal tube. However, this may
be a preservational effect. If all four specimens preserved on the same slab were feeding in the
same orientation, with the anus aimed downcurrent, then on burial all four would have
approximately the same attitude. It is therefore possible that the anal tube was apparent on the
(unknown) counterpart slab. Only a ‘complete’ crown will resolve this dilemma.

We are cautious in our classification of the Triffleton species. Without a counterpart to the
holotype, we are uncertain as to the number of plates in the basal and radial circlets, and whether
or not an anal series is present. The pattern of arm branching of the Triffleton specimens also
differs from that of all other xenocrinaceans. While tentatively suggesting that the new species
may, indeed, be a xenocrinacean, we are hesitant to postulate whether the Triffleton taxon belongs
to a known or a new family. A more complete classification requires superior material.

104

PALAEONTOLOGY, VOLUME 32
Genus celtocrinus gen. nov.

Type species. Celtocrinus ubaghsi sp. nov.

Derivation of generic name. After the ancient inhabitants of Wales.

Diagnosis. Camerate crinoid with a circular, heteromorphic ?N212, holomeric proxistele. Basal
plates pentagonal and slightly wider than high. Radial plates heptagonal, about as wide as high
and in close contact. Basal: basal: radial and radial: radial: basal sutures depressed. Interbrachial
region depressed, interbrachial plates large and radially ribbed. Arms uniserial, pinnulate, branching
isotomously at the prim- and secundaxillary but heterotomously thereafter.

Celtocrinus ubaghsi sp. nov.

Plate 14; text-figs. 1 and 2

1971 IDendrocrinus sp.; Bloxam in Owen et al., p. 40.

1988 Monobathrid gen. et sp. nov.; Donovan, p. 235, table 18.1, text-figs. 18.2 and 18.4.

Derivation of trivial name. In honour of Professor Georges Ubaghs.

Material , locality, and horizon. Four specimens on a single slab, without counterpart, preserved as external
moulds. Numbered National Museum of Wales (NMW 87.44G.ln-J) c is holotype, a partial crown with
proximal stem. Paratypes a , b , d, all partial crowns. Other arm and stem debris is preserved on the slab.
Collected from Triffleton Quarry, Dyfed, South Wales, about 10 km north of Haverfordwest (NGR
SM97752426). Locality 16 of Paul (1984). Brunei Beds, Middle Arenig sensu Fortey in Whittington et al.
(1984, pp. 20-21). We have been unable to relocate the precise horizon from which this specimen was derived
but note that dissociated plates of Cheirocrinidae sp. A. Paul, 1984 (pp. 114 116), are common in the lower
part of the section, along with a single specimen of a cyclocyclic ?crinoid columnal that is dissimilar to the
stem ossicles of C. ubaghsi (Donovan, in press).

Diagnosis. As for the genus.

Description. Stem: only the holotype retains part of the proximal column. This preserved about 24 mm of
the proxistele, which is heteromorphic, approximately N212. Secundinternodals have planar latera. Both
nodals and priminternodals have rounded to angular epifacets. In consequence, they are both broader
than secundinternodals. Nodals are taller than priminternodals. Secundinternodals about as high as
priminternodals. Latera of all columnals unsculptured. Features of the articular facet, distal column, and
attachment unknown. A dissociated pluricolumnal, close to specimen a and about 7 mm long, has a similar
morphology to the proxistele of the holotype.

Dorsal cup: seen in the holotype (PI. 14, figs. 2, 3, 4; text-fig. 1). Basals pentagonal, slightly wider than
high. Basals unsculptured, each with two parallel ridges, one derived from each of the supported radials.
Basal: basal: radial and radial: radial: basal regions depressed. Number of basal and radial plates unknown
but at least three of each apparent in the holotype. Radial plates in close contact, infolded interradially but
raised in a ridge radially. Radial plates about as wide as high, heptagonal, with convex, unsculptured latera.
No evidence of an anal series is apparent but this may be an artefact of preservation (see above).

EXPLANATION OF PLATE 14

Figs. 1-6. Celtocrinus ubaghsi gen. et sp. nov., NMW 87.44G. la-J. Whitlandian Stage, Middle Arenig,
Triffleton Quarry, Dyfed, South Wales. All latex casts whitened with ammonium chloride. 1, NMW
87.44G. 1 d, paratype, partial crown, showing branching of arms, x2-5. 2, NMW 87.44G. lc, holotype,
partial crown with proximal stem, x 3. 3, NMW 87.44G. 1 a d, complete specimen with four partial

crowns and distal parts of arms of other specimens, x 1 -25. 4, NMW 87.44G. lc, holotype, enlarged view
of part of dorsal cup to show details of interbrachial plates, x 5. 5, NMW 87.44G. 16, paratype, partial
crown displaying some interbrachial plates and pinnules on some arms, x2-5. 6, NMW 87.44G. 1 a,
paratype, partial crown showing pinnules on arms to right of figure, x 3-5.

PLATE 14

DONOVAN and COPE, Celtocrinus

106

PALAEONTOLOGY, VOLUME 32

Interbrachial plates (PI. 14, figs. 2, 3, 4, 5; text-fig. I): sutures between plates are difficult to determine but
they seem to bear a sculpture of low, radiating ribs. Plates appear to be large, with a raised, elliptical, central
region. Interbrachial plates are present at least to above the level of the secundaxillaries.

Arms (PI. 14, figs. 1-3, 5, 6; text-figs. 1 and 2): arms branch isotomously at the primi- and secundaxillaries
and heterotomously thereafter. Arms branch at least four times. Plate sutures often poorly preserved.
Apparently two large primibrachials per arm. Secundibrachials smaller than primibrachials, two per arm
branch. Branching does not appear to occur at the tertibrach level in all arm branches. Where it does occur,
the tertaxillary is at about the level of 3Br,2. More distal branches of the arm slender. Arms uniserial,
pinnulate. Pinnules more slender than the branches of the arm. Both pinnules and brachials have planar,
unsculptured latera. Adoral groove broad, U-shaped. Number of arms unknown, at least three, probably
either four or five.

Discussion. This is only the second lower Ordovician camerate crinoid to be described and is
consequently one of the oldest members of the subclass known. The other Arenig species,
Proexenocrinus inyoensis Strimple and McGinnis, is from the trilobite zone J of D. C. Ross (Ross
1966; Ausich 1986), which is approximately equivalent to the early part of the D. nitidus Biozone
(R. J. Ross et al. 1982, sheet 1). It is therefore also Middle Arenig in age. Trichinocrinus terranovicus
Moore and Laudon, 19436, was originally described as lower Ordovician (Canadian) in age but it
is most probably from the Lower Llanvirn (H. B. Whittington, written comm.). This paucity of
early Ordovician camerates is noteworthy because elsewhere in the Palaeozoic record camerate
thecae often seem to have been more durable than inadunate cups, yet at the time of writing about
eighteen species of lower Ordovician inadunates are known (Donovan 1988).

C. ubaghsi is only the sixth camerate crinoid known from the British Ordovician south of the
Iapetus suture (Colpodecrinus forbesi Donovan is now recognized to be non-camerate; Kelly 1986).
Three species are diplobathrid archaeocrinids of the genus Balacrinus Ramsbottom. B. basilis
(M‘Coy) and B. inflatus Donovan are both Caradoc and a third species, from the Lower Llanvirn,
awaits description. Two species of Xenocrinus, X.? blaenycwmensis Donovan and Xenocrinusl sp.
Donovan, have both tentatively been recognized from the Ashgill on the basis of dissociated
columnals. The fauna is thus small and also taxonomically conservative, being limited to, at most,
just three families.

Acknowledgements. We thank Professor H. B. Whittington for his comments on the stratigraphic position of
Trichinocrinus terranovicus.

REFERENCES

ausich, w. i. 1986. The crinoids of the Al Rose Formation (early Ordovician, Inyo County, California,
U.S.A.). Alcheringa , 10, 217-224.

baths, d. e. b. 1968. On Dendrocrinus cambriensis Hicks, the earliest known crinoid. Palaeontology 1 1, 406-
409.

brower, j. c. 1974. Upper Ordovician xenocrinids (Crinoidea, Camerata) from Scotland. Paleont. Contr.
Univ. Kans., Pap. 67, 1-25.

cope, j. c. w. 1988. A reinterpretation of the Arenig crinoid Ramsey ocrinus. Palaeontology , 31, 229-235.
donovan, s. k. 1984. Ramseyocrinus and Ristnacrinus from the Ordovician of Britain. Ibid. 27, 623-634.

— 1986. Pelmatozoan columnals from the Ordovician of the British Isles. Part 1. Palaeontogr. Soc.
[Monogr.l 138 (for 1984), no. 568, 1 68.

— 1988. The early evolution of the Crinoidea. In paul, c. r. c. and smith, a. b. (eds.). Echinoderm phytogeny
and evolutionary biology , 235-244. Oxford University Press, Oxford.

— In press. Pelmatozoan columnals from the Ordovician of the British Isles. Part 2. Palaeontogr. Soc.
[Monogr.].

frest, t. j., strimple, h. l. and kelly, s. m. 1976. A new Ordovician camerate crinoid from Kentucky. SEast.
Geol. 17, 139-148.

kelly, s. m. 1986. Classification and evolution of Class Crinoidea. 4th N. Am. Paleont. Conv. Abstr.. A23.
miller, j. s. 1821. A natural history of the Crinoidea or lily-shaped animals , with observation on the genera
Asteria , Eurayle, Comatula, and Marsupites, 150 pp. Bryan and Co., Bristol.

DONOVAN AND COPE: ORDOVICIAN CAMERATE CRINOID

107

moore, r. c. and laudon, l. r. 1943a. Evolution and classification of Paleozoic crinoids. Spec. Pap. geol.
Soc. Am. 46, 1 53 pp.

1943/). Trichinocrinus, a new camerate from lower Ordovician (Canadian?) rocks of Newfoundland.
Am. J. Sci. 241, 262-268.

owen, t. r., bloxam, t. w., jones, d. g., walmsley, v. g. and williams, b. p. 1971. Summer (1968) field
meeting in Pembrokeshire, south Wales. Proc. Geol. Ass. 82, 17 60.
paul, c. R. C. 1984. British Ordovician cystoids. Part 2. Palaeontogr . Soc. [Monogr 136 (for 1982), 65- 152.
ross, d. c. 1966. Stratigraphy of some Paleozoic formations in the Independence Quadrangle, Inyo County,
California. Prof. Pap. US geol. Surv. 396, 64 pp.

ross, r. j. et al. 1982. The Ordovician System in the United States. Int. Un. geol. Sci., Pub. 12, 73 pp.
strimple, h. l. and mcginnis, m. r. 1972. A new camerate crinoid from the Al Rose Formation, lower
Ordovician of California. J. Paleont. 46, 72-74.

ubaghs, G. 1978. Camerata. In moore, r. c. and teichf.rt, c. (eds.). Treatise on invertebrate paleontology.
Part T. Echinodermata 2 (2), T408-T519. Geological Society of America and University of Kansas Press,
Boulder, Colorado and Lawrence, Kansas.

wachsmuth, c. and springer, f. 1881. Revision of the Palaeocrinoidea, pt. II. Family Sphaeroidocrinidae,
with the subfamilies Platycrinidae, Rhodocrinidae and Actinocrinidae. Proc. Acad. nat. Sci. Pliilad. 175—

411.

WHITTINGTON, H. B., DEAN, W. T., FORTEY, R. A., RICKARDS, R. B., RUSHTON, A. W. A. and WRIGHT, A. D. 1984.
Definition of the Tremadoc Series and the series of the Ordovician System in Britain. Geol. Mag. 121,
17-33.

Typescript received 16 March 1988
Revised typescript received 20 May 1988

STEPHEN K. DONOVAN
Department of Geology
University of the West Indies
Mona, Kingston 7
Jamaica, WI

JOHN c. w. COPE
Department of Earth Sciences
University College
Singleton Park
Swansea SA2 8PP

SILURIAN TRILOBITES FROM THE ANNASCAUL
INLIER, DINGLE PENINSULA, IRELAND

bv DEREK J. SIVETER

Abstract. The richest and best-preserved Silurian trilobite fauna to have been discovered from Ireland, from
two localities in the Annascaul inlier, County Kerry, is described. The specimens belong to eight families,
thirteen genera, and sixteen species. Five species are new: Interproetus galvani , Conoparia hollandi , Calymene
endemopsis, Primaspis mendica, and Leonaspis parkini. The precise age of the two limestone lenses which
yielded the material is uncertain, but it is likely that they fall within a mid-Wedlock to mid-Ludlow time
interval, with several faunal elements indicating a late Wenlock age. The restricted fauna from one of the
localities is typical of the illaenid-cheirurid assemblage which typifies shallow water, nearshore carbonate
environments; the much richer fauna from the second locality may have inhabited a slightly more offshore,
open water position. At the generic level the fauna is cosmopolitan; at the specific level links are evident with
the Welsh Borderland (in particular), Bohemia, and Scandinavia.

The Silurian of Ireland is dominated both in terms of thickness and areal extent by deep water,
turbidite, and graptolite facies. Shallow water facies and associated shelly faunas are largely
confined to the Galway, Connemara, and Roscommon areas in the north-west, and to the Dingle
Peninsula, County Kerry, in the south-west. Since the mid-nineteenth century various trilobites
have been recorded from these areas but, apart from the early studies of IVTCoy (1846, 1849) and
the recent discussions of Whittington and Campbell (1967) and Thomas and Owens (1978) of
Harpidella megalops (M'Coy, 1849) and Owens (1973) of Proetus (s.l.) latifrons (IVPCoy, 1846),
they have received little attention. The taxa described here from the Annascaul inlier represent by
far the richest and best-preserved Silurian trilobite fauna to have been described from Ireland.

THE LOCALITIES AND COLLECTIONS

All the specimens come from the western slopes of Caherconree mountain, from two small
exposures separated by just over a kilometre of largely unexposed ground (text-lig. 1; Table 1 ).
These exposures were discovered in the last century by officers of the Geological Survey and
dubbed by Jukes and Du Noyer (1863, p. 12) the 'dove coloured limestone’ and 'gray crystaline
limestone’ localities. The outcrops were subsequently relocated by Dr John Parkin (1976, figs. 4
and 5) who mapped the area and included them as part of his Ballynane Member within the
Annascaul Formation, Dunquin Group, and designated them his shelly fossil localities numbers
28 and 36, respectively. The original Geological Survey material is extant: it comprises nine
specimens (GSI F00749-757) from locality 28, referred in Jukes and Du Noyer (1863) to Cheirurus
bimucronatus , Encrinurus sexcostatus , and Illaenus bowmanni , and fourteen specimens (GSI F00758-
771) from locality 36, referred in the same paper to Acidaspis jamesii. Subsequent collections by
the present author together with Parkin and Dr Cohn Harris, and by Parkin and Mr Phillip
Doughty, produced over seven hundred specimens; a few odontopleurids from this material were
figured by Siveter in Holland (1981, fig. 51). All these more recently collected specimens are now
housed in Trinity College, Dublin (TCD) and the Ulster Museum, Belfast (UM). Other material
listed in the present work includes specimens from the Museum of the British Geological Survey
(GSM), the British Museum (Nat. Hist.) (BM), the Humboldt-Universitat, East Berlin (HU), and
the Geological Survey of Canada (GSC).

| Palaeontology, Vol. 32, Part 1, 1989, pp. 109-161, pis. 15-23.|

© The Palaeontological Association

110

PALAEONTOLOGY, VOLUME 32

text-fig. 1 . Locality map. a, Ireland with inset of the Dingle Peninsula, b, geological map of the Dingle
Peninsula (after Holland 1981, fig. 48). c, geological map of the Carherconree- Derrymore Glen area, with
the two fossil localities (28 and 36) which have yielded the trilobites described herein (simplified from Parkin
1976, fig. 4). d, stratigraphical section through the Dunquin Group, Bull’s Head, Annascaul, and Derrymore
Glen inliers (after Parkin 1976, fig. 5 and Holland 1981, fig. 49).

FAUNAL COMPOSITION

The fauna (Table 1) comprises eight families, thirteen genera, and sixteen species, of which five
are new. Locality 28 has yielded three genera and three species, locality 36 eleven genera and
thirteen species. The Proetidae, Phacopidae, Calymenidae, and Cheiruridae are represented by one
genus, the Styginidae, Aulacopleuridae and Lichidae by two, and the Odontopleuridae by three.
Cheirurus , Calymene, and Leonaspis have two species, other genera have one. The fauna is
dominated both in terms of species and specimen numbers by the odontopleurids which total some
46% of the cranidia or cephala identified. Of the odontopleurids, Odontopleura ( Odontopleura )
ovata is by far the commonest species, and the most abundant in the fauna as a whole, comprising
almost 3 1 % of the cranidia or cephala. Interproetus galvani sp. nov. and Ananaspis aff. stokesii are
the second and third most common species, totalling 24% and 20% respectively of the recovered
cranidia or cephala, with the rest of the species each having less than 10%, Kosovope/tis aff. allaarti ,
Trochurusl sp. indet. and the indeterminate styginid species being represented by a single cranidium
or pygidium.

SIVETER: SILURIAN TRILOBITES FROM IRELAND

III

table I . Occurrence of trilobite fauna in localities 28 and 36 and abundance of exoskeletal parts for each
species as identified from bulk collections. Only Ananaspis is represented by cephala, the other species being

represented by cranidia.

Locality
28 36

Cephala/

cranidia

No.

%

Free cheeks

Hypostomata

Pygidia

Styginid gen. et sp. indet.

•

1

Kosovopeltis aff. allaarti

•

1

0-3

Interproetus galvani

•

83

240

40

10

46

Conoparia hollandi

•

7

21

3

1

S chary ia sp.

•

2

0-6

Cheirurus sp. nov.?

•

5

1-5

4

6

Cheirurus sp.

•

2

0-6

1

1

Calymene endemopsis

•

9

2-6

5

Calymene sp.

•

5

1-5

1

1

1

Ananaspis aff. stokesii

•

70

20-3

3

20

Odontopleura ( 0 .) ova la

•

106

30-7

97

28

88

Primaspis mendica

•

21

61

10

15

15

Leonaspis coronata coronata

•

1 1

3-1

4

2

2

L. parkini

•

19

5-5

5

3

Dicranopeltis salterP.

•

3

0-9

2

Trochurusl sp. indet.

1

0-3

Total

345

160

64

190

AGE

Parkin (1976, p. 593) considered that the Ballynane Member was most likely to be of upper
Wenlock age, citing in particular the brachiopod evidence. He noted that this member may extend
into the Ludlow, but not to any great extent as its upper limit was constrained by the overlying
Caherconree Formation which he believed to be of C. scanicus Biozone (middle Gorstian) age.
Aldridge (1980, pp. 130-131) concluded from the conodonts that the fauna from locality 28 was
most likely of Wenlock age but that it was not possible to be more specific, and that the locality
36 fauna suggested a Ludlow age; the balance of evidence indicated to him that the strata at each
of the two outcrops were deposited at different times, but that an ecological control may also have
operated.

The trilobite faunas of the two localities are mutually exclusive at the specific level and Cheirurus
is the only genus common to both. Of the taxa from locality 28 the unnamed Cheirurus species,
which is related to C. insignis and C. centralis , suggests a Wenlock age, but the very similar and
incompletely known C. strux comes from lower Ludlow strata; the relationships of the other two
species from here are indeterminate. All of the species from locality 36 are best compared with
Wenlock or Ludlow forms elsewhere, but some in particular give a better indication of its age.
The Ananaspis species seems to lie on a morphological gradient between the early Wenlock
Acernaspis rubicundula and late Wenlock material assigned to Ananaspis stokesii , though it is
clearly much closer to the latter; L. coronata coronata has a late Wenlock to earliest Ludlow range;
the new Leonaspis species is most closely related to a mid to late Wenlock form; Dicranopeltis
salterP. may be the same species as that occurring in the British late Wenlock Much Wenlock
Limestone Formation. Thus the best trilobite evidence suggests a mid/late Wenlock to earliest
Ludlow age for locality 36. The differences between the trilobite faunas of the two localities could
easily be a reflection of their somewhat different carbonate facies (see below), and possible different
palaeoslope positions, but it is not possible to discern an age difference also on the basis of

112

PALAEONTOLOGY, VOLUME 32

trilobites. Thomas (1980; in Thomas et al. 1984, pp. 54, 55) has already noted the lithofacies
control on many British Wenlock trilobite genera, and also the difficulty in determining the
Wenlock-Ludlow boundary on trilobites, as many Wenlock species range into the early Ludlow.

Limestone development of any note is uncommon in the British Silurian: it is almost absent
from the Llandovery,, with the few good examples occurring at the base and particularly the top
of the Wenlock and in the middle Ludlow. There is even less limestone in the Irish Silurian. The
faunal and lithological evidence combined suggests that the Caherconree limestones may represent
an Irish age equivalent of the Much Wenlock Limestone Formation. This was a time when
carbonates were well developed throughout western and northern Europe, for example in the
Welsh Borderland, Gotland, and the East Baltic. Even in areas subject to slightly deeper water
deposition throughout much of the Silurian, for instance the Lake District and the Long Mountain,
carbonate horizons formed in the late Wenlock (Rickards 1978; Palmer 1970).

SYSTEMATIC PALAEONTOLOGY

Family styginidae Vogdes, 1980

Discussion. For the most recent appraisal of this family, see Lane and Thomas 1983, p. 156.

styginid gen. et sp. indet.

Plate 15, figs. 7 and 8

1863 Illaenus Bowmanni; Baily in Jukes and Du Noyer, p. 12.

1878 Illaenus Bowmanni ; Baily in Kinahan, p. 40.

1976 lllaenid; Siveter in Parkin, p. 595, table 1 {pars).

1984 illaenimorph; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. Two thoracic segments, one pygidium.

Occurrence. Annascaul Formation, locality 28 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Discussion. These rare, fragmentary illaenimorph specimens have been regarded as styginids rather
than illaenids (both families sensu Lane and Thomas 1983) because, even though very incomplete,
the pygidium does not appear to possess a prominent median arch as diagnosed for the latter
family. The nature of the material and the lack of a cephalon precludes even a tentative generic
or specific assignment, especially considering the number of new illaenimorph genera introduced
in recent years (see, for example, Lane and Thomas 1983; Lane and Owens 1982; Lane and Thomas
in Thomas 1978, Howells 1982).

EXPLANATION OF PLATE 15

Figs. 1, 2, 5. Kosovopeltis sp. aff. K. allaarti Lane, 1984. Cranidium, GSI F00758, Annascaul Formation,
locality 36; dorsal stereo-pair, lateral, frontal views, x 5-5.

Figs. 3, 9 12, 15, 16, 18 21. Cheirurus sp. nov.? All specimens are from the Annascaul Formation, locality
36. 3, cranidium, internal mould, TCD 12091; dorsal view, x 1. 9, 10, 15, 16, cranidium, largely internal
mould, and silicone rubber cast of external mould; dorsal, oblique, lateral views (internal mould), dorsal
stereo-pair (cast), xl. II, pygidium, TCD 12074; ventral view, x 1-5. 12, pygidium, internal mould,

TCD 12073; dorsal view, x 1-5. 18 and 19, pygidium, internal mould, TCD 12078; dorsal stereo-pair,

lateral view, x 1-25. 20 and 21, hypostoma, TCD 12631; ventral, lateral views, x 1-25.

Figs. 4, 6, 13, 14, 17. Cheirurus sp. All specimens are from the Annascaul Formation, locality 28. 4 and 14,
cranidium, internal mould, GSI F00756; dorsal stereo-pair, lateral view, x 2. 6 and 13, pygidium, internal
mould, GSI F00753; dorsal stereo-pair, lateral view, x I -5. 17, hypostoma, largely internal mould, GSI

F00754; ventral view, x 15.

Figs. 7 and 8. Styginid gen. et sp. indet. Both specimens are from the Annascaul Formation, locality 28. 7,
pygidium, GSI F00751; ventral view, x 1. 8, two thoracic segments, GSI F00750, ventral view, x 1-5.

PLATE 15

SIVETER, Silurian trilobites

114

PALAEONTOLOGY, VOLUME 32

Genus kosovopeltis Snajdr, 1958

Type species. Kosovopeltis svobodai Snajdr, 1958, p. 178; from the Kopanina Formation (Ludlow) of Kosov,
near Kraluv Dvur, Czechoslovakia. By original designation.

Kosovopeltis sp. aff. K. allaarti Lane, 1984
Plate 15, figs. 1 , 2, 5

? 1863 Acidaspis Jamesii; Baily (pars ) in Jukes and Du Noyer, p. 12.

? 1878 Acidaspis Jamesii ; Baily (pars) in Kinahan, p. 40.

1976 scutelluid; Siveter in Parkin, p. 595, table 1.

1984 scutelluid; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. One cranidium.

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig 4, Annascaul inlier.

Description. Glabella gently convex (sag. and tr.) and 1-3 times as wide (tr.) as long (sag.) across frontal lobe
excluding occipital ring which is missing. Medial section of occipital furrow is convex forwards, abaxially
occipital furrow is not preserved. Lateral lobe lp is subrectangular in outline. Lateral furrow lp narrow
(exs.) and clearly impressed at axial furrow, runs inward and becomes wider (exs.) and shallow, not reaching
one-third the glabellar width then turning quite sharply forward for a short distance before finally swinging
abaxially and dying out on dorsal glabellar surface, though it reappears as a very shallow impression on
adaxial side of axial furrow, thus enclosing a moderately sized and inflated lobe. Furrow 2p narrow (exs.)
and slit-like, distinct, more or less transversely directed, about as wide (tr.) as posterior section of lp furrow,
isolated on dorsal glabellar surface by a distance from axial furrow equal to its own width (tr.), reaches as
far adaxially as lp furrow. Lobe 2p about as long as lp and 3p lobes. Furrow 3p distinct and slit-like,
adaxially is directed slightly forwards, is slightly deeper than, about twice as wide as, and does not reach as
far adaxially as 2p furrow, is connected abaxially to axial furrow by very weak depression. Frontal lobe
expanded abaxially, 2-3 times as wide (tr.) as glabella at lp lobes, moderately strongly convex forward in
dorsal outline.

Axial furrow at lp lobe moderately deep, is directed forward and very slightly inward, at lp furrow
becomes slightly narrower (tr.) and turns sharply outward at an angle of about 40° to the sagittal line, is
slightly abaxially convex at 2p lobe before continuing forward and slightly more outward to then widen (tr.)
and curve very sharply inward around side of frontal lobe to merge with preglabellar furrow. The latter
becomes progressively more finely marked adaxially and fades out at about one-third the width of the frontal
lobe inward from axial furrow (PI. 15, fig. 5). Anterior border short (exs.) abaxially in front of axial furrow,
runs inward below frontal lobe as a progressively more slender rim, is obsolete as an anteriorly projecting
feature anterior to central glabellar area but its position here is indicated by change in lateral profile of
cranidium (PI. 15, fig. 2). Outer face of anterior border subvertical, deepest medially, decreases gradually in
height abaxially. Rounded, smooth lateral muscle impression is independently convex of fixed cheek and
sited opposite anterior part of lobe lp, and is slightly smaller than inflation enclosed within lp furrow.
Anterior part of fixed cheek flat, slopes downward and forward. Eye ridge distinct, runs inward and forward
from anterior part of palpebral lobe to meet axial furrow at mid-length of 3p lobe. Palpebral lobe extends
(exs.) posteriorly from a point oposite most anterior part of furrow lp to a point opposite posterior margin
of lateral muscle impression. Width of fixed cheek at palpebral lobe very slightly less than that of glabella
at lp lobe. Other cranidial parts not preserved. Free cheek, hypostoma, thorax, and pygidium unknown.

Cranidial surface exclusive of all furrows and lateral muscle impressions has small to medium sized,
moderately densely scattered granules together with very fine terrace ridges, the latter becoming slightly more
conspicuous and subparallel on anterior part of frontal lobe and subvertical face of anterior border.

Discussion. The distinct, slit-like 2p and 3p lateral glabellar furrows distinguish this Kerry cranidium
from all species assigned to Kosovopeltis except K. tchernychevae Snajdr, 1960 (pi. 36, fig. 7) from
rocks of possible Ludlow age from the Urals, and K. allaarti Lane, 1984 (pi. 1, figs. 9-16) of
possible Wenlock age from north Greenland. These three species also appear to differ from other
congeneric species in having a rather narrower (tr.) anterior part to the fixed cheek in front of the
palpebral lobe, a convex forward occipital furrow behind the central glabellar area, and a more

SIVETER: SILURIAN TRILOBITES FROM IRELAND

115

clearly outlined frontal lobe abaxially. The Kerry and Urals species are, however, known from
only one specimen.

The Irish cranidium differs from allaarti in having a combination of dominantly granular and
subsidiary terrace ridge, rather than solely terrace ridge, type of sculpture (though small specimens
of allaarti have a granular sculpture), and a longer palpebral lobe reaching posteriorly to opposite
the posterior margin of, not the mid-length of, the lateral muscle impression. It differs from
tchernychevae in having granular sculpture and a preglabellar furrow present abaxially; the Russian
species appears to have only terrace ridges (though much of the glabella is in the form of an
internal mould), and as Lane (1984, p. 57) has already noted it seems to lack a preglabellar furrow,
even laterally.

Family proetidae Salter, 1864
Subfamily cornuproetinae Richter and Richter, 1956
Genus interproetus Snajdr, 1977

Type Species. Proetus intermedins Barrande, 1846, p. 63; from the Kopanina Formation (Ludlow) of Dlouha
hora, Prague district, Czechoslovakia. By original designation.

Discussion. Following the comments of Owens (1973, pp. 40 and 83), Thomas (1978, p. 42)
generically assigned a group of seven Silurian cornuproetine species to Cornuproetus (s.l ). All were
previously assigned to Cornuproetus (s.s.) but where the thorax was known it was claimed (Owens
1973; Thomas 1978) that they differed from the type of Cornuproetus , Gerastos cornutus Goldfuss,
1843, in lacking a preannulus. This group, which included P. intermedius , together with seven new
species or subspecies, were all placed by Snajdr (1977, 1980) in his new taxon Interproetus , a genus
he (1980, p. 223) claimed to be ancestral to Cornuproetus and which he diagnosed in part as having
‘a very narrow, lowered preannulus’. Snajdr (1980, p. 224) specifically described P. intermedius as
having a preannulus, though it is not apparent on the complete holotype (Snajdr 1980, pi. 45, fig.
17) due to the overlapping nature of the thoracic segments. This character is in need of further
investigation in Interproetus. In addition the preglabellar field and the pygidial doublure in
Interproetus species do not appear particularly ‘narrow’, as diagnosed (see Snajdr 1980, p. 223, pi.
49, figs. 3 and 7; pi. 44, fig. 20), especially as in discussion Snajdr (1980, p. 223) distinguished the
preglabellar field of Interproetus from that of Cornuproetus by being ‘wider’, and he described the
pygidial doublure of Interproetus numvertus Snajdr, 1980 (p. 229) as ‘very wide’. Whether
Interproetus merits full generic separation from Cornuproetus is open to question (cf. Snajdr 1980,
text-figs. 58 and 64a-c).

Interproetus is known almost exclusively from Czechoslovakia. Cornuproetus peraticus Owens,
1973 from the early Wenlock of the Welsh Borderland, which was assigned with slight doubt by
Snajdr (1980) to his new genus, must be placed there with certainty now that Thomas (1978,
p. 43) has identified the correct pygidium for this species. Cornuproetus ( Cornuproetus! ) walliseri
Alberti, 1967 from the lower Ludlow of Morocco was not discussed by Snajdr (1977, 1980), after
the remarks of Owens (1973, p. 40), but it also fits in Interproetus.

Interproetus galvani sp. nov.

Plate 16, figs. I 15, 17-29; Plate 17, figs. 16-25

1863 Acidaspis Jamesir, Baily (pars) in Jukes and Du Noyer, p. 12.

1878 Acidaspis Jamesir, Baily (pars) in Kinahan, p. 40.

1976 Decoroproetus sp.; Siveter in Parkin, p. 595, table 1.

1984 Cornuproetus sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Derivation of name. After Mr C. Galvan, nineteenth-century fossil collector for the Geological Survey of
Ireland and discoverer of the trilobite-rich limestone of shelly fossil locality 36.

Holotype. Cranidium, TCD 12107; Plate 16, figs. 1, 2, 4.

116

PALAEONTOLOGY, VOLUME 32

Type horizon and locality. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Additional material. Eighty-two cranidia, forty free cheeks, ten hypostomata, forty-six pygidia, isolated
thoracic segments.

Diagnosis. A species of Inlerproetus with a relatively long preglabellar area, 0-25 to 0-3 times as
long as glabella. Anterior border is very gently convex (sag.), slightly longer than preglabellar field;
anterior border furrow contains weak ridge centrally. Glabella is low, rises gently from axial and
preglabellar furrows, is weakly convex (tr. and sag.); lateral glabellar furrows weak or lacking,
crossed by surface sculpture. Pygidium has four (?five) axial rings and three pleural ribs.

Description. Cranidium is about as wide (tr.) as long (sag.). Glabella is generally 0-8 times as wide as long
(sag.) including occipital ring, and 0-9 to 10 times as wide excluding occipital ring; in one specimen (PI. 16,
fig. 11) these ratios are 10 and IT. Occipital ring is from 3-2 to 3-7 times as wide (tr.) as long (sag.) in all
specimens except one (PI. 16, fig. 1 1) where it is 4-5 times as wide, it forms the widest part of the glabella, is
longest behind central glabellar area, and narrows (exs.) slightly abaxially; in profile it is almost flat (PI. 16,
fig. 5) to very gently convex (PI. 16, fig. 2). Two specimens (PI. 16, figs. 19 and 21) have a very shallow furrow
abaxially on occipital ring, running adaxially from axial furrow to define a very weak occipital lobe. A small,
slightly anteriorly positioned median occipital granule is present (PI. 16, figs. 3 and 10). Occipital furrow well
defined, moderately deep, its posterior slope is longer and more gently sloping than anterior slope which is
short (sag. and exs.) and steep. Shallow, broad occipital impression present on posterior slope of occipital
furrow, behind abaxial part of lp furrow. Preoccipital part of glabella weakly convex (sag. and tr.), rises
gently from axial and preglabellar furrows (PI. 16, figs. 2 and 4). Lateral glabellar furrows weak (PI. 16, figs.
1, 8, 11) to imperceptible (PI. 16, fig. 3; PI. 17, figs. 21 and 24); where present they are crossed by glabellar
sculpture. Furrow lp runs inward and backward from axial furrow opposite a point about midway between
a and 8, expands slightly before bending a little more strongly backward to fade out before occipital furrow.
A small, rounded auxiliary impression is sited a short distance adaxially from expanded part of furrow lp.
Furrow 2p trends inward and very slightly backward from y; furrow 3p is short, does not reach axial furrow,
is transverse or trends slightly inward and forward, lies opposite a point about one-third of the distance
between y and )3.

Axial furrow sharply impressed, runs forward from posterior cranidial margin beside posterior half of
occipital ring, then turns sharply inward to meet occipital furrow, anterior to which it is gently abaxially
convex inside palpebral lobe, becoming very gently adaxially convex at y and thereafter running forward and
slightly inward into sharply defined preglabellar furrow. Preglabellar area 0-25 to 0-3 times as long as glabella
including occipital ring. Preglabellar field slightly shorter than anterior border and gently convex (sag.),
becomes gradually longer (exs.) abaxially. Anterior border furrow much broader (exs.) than preglabellar or
axial furrows, centrally it contains a very gently raised weakly anteriorly convex ridge. Anterior border very
weakly convex (sag.), projects forward and gently upward (PI. 16, figs. 2 and 5), is about 06 times the length

EXPLANATION OF PLATE 16

Figs. 1 15, 17-29. Interproetus galvani sp. nov. All specimens are from the Annascaul Formation, locality
36. 1, 2, 4, 29, holotype, cranidium, TCD 12107; dorsal stereo-pair, lateral, frontal views, x 5, sculpture,
central, posterior part of glabella, x 22. 3 and 5, cranidium, TCD 12537; dorsal, lateral views, x 5.

6 and 7, thoracic segment, UM K2824; lateral, dorsal views, x 6. 8 and 9, cranidium, TCD 12018; dorsal,
oblique views, x 5. 10, cranidium, TCD 12558; dorsal view, x 5. 11, cranidium, TCD 12015; dorsal

view, x 5. 12, free cheek, TCD 12014; ‘dorsal’ stereo-pair, x 4. 1 3, cranidium, TCD 1261 1; dorsal view,
x 10. 14, free cheek, TCD 12118; ‘dorsal’ view, x4. 15, cranidium, TCD 12001; dorsal view, x 5. 17,
pygidium, TCD 12564; dorsal view, x 5. 18, hypostoma, TCD 12050a; ventral view, x 9. 19 and 20,
pygidium, TCD 12132; dorsal, lateral views, x 5. 21, sculpture, central, posterior part of glabella, TCD
12189; dorsal view, x 22; see also Plate 17, figs. 19, 23, 24. 22, meraspis pygidium and one thoracic
segment, silicone rubber cast of external mould, TCD 12589a; dorsal view, x 10. 23, hypostoma, TCD
12554; ventral stereo-pair, x 7. 24 and 25, pygidium, TCD 11981; posterior, dorsal views, x5. 26,

pygidium, TCD 12113; dorsal stereo-pair, lateral view, x5. 27 and 28, pygidium, TCD 12113, dorsal
stereo-pair, lateral view, x 5.

Fig. 16. Trochurusl sp. indet. Cranidium, internal mould, GSI F00749, Annascaul Formation, locality 28;
dorsal view, x 2.

PLATE 16

SIVETER, Silurian trilobites

118

PALAEONTOLOGY, VOLUME 32

of the preglabellar area. Anterior cranidial margin evenly curved forward (PI. 16, fig. 8; PI. 17, fig. 21) or
slightly angular medially (PI. 16, fig. 10; PI. 17, fig. 24).

Postocular part of fixed cheek is about as wide (tr.) as occipital ring is long (sag.). Palpebral lobe moderately
narrow (tr.), extends anteriorly from occipital furrow to a point slightly in front of glabellar mid-length
excluding occipital ring, slopes adaxially initially gently and then very steeply into axial furrow. Between a
and y anterior branch of facial suture is slightly less abaxially convex than its moderately convex outline
between y and e around palpebral lobe; posterior branch between e and £ runs parallel to outward and
backwardly directed section of axial furrow around anterior part of occipital ring, at £ it swings sharply
outward to genal angle. At y and between € and £ fixed cheek is exceedingly narrow.

Lateral margin of free cheek (PI. 16, figs. 12 and 14; PI. 17, fig. 17) broadly and evenly convex abaxially;
lateral border moderately convex (tr.) anteriorly, becoming gradually less so posteriorly, at junction of lateral
and posterior border furrows it is only very slightly less wide than at anterior branch of facial suture,
posteriorly from this junction it narrows more sharply along genal spine. Lateral border furrow is distinct,
continuous, and moderately deep and broad (tr.). Posterior border more gently convex (exs.) than lateral
border. Posterior border furrow slightly narrower (exs.) and more sharply defined than lateral border furrow.
Main field of free cheek gently to moderately inflated (tr. and exs.), slopes gently into lateral border furrow,
descends more steeply and suddenly into posterior border furrow. Eye socle very narrow, separated from
posterior border furrow by a moderately narrow adaxial extension of main field; visual surface bears numerous
small facets.

Rostral plate unknown. Hypostoma (PI. 16, figs. 18 and 23) subrectangular in outline with a strongly convex
(sag. and tr.) middle body. Anterior border is a narrow, sharply upturned rim; lateral border moderately
convex (tr.); posterior and posterolateral border behind lateral shoulder flattened. Maculae weakly indicated;
median furrow is absent between maculae, and very weak running forwards abaxially to merge with lateral
border furrow. Anterior wings moderately large. A pair of very small spines present on posterior margin.

Isolated thoracic segments (PI. 16, figs. 6 and 7) have a low preannulus about half as long (sag.) as
articulating half-ring. Articulating furrow sharply defined, intra-annular furrow much less so and longer (sag.
and exs.). Articulating half-ring and axial ring extremely weakly convex and of about equal length (sag.).
Axial and pleural furrows distinctly impressed, the latter with steeper anterior than posterior slope. Anterior
pleural band convex (exs.), higher than forward and gently downward sloping posterior band.

Pygidium is from 1-8 (PI. 16, fig. 25) to 2-25 (PI. 16, fig. 19) times as wide as long. Posterior outline varies
from being evenly convex backwards (PI. 16, fig. 17) to being medially more transverse (PI. 16, fig. 25) or
slightly forwardly convex (PI. 16, fig. 27). Axis is strongly convex (tr.), 0-9 to LI times as wide as long, about
0-7 times as long as pygidial length and 0-33 to 0-4 times as long as pygidial width, bears four distinct rings
(PI. 16, fig. 27), sometimes the suggestion of a fifth (PI. 17, fig. 18), and a terminal piece. First axial ring
gently to moderately convex (PI. 16, figs. 20 and 28; PI. 17, fig. 16), more posterior ones more or less flat.
Medially, posterior margins of first two or three axial rings are projected slightly backwards. Ring furrows
sharply defined, become successively weaker and fail to reach axial furrows posteriorly. Axial furrow distinct
but shallow, essentially marked by sharp break in slope between flat inner part of pleural area and sharply
rising side of axis, fades posteriorly around terminal piece due to short postaxial ridge (PI. 17, fig. 18). Pleural
region comprises an inner and an outer part which are marked off by the inner margin of the doublure
reflected on the dorsal surface of the cuticle, this margin running forward and outward from immediately
behind the axis (PI. 16, fig. 17; PI. 17, fig. 18). Outer part of pleural region descends very gently towards
pygidial margin. There are three pleural ribs. First pleural furrow is sharply defined, curves outward and
backward, more strongly backward at inner edge of doublure; successively posterior pleural furrows are
much less distinct. Interpleural furrows are more weakly marked than pleural furrows and are most distinct
abaxially above doublure. Rib profile is of the imbricate type.

Dominant sculpture on most specimens is striae (PI. 16, figs. 1 5, 17, 19, 20, 24-29), which on cranidium
and free cheek comprises fine, moderately continuous ridges. On glabella the ridges are convex forward, on
anterior border they are subparallel with anterior margin, on preglabellar field they are directed inward and
slightly forward. On posterior part of glabella (PI. 16, fig. 29) some ridges are broken up into small, scale-
like granules. On palpebral lobe ridges run forward and slightly abaxially, on blade-like postocular part of
fixed cheek they are transverse. Ridges on lateral border of free cheek run approximately exsagittally into
border furrow where they turn sharply to join those directed into this furrow from main field; ridges on
posterior border directed transversely. Anterior part of anterior border, outer part of lateral border, and
adaxial side of genal spine have two or three terrace lines which are also present on doublure of occipital
ring (PI. 16, figs. I and 8). Hypostoma has terrace lines on anterior and posterior lobes of middle body,
anterior wing, lateral and posterior borders. On anterior lobe they diverge posteriorly, converging again

SIVETER: SILURIAN TRILOBITES FROM IRELAND

1 19

between maculae; on posterior lobe they anastomose; they run on and parallel with lateral and posterior
borders. Axial ring and preannulus have forwardly convex fine ridges, pleurae with transverse ridges;
articulating half-ring lacks ridges. Pygidium covered with ridges which appear finer than those on cranidium
or free cheek.

Some less common cranidia, free cheeks, and pygidia (PI. 16, fig. 21; PI. 17, figs. 16 25) have a sculpture
which is dominantly more granular than linear in appearance, though on glabellar side including occipital
ring and on main field of free cheek fine, short ridges are formed by merger of granules.

Discussion. The Kerry taxon conforms on the whole to Snajdr’s diagnosis of Interproetus, but with
the following differences: the cranidial anterior border is only gently convex sagittally, there are
no anterior pits or suggestions thereof in the preglabellar furrow, and the preglabellar field and
pygidial doublure are not particularly narrow (PI. 16, figs. 1, 2, 27). At least the last two of these
distinctions also apply to some of the species originally referred to Interproetus (see generic
discussion), and the first two alone seem to carry little generic significance. I have interpreted the
sculptural variation as intraspecific in nature in the Irish Interproetus specimens (see description):
in these there is a common form with a linear sculpture (PI. 16, figs. 1-15, 17, 19, 20, 24-29) and
a rarer morphotype with a more granular sculpture (PI. 16, fig. 21; PI. 17, figs. 16-25); the
uncommon form also has an additional difference— a weak furrow abaxially on the occipital ring
(PI. 17, figs. 19-21). Silurian proetid specimens from elswhere showing similar sculptural differences
have been accommodated within a single species by other authors: see, for example, Interproetus
soncobrinus Snajdr, 1980 (pi. 49, figs. 2 and 3), I. walliseri (Alberti 1969, pi. 1, fig. 9; 1981, pi. 2,
fig. 16), or Decoroproetus anaglyptus Holloway, 1980 (p. 21). Also, unusually wide specimens in
the Kerry population (PI. 16, figs. 1 1 and 19) are regarded as conspecific.

Some of the many Interproetus species described by Snajdr (1980) have been distinguished on
the basis of very fine morphological details, and they come from similar or the same horizons and
localities; recognition of all of them as independent species seems questionable. This aside, of the
species he recognized I. galvani appears closest to the Bohemian /. vertumnus (Prantl and Vanek,
1958) from the Liten Formation, M. flexilis Biozone (mid-Wenlock), to /. numvertus Snajdr, 1980
from the boundary beds of the Liter) and Kopanina formations (late Wenlock to early Ludlow),
and to I. venustus (Barrande, 1846) from the Kopanina Formation, Ananaspis fecunda horizon
(mid to late Ludlow). I. vertumnus (a cast of the holotype of which I have examined) has a more
raised glabella and a shorter preglabellar area (from 0-2 to 0-25 times its glabellar length; Snajdr
1980, pi. 47, figs. 14 and 15); I. numvertus and /. venustus differ in having a shorter preglabellar
area, of the same ratio to their glabellae as that in vertumnus (Snajdr 1980, pi. 47, figs. 18-20;
pi. 48, fig. 8), and venustus also has wider palpebral lobes (Snajdr 1980, text-fig. 64c).

Morphologically the closest species to I. galvani appears to be I. walliseri (casts of the type
specimens of which I have examined); in particular both have a rather long preglabellar area. The
Irish and Moroccan material have been considered different because in walliseri the glabella stands
more proudly, rising more abruptly from the axial furrow, and the glabellar furrows are better
marked, especially lp furrow across which the sculpture is effaced. In the geographically close I.
peraticus the preglabellar area is shorter (0-2 times its glabellar length), the preglabellar field
relatively shorter and anterior border relatively longer and more convex sagittally, the glabellar
furrows seem better developed and lack sculpture, and there is apparently no ridge-like feature in
the anterior border furrow (cf. PI. 16, figs. 1, 2, 4 with Thomas 1978, pi. 11, fig. la-c).

Occurrence. Known only from the type horizon and locality.

Family aulacopleuridae Angelin, 1854

Discussion. Thomas and Owens ( 1978) reviewed this family which they regarded as comprising the
Scharyiinae Osmolska, 1957 in addition to the nominate subfamily and, following Bergstrom
(1973) and Fortey and Owens (1975), they considered the Otarionidae Richter and Richter, 1926
to be a junior synonym. Owen and Bruton (1980), Holloway (1980), and Chatterton and Wright
(1986) subsequently agreed with this synonymy but Pribyl and Vanek (1981) have used the junior

120

PALAEONTOLOGY, VOLUME 32

name Otarionidae to encompass the Aulacopleurinae Angelin, 1854, the Cyphaspidinae Pribyl,
1947, the Brachymetopinae Prantl and Pribyl, 1950, and the nominate subfamily. Pribyl and Vanek
(1981) did not comment on the earlier work of Thomas and Owens (1978). The phylogeny of the
Aulacopleuridae (= Otarionidae) and recognition of its constituent genera and subgenera are also
considerably different as portrayed by Thomas and Owens (1978, text-fig. 1 ) and Pribyl and Vanek
(1981, text-fig. 1). An assessment of such discrepancies lies outside the scope of the present paper,
in which I have preferred to adopt in general the outline higher classification of the former authors.

Subfamily aulacopleurinae Angelin, 1854
Genus conoparia Hawle and Corda, 1847

Type species. Conoparia convexa Hawle and Corda, 1847, p. 83; from the Devonian (Pragian) of Czechoslo-
vakia. By subsequent designation of Pribyl and Vanek 1981, pp. 173-174.

Discussion. The 1978 work of Thomas and Owens attempted, as far as was possible, to revise
aulacopleurine genera and subgenera; of these, Otarion (Otarion) Zenker, 1833 and their Cyphaspis
s.l. best potentially embrace the new aulacopleurine species described below.

Pribyl and Vanek (1981), as similarly advocated by Thomas and Owens (1978), restricted the
formerly widely conceived Siluro-Devonian Otarion (Otarion) to species corresponding closely to
O. (O.) diffraction Zenker, 1833, the type species. They regarded Cyphaspis Burmeister, 1843 as a
subgenus of Otarion , and resurrected Conoparia Hawle and Corda, 1847 as a subgenus of Otarion
for numerous Ordovician to Devonian species previously placed mainly in Otarion or Cyphaspis.

Thomas and Owens (1978, p. 1 1) considered that there is a group of Ordovician to Devonian
species related to Cyphaspis which possibly comprises several undifferentiated genera or subgenera.
It was such species that they referred to Cyphaspis s.l. until they became better known. It now
seems that Conoparia is the available name for at least some of these species. Pribyl and Vanek
(1981, pp. 174-175) included in Conoparia: O. (Otarion) horani Chatterton, 1971, O. (Otarion)
n. sp. of Chatterton 1971, and Otarion tridens Ingham, 1970. The first two of these species
were placed in Cyphaspis s.l. by either Thomas and Owens (1978, pp 71, 76-78) or Thomas
(1978, p. 31; the new species left unnamed by Chatterton being established therein as C. (s.l.)
elachopos), and tridens was implied by them to fall within this group or to be a precursor of it.

Conoparia hollandi sp. nov.

Plate 17, figs. 115

1976 Otarion sp.; Siveter in Parkin, p. 595, table 1 (pars).

1984 Cyphaspis (s.l.) sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Derivation of name. For Professor C. H. Holland, student of the Silurian geology of Ireland.

EXPLANATION OF PLATE 17

Figs. 1-15. Conoparia hollandi sp. nov. All specimens are from the Annascaul Formation, locality 36.
1-3, 12, holotype, cranidium, TCD 12587; dorsal stereo-pair, oblique, lateral views, x 5, frontal view, x 4.
4 and 8, cranidium, TCD 12529; lateral, dorsal views, x 5. 5, 9, 10, cranidium, TCD 12586; frontal,
dorsal, lateral views, x 5. 6, cranidium, TCD 12526; dorsal view, x 5. 7, cranidium, TCD 12527; dorsal
view, x 10. I I , free cheek, TCD 12531; ‘dorsal’ view, x 7. 13, pygidium, UM K28 16; dorsal stereo-pair, x 7.
14, free cheek, TCD 12627; ‘dorsal’ stereo-pair, x 7. 15, cranidium, TCD 12528; dorsal view, x5.

Figs. 16-25. Interproetus galvani sp. nov. All specimens are from the Annascaul Formation, locality 36.
16 and 20, pygidium, TCD 12191; lateral view, dorsal stereo-pair, x 5. 17, free cheek, TCD 12141; ‘dorsal’
stereo-pair, x4. 18, pygidium, TCD 11980; dorsal stereo-pair, x5. 19, 23, 24, cranidium, TCD 12189;
oblique, lateral, dorsal views, x 5; see also Plate 16, fig. 21. 21, 22, 25, cranidium TCD 12117; dorsal
stereo-pair, lateral, frontal views, x 5.

PLATE 17

SIVETER, Silurian trilobites

122

PALAEONTOLOGY, VOLUME 32

Holotype. Cranidium, TCD 12587; PL 17, figs. 1-3, 12.

Type horizon and locality. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.
Additional material. Six cranidia, three free cheeks, one pygidium.

Diagnosis. A Conoparia species with a relatively large basal glabellar lobe which is about one-third
as long as the strongly convex (sag. and tr.) glabella; mid-length of palpebral lobe opposite mid-
length of 2p lobe; preglabellar field very steeply to vertically sloping, very short (sag.) in dorsal
view, moderately high in frontal view. Pygidium has five or ?six axial rings, four or five pleural
ribs and a well-defined border furrow. Sculpture of small to medium-sized granules and small pits
on main field of free cheek and preglabellar field.

Description. Cranidium is about 1-9 times as wide as long. Glabella is about 0-9 times as wide as long and
0-4 times as wide as cranidium; it is strongly convex (sag. and tr.) and subparabolic in outline. Occipital ring
bears an extremely weak median tubercle in some specimens (PI. 17, fig. 7), is of more or less constant width
(sag. and exs.) behind central glabellar area; in small specimens (PI. 17, figs. 7 and 15) il is wider (sag.) than
anterior border, in larger specimens (PI. 17, figs. 1 and 8) they are more subequal in width. Behind basal
glabellar lobe occipital ring narrows (exs.) slightly and turns forward a little. Occipital furrow well defined
and moderately deep behind central glabellar area, deeper abaxially behind basal glabellar lobe. The latter
is well inflated, about one-third as long as glabella, tear-shaped. Furrow lp is moderately deep at axial
furrow, trends acutely backward and inward becoming shallower, is still distinctly impressed across neck of
lobe lp to occipital furrow, ‘Lobe’ 2p indicated by the slightest abaxial convexity of glabellar outline in
dorsal view (PI. 17, figs. 1, 6-9) and an exceedingly weak furrow 2p which is confined to the lower side of
glabella (PI. 17, fig. 3). Frontal glabellar lobe is well rounded in dorsal outline, strongly convex (tr. and sag.).
In lateral profile dorsal surface of glabella runs forwards for a short distance from occipital furrow before
descending in a strongly, evenly convex arc to preglabellar furrow which it slightly overhangs (PI. 17,
figs. 4, 10, 12).

Axial furrow wide (tr.) beside occiptal ring, slightly narrower (tr.) and deeper around lobe lp, is deepest
in front of lobe lp. Sagittally, preglabellar furrow is about as wide as occipital furrow, but is deeper here,
being sharply incised and undercutting the frontal glabellar lobe. Preglabellar area is just less than 0-25 times
as long as glabella in dorsal view (PI. 17, fig. 1). Preglabellar field is very short in dorsal view, about one-
fifteenth to one-twentieth as long as glabella and always shorter than the length of the occipital ring; in
frontal view (PI. 17, figs. 2 and 5) it is moderately high and in lateral profile slopes very steeply (PI. 17,
fig. 4) to almost vertically (PI. 17, figs. 10 and 12) to anterior border furrow. Abaxially the preglabellar field
merges imperceptibly into anterior part of fixed cheek (PI. 17, fig. 3). Anterior border furrow is clearly
defined. In dorsal view anterior border is shorter (sag.) than preglabellar field in small specimens (PI. 17,
figs. 7 and 15) and longer than preglabellar field in largest specimens (PI. 17, figs. 1 and 8), is flat to slightly
convex in profile and projects forward and very slightly downward from preglabellar field. Anterior margin
broadly convex (PI. 17, fig. 9) to weakly angular (PI. 17, fig. 15) in dorsal outline.

Posterior border of fixed cheek narrow (exs.) and moderately convex at axial furrow, widens gradually
and becomes more strongly convex (exs.) abaxially to facial suture. Posterior border furrow broad (exs.) and
moderately deep, rises less steeply anteriorly than posteriorly. Posterior part of fixed cheek gently convex
(exs.), anterior part descends very steeply forwards into preglabellar field. Adaxially from palpebral lobe,
fixed cheek is moderately convex and falls steeply into axial furrow. Palpebral lobe rises steeply from fixed
cheek, its mid-length lies oposite mid-length of 2p ‘lobe’, its posterior margin is opposite anterior abaxial
end of furrow lp, and anterior margin opposite furrow 2p. Weak, narrow eye ridge present on some specimens
(PI. 17, figs. 5 and 9). Shallow but distinct palpebral furrow present. Posterior branch of facial suture descends
from palpebral lobe to run very slightly sinuously outward and strongly backward to meet posterior margin
at an acute angle; anterior branch trends forward and very gently outward to anterior margin.

Rostral plate, hypostoma and thorax unknown. Visual surface of eye is semiglobular and high (PI. 17, fig.
14). Main field of free cheek is gently convex, descends into shallow lateral border furrow which is wider
and most distinct anteriorly, weakest posteriorly. A short (tr.) section of the posterior border furrow runs
into base of genal spine on free cheek. Lateral border rises quite sharply from border furrow, is tightly convex
(tr.). Lateral margin of free cheek moderately (PI. 17, fig. 14) to strongly (PI. 17, fig. 11) abaxially convex.
Genal spine deflected sharply outwards from curve of lateral margin, is slender, gently arched, and about as
long as main field of cheek, subcircular in section.

Pygidium (PI. 17, fig. 13) is bow-shaped in outline, about 3 0 times as wide as long. Axis is 0-3 times as

SIVETER: SILURIAN TRILOBITES FROM IRELAND

123

wide as pygidium, it is moderately convex (tr.), has five or possibly six axial rings (preservation imperfect in
this region on the one known pygidium) and a tiny terminal piece. Ring furrows become weaker posteriorly,
are complete across axis. Axial furrow most distinctly impressed and widest anteriorly, becomes narrower
and shallower posteriorly, scarcely continues behind terminal piece. Pleural region is 1-2 times as wide as
axis and has at least four pairs of ribs, anterior pleural band of each rib is narrower (exs.) than posterior
band, both pleural and interpleural furrows reach clearly marked border furrow inside narrow border.

Cranidium has medium to large sized, moderately closely (PI. 17, fig. 1) to sparsely (PI. 17, fig. 6) distributed
granules; posterior and anterior borders have fewer granules than other parts of the cranidium; granules
lacking in all furrows. Preglabellar field also covered with small pits which are slightly smaller and more
densely distributed than adjacent granules. Main field of free cheek has similar distribution and size of pits
and granules as preglabellar field. Very rare, small granules on lateral border and genal spine. Pygidium has
small granules along anterior and posterior pleural bands and axial rings.

Discussion. The generic assignment of this new species is somewhat equivocal as ideas on the
phylogeny, classification, and definition of aulacopleurid genera are in a state of flux. Species of
the Devonian Cyphaspis as defined by Thomas and Owens (1978, pp. 67, 70-71), whilst being not
far removed from the Kerry form, can be generically separated by their even more strongly swollen
glabella which markedly overhangs the preglabellar field and their wider pygidial axis which is
anteriorly as wide as the pleural region.

On the available morphology the new species (thorax unknown) shows quite a strong similarity
to species of Otarion ( Otarion ) as restricted by Thomas and Owens (1978, pp. 66-68). However,
in Otarion ( Otarion ) the inner edge of the articulating facet is diagnosed as being about two-thirds
of the pleural width abaxially from the axial furrow, but in the Irish taxon, based on the position
of the fulcrum on the posterior cephalic margin, it is only two-fifths abaxially from the axial furrow
(PI. 17, figs. 1, 8, 9). Although the pygidium of the new species is generally similar to that of O.
(O.) dijfractum (very wide and short, with a moderately large number of axial rings— five or six
and five to eight respectively — and a clearly marked narrow border and border furrow), in the
latter the pygidial pleural region anteriorly is about 1-5 times the width of the anterior part of the
pygidial axis, whereas in the former it is about 1-2 times as wide. In this respect the new species
stands approximately midway between Otarion ( Otarion ) and Cyphaspis (cf. PI. 17, fig. 13 with
Thomas and Owens 1978, pi. 7, figs. 3 and 46). Certain cephalic features of the Kerry species are
also different from those of O. ( O .) diffraction and from this subgenus as diagnosed by Pribyl and
Vanek (1981), which typically has a more slender glabella particularly in front of the basal lobes,
and a longer, deeper preglabellar field with numerous distinct genal cacae. Acceptance of the
phylogeny proposed by Thomas and Owens (1978, text-fig. 1), together with a Wenlock or earliest
Ludlow age for the Irish species, makes it seem even more unlikely that it belongs in Otarion
(Otarion). It would be the earliest member of the subgenus, and this would not fit in well with
their postulated derivation of Otarion ( Otarion ) from Otarion ( Aulacopleura ), as especially in its
glabella and preglabellar field the Irish species is not very close to O. (Aulacopleura) and it is
unlikely to stand phylogenetically between the latter (Ordovician to Devonian) and undoubted
species of the nominate subfamily (all being no older than mid-Ludlow in age), such as O. (O.)
dijfractum (cf. PI. 17, fig. 1 with Thomas and Owens 1978, pi. 7, figs. 2 a and 9).

Cyphaspis s.l. was not characterized by Thomas and Owens (1978), nor by Thomas (1978), other
than to say that it embraced species which differed in ‘various details’ from Cyphaspis s.s. However,
the morphology represented in this informal grouping can be gauged from that of C. (s.l.) elachopos
Thomas, 1978 (pi. 7, figs. 5, 8-13) from the Wenlock of the Welsh Borderland. I am unable
generically to divide the Kerry species from elachopos which now appears best placed in Conoparia,
as assigned by Pribyl and Vanek (1981). In slight variance to the diagnosis of Conoparia of the
latter authors, the pygidium of hollandi and of elachopos is not narrower than that of Otarion
(Otarion), and at least in hollandi there is a well-defined (not slight) pygidial border furrow. Whether
Conoparia becomes re-established as a useful genus or subgenus depends on future revision of
many of the species placed therein by Pribyl and Vanek (1981).

The most obvious specific differences of Conoparia elachopos from C. hollandi include it having

124

PALAEONTOLOGY, VOLUME 32

fewer pygidial axial rings (two to three) and pleural ribs (two). Otarion brauni Perry and Chatterton,
1979 from the Wenlock of north-west Canada is similar to the Irish species in gross morphology
and in particular both have a combination of pits and granular sculpture on the field of the free
cheek. The Canadian species differs because it has a straighter genal spine, only three to four
pygidial axial rings, three pairs of pygidial pleural ribs, and a much weaker pygidial border furrow.
Of the several aulacopleurid species described in recent years from the Siluro-Devonian of Australia,
the poorly known Otarion bowningensis (Etheridge and Mitchell, 1893) from the Devonian and
possibly Ludlow of New South Wales, redescribed by Chatterton (1971, p. 9, pi. 24, figs. 8-10),
seems nearest to the Kerry species, but it has four pygidial axial rings, a more weakly defined
pygidial border furrow and a gently abaxially convex lateral margin to the free cheek from which
the genal spine is apparently only weakly deflexed laterally. Otarion ( Conoparia ) inculpation Pribyl
and Vanek, 1981 (p. 195, pi. 8, figs. 7-8, pi. 9, fig. 5) from the Wenlock of Bohemia may be close
to C. hollandi , but detailed comparison is precluded on account of the poor steinkern preservation
of the three specimens belonging to the type suite, and Snajdr (1984c, p. 287) has subsequently
questioned the validity and type horizon of this species, though this claim has been repudiated by
Pribyl, Vanek and Horbinger (1985, p. 242). The complete steinkern figured by Snajdr (1984c, pi.
2, fig. 8) as C. cf. inculpation (Pribyl and Vanek), which was assigned with certainty to this species
by Pribyl, Vanek and Horbinger (1985, p. 243), appears to differ from the Irish species by its
relatively longer preglabellar field.

Subfamily scharyiinae Osmolska, 1957
Genus scharyia Pribyl, 1946

Type species. Proetus micropygus Hawle and Corda, 1847, p. 78; from the Kopanina Formation (Ludlow),
Czechoslovakia. By original designation.

Scharyia sp.

Plate 22, figs. 14, 17, 19

1976 Scharyia sp.; Siveter in Parkin, p. 595, table 1.

1984 Scharyia sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material Two cranidia.

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Description. Glabella subparabolic in outline, about 0-9 times as wide as long (sag.) including occipital ring.
Occipital ring has slight median swelling, which may represent an occipital tubercle (preservation imperfect),
sited closer to posterior margin than to occipital furrow. No trace of lateral glabellar furrows or lobes.
Glabella gently convex (sag.), in profile descends evenly anteriorly from occipital to preglabellar furrow (PI.
22, fig. 17).

Axial furrow shallow and widest beside occipital ring, narrows and becomes deeper at posterior one-third
of glabella exclusive of occipital ring, anteriorly it shallows before running into the preglabellar furrow where
it deepens a little. Preglabellar area is about 0-45 times as long as glabella and 0-3 times as long as cranidium;
preglabellar field very gently convex (sag.), one-third as long as glabella. Anterior border is about 0-25 times
as long (sag.) as preglabellar area, rises quite steeply from border furrow which is moderately deep and wider
(sag. and exs.) than axial, preglabellar, or occipital furrows. Anterior cranidial margin is broadly convex
forwards. Anterior branch of facial suture diverges forward and outward to abaxial part of preglabellar
furrow at an angle of about 30° to the sagittal line, course thereafter uncertain.

Discussion. This Scharyia material does not fit that of any described species and the features
displayed by the best preserved cranidium (PI. 22, figs. 17 and 19) suggest that a new species could
be established if this morphology were found to be constant in any future collections. The
combination of a relatively short preglabellar field, rather weakly diverging anterior branches of
the facial suture, and a relatively short anterior border (sag.) is distinctive. Both S. micropya

SIVETER: SILURIAN TRILOBITES FROM IRELAND

125

(Hawle and Corda, 1847) wenlockiana Pribyl, 1967 from the late Wenlock of Bohemia (see Owens
1974, pi. 98, fig. 13; Snajdr 1980, pi. 30, figs. 6-10) and S. siceripotrix Owens, 1974 (pi. 98, figs.
1-9, text-figs. 1a, c and 3) from the lowest Ludlow of Shropshire are not too dissimilar from the
Kerry species, but are immediately distinguished by having a shorter preglabellar field and more
widely anteriorly diverging facial sutures. The Scharyial sp. of Lane (1972, p. 352, pi. 61, fig. 12)
from the Wenlock (?Llandovery; see Perry and Chatterton 1979, p. 570) of Greenland also has a
moderately short preglabellar field and, if anything, even less diverging facial sutures anteriorly
than the Irish form, but here the preglabellar area is transversely narrower, it has a very large
occipital tubercle, a longer, sagittally convex anterior border with terrace lines, and a coarser
granular sculpture.

Family cheiruridae Hawle and Corda, 1847
Subfamily cheirurinae Hawle and Corda, 1847
Genus cheirurus Beyrich, 1845

Type species. Cheirurus insignis Beyrich, 1845, p. 12; from the Liten Formation (Wenlock) of Czechoslovakia.
By subsequent designation of Barton 1916, p. 129.

Cheirurus sp. nov.?

Plate 15, figs. 3, 9 12, 15, 16, 18-21

1976 Cheirurus sp.; Siveter in Parkin, p. 595, table 1 (pars).

1984 Cheirurus sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. Five cranidia, four hypostomata, six pygidia.

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Description. Cranidium (PI. 15, fig. 9) estimated at F6 times as wide as long (sag.). Glabella (PI. 15, figs. 9
and 15) about 0-7 times as wide at frontal lobe as long (sag.), and I I to F2 times as wide at frontal lobe
than at basal lobes, appearing to widen (tr.) only weakly anteriorly. Occipital ring (PI. 15, fig. 15) longest
(sag.) medially where it is convex and slopes quite strongly downwards into occipital furrow; laterally it has
a flatter dorsal surface and narrows (exs.) sharply to axial furrow where it is about half its sagittal length.
Occipital furrow strongly convex forwards behind central glabellar area where it is moderately deep, becomes
deeper trending backwards abaxially. Lobe lp subtriangular, slightly greater than one-third the glabellar
width at this point. Furrow lp is moderately deep, curves gently posteriorly from axial furrow to meet
occipital furrow in medially depressed area which is narrower (tr.) than basal lobe. Lobe 2p transversely
elongate, longer (exs.) adaxially than at axial furrow. Furrow 2p more or less transverse or is very gently
convex forwards, is deepest at axial furrow, reaches about one-third across glabella. Lobe 3p subrectangular;
furrow 3p very gently curved forwards, directed transversely overall, deepest at axial furrow where it meets
anterior pit, adaxially falls just short of inner point of furrow 2p. Frontal lobe about 0-4 times as long as
glabella including occipital ring, with strongly rounded outline (PI. 15, fig. 9) or with anterolateral margins
obliquely flattened (PI. 15, fig. 3), falls anteriorly in a steeply convex arch (PI. 15, fig. 16), is strongly convex
transversely. Axial furrow narrow and moderately deep at lp lobe anterior to which it becomes slightly wider
and shallower except at very deep anterior pit. Anterolaterally, preglabellar furrow is shallow but distinct
and merges with axial furrow; adaxially it becomes gradually narrower and more weakly impressed, not
reaching sagittal line. Anterior border is a narrow, tightly convex rim below anterolateral part of glabella,
becomes gradually finer and much less prominent adaxially.

Posterior border of fixed cheek (PI. 15, figs. 9, 10, 16) narrow (exs.) and tightly convex (exs.) adaxially,
widens slightly abaxially; border furrow narrow and rille-like, narrowest at genal angle, moderately deep.
Lateral border of fixed cheek tightly convex (tr.), widens (tr.) slightly posteriorly and is produced into a
short genal spine which on available material appears slightly incomplete but is about as long (exs.) as lobe
lp or one-fifth to one-sixth as long as glabella, and is directed more or less exsagittally. Outer, Posterior
part of fixed cheek descends steeply to shallow lateral border furrow; preocular part of fixed cheek is
small and narrow (tr.), bears a palpebral ridge which trends inward and slightly forward and at axial furrow

126

PALAEONTOLOGY, VOLUME 32

protrudes towards anterior pit as a short, sharply rounded projection. Palpebral lobe sited slightly less
than half-way across fixed cheek, posterior margin lies about opposite mid-length of lobe 2p, length
and form unknown; traces of palpebral furrow preserved. Posterior branch of facial suture runs outward
and very slightly forward abaxially before turning sharply backward for a short distance on lateral
border to meet lateral margin opposite about anterior part of lobe lp. Free cheek, rostral plate, and
thorax unknown.

Hypostoma (PI. 15, figs. 20 and 21) estimated at about 0-75 to 0-8 times as wide as long (sag.) with
maximum width at lateral shoulder. Anterior lobe of middle body very strongly convex (sag. and tr.), roughly
pear-shaped in outline, about 0-75 times as wide as long with a very strongly forwardly convex anterior
margin. Median furrow weak laterally, runs forward and outward from large, ovate, convex maculae to meet
lateral border furrow at anterior margin of lateral shoulder. Posterior lobe crescentic in outline, independently
convex (sag. and exs.) of and 0-2 times as long as anterior lobe, produced anterolaterally into long, gradually
and finely tapering extensions which end at lateral shoulder. Anterior border is extremely narrow and low
medially, not visible medially in ventral view due to highly convex anterior lobe, gradually widens (tr.)
posterolaterally and centrally contains a shallow furrow. Anterior border furrow medially is a shallow nick,
becomes broader posterolaterally where it runs into lateral border furrow. Lateral border at shoulder,
posterior to well-developed notch, is gently convex (tr.) and moderately wide (tr.), narrows rapidly posteriorly
before broadening again towards posterior border; border furrow is shallow and wide (tr.). Lateral margin
posterior to shoulder is broadly concave abaxially. Posterior border furrow shallow and wide (sag.); posterior
border flat, posterior extent unknown.

Dorsal morphology of pygidia known essentially from internal moulds only (PI. 15, figs. 12, 18, 19), ventral
morphology from one incomplete specimen with cuticle remaining (PI. 15, fig. 11). Axis has three rings and
a terminal piece. First ring imperfectly preserved, narrows adaxially, abaxial margin directed exsagittally.
Second ring has abaxial margin obliquely directed forward and slightly outward, narrows adaxially, has
ankylosed half-ring. Third ring apparently longer (sag. and exs.) than first two with what seems to be a weak,
ankylosed half-ring. Terminal piece is strongly convex, forms a half-cone with a shorter, more steeply
posteriorly descending anterior part. Postaxial sector is drawn posteriorly into a short, obtusely pointed
mucronation. Axial furrow distinct and narrow beside first two axial rings, shallower and broader beside
third ring; beside each ring furrow is a deep pit. First pleural furrow distinctly impressed, narrow, runs
obliquely outward and backward to divide two small, transversely elongate, boss-like dorsal projections of
the first pleura which represent the anterior and posterior pleural bands; anterior band is the larger and the
more prominent and expands (exs.) abaxially, posterior band expands adaxially. First interpleural furrow
distinct at axial furrow, shallows abaxially. Second pleural furrow shallower, shorter (tr.), and runs more
obliquely backwards than the first, divides a lower anterior from posterior pleural band. Second interpleural
furrow broader and shorter than first. Third pleural furrow lacking; shallow depression which represents
third interpleural furrow separates base of third marginal spine from median mucronation. Marginal spines
are slender, each is in the form of a flat ellipse in section, they taper gradually distally and are radially
disposed. First spine curves outward to opposite (tr.) posterior part of terminal piece, and moderately (PI.
15, fig. 12) to strongly (PI. 15, fig. 18) backward, at its base it is distinctly swollen; second spine slightly less
wide, of about the same length (the tips of all three spines in PI. 15, fig. 18 are missing), extends just behind
median mucronation, curves more backward than outward; third spine only slightly narrower than the
second, curves strongly backward. Between third pair of spines, along a posteriorly convex, median
embayment, doublure (PI. 15, fig. 11) is acutely reflexed dorsally from ventral facing part of pygidial ‘border’,
this reflex forming a sharp ridge. This ridge continues abaxially but at second spine is obtusely angled, the
doublure here being horizontal.

Glabella covered with small to medium sized, moderately densely scattered granules which become more
sparsely distributed on abaxial parts of lateral glabellar lobes. Fixed cheek has scattered large pits with a
row of pits alongside proximal part of posterior branch of facial suture. Flypostoma has anterior lobe and
lateral and posterior border with very closely spaced small granules; one or two larger, low granules present
near centre of anterior lobe. Border furrows and median furrow with very small sparser granules. Sculpture
of pygidium largely unknown; a few very small granules on ventral surface of second and third marginal
spines and doublure.

Discussion. Specimens of this species are incomplete and many are mainly in the form of internal
moulds. It is characterized by its glabella which apparently expands forwards relatively weakly
compared with other species, and by its slender pygidial marginal spines. The latter invite
comparison with those of Cheirurus infensus Campbell, 1967 from the late Wenlock to early Ludlow

SIVETER: SILURIAN TRILOBITES FROM IRELAND

127

Henryhouse Formation of Oklahoma but in infensus the spines, particularly the second and third,
are directed distinctly upwards posteriorly, they have a more nearly circular outline in section and
the first spine curves slightly forwards proximally before swinging backwards. Additionally in
infensus the glabella expands more rapidly forwards, having a ratio of glabellar width across the
frontal to basal lobes of 1-3 to 1-4:10 (own measurements based on the figures of Campbell),
compared with that estimated for the Irish species of I I to 1-2 : 10; the genal spine is longer; the
glabellar granules are generally finer; the hypostoma is slightly wider relative to its length and it
has better developed anterior wings which project well beyond the lateral shoulder, a wider lateral
border, a less anteriorly extending lateral part to the crescentic posterior lobe, weaker maculae,
and even finer granulation over most of the middle body.

The pygidial marginal spines of Cheirurus phollikodes Holloway, 1980 from the Wenlock of
Arkansas and Oklahoma are in some specimens (Holloway 1980, pi. 6, fig. 22; pi. 7, fig. 21) as
slender as those of the Irish species, though there are other pygidial, cranidial, and hypostomal
differences, too numerous to list, between the two. Cheirurus prolixus Holloway, 1980 from the
same general provenance and horizon is equally distinct, most obviously in its widely diverging
genal spines.

Cheirurus tarquinius Billings, 1869 from the West Point Formation (Ludlow) of Gaspe remains
essentially unrevised in modern terms, although it has been commented on by Raymond (1916),
Campbell (1967), Lane (1971), and Holloway (1980), and re-illustrated by Northrop (1939) and
Kindle (1945). Billings (1869, pi. 3, fig. 22; see also Northrop 1939, pi. 27, fig. 3) depicted the eye
in his line-drawing of the lectotype (Raymond 1916, p. 38) as being wholly opposite the 3p lobe
and he regarded this as an important specific character. My examination of the lectotype (GSC
3081) confirms that the illustration of Billings is accurate, and the one other syntype cranidium
with the eye preserved (GSC 308 \d) has it in a similar position, as does one well-preserved
cranidium figured by Kindle (1945, pi. 68, fig. 3; right cheek). This feature separates tarquinius
and its genal spine seems longer and more outwardly directed, and its pygidium to have more
broadly based border spines and a longer, more acutely pointed terminal mucronation (Northrop
1939, pi. 27, figs. 1 and 2; pi. 28, figs. 3-5; Kindle 1945, pi. 68, fig. 5).

Cheirurus centralis Salter, 1853 and the closely related C. insignis from the Wenlock of the
Anglo-Welsh area and Czechoslovakia respectively, have at least more anteriorly expanded glabellae
and broader, shorter pygidial spines (see Lane 1971, pi. 1, figs. 1-6; text-fig. 3; Pribyl, Vanek and
Pek 1985, pi. 6, figs. 1 and 2). The low Wenlock C. cf. centralis of Thomas (1981, p. 57, pi. 15,
figs. 3, 4, 7) has slender pygidial spines, but in this material a distinct second pleural furrow is
lacking and the glabella widens strongly at the frontal lobe as in centralis. Incidentally, the neotype
selected for insignis by Prantl and Vanek in Horny el al. (1958) fr-om Barrande’s Collection is
invalid; Dr H. Jaeger (Humboldt-Universitat, Berlin) has commented that Beyrich’s type of C.
insignis ‘is well preserved in our collection’ (pers. comm. 29 Dec. 1976; this contradicts Pribyl and
Vanek 1975, p. 48).

The eye ridge on the Irish material is similar to that on Cheirurus postremus Lane, 1971 from
the Ludlow of the Lake District, which its author claimed was unique for a Cheirurus species. But
often such features are more prominent on internal moulds, like the only two known specimens
of postremus and the Irish cranidia, and the same albeit more subdued feature appears to be
present on other Cheirurus species known from testiferous specimens, such as C. infensus (see
Campbell 1967, p. 18, pi. 6, fig. 4; pi. 7, fig. 1), or C. phollikodes (see Holloway 1980, pi. 7, figs.
1 and 8). The usefulness of this feature as the main diagnostic character of postremus would seem
in doubt, but the British species appears to be distinct from the Irish in its more anteriorly widening
glabella (1-3 to 1-4 times as wide at the frontal compared with basal glabellar lobes: own
measurements).

The Irish Cheirurus species appears to be new, but I hesitate formally to establish it because of
the incomplete nature of the material.

128

PALAEONTOLOGY, VOLUME 32

Cheirurus sp.

Plate 15, figs. 4, 6, 13, 14, 17

1863 Cheirurus bimucronatus; Baily in Jukes and Du Noyer, p. 12.

1878 Cheirurus bimucronatus ; Baily in Kinahan, p. 40.

1976 Cheirurus sp.; Siveter in Parkin, p. 595, table 1 (pars).

1984 Cheirurus sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. Two cranidia, one hypostoma, one pygidium.

Occurrence. Annascaul Formation, locality 28 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Discussion. This Cheirurus material can easily be distinguished from Cheirurus sp. nov.? by, amongst
other features, its much shorter, broader pygidial spines, more weakly transversely and sagittally
convex middle body of the hypostoma and, apparently, more anteriorly expanding glabella which
is slightly less than 1-3 times as wide at the frontal lobe than at lp lobe. It belongs within the
Wenlock -lower Ludlow group which includes C. insignis , C. centralis , and the Moroccan C. strux
Alberti, 1970, but without knowledge of its variation its specific identity remains uncertain. The
cranidia of all these are very similar. The pygidial spines of the Kerry form are similar to those
of centralis but in the latter the spine tips appear to lie on a more posteriorly gentle arc, as the
first spine extends backwards beyond and the second spine well beyond the median mucronation,
with the tips of the second and third spines being more nearly transversely in line (Lane 1971,
pi. 1, figs. 2, 6-8, 11, 12; Thomas 1981, pi. 15, fig. 2b). In insignis the pygidial spines appear
generally slimmer than in the Irish specimen, though in both the tips seem more strongly radially
disposed than in centralis (see Prantl and Vanek in Horny et al. 1958, pi. 7, fig. 2; Horny and
Bastl 1970, pi. 14, fig. 5; Lane 1971, p. 15, text-fig. 3 d, h; Pribyl, Vanek and Pek 1985, pi. 6,
fig. 2). The pygidium of C. strux is unknown.

Family calymenidae Milne Edwards, 1840
Subfamily calymeninae Milne Edwards, 1840
Genus calymene Brongniart in Brongniart and Desmarest, 1822

Type species. Calymene blumenbachii Brongniart in Desmarest, 1817, p. 517; from the Much Wenlock
Limestone Formation, Homerian Stage, Wenlock Series, Dudley, West Midlands, UK. By subsequent
designation of Milne Edwards in Cuvier 1844, pi. 80, fig. la, b. Siveter (1986, p. 785) and Whittington and
Siveter (1986, pp. 105 106) thought Shirley (1933) was the first author to validly designate the type species
of Calymene , but Dr L. B. Holthuis (Rijksmuseum van Natuurlitke Historie, Leiden, The Netherlands)
subsequently informed (June 1986) the International Commission on Zoological Nomenclature that Milne
Edwards was the first. The prior designation by Milne Edwards, together with the central nomenclatural
proposals of Whittington (1983) and Whittington and Siveter (1986) regarding Calymene , have now been
accepted by the ICZN (1987).

Calymene endemopsis sp. nov.

Plate 18, figs. I 7, 15 17, 20, 21, 23

1976 Calymene sp.; Siveter in Parkin, p. 595, table 1 (pars).

1984 Calymene sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Derivation of name. From the Greek, endemos , living in, and apopsis , a lofty spot that gives a commanding
view. Alluding to its locality high on the flank of Caherconree.

Holotype. Incomplete cranidium, TCD 12095; PI. 18, figs. 1-3.

Type horizon and locality. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.
Additional material. Eight cranidia, five pygidia.

SIVETER: SILURIAN TRILOBITES FROM IRELAND

129

Diagnosis. A species of Calymene with a very strongly bell-shaped glabella which becomes distinctly
narrower in front of lobe 2p; frontal lobe 0-6 times as wide as glabella at lobe lp; 3p lobe small.
Axial furrow narrow around lp lobe, wide beside 3p and frontal lobe. Preglabellar area relatively
long (sag.), estimated at 0-25 to almost 0-3 times the glabellar length (including occipital ring).
Moderately long (sag.), deep preglabellar furrow and subequally long, high anterior border,
Pygidium with eight axial rings and five pleural furrows. The latter are deeply impressed and
wide over inner pleural region and abruptly die out well before lateral margin; interpleural
furrows very much weaker and longer than pleural furrows, sharply and narrowly defined at
axial furrow, scarcely impressed on inner pleural region, clearly and finely impressed on outer
pleural region.

Description. Glabella very strongly bell-shaped, about 0-8 (PL 18, fig. 7) to 10 times (PI. 18, fig. 1) as long
as wide. Occipital ring (PI. 18, fig. 7) longest (sag. and exs.) behind central glabellar area, narrows gradually
abaxially. Occipital furrow longest and shallowest behind central glabellar area, deep and slit-like abaxially.
Lobe lp large, well inflated, subquadrate in outline with abaxial anterolateral corner obtusely pointed, joined
by a rather narrow (exs.), slightly depressed neck to central glabellar area. Furrow 2p is deep and narrow at
axial furrow, runs inward and backward, forks at anterolateral corner of lobe lp, posterior branch directed
more sharply backwards then curves more or less transversely inwards for a short distance and fades out,
anterior branch trends forward and inward; within this fork is a distinct intermediate lobe. Glabellar lobe
2p elongate (tr.), trends slightly forward abaxially, divorced from central glabellar area by moderately shallow,
broad (tr.) furrow; 2p furrow short (tr.), confined largely to side of glabella. Lobe 3p gently to moderately
inflated, small in comparison to 2p lobe, confined to dorsolateral surface of glabella, marked anteriorly by
narrow, shallow 3p furrow (PI. 18, fig. 3). Frontal lobe 0-6 times as wide as glabellar width at lp lobes,
moderately convex forward anterior outline, well-rounded anterolateral corners, falls very steeply to and is
slightly undercut by preglabellar and axial furrows (PI. 18, figs. 2 and 6), projects well in front of anterior
part of fixed cheek.

Axial furrow shallow beside occipital ring, is narrow and becomes progressively deeper anteriorly around
lp lobe, runs under bridge of lobe 2p and buttress from fixed cheek, is deep and becomes very much wider
(tr.) in front of 2p lobe due to glabella becoming distinctively narrower (tr.); at 2p lobe the base and lower
abaxial side of axial furrow is interrupted by a ridge. Anterior pit well marked, sited on adaxial side of base
of axial furrow at about mid-length of frontal lobe. Preglabellar area relatively long, estimated at 0-25 to
almost 0-3 times as long as glabella. Preglabellar furrow is very deep, moderately long, anteriorly it curves
very steeply upwards or upwards and slightly backwards and runs into inner face of anterior border; it
extends laterally to separate anterior border from fixed cheek where it is slightly less deep. Dorsal surface of
anterior border is slightly shorter than or is about the same length (sag.) as preglabellar furrow, of more or
less constant length (exs.) abaxially until shortening near facial suture, reaches high up the face of frontal
lobe, has a moderately to strongly convex surface which curves over anteriorly into a near vertical anterior
face that trends downward then slightly backward (PI. 18, figs. 2, 6, 16).

Posterior border of fixed cheek imperfectly preserved, lengthens (exs.) abaxially to fulcrum; border lens-
like with shallower anterior slope. Moderately sloping palpebral lobe is about as long or slightly longer than
lobe 2p, mid-length just anterior to mid-length of lobe 2p. Width of cranidium at posterior margin of
palpebral lobe about 1-7 times that of glabella at lobe 2p. Anterior part of fixed cheek is quite strongly
convex (exs.) and descends steeply forwards. Posterior branch of facial suture runs outward and slightly
backward from eye, then bends more sharply backwards to run to and across lateral border furrow after
which it turns sharply more exsagittally; course thereafter uncertain. Anterior branch runs forward and very
slightly inward to anterior border, very sharply inward on outer face of anterior border towards broadly
arched rostral suture (PI. 18, fig. 16). Free cheek, hypostoma, rostral plate, and thorax unknown.

Pygidial axis has seven distinct and one indistinct axial rings and a terminal axial piece (PI. 18, figs. 17,
20, 21, 23); the eighth axial ring is marked posteriorly by a very weak ring furrow. Axial furrow well marked
beside axis, shallow anteriorly, deepest posteriorly abaxial to terminal piece, shallowest behind terminal piece.
Inner part of pleural region descends relatively steeply, outer part very steeply, to pygidial margin. The five
pleural furrows are very well incised and abaxially become rather wide (exs.) across inner pleural region, they
abruptly become very weak and narrow and fade out rapidly on outer part of pleural region. Interpleural
furrows distinct immediately adjacent to axial furrow, very narrow and weak adaxially on inner pleural
region, becoming gradually slightly more distinct abaxially across inner pleural region, are best developed
on outer pleural region where they extend well beyond pleural furrows, but do not reach pygidial margin.

130 PALAEONTOLOGY, VOLUME 32

Anterior pleural band behind lifth interpleural furrow very slightly raised, forms abaxial margin of posterior
sector.

At least the glabella and inner part of fixed cheek have small to large granules, moderately densely scattered;
on dorsal and outer face of anterior border, and abaxial part of fixed cheek, the larger granules appear to
be absent. Preglabellar and axial furrows smooth. Posterior border furrow with fine granules. Pygidium with
small granules close together on central part of axis, more loosely scattered on inner pleural region, very
densely distributed fine granules on outer pleural region and border roll.

Discussion. This new taxon is closest to Wenlock and Ludlow species of the C. blumenbachii plexus,
and certain of its diagnostic characters are similar also to those of some other Llandovery and
Pfidoli Calymene species. However, in particular its very strongly bell-shaped glabella which is
wide across the basal glabellar lobes and narrows sharply in front of the 2p lobes, its relatively
long and deep preglabellar furrow and subequally long, high anterior border, the disparity in
definition of its pleural and interpleural furrows, and its style of sculpture all combine to make it
different. The Wenlock age C. blumenbachii blumenbachii , C. blumenbachii neotuberculata Schrank,
1970, and C. blumenbachii subsp. nov. of Siveter 1980 all have a similar type of sculpture, but they
are immediately distinguished by their shorter preglabellar furrow and preglabellar area, and they
have the anterior border slightly swollen opposite the axial furrow (PI. 17, fig. 1, cf. Schrank 1970*7,
pi. 8, fig. 1; Siveter 1980, pi. 100, fig. 12; 1986, pi. 90, fig. 1, pi. 91, figs. 3 and 6). Also in
blumenbachii and its subspecies there is not such a marked difference between the depth of the
pleural and interpleural furrows, and the pleural furrows are longer (PI. 18, fig. 23, cf. Siveter
1980, pi. 100, figs. 1 1 and 16; 1986, pi. 90, fig. 5; pi. 91, figs. 4 and 5). Some of the varieties of C.
tentaculata (Schlotheim, 1820) from the Pfidoli of the Baltic area figured by Schrank ( 1 970a, b)
appear similar to the new Irish species in the form of their preglabellar area, but all of them have,
amongst other differences, a less variably sized, finer, and more closely-knit glabellar granulation
and a slightly abaxially convex outline to the pygidial axial furrow.

Of the large number of Llandovery to Pfidoli age East Baltic Calymene species described by
Mannil (1977, 1983), the Wenlock age C. restevensis Balashova, 1968 (= senior synonym of C.
mimaspera livonica Mannil, 1977; pers. comm. Dr Reet Mannil, 1985; see also Mannil 1982, 1986)
shows some similarity in the form of its preglabellar area to C. endemopsis, though in restevensis
the glabella has a more evenly tapering glabellar outline and lacks a strong intermediate lobe and
it has a relatively wider axial furrow beside the basal glabellar lobe. Calymene flabellata Mannil,
1983 (pi. 1, figs. 1-7) from the Ludlow age Kuressaare Stage has a glabellar outline reminiscent
of the Irish species, but it has a shorter preglabellar area and preglabellar furrow, four lateral
glabellar lobes, and longer, narrower pleural furrows.

Calymene planicurvata Shirley, 1936 from the Llandovery Aeronian Stage of the Shelve area,
Welsh Borderland, resembles C. endemopsis in its relatively long preglabellar area but in this older
species the preglabellar furrow is shallower, the glabella tapers more evenly and has a better

EXPLANATION OF PLATE 18

Figs. 1-7, 15-17, 20, 21, 23, Calymene endemopsis sp. nov. All specimens are from the Annascaul Formation,
locality 36. 1-3, holotype, cranidium, TCD 12095; dorsal stereo-pair, lateral, oblique views, xl-25.

4 and 7, cranidium, TCD 12093; oblique, dorsal views, x 1 -25. 5 and 6, cranidium, TCD 12071; dorsal,
lateral views, xl-25. 15 and 16, cranidium TCD 12094u; dorsal, frontal views, xl-25. 17, 20, 21,

pygidium, TCD 12067; dorsal view, posterior stereo-pair, lateral view, x 1 -25. 23, pygidium, TCD 12064;
oblique view, xl-25.

Figs. 8-14, 18, 19, 22, 24-27. Calymene sp. All specimens are from the Annascaul Formation, locality 36.
8, 14, 18, 27, cranidium, UM K2819; frontal, dorsal, lateral, oblique views, x4. 9 and 22, hypostoma,
TCD 12629; ventral, lateral views, x 7. 10, cranidium, TCD 12627; dorsal view, x4. 1 1, cranidium, UM
K2820; dorsal view, x4. 12, pygidium, LIM K2818; posterior view, x 4. 13, rostral plate, UM K2817;
ventral view, x 7. 19, cranidium, TCD 12626; dorsal view, x 4. 24 and 25, cranidium, UM K2803; dorsal
stereo-pair, lateral view, x3. 26, free cheek, TCD 12625; ‘dorsal’ view, x4.

PLATE 18

SIVETER, Silurian trilobites

132

PALAEONTOLOGY, VOLUME 32

developed 3p lobe, a 4p lobe, and no intermediate lobe. The cranidium of Calymene kokbaitalensis
Maksimova, 1968 (pi, 8, figs. 1 and 2) from the Devonian of Kazakhstan looks very like that of
the Irish species as far as can be determined from the internal moulds of the Russian form, but
the much longer pygidial pleural furrows of kokbaitalensis immediately distinguish it (Maksimova
1968, pi. 8, figs. 2-4, 6, 7).

Occurrence. Known only from the type horizon and locality.

Calymene sp.

Plate 18, figs. 8 14, 18, 19, 22, 24-27

1976 Calymene sp.; Siveter in Parkin, p. 595, table I (pars).

1984 Calymene sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. Five cranidia, one free cheek, one rostral plate, one hypostoma, one pygidium.

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Discussion. The material of this second species of Calymene is mostly small in size, rather
incomplete, and one of the cranidia (PI. 18, figs. 24 and 25) has apparently suffered vertical
compaction. The morphology of the cranidium figured on Plate 18, fig. 14 is probably the closest
to that of the original form. This species is easily separated from C. endemopsis sp. nov. by its
much shorter preglabellar area, more evenly tapering glabella, relatively larger 3p lobe, the presence
of a 4p lobe, and the slightly denser, relatively smaller granular sculpture. The size of all the
material ascribed to Calymene sp. is much smaller than that of C. endemposis but the length and
form of preglabellar area are so different between the two groups of specimens that their
morphological differences are not thought to be ontogenetic in nature.

The short preglabellar area of Calymene sp., with its short, shallow preglabellar furrow and
short anterior border, invites comparison with such species as C. blumenbachii, C. minimarginata
Schrank, 1970, and C. tuberculosa Dalman, 1827 from the Wenlock of the Anglo-Baltic area, and
the Pridoli C. chica Snajdr, 1982 from Czechoslovakia. Of these both minimarginata, which has
palpebral lobes wider apart compared with its glabellar width at 2p lobes together with other
differences, and tuberculosa, which has a subconical (not elongate) protuberance on the middle
body of the hypostoma (Siveter, in prep.), a more rounded outline to the frontal glabellar lobe,
and which lacks an intermediate lobe within the fork of lp furrow (Schrank \910a), are clearly
different; blumenbachii also appears distinct in having a deeper preglabellar furrow and more
upturned anterior border. The poorly defined C. chica, which may be synonymous with C. nabrici
Snajdr, 1982 also from the Pridoli of Bohemia, is not well known though the age difference between
this species and the Irish Calymene material suggests that they are unlikely to be conspecific.
Further comparison is unwarranted on the basis of the present material from Kerry.

Family phacopidae Hawle and Corda, 1847
Subfamily phacopidae Hawle and Corda, 1847
Genus ananaspis Campbell, 1967

Type species. Phacops fecundus Barrande, 1846, p. 46; from the Kopanina Formation (Ludlow), Kolednik
near Beroun, Czechoslovakia. By original designation.

Discussion. In establishing Acernaspis (mainly Llandovery in age) and Ananaspis (mainly Ludlow
to Pridoli in age), Campbell (1967; see also 1977) recognized that there were transitional (Wenlock)
species in the evolutionary plexus between the two genera; these species could be expected to show
differential rates of development of the features characterizing typical members of each genus and
they would therefore be difficult to unequivocally assign generically. Ramskold (1985) has
subsequently highlighted such a gradual transformation from typical Acernaspis to typical Ananaspis

SIVETER: SILURIAN TRILOBITES FROM IRELAND

133

species showing, for example, that his new Gotland species Acernaspis rubicundula has, albeit much
more weakly developed, the type of glabellar granulation that has been regarded as characteristic
of Ananaspis. The Irish phacopid described below has glabellar sculpture which is similar to that
of A. rubicundula but is more primitive than that in late Wenlock specimens of A. stokesii (Milne
Edwards, 1840), the latter being regarded by Ramskold as the first definite Ananaspis. I agree with
Ramskold that to put all the Acernaspis and Ananaspis species into a single genus is too unwieldy
and so have retained them as separate. The Irish phacopid has been placed in Ananaspis mainly
on account of its distinct lateral border furrow, which seems to me one of the most reliable ways
of distinguishing between the groups of species assigned to each genus. On the latest Llandovery
to latest Wenlock morphological series indentified by Ramskold (1985, p. 13), Acernaspis sororia
Ramskold, 1985-/4. rubicundula- A. stokesii- Ananaspis amelangi Ramskold, 1985, the Irish species
fits best between rubicundula and stokesii.

Ananaspis sp. aff. A. stokesii (Milne Edwards, 1840)

Plate 19, figs. 1 27; Plate 20, figs. 8-10

1976 phacopids; Siveter in Parkin, p. 595, table I (pars).

1984 Ananaspis sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. Seventy cephala, three hypostomata, twenty pygidia.

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Description of small-eyed dimorph. Cephalon 1-8 (PI. 19, fig. 9) to 2-25 (PI. 19, fig. 1) times as wide as long.
Glabella 0-9 (PI. 19, fig. 7) to 1-2 (PI. 19, fig. 1) times as wide across frontal lobe as long. Occipital ring
moderately convex (PI. 19, fig. 24) to flat-topped (PI. 19, fig. 6) in profile, descends moderately to strongly
anteriorly into occipital furrow, abaxially behind lobe lp it bends gently forward and is slightly greater than
half its sagittal length. Occipital furrow distinct, shallowest medially, deepest abaxially, with a very short,
subsidiary posterior notch which indents anterolateral margin of occipital ring posterior to inner margin of
lobe lp. The latter is discrete, swollen, transversely elongate, and clearly marked off from intercalating ring
by a distinct furrow. Furrow lp deepest in front of Ip lobe, becomes shallower adaxially where it runs
parallel with occipital furrow (PI. 19, fig. 25) or swings more strongly forward (PI. 19, fig. 16); the two lp
furrows join medially to form an intercalating ring. On most small to medium sized specimens furrows 2p
and 3p are well marked (PI. 19, figs. 1 and 9), on large specimens they are weak (PI. 19, figs. 7, 14, 16).
Furrow 2p gently (PI. 19, fig. 9) to moderately (PI. 19, fig. 1) arched forward, isolated from axial furrow.
Posterior section of furrow 3p weakly convex forward, is of about the same definition or is slightly less well
defined than furrow 2p; straighter anterior section trends forward and outward towards axial furrow, fades
out high on glabellar side about opposite anterior margin of palpebral lobe. ‘Lobe’ 3p slightly longer than
iobe’ 2p. Frontal glabellar lobe broadly convex to subangular in anterior outline (PI. 19, figs. 7 and 22),
extends slightly anterolaterally. In lateral profile glabella is moderately convex dorsally from occipital furrow
to anterior face of frontal lobe where it curves relatively sharply to descend vertically to preglabellar furrow.

Axial furrow weakest beside occipital ring, becomes a little deeper and more sharply defined around lobe
lp, at furrow lp turns sharply outwards at an angle of about 30° (PI. 19, fig. 7) to 35° (PI. 19, fig. 1) to the
sagittal line and is shallow to moderately deep from here anteriorly to anterior margin of palpebral lobe
where it becomes slightly deeper before merging with lateral border furrow. Preglabellar furrow shallow but
sharply defined, in frontal view it dips slightly ventrally medially. Anterior border is a very narrow, exceedingly
short rim.

Posterior border narrow and tightly convex (exs.) to fulcrum, widens and becomes gradually less convex
from fulcrum to genal angle; border furrow very narrow and rille-like throughout its length (tr. ). Length of
fixed cheek between posterior border furrow and posterior margin of palpebral lobe is equal to or slightly
less than length of lp lobe. Palpebral lobe strongly abaxially convex, in dorsal view trends overall obliquely
forward and inward and in lateral and frontal views slopes forward and downward; palpebral furrow is
subparallel and well marked, Just adaxial to outer margin of palpebral lobe there is a shallow, parallel furrow
which confines a very narrow, abaxial border to palpebral lobe; this furrow pinches out anteriorly and
posteriorly at visual surface of eye. Facial suture ankylosed, posterior branch well marked, running moderately
strongly forward to lateral border furrow, very faintly present in some specimens on dorsal part of lateral

134

PALAEONTOLOGY, VOLUME 32

border (PI. 19, fig. 6); anterior branch cuts across axial furrow and anterolateral corner of frontal lobe to
run along lower side of this lobe. Eye moderately large in medium and large sized holaspides, visual surface
has from seventeen lens files with a total of eighty-six lenses (PI. 19, fig. 10) to twenty files and ninety-two
lenses (PI. 19, fig. 1), the maximum number of lenses in any file being six. A small holaspid (PI. 19, fig. 5;
glabellar length 3-7 mm) has seventeen files with a total of fifty-seven lenses; maximum number of lenses in
any file is, exceptionally, five, normal complement is four. Beneath visual surface some specimens show a
very narrow, weakly defined eye socle from below which the narrow, posteriorly expanding main field of the
cheek descends more or less vertically to lateral border furrow. The latter is shallow anteriorly, becomes
slightly less so and a little wider towards junction with facial suture posterior to which it becomes narrower
and more sharply defined to posterior border furrow. Below anterior part of eye lateral border is gently to
moderately convex (tr.) and relatively wide, gradually expands in width posteriorly to genal angle (PI. 19,
fig. 10).

Vincular furrow continuous and shallow medially, more deeply notched laterally (PI. 19, fig. 13). Medially,
doublure is wide (sag.) and hypostomal suture extremely weakly convex posteriorly. Hypostoma known
completely only in internal mould form, is shield-shaped and about 0-9 times as wide as long exclusive of
anterior wings. Convex middle body falls steeply abaxially and posteriorly to border furrows. Lateral border
narrow (tr.), posterior border only slightly wider (sag.). At junction of lateral and posterior margins and
sagittally the hypostomal outline is angular, but without evidence on the internal mould material available
of discrete spines. Lateral and posterior border furrows shallow and narrow.

Thorax unknown. Pygidium (PI. 19, figs. 15, 26, 27) is about 2T times as wide as long with a subrounded,
bow-like posterior outline. Axis about 13 times as long as wide and about 0-25 times as wide and 0-85 times
as long as pygidium, contains seven to eight rings. Axis is moderately convex (tr.) anteriorly, progressively
less so posteriorly where the axial rings are flat-topped. At least the second to sixth axial rings have ankylosed
half-rings. Lirst axial ring moderately convex (sag.), narrowest medially, at axial furrow lengthens to about
double its sagittal length; difference in length between sagittal and abaxial parts of axial rings becomes less
marked posteriorly. More posterior ring furrows fail to reach axial furrow. Terminal axial piece is extremely
short (sag.) and bluntly rounded. Axial furrow very broad and shallow at first ring, becomes slightly narrower
and better marked from second axial ring to terminal piece around which it becomes shallower and weaker
again. Wide pleural region descends away from axial furrow in an even, moderately strong, dorsally convex
curve. There are five, finely incised pleural furrows which become progressively shorter and weaker posteriorly;
first furrow ends abruptly on reaching the well-developed articulating facet. Interpleural furrows are much
broader than pleural furrows and are extremely shallow. Both pleural and interpleural furrows stop well
short of pygidial margin to leave a wide, smooth ‘border’ around outer part of pleural region.

Glabella and palpebral area have very subdued, scattered, medium-sized granules which themselves are
composed of several minute granules (PI. 19, figs. 1 and 12); in large specimens the medium-sized granules
are less frequent and of exceedingly weak relief. On the palpebral area of some of the larger cephala there
are a few very shallow ‘pits’. Internal moulds of glabellae show a suboval arrangement of raised auxiliary
muscle impressions medially on the frontal lobe (PI. 19, fig. 18). Internal moulds of hypostoma have a few
dispersed medium-sized granules on anterior and central part of middle body. Pygidium has no obvious
granulation.

EXPLANATION OF PLATE 19

Pigs. 1-27. Ananaspis sp. aff. A. stokesii (Milne Edwards, 1840). All specimens are from the Annascaul
Formation, locality 36. 1, 5- 10, 12-16, 18 20, 22, 24-27, small-eyed dimorph. 1 and 6, cephalon, UM

K2800a; dorsal, lateral views, x 4. 5, 9, 12, 13, 24, cephalon, TCD 12175; oblique, dorsal, frontal, ventral,
lateral views, x 6. 7 and 8, cephalon, UM K2827; dorsal, oblique views, x 2. 10 and 25, cephalon, partial
internal mould, TCD 12061; oblique view of cheek, x 7, dorsal view, x 2. 14, cephalon, TCD 12166;

dorsal view, x 2. 15, 26, 27, pygidium, TCD 12179; lateral, dorsal, posterior views, x 3. 16, cephalon,

TCD 12169; dorsal view, x 3. 18 and 22, cephalon, largely internal mould, TCD 12633; frontal, dorsal

views, x 2. 19, hypostoma, internal mould, TCD 12146; ventral view, x 5. 20, pygidium, internal mould,
TCD 12185; dorsal view, x 3. 2-4, 17, 23, ?11, ?21, large-eyed dimorph. 2-4, 23, cephalon, largely internal
mould except for palpebral area and visual surface of eye, UM K2826; oblique, dorsal, lateral views, x 2,
eye and cheek, x 7. 17, cephalon, TCD 12156; dorsal view, x 2. II and 21, pygidium, UM K2802a;

dorsal, oblique views, x 3.

PLATE 19

SIVETER, Silurian trilobites

136

PALAEONTOLOGY, VOLUME 32

Description of the large-eyed dimorph. This takes the form of a differential comparison with the small-eyed
morph (cf. PI. 19, figs. 2, 3, 7, 8, 10, 17, 23). The large-eyed morph differs in having: 1, an eye with at least
twenty-one lens files and on the one specimen of this dimorph with the eye preserved (PI. 19, fig. 23) there
is room on the incompletely preserved visual area for an extra file, making the probable total twenty-two.
The maximum number of lenses in any file is nine and the estimated total number of lenses is 165-170; 2,
a much shorter postocular section on the cheek, the eye almost reaching the posterior border furrow; 3, a
narrow cheek field which below the visual surface descends very strongly inward to a very narrow, deep, slit-
like lateral border furrow; 4, a lateral border which below the anterior part of the visual surface is much
narrower, and expands in width posteriorly at a much greater rate.

One large phacopid pygidium (PI. 19, figs. 1 1 and 21) differs from the others in having: 1, a straighter,
obliquely directed lateral margin; 2, one more (9) axial ring; 3, one more (6) pleural furrow; 4, a more pointed
posterior outline to the terminal axial piece; 5, a pleural region which has a more abrupt change in slope
between its inner and outer parts descending abaxially from the axial furrow, so that the transversely wider
outer part falls a little more steeply towards the lateral margin. This pygidium may represent that of the
large-eyed morph.

Discussion. Two distinct forms have been identified in the Caherconree phacopid material; they
are distinguished essentially on the morphology of the eye and associated characters of the cheek,
though there may be pygidial differences also (see description). Of the fifteen prepared phacopid
cephala two specimens (PI. 19, figs. 2-4, 17, 23) are of the large-eyed morph which, following the
work of Eldredge (1973), Campbell (1977), and others on this family, I have interpreted as a
sexually dimorphic form. Sexual dimorphism has been increasingly invoked in the literature as a
feature of Devonian phacopids; it has not been recognized hitherto in Ananaspis with the exception
of the suggestion of Holloway (1980) that the co-existing A. fecunda fecunda and A.fecunda aspera
from the Ludlow of Czechoslovakia, which differ in the size of their eyes and in other features
(see Chlupac 1977), are probably sexual dimorphs. Non-dimorphic variability in the Kerry material
which is considered to be intraspecific includes the definition of the 2p and 3p glabellar furrows,
the degree of divergence of the axial furrows, and the distinctness of the larger glabellar granules
(see description). The latter, at least, may be ontogenetic in nature with the larger specimens having
the weaker sculpture, though this variability may be controlled by specimen preservation as may
that of the lateral glabellar furrow definition.

The Kerry taxon is most clearly related both to A. amelangi from the latest Wenlock Mulde
Beds of Gotland, and to similarly aged British specimens that I have examined which have been
referred to the unrevised A. stokesii , a species established on the basis of material from England,
Bohemia, and the United States. A. amelangi differs from the small-eyed dimorph of the Irish
species in having from one to four less lens files (sixteen) in equivalent, adult-sized holaspides, and
from the large-eyed dimorph it has at least five files less; it also has a prominent median node on
the occipital ring, more conspicuous glabellar granules, a 2p furrow positioned closer to lp furrow
thus forming an exsagittally shorter lp lobe (cf. PI. 19, figs. 1 and 9 with Ramskold 1985, pi. 4,
figs. 2a and 3 b), and less definite eighth and ninth ring furrows (cf. PI. 19, figs. 11 and 26 with
Ramskold 1985, p. 20, pi. 4, fig. 9) though this difference may be due to the amelangi pygidia
being smaller. Also in the Swedish form the lateral border furrow runs immediately below the eye,
contrasting at least with the small-eyed dimorph of the Irish species which has a narrow but definite
strip of the cheek field separating this furrow from the eye base (cf. PI. 19, fig. 6 with Ramskold
1985, pi. 4, fig. 2b); in the large-eyed dimorph of the Irish form the border furrow runs very close
to the eye base similar to that in amelangi , but the latter still differs in not having a sharply
inwardly descending narrow strip of the cheek field below the eye together with having a much
shallower lateral border furrow.

A. stokesii has not received attention since Salter (1864, p. 21); a revision is being prepared by
Dr A. Thomas but meanwhile the nature of the type material, horizon, and locality and the
variability within the species remains either unknown or imprecisely defined. Ramskold (1985,
pi. 4, fig. 7a, b ) has figured a specimen attributed to stokesii from the Much Wenlock Limestone
Formation of the Malverns and I have looked at material assigned to stokesii, housed in the British

SIVETER: SILURIAN TRILOBITES FROM IRELAND

137

Museum (Nat. Hist.) from the same formation of the Dudley area, the latter occurrence forming
one of the type localities of Milne Edwards in his account of this species. Comparison of the Kerry
phacopid with stokesii is based on these Dudley and Malverns specimens. The British species
appears slightly closer than amelangi to the Irish forms, but differs in its slightly stronger glabellar
granulation, deeper cephalic axial furrow, and exsagittally shorter 2p lobe; additionally, comparing
stokesii to equal-sized specimens of the small-eyed Irish dimoph, the cheek field below the eye is
narrower and falls sharply inward to a deep lateral border furrow which undercuts the base of the
eye, allied to which the anterior part of the lateral border is narrower, the border expands in width
more rapidly posteriorly and is more tightly convex. These differences are well exhibited on
comparing UM K2800r/ (PI. 19, figs. 1 and 6) with BM 59043 and BM 1310 (one of two specimens
with this number, the other being Eophacops musheni). Undercutting of the base of the eye by the
lateral border furrow is common to both stokesii and the large-eyed Kerry dimorph, but in the
former this furrow is less sharply incised and very much wider, and the eye is much smaller having
at least three fewer lens files (seventeen to eighteen). There are at least seven axial rings and five
pleural furrows in the pygidium of stokesii. Pending a full revision of the British species I prefer
to refer to the Irish material as aff. stokesii.

Family odontopleuridae Burmeister, 1843
Subfamily odontopleurinae Burmeister, 1 843
Genus odontopleura Emmrich, 1839

Discussion. The concept of Odontopleura and the number of species it contained (two) remained
stable up to and including the work of Thomas (1981), except for the brief note of Snajdr (1979)
resurrecting O. prevosti Barrande, 1846. Subsequently, Chatterton and Perry (1983) erected from
the early Wenlock of north-west Canada O. brevigena which, like the long established O. ovata
Emmrich, 1839 and O. dufrenoyi Barrande, 1846, also has paired occipital spines. They further
introduced four new Llandovery species which in overall morphology fitted best in Odontopleura ,
but differed from other taxa formerly placed therein in lacking paired occipital spines; additionally
they allied Taemasaspis Handover yiana Snajdr, 1975 (see also Snajdr 1978) from the Llandovery of
Bohemia to the latter group and placed it, also, in Odontopleura. I agree with Chatterton and
Perry (1983) that the Canadian and Bohemian Llandovery forms lacking paired occipital spines
are best regarded, if somewhat uneasily, as early representatives of Odontopleura. Ramskold (1984,
p. 249) independently proposed that T. llandoveryiana is morphologically (and stratigraphically)
intermediate between Primaspis and Odontopleura (see also Chatterton and Perry 1983, p. 23),
whilst Pribyl et al. (1986, p. 268) have most recently suggested that the material Snajdr (1975,
1978) placed in llandoveryiana belongs to both Primaspis ( Meadowtownella ) Pribyl and Vanek,
1965 and to Odontopleura.

Snajdr (1979, 1984 a, b , c) re-examined Bohemian Odontopleura specimens in combination with
his review of the type material of Hawle and Corda (1847), and he recognized thirteen species or
subspecies (see species discussion below), all with occipital spines, ranging in age from the late
Aeronian sedgwickii Zone of the Llandovery to the lower Ludlow nilssoni Zone. A new genus was
set up, Ivanopleura , containing the Wenlock flexilis Zone O. dufrenoyi , type species, together with
Miraspis rarissima Snajdr, 1975 from the sedgwickii Zone. The monotypic Borkopleura was also
introduced, based on a new late Wenlock species B. gorella.

Whereas O. dufrenoyi and M. rarissima differ collectively from Odontopleura in the characters
listed in the remarks of Snajdr (1984u, p. 49), it seems to me that in these two species the overall
form of the cephalon, thorax, and pygidium is so like that of Odontopleura that a division at the
subgeneric level is more appropriate. Pribyl et al. (1986, p. 267) also believed M. rarissima belonged
to Odontopleura. Thus in Bohemia early (Llandovery) species of Odontopleura seem to have
occipital morphologies which range from lacking paired occipital spines, to possessing either short-
or long-paired occipital spines, these different types being represented by O. ( Odontopleura )

138

PALAEONTOLOGY, VOLUME 32

Handoveryiana , O. ( Odontopleura ) perpeta Snajdr, 1984/?, and O. ( Ivcinopleura ) rarissima respectively,
all from the sedgwickii Zone of the Hyskov area.

Subgenus odontopleura (odontopleura) Emmrich, 1839

Type species. Odontopleura ovata Emmrich, 1839, p. 53; from a late Wcnlock to lower Ludlow age
Graptolithengestein glacial erratic of Niederkunzendorf, Silesia. By monotypy.

Odontopleura ( Odontopleura ) ovata Emmrich, 1839

Plate 20, figs. 1-7, 11-21; Plate 21, figs 1, 3, 4, 8

1839 Odontopleura ovata Emmrich, p. 53, pi. 1, fig. 3.

1847 Odontopleura Siemangii Hawle et Corda; Hawle and Corda, p. 147.

1847 Odontopleura Neumanni Hawle et Corda; Hawle and Corda, p. 151.

1847 Odontopleura tenuicornis Hawle et Corda; Hawle and Corda, p. 155.

1863 Acidaspis Jamesii', Baily (pars) in Jukes and Du Noyer, p. 12.

1878 Acidaspis Jamesii ; Baily (pars) in Kinahan, p. 40.

? 1930 Ceratocephala (Acidaspis) ovata Emmrich; Gaertner, p. 196, pi. 24, fig. 11.

1966 Odontopleura ovata Emmrich, 1839; Pribyl and Vanek, p. 294, pi. 6, fig. 5; ?pl. 9, figs. 2 4.

1967 Odontopleura ovata Emmrich, 1839; Bruton, p. 216, pi. 30, fig. 1 (with synonymy).

1967 Odontopleura ovata (Emmrich, 1834); Hucke and Voigt, pi. 28, figs. 13 and 14.

1968 Odontopleura ovata Emmrich, 1839; Bruton, p. 8, pi. 1, figs. 1-4, 6, 7.

1969 Odontopleura ovata Emmrich, 1839; Schrank, p. 707, pi. 1, figs. 1 -7; pi. 2, figs 1-5.

1970 Odontopleura ovata Emmrich, 1839; Alberti, p. 129, pi. 19, figs. 1-5.

1972 Odontopleura ovata ; Schrank, p. 7.

1973 Odontopleura ovata Emmrich, 1839; Pribyl and Vanek, p. 302, text-fig. 1-1.

1973 Odontopleura ovata Emmrich, 1839; Neben and Kreuger, pi. 103, figs. 1-4.

1976 Odontopleura ovata Emmrich; Siveter in Parkin, p. 595, table 1.

1979 Odontopleura ovata Emmrich, 1839; Snajdr, p. 171.

? 1979 Odontopleura prevosti Barrande, 1846; Snajdr, p. 171, pi. 1, fig. 1.

1981 Odontopleura ovata Emmrich, 1839; Thomas, p. 80, pi. 22, figs. I 3 (with synonymy).

1981 Odontopleura ovata Emmrich, 1839; Siveter in Holland, p. 78, fig. 51.2, 4, 5.

? 1983 Odontopleura ovata Emmrich, 1839; Chatterton and Perry, p. 18.

1984a Odontopleura ovata Emmrich; Snajdr, p. 50.

? 19846 Odontopleura prevosti prevosti Barrande, 1846; Snajdr, p. 97, pi. 4, figs. 1-4.

19846 Odontopleura prevosti tenuicornis Hawle et Corda, 1847; Snajdr, p. 98, pi. 4, figs. 5 and 6.
19846 Odontopleura prevosti adriana subsp. nov.; Snajdr, p. 99, pi. 4, figs. 7 and 8.

? 19846 Odontopleura salma sp. nov.; Snajdr, p. 99, pi. 3, figs. 8 and 9.

19846 Odontopleura siemangi Hawle et Corda, 1847; Snajdr, p. 100, pi. 2, figs. 5-8.

19846 Odontopleura omega omega subsp. nov.; Snajdr, p. 100, pi. 1, figs. 1-5.

19846 Odontopleura omega kalidasa subsp. nov.; Snajdr, p. 101, pi. 2, fig. 9.

EXPLANATION OF PLATE 20

Figs. 1-7, 11 -21. Odontopleura (Odontopleura) ovata Emmrich, 1839. All specimens are from the Annascaul
Formation, locality 36. 1 3, 13, cranidium, TCD 12245; dorsal stereo-pair, frontal view, x 3, lateral,

oblique views, x 4; figured Siveter in Holland 1981, fig. 51.4 as O. ovata. 4, 14, 19, 21, cranidium, TCD
12217; posterior oblique view, x 3-5, dorsal view, x 3, occipital organ, x 30, lateral view, x 3. 5, 17, 20,

cranidium, TCD 12048; dorsal view, x 3, frontal, oblique views, x 4. 6, hypostoma, TCD 12320; ventral
view, x 8. 7, cranidium, TCD 12592; dorsal view, x4. 11, free cheek, TCD 12328; ‘dorsal’ view, x 4;

figured Siveter in Holland 1981, fig. 51.2 as O. ovata. 12, pygidium, TCD 12362; dorsal view, x3. 15,

thoracic segment, TCD 12202; dorsal view, x4. 16, pygidium, TCD 12543; dorsal view, x3. 18,

hypostoma, TCD 12243; ventral stereo-pair, x 5.

Figs. 8 10. Ananaspis sp. aff. A. stokesii (Milne Edwards, 1840). Small-eyed dimorph, cephalon, UM K2801,
Annascaul Formation, locality 36; lateral, dorsal, oblique views, x 4.

PLATE 20

MfgP'

SIVETER, Silurian trilobites

140

PALAEONTOLOGY, VOLUME 32

19846 Odontopleura betina sp. nov.; Snajdr, p. 101, pi. 3, figs. 5-7.

19846 Odontopleura rataji sp. nov.; Snajdr, p. 102, pi. 3, figs. 2-4.

19846 Odontopleura cf. ovata Emmrich, 1839; Snajdr, p. 103, pi. 2, figs. 3 and 4.

1984c Odontopleura prevosti tenuicornis Hawle et Corda, 1847; Snajdr, p. 198, pi. 8, fig. 3.
1984c Odontopleura prevosti neumanni Hawle et Corda, 1847; Snajdr, p. 175, pi. 6, fig. 7.
1984c Odontopleura siemangii Hawle et Corda, 1847; Snajdr, p. 192, pi. 12, fig. 1.

1984 Odontopleura sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Holotype. Articulated specimen comprising incomplete cephalon, partial thorax and pygidium, HU K162;
figured Emmrich 1839, pi. 1, fig. 3; Bruton 1967, pi. 30, fig. 1; Hucke and Voigt 1967, pi. 28, fig. 13; Schrank
1969, pi. 1, fig. I; Neben and Krueger 1973, pi. 103, fig. 1.

Type horizon and locality. As for type species. See Schrank 1969, p. 726.

Material. 106 cranidia, 97 free cheeks, 28 hypostomata, 88 pygidia, isolated thoracic segments.

Description. See Bruton 1967, p. 217; Schrank 1969, p. 707.

Discussion. The type specimen of O. ( O .) ovata has the shortcomings of occurring in a glacial
erratic, having each paired occipital spine broken off at its base and, it seems from the illustrations,
having its glabellar sculpture somewhat abraded. With the considerable increase recently of species
assigned to O. ( Odontopleura ) (see generic discussion), it has become more important to recognize
the amount of variation within ovata , a task made difficult by the lack of assignable topotype
material. Therefore, for practical purposes, the extra Graptolithengestein erratic material of O.
( Odontopleura ) assigned to ovata by Schrank (1969) from northern Germany assumes greater
significance; I agree with this assignment and have used this material to assess the likely range of
variation in the species.

The Kerry O. ( Odontopleura ) material shows clear variation in the following features (all
considered to be intraspecific and the material to belong to ovata): the angle of divergence of the
paired occipital spines (PI. 20, figs. 5 and 14), the single or bicomposite eye ridge (PI. 20, figs. 13
and 20), the density and size of cranidial granules (pi. 20, fig. 7; PI. 21, fig. 1), the number (four
or five) of subsidiary pygidial spines external to the major border spines (PI. 20, fig. 12; PI. 21,
fig. 4).

The synonymy above reflects the extent to which I have accepted, at least on the basis of the
present evidence, that the many Bohemian O. ( Odontopleura ) species introduced or rehabilitated
by Snajdr (19846) deserve separate recognition from ovata. Many of these have been exceedingly
finely drawn, very often being based largely on minute sculptural differences, and have similar or
the same occurrences. In contrast, Snajdr (19846), who did not refer to Schrank's (1969) work,
did not identify ovata with certainty in Bohemia. O. (O.) prevosti prevosti Barrande, 1846 and O. (O.)

EXPLANATION OF PLATE 21

Figs. 1, 3, 4, 8. Odontopleura ( Odontopleura ) ovata Emmrich, 1839. All specimens are from the Annascaul
Formation, locality 36. 1, cranidium, TCD 12252m dorsal view, x4. 3, cranidium, TCD 12219; dorsal
view, x4. 4, pygidium, TCD 12050u; dorsal stereo-pair, x 3; figured Siveter in Holland 1981, fig. 51.5 as
O. ovata. 8, hypostoma, UM K2813a; ventral view, x 8.

Figs. 2, 5-7, 9-21. Leonaspis parkini sp. nov. All specimens are from the Annascaul Formation, locality 36.
2, cranidium, TCD 12275; dorsal view, x 5. 5-7, 9, holotype, cranidium, TCD 12207; oblique view, x 6,
frontal, lateral views, dorsal stereo-pair, x 5. 10, 12, 19, dorsal view, TCD 12624; dorsal, frontal, lateral
views, x 6. 1 1, pygidium, TCD 12203; dorsal stereo-pair, x 6. 13, free cheek, TCD 12576; ‘dorsal’ view,
x4. 14, thoracic segment, UM K2822; dorsal view, x 4. 15, cranidium, TCD 12206; dorsal view, x 5.

16, pygidium, TCD 12599; dorsal view, x 6. 17, pygidium, UM K2809; dorsal view, x 6. 18, free cheek,
TCD 12327; ‘dorsal’ stereo-pair, x4. 20, cranidium, TCD 12202; dorsal view, x6. 21, free cheek, TCD
12301; ‘dorsal’ view, x 6.

PLATE 21

SIVETER, Silurian trilobites

142

PALAEONTOLOGY, VOLUME 32

salma Snajdr, 1984/) (see also Snajdr 1979, pi. 1, fig. 1) may differ from ovata but both are based
on compressed mould specimens and are questionably put in synonymy. Both prevosti prevosti and
salma occur in the testis Biozone at the same localities and seem to me synonyms; prevosti prevosti
was considered a junior synonym of ovata by Prantl and Pfibyl (1949), Bruton (1967, 1968), and
subsequent authors until Snajdr (19846) resurrected it. The general lack of glabellar granules except
for the larger paired granules in the lower Ludlow O. ( O .) palava Snajdr, 19846, and the genuinely
short occipital spines of O. (O.) brevigena and O. ( O .) perpeta distinguish these from ovata.
Incidentally, O. ( O .) perpeta may be a junior synonym of brevigena , but the nature of the Bohemian
material makes comparison with the excellent Canadian specimens difficult.

In some of the Irish ovata specimens (PI. 20, figs. 1,5, 13, 20) the lp glabellar lobe is somewhat
composite in nature due to the presence anteriorly of two secondary furrows. These originate from
the lp furrow just either side of the point at which it turns more sharply backward at the median
glabellar lobe, and they run outward and backward to almost circumscribe a small raised area on
the anterolateral, adaxial part of the lp lobe. These furrows are also clearly present on the
Bohemian ovata cranidium figured by Bruton (1968, pi. 1, fig. 1), and a similar morphology was
noted by Whittard (1961, p. 200, pi. 27, fig. 1) in Primaspis whitei, which shows (at least) a single,
secondary subtransverse furrow on the basal lobe. These secondary furrows are probably atavistic
features. Bruton (1983, p. 882, text-fig. 2) has recently suggested that a likely Cambrian ancestor
of the Odontopleurinae, of which Odontopleura is a relatively late representative and Primaspis a
root stock, is Acidaspides Lermontova, 1951. Acidaspides precurrens, Lermontova, 1951, the type
species, appears to show such furrows invading the lp lobe (see Bruton 1983, pi. 88, figs. 12
and 15).

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier. Also from the
Wenlock to lower Ludlow of Czechoslovakia (Bruton 1968; Snajdr 19846, e), Poland (Tomczykowa 1957),
German Democratic Republic (Schrank 1969, 1972), German Federal Republic (Alberti 1970), and Anglo-
Welsh area (Thomas 1981). Possibly from the late Wenlock of the southern Urals (Bruton 1968), the late
Wenlock to lower Ludlow of north-west Canada (Chatterton and Perry 1983), and the Silurian of the Carnic
Alps (Gaertner 1930).

Genus primaspis Richter and Richter, 1917

Type species. Odontopleura primordialis Barrande, 1846, p. 29; from the middle Ordovician, Ded near Beroun,
Czechoslovakia. By original designation. Bruton (1968, p. 12) gives the type horizon as the Liben Formation,
whereas Horny and Bastl (1970, p. 249) and Snajdr ( 1 984z/, p. 50) give it as the Letna Formation.

Discussion. Since Primaspis was established a subdivision of it has been attempted by Pfibyl and
Vanek (1965) who proposed P. ( Meadowtownella ), Chatterton (1971) who set up P. ( Taemasaspis ),
and Snajdr (1984c/) who erected P. ( Bojokoralaspis ). Bruton (1968) rejected P. ( Meadowtownella )
and subsequently this subgenus has not been formally used by other authors, though Ramskold
(1984) has recently suggested that it is a coherent group. P. ( Taemasaspis ) was synonymized by
Thomas (1981) with Dudleyaspis , a procedure accepted by Ramskold (1984), by the present author,
and also, at least on present knowledge, by Chatterton and Perry (1983, p. 45) and by Chatterton
and Wright (1986, p. 293). Of the characters listed by Snajdr (1984c/, p. 54) differentiating P.
(Bojokoralaspis) from P. ( Primaspis ), it seems to me from his figures that only the absence of
paired occipital spines in his new subgenus is a potentially usable character and, in any case, if a
subgeneric distinction is made on this basis, P. ( Bojokoralaspis ) is really best regarded as a junior
synonym of P. ( Meadowtownella ), type species P. whitei Whittard, 1961.

The validity of separating odontopleurid subgenera on the basis of posterior occipital or cephalic
spines is debatable. Spinose and non-spinose species have, for example, been placed in the nominate
subgenera of Odontopleura (see generic discussion above), Diacanthaspis (Chatterton and Perry
1979, pp. 1333-1334) and Dudleyaspis (Chatterton and Perry 1983, p. 44); also Whittington (1956 a),
Bruton (1968), and others have considered the paired occipital spines in the type species of
Primaspis to be of no greater than specific value, though admittedly in Primaspis only the type

SIVETER: SILURIAN TRILOBITES FROM IRELAND

143

species of the genus was then known to have this feature, and in his recent work Snajdr (1984c/)
assigned to the nominate subgenus three other species with such spines. For the present I take a
similar view about Primaspis as did Chatterton and Perry (1983) and Chatterton and Wright (1986)
with respect to the Taemasaspis- Dudley aspis situation; unless the spinose and non-spinose forms
of Primaspis can be shown to belong to separate lineages, which can then be considered distinct
subgenera or even genera, I prefer to regard all as belonging to Primaspis (Primaspis). It appears
to me, in conjunction, that there is difficulty in recognizing as specifically distinct at least two of
the four species Snajdr (1984 <7) placed in Primaspis ( Primaspis ), namely P. propriofan and P.
oxitron, both described by him as new. Acceptance of the non-spinose Primaspis species as a
distinct higher taxon would be made easier if this were not the case.

Primaspis mendica sp. nov.

Plate 22, figs. 1-13, 15, 16, 18, 21

1863 Acidaspis Jamesii; Baily (pars ) in Jukes and Du Noyer, p. 12.

1878 Acidaspis Jamesii ; Baily (pars) in Kinahan, p. 40.

1976 Primaspis sp.; Siveter in Parkin, p. 595, table I

1981 Primaspis sp.; Siveter in Holland, p. 78, figs. 51. 1, 3, 6.

1984 Primaspis sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Derivation of name. From the Latin, mendicus , beggarly, reflecting the coarse appearance of the cuticular
sculpture.

Holotype. Cranidium lacking outer part of left fixed cheek, TCD 12254; Plate 22, figs. 1, 2, 5, 11.

Type horizon and locality. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.
Additional material. 20 cramdia, 10 free cheeks, 15 hypostomata, 15 pygidia, isolated thoracic segments.

Diagnosis. A species of Primaspis with glabellar outline subparabolic to horseshoe shaped.
Preglabellar area extremely short (sag.). Cranidium with mid-length of palpebral lobe opposite a
point slightly anterior to the mid-length of basal glabellar lobe. Median occipital spine sited
anteriorly, two-thirds of the length from posterior margin to occipital furrow. Four internal and
two external secondary spines on pygidium; inner spine adjacent to major spine is not fused to it
at its base. Sculpture of cranidium, free cheek, thoracic segments, and pygidium dominated by
large, moderately to densely distributed granules; main field of free cheek has pitting.

Description. Cranidium F9 to 2-0 times as wide as long. Glabella including occipital ring 0-8 to 0-9 times as
wide across basal lobes as long (sag.), subparabolic to horseshoe shaped in outline, very convex (sag. and
tr. ), in dorsal view projects anteriorly fractionally in front of anterior cranidial margin (PI. 22, fig. 1 ). Occipital
ring 3-5 times as wide as long (sag.) with median occipital spine sited anteriorly two-thirds of the distance
between posterior margin and occipital furrow. Lateral occipital lobe defined anterolaterally by short furrow
extending posteriorly one-third the length (exs.) of occipital ring from opposite inner margin of lp lobe.

Occipital furrow widest (sag. and exs.) and shallowest behind central glabellar area, posterior to lp lobe
it narrows and deepens sharply. Lobe Ip slightly less than one-third as wide as glabellar width, cat’s-ear
shaped, drawn out and pointed anterolaterally with its anterior, abaxial margin being slightly sinuous, has
independent inflation to central glabellar area from which it is completely separated by shallow, broad
posterior extension of Ip furrow. The latter is moderately deep at axial furrow, trends gently backward and
becomes deeper adaxially between lp and 2p lobes, turns sharply backward and shallows at central glabellar
area. Lobe 2p subquadrate in dorsal view (PI. 22, fig. 1) and cat’s-ear shaped in oblique view (PI. 22, fig. 5),
about two-thirds the size of lp, strongly convex (exs. and tr.), clearly divorced from central glabellar area by
shallow, broad furrow connecting furrows lp and 2p. Furrow 2p well defined, moderately deep, curves
inward and backward. Lobe 3p very small compared with lobe 2p, clearly delimited anteriorly by sharply
incised, short 3p furrow which runs into 2p furrow on side of glabella (PI. 22, figs. 5 and 11). Frontal
glabellar lobe quite strongly convex in outline, falls steeply anteriorly. Central glabellar area widest between
lp furrows.

Axial furrow broad and shallow and trends forward and inward at side of occipital ring, is slightly deeper

144

PALAEONTOLOGY, VOLUME 32

and narrower curving around base of lobe lp and running sinuously forward, shallowing slightly at
anterolateral part of lobe lp, becomes slightly deeper around 2p lobe before trending inward below frontal
lobe to meet preglabellar furrow. Preglabellar area extremely short sagittally, lengthens (exs.) toward facial
suture. Preglabellar furrow is clearly defined. Anterior cranidial margin very broadly convex forward.

Outer part of fixed cheek about 0-8 times as wide as occipital ring. Posterior border narrow and tightly
convex adaxially, gradually expands in width (exs.) abaxially; border furrow widest (exs.) about midway
between axial furrow and facial suture, moderately deep. Fixed cheek adaxial to eye is strongly convex (exs.
and tr.), about as wide as lobe lp, anteriorly it is drawn inward, narrows rapidly, and falls steeply. Very
strong eye ridge curves inward to meet axial furrow just in front of base of 2p furrow. Small, sunken
subtriangular area between anterior margin, suture, and eye ridge. Low, steeply raised palpebral lobe is about
one-third as long as lp lobe, it is positioned half-way or slightly less across fixed cheek abaxially from axial
furrow and its mid-length lies opposite a point just anterior to the mid-length of lp lobe; palpebral furrow
distinct. Anterior branch of facial suture runs forward and inward to cranidial margin; posterior branch
trends outward and backward from eye, turns exsagittally at genal angle to meet posterior border; both
branches run on a weak sutural ridge.

Lateral border of free cheek (PI. 22, figs. 6 and 15) moderately convex, broadens posteriorly. Lateral
margin has an estimated twelve to fourteen fringing marginal spines (the remains of at least eleven are
present) that increase in length gradually posteriorly until the last one, sited on base of genal spine, which
is slightly shorter than the penultimate spine. Genal spine short, about one-third the length of the rest of the
free cheek; it is more or less straight (PI. 22, fig. 6) or at most is extremely gently curved (PI. 22, fig. 15).
Lateral border furrow broad, slightly deeper anteriorly than posteriorly. Main field of cheek relatively wide
and broadly convex; eye socle very narrow. Visual surface kidney-shaped with numerous small lenses.

Hypostoma (PI. 22, figs. 3 and 16) subrectangular in outline, about 1-4 times as wide as long. Anterior
border lacking medially, widens (exs.) quickly abaxially; laterally, border widens (tr.) gradually posteriorly,
is moderately convex and falls steeply into border furrow; posterior border gently convex (sag. and exs.).
Notch and shoulder strongly developed on lateral margin, posterior margin transversely directed, anterior
margin broadly convex forwards. Anterior border furrow scarcely present sagittally, broadens and deepens
towards anterior wing; lateral border furrow deepest, well marked along its whole length; posterior border
furrow weak medially, becomes better defined abaxially. Middle body moderately convex (tr. and sag.);
anterior lobe oval (tr.), posterior lobe crescentic; finely incised median furrow trends backward and inward
for a short distance before fading out. Maculae small, oval, weakly inflated, positioned close together on
projected adaxial path of median furrow just either side of median line.

Isolated thoracic segments (PI. 22, figs. 7 9) show moderately convex axial ring. Articulating furrow divided
into two medially by a narrow, gently raised strip which lenses out either side the sagittal line towards the
weak axial furrow. Articulating half-ring well developed. Pleural furrow narrow (exs.), well incised. Posterior
pleural band about three times as wide (exs.) as anterior band, developed abaxially into relatively short, stout
spine; form of anterior pleural spine unknown. Very narrow facet behind posterior band.

Pygidium (PI. 22, figs. 4, 12, 21) 3-4 to 3-7 times as wide as long excluding articulating half-ring; posterior
margin excluding spines is saucer-shaped. Axis has two rings and a very short terminal piece. Second axial

EXPLANATION OF PLATE 22

Figs. 1-13, 15, 16, 18, 21. Primaspis mendica sp. nov. All specimens are from the Annascaul Formation,
locality 36. 1, 2, 5, 1 1, holotype, cranidium, TCD 12254; dorsal stereo-pair, lateral, oblique, frontal views,
x4; figured Siveter in Holland 1981, fig. 51.1 as Primaspis sp. 3, hypostoma, TCD 12318; ventral stereo-
pair, x 8. 4, pygidium, TCD 1221 1; dorsal view, x 5. 6, free cheek, UM K2807; ‘dorsal’ view, x4. 7 and
9, thoracic segment, TCD 12309; frontal, dorsal views, x4. 8, thoracic segment, TCD 12650; oblique
view, x4. 10 and 13, cranidium, TCD 12622; lateral, dorsal views, x4. 12, pygidium, TCD 12202;

dorsal stereo-pair, x 5; figured Siveter in Holland 1981, fig. 51.6 as Primaspis sp. 15, free cheek, TCD
12596; ‘dorsal’ stereo-pair, x 5; figured Siveter in Holland 1981, fig. 51.3 as Primaspis sp. 16, hypostoma,
TCD 12331; ventral view, x 8. 18, cranidium, TCD 12255; dorsal view, x4. 21, pygidium, TCD 12210;
dorsal view, x 5.

Figs. 14, 17, 19. Scharyia sp. Both specimens are from the Annacaul Formation, locality 36. 14, cranidium,
TCD 12145; dorsal view, x 10. 17 and 19, cranidium, TCD 12144; lateral view, dorsal stereo-pair, x 10.

Fig. 20. Leonaspis coronata coronata (Salter, 1853). Cranidium, TCD 12197, Annascaul Formation, locality
36; dorsal view, x 5.

PLATE 22

SIVETER, Silurian trilobites

146

PALAEONTOLOGY, VOLUME 32

ring has ankylosed half-ring. First axial ring is narrowest sagittally, becomes wider (exs.) abaxially. Axial
furrow weakly developed beside first axial ring and centrally behind terminal piece, best defined beside
articulating half-ring, second axial ring, and the side of terminal piece. Pleural area sunken beside second
axial ring and terminal piece, dominated by strongly convex (tr.) ridge which gradually expands (tr.) abaxially
from first axial ring and is most swollen just inside posterior margin; this ridge then narrows gradually into
gently curved, stout, major border spine. Very slight ridge runs obliquely backward and outward across
pleural area to connect with first secondary spine outside major spine; a second external secondary spine is
also present. There are two pairs of inner secondary spines, estimated to be about half as long as major
spines. Indistinct posterior border present between major spines, becomes lost abaxially to them.

Cranidium, excluding all furrows and to a lesser extent anterior border and sublriangular area, covered
with large to medium sized granules which are moderately loosely (PI. 22, fig. 1) to densely distributed (PI.
22, figs. 13 and 18). Whole cranidium has a very finely granulated groundmass. Cranidial sculpture thus
essentially bimodal. Free cheek has medium to large granules moderately widely to loosely scattered on main
field, lateral border, and base of genal spine. Main field and lateral border furrow also has moderately to
loosely distributed pits. All the free cheek has a fine, granulate groundmass. Hypostoma has large to medium
sized granules on central body either side of median furrow, medium granules on centre of anterior lobe and
inner part of posterior border near border furrow, and a dense overall covering of fine granules except for
the border furrows. Axial ring and posterior pleural band have large to medium granules; anterior pleural
band, articulating half-ring and furrow lack coarse granules; posterior pleural spine (PI. 22, fig. 8) has small
granules. Pygidial axial rings excluding half-rings, terminal piece, pleural ridges, and outer pleural area inside
margin have large to medium sized granules; large granule on each spine base; large, paired granules on
adaxial part of major pleural ridge and axial rings. Pygidial spines have small granules. Whole pygidium
finely granulated.

Discussion. This new species appears to be the youngest member of the genus. The combination
of the characters given in the diagnosis, particularly the anterior position of the eye, easily
distinguish it from other Primaspis species, of which there are very few recorded or generally
agreed upon from the Silurian. P. mackenziensis Chatterton and Perry, 1983 (pi. 5, figs. 1-24) from
the early Llandovery of the Mackenzie Mountains, north-west Canada has, amongst other things,
the eye placed opposite the occipital furrow and basal part of lobe lp, a more sharply tapering
glabellar outline, a relatively longer, posteriorly extended occipital ring, and longer genal and
major pygidial spines.

Primaspis kreugeri Schrank, 1969 from the Baltic Graptolithengestein glacial erratics can easily
be distinguished from the Irish species and in any case belongs to Anacaenaspis , as Thomas (1981)
and Ramskold (1984) have noted. P. mendica, though being clearly specifically distinct from
Ordovician Primaspis species, has a reasonably similar glabella to several of the earliest of them,
all of middle Ordovician age, for example P. whitei and P. rorringtonensis Whittard, 1961 from
the Shelve inlier, P. ascitus Whittington, 1956« from Virginia, P. koral Snajdr, 1984c/ from
Czechoslovakia, and P. multispinosa Bruton, 1965 from Norway.

Occurrence. Known only from the type horizon and locality.

Genus leonaspis Richter and Richter, 1917

Type species. Odontopleura leonhardi Barrande, 1846, p. 58; from the Kopanina Formation (Ludlow),
Kolednik, Beroun district, Czechoslovakia. By original designation.

Discussion. Over the past thirty years there have been many comments in the literature on the
origin of Leonaspis and how clearly to distinguish it from Primaspis. Investigation of such problems
requires examination of all Silurian and Ordovician species of both genera. Comparison of the
Leonaspis and Primaspis species described in this paper endorses the need for such a study.

Prantl and Pfibyl (1949, p. 146), as commented on by Whittington (1956«, p. 506; 19566,
p. 199), noted the Primaspis- like features displayed by Leonaspis coronata (Salter, 1853) and
Leonaspis deflexa (Lake, 1896). Whittington (19566) maintained both these species in Leonaspis ,
which he suggested was derived from Primaspis , and he provided new diagnoses for both genera.
Bruton (1968, p. 1 1) later said that the most reliable way of distinguishing the two was the presence

SIVETER: SILURIAN TRILOBITES FROM IRELAND

147

in Primaspis of a 3p lobe, other features having proved to be less useful. Chatterton and Perry
(1979) claimed that Leonaspis was derived from Diacanthaspis and they later (1983) suggested that
Leonaspis might be diphyletic, those species with two secondary pygidial spines between the major
border spines having come from Diacanthaspis , and those with four inner secondary spines having
their origin in Primaspis. The two-spined forms are overwhelmingly the most common type in
North America, only L. churkini Chatterton and Perry (1979) from California having four, but
they occur also in Europe, Asia, and Australia; the four-spined forms are abundant in Europe and
include the type species, L. leonhardi. This dichotomy of Leonaspis taxa had earlier been commented
on by Campbell (1977, p. 1 13).

The overall morphology of the L. coronata coronata specimens figured in this study, particularly
the form of the glabella and lp lobe, seems closer to that of Primaspis mendica sp. nov. and other
Primaspis species such as the late Ordovician P. hacculenta McNamara, 1979 (see Owen 1981,
pi. 17, figs. 6-13), rather than to Leonaspis parkini sp. nov. and many other Leonaspis species, for
example L. williamsi Whittington, 1956 (see Campbell 1977, pi. 32). The Kerry coronata material
lacks a 3p lobe and I have assigned it to Leonaspis (cf. PI. 21, fig. 5; PI. 22, fig. 5; PI. 23, fig. 9),
though it seems questionable to place generic emphasis on such a small feature, whose presence is
sometimes difficult to determine unequivocally. Thomas (1981) does not describe a third lobe in
English material of coronata , though there appears the suggestion of one on one of his figured
specimens (pi. 23, fig. 176).

Leonaspis coronata coronata (Salter, 1853)

Plate 22, fig. 20; Plate 23, figs. 9 23
1976 Leonaspis sp.; Siveter in Parkin, p. 595, table 1 (jiars ).

1981 Leonaspis coronata (Salter, 1853); Thomas, p. 88, pi. 23, tigs. 10, 12-14, 16 18, 20, 22-26
(with full synonymy).

1984 Leonaspis coronata (Salter); Thomas in Thomas, Owens, and Rushton, p. 53.

1984 Leonaspis sp.; Siveter in Thomas, Owens, and Rushton, p. 55 (pars).

1984 Leonaspis coronata coronata ; Ramskold, p. 257.

Lectotype. Selected Whittard 1938, p. 109, internal mould of incomplete cephalon, GSM 36734; see Thomas
1981, p. 88, pi. 23, fig. 13.

Type horizon and locality. Ludlow Series, lower Gorstian Stage, Vinnal Hill, Ludlow district, Shropshire.

Material. Eleven cranidia, four free cheeks, two hypostomata, two pygidia, isolated thoracic segments.

Description. Thomas (1981) has given a full description of this subspecies, which can apply in general to the
present Irish material, apart from the points raised below.

Discussion. The Kerry material clearly belongs to the late Wenlock-early Ludlow L. coronata. In
addition to the nominate subspecies Ramskold (1984) has recently introduced L. coronata hufo
from the late Wenlock of Gotland and he regarded the coeval L. marklini (Angelin, 1854), which
was also based on Gotland material, as a nomen duhium. Given that it is impossible to re-collect
topotype material of marklini I agree with this action, but it is unfortunate because marklini has
become entrenched in the literature for over a century and it is quite possible, if not probable,
that coronata hufo is its junior synonym. Certainly coronata hufo is of a similar morphology to
that ascribed over the years to marklini , as is evidenced by Ramskold placing in the synonymy list
of coronata hufo the late Wenlock Graptolithengestein erratic material from the Baltic described
by Schrank (1969, pi. 4, figs. 9 and 10; pi. 5, figs. 1-6; pi. 6, figs. 1, 2, 4-7) under the name of
Angelin’s species.

There appears little variation in the Kerry material, the most obvious being the sparseness of
larger granules on one of the specimens (PI. 23, figs. 9, 22, 23), but this could be preservational,
and m any case should be considered intraspecific. From comparing coronata coronata (see Thomas
1981, p. 88, pi. 23, figs. 10, 12-14, 16-18, 20, 22-26; Whittington 1956a, pi. 59, fig. 12) with

148

PALAEONTOLOGY, VOLUME 32

coronata bufo (Ramskold 1984, p. 257, pi. 30, figs. 6, 14-16, pi. 31, figs. 1-6; Schrank 1969) the
following differences separate the subspecies. The genal spine is longer in coronata bufo and reaches
back to about the seventh thoracic segment, whereas in coronata coronata it extends back to the
fourth. In coronata bufo the posterior pleural spines are slender and pointed, longest on the sixth
segment, short on the anterior three segments, but distinctly shorter only on the first segment; in
coronata coronata the posterior pleural bands are extended into thick pleural spines which on the
anterior four segments are short, but on the posterior six segments are long. The major pygidial
border spines of coronata bufo are diagnosed as being directed exsagittally, or are slightly divergent
posteriorly; in the nominate subspecies they incurve posteriorly.

The length of the genal spine of the Irish material (PI. 23, figs. 19 and 20) seems closer to that
of coronata coronata , though I have no articulated material to assess its posterior extension
alongside the thorax; isolated thoracic segments (PI. 23, fig. 14) also seem to be of the less slender
coronata coronata type. Only two Irish pygidia (PI. 23, figs. 15 and 16) are known, and only one
of these has a single, complete major border spine, which is directed more or less exsagittally. It
is questionable which subspecies this single specimen most resembles, especially as in one figured
specimen (Schrank 1969, pi. 6, fig. 1) put (Ramskold, 1984) in synonymy of coronata bufo one of
the major spines is slightly incurved, and in another (Schrank 1969, pi. 6, fig. 2) the spines also
appear inwardly directed.

In both coronata coronata and coronata bufo three prominent granules at the base of each major
border spine are described and a single one at the base of each inner secondary spine. One of the
Irish pygidia (PI. 23, fig. 16) apparently shows just one larger granule at the base of the only major
border spine preserved on this specimen, and lacks a larger granule inside the base of each of the
innermost pair of secondary spines, though one topotype specimen of coronata bufo (Ramskold
1984, pi. 31, fig. 2) appears the same in this latter respect. The second Irish pygidium (PI. 23,
fig. 15) may have the three larger granules at the major spine base, but they are only slightly larger
(not distinctly larger as in typical coronata coronata or coronata bufo) than the fine, dense,
background granulation. Also in both coronata coronata and coronata bufo there is a prominent
granule at the fulcrum of each thoracic segment and one half-way across each pleura. The first of
these, though again somewhat less prominent, is present in the Irish material (PI. 23, fig. 14), yet
the second one appears absent. Thomas (1981) did not describe any pitting which might represent
rudimentary genal cecae on the free cheek of coronata coronata ; this feature is present in the Irish
specimens, but Thomas’s illustrations of free cheeks are small and these structures are not easily
picked out if present. However, they do exist in coronata bufo (Ramskold 1984, pi. 30, fig. 6a, b).
Also Thomas (1981) described, but did not figure, the hypostoma of coronata coronata. The Irish
hypostoma fits his description except it (PI. 23, figs. 10 and 21) has a slightly posteriorly convex
posterior margin, not ‘weakly bifurcate’ as described for the English material. The Irish specimens
agree entirely with those figured by Ramskold (1984, pi. 30, figs. 15 and 16) for coronata bufo.

Both the English and Baltic material of coronata seem to have a better defined contact of the
lateral and posterior border furrows on the free cheek inside the genal spine base, and a slightly
better impressed lateral border furrow than is the case in the Irish free cheek, but these minor
differences appear to be the only non-sculptural distinctions setting the non-Irish material
(collectively) apart.

Because of the form of its genal spine and thoracic segments, the Irish material appears to be
excluded from coronata bufo. I have, considering the sample size, not placed too much weight on
the slight differences of the Irish examples from the English specimens of coronata coronata , and
on balance placed them without qualification in this subspecies. Closely related species include L.
mutica (Emmrich, 1844), L. muldensis Bruton, 1967, and L. varbolensis Bruton, 1967 (see Bruton
1967; Schrank 1969; Temple 1970; Thomas 1981; Howells 1982; Ramskold 1984).

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4; late Wenlock part of the
Coalbrookdale Formation, Dudley, West Midlands, England; Much Wenlock Limestone Formation of the
West Midlands and Welsh Borderland; the late Wenlock of South Wales; the Ludlow, lower Gorstian Stage,
of the Welsh Borderland (see Thomas 1981, p. 88).

SIVETER: SILURIAN TRILOBITES FROM IRELAND

149

Leonaspis parkini sp. nov.

Plate 21, figs. 2, 5-7, 9-21

1863 Acidaspis Jamesii, Baily (pars ) in Jukes and Du Noyer, p. 12.

1878 Acidaspis Jamesii ; Baily (pars) in Kinahan, p. 40.

1976 Leonaspis sp.; Siveter in Parkin, p. 595, table I ( pars ).

1984 Leonaspis sp.; Siveter in Thomas, Owens, and Rushton, p. 55 (pars).

Derivation of name. For Dr John Parkin, geological mapper of the Annascaul inlier.

Holotype. A complete cranidium, TCD 12207; Plate 21, figs. 5-7, 9.

Type horizon and locality. Annascaul Formation, locality 36 of Parkin 1976, p. 587, fig. 4, Annascaul
inlier.

Additional material. Eighteen cranidia, five free cheeks, three pygidia, isolated thoracic segments.

Diagnosis. A species of Leonaspis with obliquely elongate lp lobe; axial furrow at anterolateral
corner of this lobe extremely weak or lacking, so that here lp lobe tends to merge with inner part
of fixed cheek which is weakly developed or pinched out by convergence of the lobe and eye ridge.
Lateral and posterior border furrow on free cheek well developed; genal spine about as long or
slightly longer than main part of free cheek. Pygidium with two inner secondary spines less than
half as long as major border spines; one external secondary spine of about the same or slightly
shorter length plus a second very small external secondary spine. Whole dorsal surface of
exoskeleton extremely finely and densely granulate. Anterior part of glabella, 2p lobe, eye ridge,
anterior cranidial margin, and inner part of main field of free cheek with medium sized, moderately
densely distributed granules; posterior part of glabella has a few, very widely scattered small to
medium granules; granules sparse on central glabellar area. Occipital ring has median organ with
four symmetrically arranged pits, and two granule pairs, the posterior pair of which is the larger.
Lateral border of free cheek with a single row of five, isolated medium sized granules. Two granule
pairs on each pygidial axial ring, the inner pair of each the larger.

Description. Cranidium 2 0 (PI. 21, fig. 15) to 2-3 times (PI. 21, fig. 9) as wide as long. Glabella including
occipital ring 0-8 to 0-9 times as wide as long (sag.), bell-shaped in outline, in sagittal profile gently to
moderately convex from occipital furrow to frontal lobe, falls more steeply to preglabellar furrow. Occipital
ring extended posteriorly behind central glabellar area so that here it is longest (sag. and exs.), becomes
shorter (exsag.) behind lp lobe due to sharp forward flexure of posterior margin; median occipital organ
with four subquadrately arranged pits is sited slightly anteriorly (PI. 21, figs. 2 and 10); abaxially, occipital
ring has a weakly inflated occipital lobe which is outlined anterolaterally by short, broad posterior extension
of longitudinal glabellar furrow. Occipital furrow is very broad (sag.) and shallow behind central glabellar
area, deeper and narrower behind lp lobe which is very elongate and trends forward and outward, becoming
more or less confluent with fixed cheek anteriorly (PI. 21, fig. 20), separated from central glabellar area by a
broad (tr.), shallower part of the longitudinal furrow. Furrow 2p deep, runs inward and backward. Lobe 2p
subcircular and strongly convex (tr. and exs.), very clearly separated from central gabellar area by distinct
furrow. Frontal lobe moderately convex in dorsal outline.

Axial furrow extremely broad and shallow and runs inward from posterior cranidial margin to
occipital furrow, becomes narrower and slightly better defined as it trends outward alongside lobe lp,
shallows and more or less disappears at anterolateral margin of this lobe (PI. 21, fig. 5), is well defined
around 2p lobe where it merges with anterior extension of furrow which runs inward from palpebral
furrow on the inside of eye ridge. Preglabellar furrow shallow; preglabellar area short (sag.), lengthens
(exs.) abaxially. Anterior cranidial margin straight (tr.) (PI. 21, fig. 9) to extremely gently convex forwards
(PI. 21, fig. 20).

Outer part of fixed cheek 0-9 times as wide as occipital ring, very narrow (exs.); posterior border quite
strongly convex and narrow adaxially, becomes flatter and very much wider (exs.) near facial suture; posterior
border furrow is long, narrow, rille-likc, fades abaxially. Inner part of fixed cheek adaxially from eye is about
as wide or slightly wider that lp lobe, it falls steeply posteriorly, less so anteriorly, narrows rapidly forwards
alongside lobe lp just anterior to which it pinches out. Palpebral lobe is high (PI. 21, fig. 6), rises up almost

150

PALAEONTOLOGY, VOLUME 32

vertically from fixed cheek before becoming less steeply inclined towards its outer margin; mid-length of lobe
is opposite posterior margin of lobe lp or outer part of occipital furrow; length of lobe is about two-thirds
to three-quarters that of lp lobe. Eye ridge is pronounced, curves strongly forward and inward in front of
lobe 2p. Relatively large subtriangular area between facial suture, anterior cranidial margin, and eye ridge.
Anterior branch of facial suture runs forwards for a short distance before turning inward to anterior margin;
posterior branch trends transversely from eye then curves broadly backward to posterior margin (PI. 21, figs.
5 and 6).

Main field of free cheek (PI. 21, fig. 18) relatively small in area, moderately strongly convex from visual
surface to lateral border furrow, which is well developed and moderately wide (tr.) from anterior branch of
facial suture posteriorly to transversely directed, equally deep posterior border furrow. Lateral border
becomes slightly more convex (tr.) posteriorly. Fringing border spines incomplete in one small specimen (PI.
21, fig. 21) where they are relatively short; at least eleven present with the most posterior spine sited at or a
short distance posteriorly from base of genal spine. The latter is moderately broadly based, gently (PI. 21,
figs. 13 and 21) to moderately (PI. 21, fig. 18) curved, about as long (PI. 21, fig. 13) or slightly longer than
(PI. 21, fig. 18) main part of free cheek from which it bends outwards and downwards.

Hypostoma unknown. Isolated thoracic segments (PI. 21, fig. 1 1) show wide (sag.) axial ring and articulating
half-ring. Articulating furrow weak. Axial furrow and pleural furrow very weak. Anterior pleural band
slender, weakly convex, produced abaxially into short, posteriorly directed spine; posterior band wider (exs.)
and more prominent, swollen slightly at fulcrum, extended into long, exsagittally directed pleural spine.
Relatively wide (exs.), flat, facet behind posterior pleural band.

Pygidium 2-35 (PI. 21, fig. 17) to 2-5 (PI. 21, fig 11) times as wide as long (sag.) excluding articulating half-
ring. Axis has two rings and a short (sag.), bluntly rounded terminal piece. Second axial ring has half-ring
fused to first axial ring. Ring furrow between second axial ring and terminal piece is incomplete. Axial furrow
weakest beside first axial ring and behind terminal piece, strongest beside second axial ring and side of
terminal piece. Pleural area small. Pleural ridge runs outward and almost immediately backwards from first
axial ring, expands in width near posterior margin and is extended into major spine which in lateral profile
bends gently upward posteriorly. Two inner secondary spines are less than half the length of major spines
(PI. 21, fig. 17); first secondary spine outside major spine about as long or slightly shorter than internal
secondary spines, a second very much shorter external secondary spine present. Anterolateral pygidial margin
with an obliquely directed articulating facet.

Frontal glabellar lobe and 2p lobe have moderately densely distributed medium sized granules which are
also present but much more loosely scattered on central glabellar area between lp and 2p furrows, and are
present though somewhat more concentrated on eye ridge and anterior cranidial margin. Posterior part of
central glabellar area with extremely sparse, small to medium granules or lacking in granules. On occipital
ring are two small to medium sized granule pairs arranged either side of the pitted median occipital organ;
the anterior pair are more nearly in transverse line with the median tubercle and tend to be slightly smaller
and wider apart than the posterior pair (PI. 21, figs. 9 and 10). Small to medium sized granules present on
posterior and adaxial, anterolateral angle of lp lobe, two or three more concentrated on anterior part of this
lobe and two or three widely dispersed in a line (exs.) on fixed cheek between palpebral lobe and axial furrow.
Whole of the cranidial surface is extremely finely and densely granulate, as is that of free cheek, thoracic
segments, and pygidium. Free cheek has a row of five, widely separated, medium sized granules along lateral
border and a moderately dense concentration of these on the inner, mainly anterior, part of main field.
Thoracic segments have a medium to large granule on posterior pleural band just inside fulcrum and a few
small granules and one medium sized granule (?one of a pair) on axial ring (PI. 7, fig. 14). Each pygidial
axial ring has two pairs of granules, the inner pair medium to large, the outer pair of medium size. Terminal
piece shows a single granule pair. Distinct granule on pleural ridge at base of each major spine; less consistent
granule at base of each internal secondary spine and on pleural area near base of larger external secondary
spine.

Discussion. There are a large number of described Leonaspis species, at least fifty on my own
records not counting synonyms. A few have unique features, but the majority have been established
on the basis of different combinations of characters. L. parkini falls into the latter category, and
although there are several existing species which are similar to it, I can find none which is exactly
the same.

The closest well-known species to parkini is Leonaspis lenzi Chatterton and Parry, 1983 from
the middle to late Wenlock, of the Mackenzie Mountains, one of eight new Leonaspis species

SIVETER: SILURIAN TRILOBITES FROM IRELAND

151

described by Chatterton and Perry (1983) on the basis of excellent silicified material from the
Llandovery to Ludlow of north-west Canada. Were it not for the fine preservation of the specimens
of both species, it would be difficult to separate them. The large number of observed differences,
mostly concerning fine detail, is also a reflection of this preservation, rather than a measure of the
taxonomic distance between the two, which I would recognize at the subspecific level were it not
for distinctions in the glabellar lobation, fixed cheek, and number of pygidial spines. In the
cranidium of lenzi, the fixed cheek between the anterior part of lp lobe and the eye ridge is more
anteriorly continuous and better developed and the anterolateral part of this lobe is better separated
from the cheek by a more distinct axial furrow; the four pits in the median occipital organ are
lacking; there are a few more granules on the posterior part of the glabella and small specimens
have distinctly paired granules on the central glabellar area, a feature not evident in the only
known small specimen of L. parkini (PI. 21, fig. 15) which has a sculpture like that of the more
mature Irish cranidia (PI. 21, figs. 2, 9, 10, 15, 20; cf. Chatterton and Perry 1983, pi. 8, figs. 1, 6,
7, 9, 14, 19; pi. 9, figs. 2-4). The number of granules on the lateral border of the free cheek in
lenzi is much greater (PI. 21, figs. 13, 18, 21; cf. Chatterton and Perry 1983, pi. 8, figs. 5, 32, 33,
37, 38). The pygidium of lenzi has only one secondary pygidial spine outside the major border
spine (in the diagnosis of lenzi it says ‘one or sometimes two’, but not one of the twenty figured
pygidia shows a second spine and in their discussion of the closely related L. jaanusoni, Chatterton
and Perry (1983) cite the lack of a second spine as a feature of lenzi), the inner secondary spines
are longer than half the length of the major spines (own measurements), and each axial ring has
only one pair of granules (PI. 21, figs. 11, 16, 17; cf. Chatterton and Perry 1983, p. 27, pi. 8,
figs. 18, 22-24, 34; pi. 9, fig. 8).

L. crenata crenata (Emmrich, 1844) and L. crenata angelini (Prantl and Pribyl, 1949) from
Gotland, and L. crenata brutoni Thomas, 1981 from Britain, all of late Wenlock age, are the closest
named Anglo-Baltic taxa to the Irish species. This group has recently received detailed comparison
by Thomas (1981) and by Ramskold (1984). All of them differ in, for example, having only one
secondary pygidial spine external to the major border spine, a much longer and more slender genal
spine, a less well-developed lateral border furrow on the free cheek and in the detailed form of the
granular sculpture.

Leonaspis is well represented in the Siluro-Devonian of Czechoslovakia, though the range of
variation has only been adequately illustrated in a very few species and there is contradiction in
the literature over possible synonyms and the status of type material (compare, for example, the
Leonaspis species in Prantl and Pribyl 1949; Bruton 1968; Horny and Bastl 1970; Snajdr 1984c).
My survey of this taxonomic minefield suggests that the middle to late Wenlock L. dormitzeri
(Hawle and Corda, 1847) (see Bruton 1968, pi. 3, figs. 5 and 10, and Snajdr 1984c, p. 152, pi. 13,
fig. 5 for new lectotype selection) is fairly close to L. parkini but differs most obviously in having
a very long, slender genal spine. The uncertainty over the possible synonymy of L. leonhardi with
L. geinitziana (Hawle and Corda, 1847), both from the Kopanina Formation, remains (see Pribyl
and Vanek 1966, pp. 294-295, pi. 2, fig. 1; pi. 3, figs. 1, 3, 4; Bruton 1968, p. 20, pi. 2, figs. 6-8,
10; Ramskold 1984, p. 260; Snajdr 1984c, p. 157, pi. 7, fig. 11, new lectotype selected). As far as
can be determined, L. parkini is similar to the figured cephalic parts of both leonhardi and
geinitziana , which Bruton (1968) regarded as synonyms, though the leonhardi material, at least,
differs from the Irish species in the nature of its occipital granular sculpture (Bruton 1968, p. 19).
L. leonhardi is also clearly distinct in having four inner, secondary pygidial spines, whereas the
pygidia assigned to geinitziana have only two. An articulated specimen of geinitziana is needed
before closer comparison of the Bohemian and Irish species can be made.

L. clavatns Chatterton, 1971, L. rattei (Etheridge and Mitchell, 1896) and L. wellingtonensis
Chatterton, Johnson and Campbell, 1979 from the Devonian of Australia all show resemblances
to L. parkini , particularly clavatns in its pygidium and wellingtonensis in its cranidial sculpture,
but all have a different combination of other characters.

Occurrence. Known only from the type horizon and locality.

152

PALAEONTOLOGY, VOLUME 32

Family lichidae Hawle and Corda, 1847
Subfamily lichinae Hawle and Corda, 1847
Genus dicranopeltis Hawle and Corda, 1847

Type species. Lichas scabra Beyrich, 1845, p. 28; from the Liten Formation (Wenlock), Svaty Jan pod Skalou,
Czechoslovakia. By subsequent designation of Reed 1902, p. 61.

Dicranopeltis salteri! (Fletcher, 1850)

Plate 23, figs. 1 -8

1976 lichids; Siveter in Parkin, p. 595, table 1 (pars).

? 1981 Dicranopeltis salteri (Fletcher, 1850); Thomas, p. 68, pi. 19, figs. 4, 5, 9 12, non figs. 6-8 (with
synonymy).

1984 Dicranopeltis sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. Three cranidia, two pygidia.

Occurrence. Annascaul Formation, locality 36 of Parkin 1976, p. 581, fig. 4, Annascaul inlier. Possibly also
the Wenlock and ?low Ludlow of the Anglo-Welsh area, and the Silurian of ?Scandinavia (see below and
Thomas 1981, pp. 68-69).

Description. For the cranidium see the description of D. salteri by Thomas (1981, p. 69), which largely applies
to that of the Irish Dicranopeltis , and the discussion below. Free cheek, hypostoma, and thorax unknown.

Pygidium (PI. 23, figs. 4 and 7) is 1-8 times as wide (tr.) as long sagittally and T5 times as wide as pygidial
length (exs.) along posterior pygidial spine. Axis is broadly convex anteriorly, has two rings and a terminal
piece; it is slightly less than 0-3 times as wide anteriorly as width of pygidium and is 0-4 times as long as
pygidial length (sag.). Ring furrows continuous across axis, second furrow weakest. Terminal piece slightly
retrousse in its medial posterior part, falls steeply posteriorly to shallow, gently backwardly convex posterior
section of axial furrow which continues anteriorly to become much more sharply defined along side of
terminal piece and axial rings. Postaxial sector slopes very gently posteriorly or is horizontal, narrows (tr.)
gradually posteriorly where it is about 0-4 times as wide as anteriorly behind terminal piece. Pleural region
horizontal, has three well-marked pleural and interpleural furrows. First pleural furrow curves broadly
outward and backwards abaxially, fades on pleural spine, second swings very gently more backwards, distal
part unknown; third runs backward and slightly outward, falls well short of pygidial margin. First interpleural
furrow swings outward and backward, distal part unknown; second trends gradually backward and slightly
outward, swings gently abaxially and becomes much shallower a short distance before reaching lateral margin;
third delimits side of postaxial sector, is very weakly present for a short distance on most posterior part of
pleural region running backward and slightly inward to posterior margin sagittally. Lateral margin is
produced into three pairs of long, sharply pointed, backwardly directed spines between posterior pair of
which it is acutely incised. Doublure is wide with subparallel terrace lines.

EXPLANATION OF PLATE 23

Figs. 1-8. Dicranopeltis salteril (Fletcher, 1850). All specimens are from the Annascaul Formation, locality
36. 1-3, cranidium, silicone rubber cast of external mould, TCD 12533; dorsal stereo-pair, lateral, frontal
views, x3. 4, pygidium TCD 12590a; dorsal stereo-pair, x3. 5 and 6, cranidium, TCD 12589a; lateral,
dorsal views, x 6. 7, pygidium, TCD 12534a; dorsal view, x4. 8, cranidium, TCD 12532; dorsal view,
x 4.

Figs. 9-23. Leonaspis coronata coronata (Salter, 1853). All specimens are from the Annascual Formation,
locality 36. 9, 22, 23, cranidium, TCD 12198; oblique view, x 4, lateral, dorsal views, x 3. 10, hypostoma,
TCD 12213; ventral stereo-pair, x 8. 11, cranidium, TCD 19598; dorsal view, x 6. 12, 13, 17, cranidium,
TCD 12196; frontal view, dorsal stereo-pair, lateral view, x 6. 14, thoracic segment, TCD 12272; dorsal
view, x 4. 15, pygidium, TCD 12209; dorsal view, x 5. 16, pygidium, UM K2808; dorsal stereo-pair,
x4. 18, cranidium, TCD 12201; dorsal view, x 5. 19, free cheek, TCD 12216; ‘dorsal’ stereo-pair, x3.
20, free cheek, UM K2806a; ‘dorsal’ view, x4. 21, hypostoma, UM K2811; ventral view, x 8.

PLATE 23

SIVETER, Silurian trilobites

154

PALAEONTOLOGY, VOLUME 32

Pygidial sculpture comprises small to medium sized granules over whole pleural region and axis
exclusive of furrows. Axis has a pair of medium sized granules on central, posterior part of each axial ring,
and one pair on central part of terminal piece about one-third of its length (sag.) posteriorly from second
ring furrow.

Discussion. Thomas (1981, p. 68) has recently revised D. salteri, his description of the cranidium
being based on the type and topotype specimens from the late Wenlock Much Wenlock Limestone
Formation of Dudley (Thomas 1981, pi. 19, figs. 4, 11, 12), though the thorax and pygidium were
described from non-topotype Malverns’ material from the slightly older Coalbrookdale Formation
(Thomas 1981, pi. 19, figs. 6-8). Isolated pygidia and thoracic segments of Dicranopeltis are known
from Dudley but Thomas assigned these to other species of this genus; unlike them the Malverns’
material, which includes one specimen with pygidium, thorax, and (partial) cranidium all articulated
(Thomas 1981, pi. 19, fig. 6), shows the same paired arrangement of larger granules on the pygidial
and thoracic axes and median glabellar lobe as do topotype salteri cranidia. In particular, topotype
salteri , when well preserved (Thomas 1981, pi. 19, fig. 11), show four pairs of larger granules on
the median lobe.

The three Kerry Dicranopeltis cranidia (PI. 23, figs. 1-3) appear similar to the cranidium of
salteri except that on them the lp glabellar, preglabellar, and axial furrows seem more distinct
than those described (Thomas 1981, p. 68) for the English material, but the illustrations of the
latter show that the Irish and English specimens cannot be distinguished on these features. However,
on the two Irish cranidia which show a complete frontomedian glabellar lobe (PI. 23, figs. 1-3, 5,
6), there are only three pairs of larger granules. The significance of this is debatable, since the
sample size is small and, moreover, four pairs cannot always be confirmed on English specimens
due to preservational shortcomings, this apparently applying even to the lectotype and paralectotype
(see Thomas 1981, pi. 19, figs. 4 and 12).

The pygidium of the Irish form is certainly distinct from the non-topotype Malverns’ material
ascribed to salteri. The axis is relatively smaller with respect to the size of the pleural region, being
only 0-3 times as wide as the pygidial width anteriorly. In the Malverns’ material the axis appears
to be about 04 times as wide; Thomas (1981, p. 69) said about 05 times as wide but this figure
seems excessive for the two reasonably complete pygidia he figured (1981, pi. 19, figs. 6 and 7).
The postaxial sector narrows less rapidly posteriorly. The pygidial spines are much longer. The
axis has only two axial rings, though admittedly the two extra incomplete rings in the Malverns’
pygidia are only indicated in some specimens (Thomas 1981, p. 69; even though he gave the
number of axial rings as a diagnostic character for salteri). The paired larger granules on the first
axial ring are much less widely separated compared with those on the second ring.

It is clear that there is more than one Dicranopeltis taxon exhibiting paired, granular, axial
sculpture (see also below) and due to the uncertainties involved I have questionably assigned the
Irish material to saheri , and consider that the different Malverns’ material on present evidence
should also be questionably referred to this species. Dicranopeltis woodwardi (Reed, 1903) from
the Much Wenlock Limestone Formation of Dudley, which is known from pygidia only, differs
from the Irish species in showing only one complete axial ring, a subparallel posterior part to the
postaxial sector, and the lack of paired granular sculpture.

Paired rows of larger granules have also been indicated in some cranidia attributed to D. scabra
scabra and D. scabra propinqua (Barrande, 1846) from the Wenlock and low Ludlow of Bohemia
(Vanek 1959, text-fig. 13; Prantl and Vanek in Horny el al. 1958, p. 271; contrary to the
comments of Thomas 1981, p. 69). There is need of careful revision of the scabra material and of
first-hand comparison of scabra and salteri specimens to clarify the relationship of the two
species.

The synonymy of Lichas laticeps Angelin, 1854 from the Silurian of ?Scandinavia with D. salteri
(see Thomas 1981, pp. 68, 69) seems questionable until the former is revised: it has not
been refigured since being established and Angelin’s illustration is inadequate for modern
comparison.

SIVETER: SILURIAN TRILOBITES FROM IRELAND

155

Subfamily ceratarginae Tripp, 1957
Genus trochurus Beyrich, 1845

Type species. Trochurus speciosus Beyrich, 1845, p. 31, unnumbered plate, fig. 14; from the Liten Formation
(Wenlock), Svaty Jan pod Skalou, Czechoslovakia. By monotypy.

Trochurus ? sp. indet.

Plate 16, fig. 16

1863 Encrinurus sexcostatus\ Baily in Jukes and Du Noyer, p. 12.

1878 Encrinurus sexcostatus\ Baily in Kinahan, p. 40.

1976 Primaspisl sp.; Siveter in Parkin, p. 595, table 1.

1984 Dicranopeltisl sp.; Siveter in Thomas, Owens, and Rushton, p. 55.

Material. One cranidium.

Occurrence. Annascaul Formation, locality 28 of Parkin 1976, p. 587, fig. 4, Annascaul inlier.

Discussion. This fragmentary cranidium is apparently a lichid but is different to the Dicranopeltis
species described above in having the lp furrows extended and meeting sagittally to isolate a
posterior section of the median lobe, and in having the remnant of what is probably an occipital
lobe. Sagittally confluent lp furrows and a transversely, unequally divided median lobe are present
in the specimens of Trochurus sp., Acanthopyge hirsuta (Fletcher, 1850), and Hemiarges bucklaudii
(Milne Edwards, 1840) recently described from the late Wenlock of Britain by Thomas (1981).
However, these Acanthopyge and Hemiarges taxa do not have an occipital lobe and in H. bucklaudii
the posterior section of the median lobe is much longer than that of the Irish specimen. The
remains of the Irish cranidium show no differences to the similarly aged Trochurus sp. of Thomas,
or to some specimens of T. speciosus (see, for example, Vanek 1959, pi. 9, fig. 1; Pfibyl and Vanek
1975, pi. 1, fig. 6), and I have therefore questionably assigned it to this genus, which is typically
a ‘middle’ Silurian taxon, whereas Acanthopyge is typically a Devonian genus.

Watkins (1978, fig. 10a) illustrated a partial cranidium of an unnamed lichid from the upper
Wenlock (Homerian Stage) Ferriters Cove Formation in the west of the Dingle Peninsula. This
specimen appears to have a transversely continuous lp furrow as in the Caherconree specimen of
Trochurus ? sp. indet.

PALAEOECOLOGY

The thinly bedded limestone unit at locality 28 may be up to 20 m thick and contains alternations
of purer carbonate and more silty horizons; in addition to the trilobites it has yielded about ten
species of brachiopods, indeterminate bivalves, fenestellid bryozoans, crinoid ossicles, and is
particularly rich in the corals Coenites repens Wahlenberg and Favosites sp. (Parkin 1976, p. 588,
table 1). The trilobites, an illaenimorph, a cheirurid, and a lichid, are sparse in terms of taxa and
specimens, and though all these families are of the more robust morphological type, they are
represented here by incomplete exoskeletal parts. All the evidence suggests shallow water (see also
below).

The limestone of locality 36 is about 3 m thick and consists essentially of small, light-grey
coloured carbonate nodules enveloped within generally finer grained sediments (Parkin 1976,
p. 588). On sectioning the contact of several nodules with the surrounding sediment, the nodules
are found to consist largely of crinoid debris and small trilobite parts in what may be termed a
silty crinoidal packstone. Contact with the surrounding sediment, a silty wackestone, is very sharp,
with the silt in the wackestone being finer and more abundant. The silty wackestone also contains
trilobites, but these are large specimens of Cheirurus sp. nov.? and Calymene endemopsis sp. nov.,
whereas the crinoidal packstone contains the other taxa from this locality which are all smaller in
size. In addition to the trilobites only athyrid brachiopods and crinoid ossicles were recorded from

156

PALAEONTOLOGY, VOLUME 32

locality 36 by Parkin (1976, p. 588, table 1). The sedimentological evidence suggests that the
crinoidal packstone was already cemented when it was deposited in the finer host-rock wackestone,
but this would not necessarily imply a large age difference between the two or that they originated
any great distance apart.

The abundant trilobite material from locality 36 as a whole represents a death assemblage,
consisting entirely of disarticulated material. Some movement and concentration of the exoskeletal
parts with the crinoid debris must have occurred, but the excellent preservation of the trilobites
in general and the intact preservation of the odontopleurid spines in particular suggests that any
transportation was limited. A lower energy, possibly slightly more offshore and deeper water
environment is suggested for this muddier, coral-free limestone with its diverse trilobite fauna,
compared to that of locality 28. On a regional basis, in upper Wenlock times Parkin (1976,
pp. 602-603) envisaged a palaeoslope dipping approximately east-north-east, away from the Blasket
Islands and Dunquin volcanic centres in the west and towards the Bull’s Head and Annascaul
inkers in the east.

In recent years, Thomas (1980), Mannil (1982, 1986), and Chlupac (1987) have respectively
plotted the occurrence of trilobites with respect to lithofacies distribution across the palaeoslope
of the Welsh Basin during the Wenlock, the Baltic Basin during the Silurian, and the Barrandian
area of Czechoslovakia during the Silurian. The illaenimorph-cheirurid-lichid assemblage from
locality 28 includes the same major groups that form part of the fauna inhabiting the nearshore
shelf limestone lithofacies of the Welsh Basin during the Wenlock (Thomas 1980, fig. 2); moreover
these three groups are typical of shallow water bioclastic limestones from the Ordovician through
to the Devonian (Lane 1972; Fortey 1975; Lane and Thomas 1983). The locality 28 fauna can also
be compared to part of the Bumastus-Sphaerexochus-Cheirurus assemblage occurring in the
volcanic-carbonate facies of the Bohemian Wenlock, which also occupies a similar nearshore
palaeoslope position (Chlupac 1987, p. 172, fig. 2).

The trilobite families from locality 36 compare well with those inhabiting the shelf limestone
and adjacent offshore shale belt on the Welsh palaeoslope of Wenlock age. The restricted presence
there of the Styginidae (= Scutelluidae of Thomas 1980, fig. 2) on the outer part of the shelf
limestones, close to where they interfinger with the shale belt, may by analogy suggest a similar
position for the Kerry fauna which also contains the styginid Kosovopeltis, though only one Irish
specimen is known. Consideration of the sedimentological factors outlined above may also suggest
a more distal rather than proximal position on the shelf area. At the generic level the fauna from
locality 36 does not compare closely with the Proetus-Warburgella association or Dalmanites-
Raphiophorus association which Thomas (1980, fig. 3) used to typify, respectively, the shelf
limestone and the shale belt lithofacies of the Welsh Basin. In the Wenlock of Bohemia the
Bumastus-Cheirurus-Sphaerexochus, the Miraspis, and the Aulacopleura konincki assemblages,
which in total encompass the nearshore volcanic-carbonate to more offshore calcareous and
tuffaceous shale facies, all have genera in common with the locality 36 fauna.

A comparison of the Irish trilobite fauna with the Wenlock palaeoslope associations identified
by Mannil (1982, 1986), shows that the generic diversity of the East Baltic trilobites is much lower
and the composition different. There is no species the same, and generically only Leonaspis and
Calymene are common to both areas at this time, the latter genus occurring right across the East
Baltic palaeoslope from nearshore to the central part of the 'basin’, and the former being found
in the slope facies.

BIOGEOGRAPHY AND PALAEOGEOGR APH Y

Genera which occur elsewhere in the Annascaul Formation but which are unknown from the two
limestone outcrops include Encrinurus , Dalmanites, and Proetus (Siveter, in prep.). The genera
described herein are mostly cosmopolitan, with the exception of the relatively newly defined
Interproetus , which also occurs in Bohemia, Morocco, and England. At the specific level faunal
links with several areas are evident. Many of the Annascaul species are closely related to forms in

SIVETER: SILURIAN TRILOBITES FROM IRELAND

157

the Welsh Borderland, especially A. aff. stokesii, D. salteril , and Cheirurus sp., whilst L. coronata
coronata and O. ( O .) ovata actually occur there. The Irish species also show some similarity with
those from Bohemia: O. (O.) ovata , the most common species in the Irish fauna, occurs there,
Interproetus species are best represented there, and D. salteril is very close to D. scabra , as is
Cheirurus sp. to C. insignis. Species’ connections with Scandinavia appear less strong but a
subspecies of L. coronata occurs in the late Wenlock of Gotland as does A. amelangi , a form very
close to A. aff. stokesii. North American species have not been recognized but L. parkini shows
the closest connection, standing very near to L. lenzi from north-west Canada. The Irish Kosovopeltis
species has affinities with a Greenland form.

Thus species’ connections can be made with the Welsh Borderland and Bohemia to the east,
where they are strongest, and Scandinavia to the north-east. America, to the west, has some very
similar forms without having any species in common. This pattern conforms closely to the triparite
origin of Silurian faunal elements in the British Isles, anticipated by Holland (1969u, h) following
the suggestions of Holland and Lawson (1963) and Turner (1935). Bassett (1974), Holland (19696),
and others have commented on the Wenlock age rocks that occur under eastern England and their
connection with those in central Europe; this indicates there was a possibility for migration between
Ireland and Bohemia via the Welsh Borderland. The faunal connection of the Dingle trilobites
with those of the Welsh Borderland does not support the idea that these two areas were separated
by a southerly extension of the Irish Sea landmass from middle Wenlock to early Ludlow times
(Hirst et al. 1978, text-figs. 8-10). An alternative migration pathway between Ireland and Bohemia
bypassing the Welsh Borderland may have been along the east-west trending Rheic Ocean just to
the south, which was the site of the sediments involved in the late Palaeozoic orogeny, but which
has been postulated to be in existence at least as early as the late Llandovery (McKerrow and
Cocks 1977, text-fig. 1; Cocks and Fortey 1982, text-fig. 5).

Acknowledgements. The following generously loaned material in their care: Dr J. Archer and Dr A. Sleeman
(Geological Survey of Ireland, Dublin), Mr P. Doughty (Ulster Museum, Belfast), Dr J. Destombes (Ministere
de l’Energie et des Mines, Rabat), Dr A. Fortey and Mr S. F. Morris (British Museum, Natural History),
Dr R. M. Owens (National Museum of Wales), Mr H. P. Powell (Oxford University Museum), Dr R. B.
Rickards (Sedgwick Museum, Cambridge), and Dr W. Struve (Forschungsinstitut Senckenberg, Frankfurt).
Drs B. D. E. Chatterton (University of Alberta), R. A. Fortey, P. D. Lane (Keele University), R. M. Owens
and A. T. Thomas (Aston University) discussed various aspects of the systematics with the author, and Dr
M. Pedley (Hull University) the nature of the carbonate sediments. Dr Thomas kindly showed the author a
manuscript he has prepared with Dr D. J. Holloway (Museum of Victoria) on lichids. I am grateful to Dr
John Parkin and Dr Colin Harris (Gearhart, Aberdeen) for helping to collect bulk material. Much of the
work was undertaken in the Department of Geology, Trinity College, Dublin, during an Irish Department
of Education Fellowship, which I acknowledge; Professor C. H. Holland is thanked for providing research
facilities. Mr Paul McSherry (Hull University) drafted the figure.

REFERENCES

alberti, G. k. b. 1967. Neue obersilurische sowie Unter- und Mitteldevonische Trilobiten aus Marokko,
Deutschland und einigen anderen europaischen Gebieten. 2. Senckenberg. ieth. 48, 48 1 509.

— 1969. Trilobiten des jungeren Siluriums sowie des Unter und Mitteldevons. I. Mit Beitragen zur Silur-
Devonstratigraphie einiger Gebiete Marokkos und Oberfrankens. Abb. senckenb. naturforsch. Ges. 520,
I 692.

— 1970. Trilobiten des jungeren Siluriums sowie des Unter- und Mitteldevons. II. Ibid. 525, 1-233.
aldridge, r. j. 1980. Notes on some Silurian conodonts from Ireland. J. Earth Sci. R. Soc. Dublin , 3, 127

132.

balashova, e. a. 1968. Trilobites of the Skala and Borschev horizons in Podolia, In Siluro-Devonian faunas
of Podolia, 95-124. Leningrad University. [In Russian ]
barrande, J. 1846. Notice preliminaire sur le Systeme Silurien et les trilobites de Boheme , vi+97 pp. C. L.
Hirschfeld. Leipzig.

barton, d. c. 1916. A revision of the Cheirurinae, with notes on their evolution. Wash. Univ. Stud, scient.
Ser. 3, 101 152.

158

PALAEONTOLOGY, VOLUME 32

bassett, m. g. 1974. Review of the stratigraphy of the Wenlock Series in the Welsh Borderland and South
Wales. Palaeontology., 17, 745-777.

Bergstrom, J. 1973. Organisation, life and systematics of trilobites. Fossils Strata , 2, 1-69.
beyrich, E. 1845. Ueber einige bohmischen Trilobiten , 47 pp. Reimer, Berlin.

billings, e. 1869. Descriptions of some new species of fossils with remarks on others already known, from
the Silurian and Devonian rocks of Maine. Proc. Portland Soc. nat. Hist. 1, 104-126.
bruton, d. l. 1965. The Middle Ordovician of the Oslo Region, Norway. 19. The trilobite family
Odontopleuridae. Norsk, geol. Tidsskr. 45, 339-356.

— 1967. Silurian odontopleurid trilobites from Sweden, Estonia and Latvia. Palaeontology, 10, 214-
244.

— 1968. A revision of the Odontopleuridae (Trilobita) from the Palaeozoic of Bohemia. Skr. norske
Vidensk-Akad. Mat. -nature. Kl. (ns), 25, 1 73.

1983. Cambrian origins of the odontopleurid trilobites. Palaeontology, 26, 875-885.

Campbell, K. s. w. 1967. Henryhouse trilobites. Bull. Okla. geol. Surv. 115, 1-68.

— 1977. Trilobites of the Haragan, Bois d’Arc and Frisco formations (early Devonian), Arbuckle Mountain
region, Oklahoma. Ibid. 123, 1-227.

chatterton, b. d. e. 1971. Taxonomy and ontogeny of Siluro-Devonian trilobites from near Yass, New
South Wales. Palaeontographica, A 137, 1-108.

— Johnson, b. d. and Campbell, k. s. w. 1979. Silicified Lower Devonian trilobites from New South Wales.
Palaeontology, 22, 799 837.

— and perry, d. g. 1979. Acanthalomina Prantl and Pfibyl, a valid subgenus of the trilobite genus
Diacanthaspis. J. Paleont. 53, 1327 1342.

— 1983. Silicified odontopleurid trilobites from the MacKenzie Mountains. Palaeontographica
Canadiana, 1, 1-127.

— and wright, a. j. 1986. Silicified Early Devonian trilobites from Medgee, New South Wales. Alcheringa,
10, 279-296.

chlupac, i. 1977. The phacopid trilobites of the Silurian and Devonian of Czechoslovakia. Rozpr. ustfed.
Ust. geol. 43, 1-172.

— 1987. Ecostratigraphy of Silurian trilobite assemblages of the Barrandian area, Czechoslovakia. News!.
Stratigr. 17, 169 186.

cocks, L. r. m. and fortey, r. a. 1982. Faunal evidence for oceanic separations in the Palaeozoic of Britain.
J. geol. Soc. Loud. 139, 451-454.

cuvier, G. L. c. F. D. 1844. Le Regne animal distribue d'apres son organisation, pour servir de base a L'histoire
naturelle des animaux et d' introduction a Vanatomie comparee, par Georges Cuvier. Edition accompagnee de
planches gravees representant les types de tous les genres , les characteres distinctifs des divers groupes et les
modifications de structure sur lesquelles repose cette classification; par une reunion de disciples de Cuvier. Les
Crustaces avec un atlas par H. Milne Edwards, 278 pp. Portin, Masson and Co., Paris.
desmarest, a. g. 1817. In Nouv. Diet. Hist. nat. (2nd edn.), 8, 517-518.

eldredge, n. 1973. Systematics of Lower and lower Middle Devonian species of the trilobite Phacops
Emmrich in North America. Bull. Am. Mus. nat. Hist. 151, 285-337.
emmrich, h. f. 1839. De trilobitis, 56 pp. Berolini, Berlin.

etheridge, r. and mitchell, j. 1893. The Silurian trilobites of New South Wales, with references to
those of other parts of Australia. Part II. The genera Proetus and Cyphaspis. Proc. Linn. Soc. N.S.W. 8,
169- 178.

— 1896. The Silurian trilobites of New South Wales, with references to those of other parts of
Australia. Part IV. The Odontopleuridae. Proc. Linn. Soc. N.S.W. 11, 694 721.

Fletcher, T. w. 1850. Observations on Dudley trilobites. Q. Jl geol. Soc. Lond. 6, 235-239.
fortey, r. a. 1975. Early Ordovician trilobite communities. Fossils Strata, 4, 331-352.

— and Owens, R. m. 1975. Proetida— a new order of trilobites. Ibid. 227-239.

gaertner, h. r. von. 1930. Silurische und tiefunter-devonische Trilobiten und Brachiopoden aus den
Zentralkarnischen Alpen. Jber. preuss. geol. Landesanst . 51, 188 252.
hawle, i. and CORDA, A. J. c. 1847. Prodrom einer Monographic der bohmischen Trilobiten, 176 pp. J. G.
Calve, London.

hirst, j. m., Hancock, n. j. and mckerrow, w. s. 1978. Wenlock stratigraphy and palaeogeography of Wales
and the Welsh Borderland. Proc. Geol. Ass. 89, 197-226.

Holland, c. h. 1969a. The Irish counterpart of the Silurian of Newfoundland. In kay, m. (ed.). North
Atlantic geology and continental drift, a symposium. Mem. Am. Assoc. Petrol. Geol. 12, 298-308.

SIVETER: SILURIAN TRILOBITES FROM IRELAND

159

— 19696. The Welsh Geosyncline in its regional context. In wood, a. ( ed . ) . The Pre-Cambrian and Lower
Palaeozoic rocks of Wales , 203-217. University of Wales Press, Cardiff.

1981. Silurian. In Holland, c. h. (ed.). A geology of Ireland, 65 81. Scottish Academic Press, Edin-
burgh.

— and lawson, j. d. 1963. Facies patterns in the Ludlovian of Wales and the Welsh Borderland. Lpoo!
Manchr geol. J. 3, 269 288.

holloway, D. J. 1980. Middle Silurian trilobites from Arkansas and Oklahoma, U.S.A. Palaeontographica,
A 170, 1-85.

horny, r. and bastl, f. 1970. Type specimens of fossils in the National Museum , Prague, Volume 1. Trilobita,
354 pp. Prague.

— prantl, f. and vanek, j. 1958. On the limit between the Wenlock and the Ludlow in the Barrandian.
Sb. ustfed. Ust. geol. (Paleont.), 24, 217-278. [In Czech with English summary.]

howells, Y. 1982. Scottish Silurian trilobites. Palaeontogr. Soc. [Monogr.], 1-76.

hucke, k. and voigt, e. 1967. Einfiihrung in die Geschiebeforschung (Sedimentargeschiebe). Ned. geol. Veren.
1-132.

ingham, j. k. 1970. A monograph of the upper Ordovician trilobites from the Cautley and Dent districts of
Westmorland and Yorkshire, part 1. Palaeontogr. Soc. [Monogr.], 1-58.
international commission on zoological nomenclature. 1987. Opinion 1433. Bull. zool. Nom. 44, 132-
133.

jukes, d. b. and du noyer, g. v. 1863. Explanation of sheets 160, 161, 171 and part of 172, and of the
engraved section, sheet no. 15. Mem. geol. Surv. Ireland , 57 pp. Dublin and London.
kinahan, g. h. 1878. Manual of the geology of Ireland , 444 pp. C. Kegan Paul, London.
kindle, c. h. 1945. Some Silurian trilobites from Port Daniel, Quebec. J. Paleont. 19, 529-533.

LANE, P. D. 1971. British Cheiruridae (Trilobita). Palaeontogr. Soc. [Monogr.], 1-95.

— 1972. New trilobites from the Silurian of north-east Greenland, with a note on trilobite faunas in pure
limestones. Palaeontology, 15, 336-364.

— 1984. Silurian trilobites from Hall Land and Nyeboe Land, western North Greenland. Rapp. Gronlands
geol. Unders. 121, 53-75.

— and owens, R. m. 1982. Silurian trilobites from Kap Schuchert, Washington Land, western North
Greenland. Rapp. Gronlands geol. Unders. 108,41-69.

— and thomas, a. t. 1983. A review of the trilobite suborder Scutelluina. In briggs, d. e. g. and lane,
p. d. (eds.). Trilobites and other arthropods: papers in honour of Professor H. B. Whittington, F.R.S.
Spec. Pap. Palaeont. 30, 141-160.

m‘coy, f. 1846, A synopsis of the Silurian fossils of Ireland, 72 pp. University Press, Dublin.

— 1849. On the classification of some British fossil Crustacea, with notices of new forms in the university
collection at Cambridge. Ann. Mag. nat. Hist. (2), 4, 161 179, 330 -335, 392-414.

mckerrow, w. s. and cocks, l. r. m. 1977. The location of the Iapetus Ocean suture in Newfoundland. Can.
J. Earth Sci. 14, 488-495.

maksimova, z. 1968. Middle Palaeozoic trilobites of Central Kazakhstan. Trudy vses. nauchno-issled. geol.
Inst, ns, 165, 1 208. [In Russian.]

mannil, R. 1977. East Baltic Lower Silurian Calymenidae (Trilobita). In kaljo, d. (ed.). Facies and fauna of
the Baltic Silurian. Akad. Nauk. Eston. SSR Inst. Geol. Tallin , 240 258. [In Russian with English and
Estonian summaries.]

— 1982. Wenlock and late Silurian trilobite associations of the East Baltic area and their stratigraphical
value. In kaljo, d. and klaamann, e. (eds.). Ecostratigraphy of the East Baltic Silurian. Ibid. 63-70. [In
English with Russian summary.]

1983. Upper Silurian Calymenidae (Trilobita) of the East Baltic. In Palaeontology of the ancient
Palaeozoic of the Prebaltic and Podolia. Ibid. 72-100. [In Russian with Estonian and English summaries.]

— 1986. Distribution of trilobites in different facies of the East Baltic Silurian. In kaljo, d. and klaamann,
e. (eds.). Theory and practice of ecostratigraphy. Ibid. 99 109, 271. [In Russian with English summary ]

neben, w. and kreuger, h. h. 1973. Fossilien Ordovischer und Silurischer Geschiebe. Staringia, no 2, Ned.
geol. Veren. 1-10 [unnumbered].

Northrop, s. a. 1939. Paleontology and stratigraphy of the Silurian rocks of the Port Daniel-Black Cape
region, Gaspe. Spec. Pap. geol. Soc. Am. 21, 1-302.
owen, a. 1981. The Ashgill trilobites of the Oslo Region, Norway. Palaeontographica , A 175, 1-88.

— and bruton, d. l. 1980. Late Caradoc-early Ashgill trilobites of the central Oslo Region, Norway.
Paleont. Contrib. Oslo Univ. 245, 1-62.

160 PALAEONTOLOGY, VOLUME 32

OWENS, R. M. 1973. British Ordovician and Silurian Proetidae (Trilobita). Palaeontogr. Soc. [Monogr.],

I 98.

- 1974. The affinities of the trilobite genus Scharyia , with a description of two new species. Palaeontology ,

17, 685-697.

parkin, j. 1976. Silurian rocks of the Bull’s Head, Annascaul and Derrymore Glen inkers, Co. Kerry. Proc.
R. Ir. Acad. 76(B), 577-606.

palmer, d. c. 1970. A stratigraphical synopsis of the Long Mountain, Montgomeryshire. Proc. geol. Soc.
Loud. 1660, 341-346.

perry, d. g. and chatterton, b. d. e. 1979. Wenlock trilobites and brachiopods from the Mackenzie
Mountains, north-west Canada. Palaeontology , 22, 569 607.
prantl, F. and pribyl, a. 1949. A study of the new superfamily Odontopleuracea nov. superfam. (Trilobites).

Rozpr. ustfed. Ust. geol. 12, 1-221. [In Czech and English with Russian Summary.]
pribyl, a. and vanek, J. 1965. Neue Trilobiten des bohmischen Ordoviziums. Vest, ustfed. Ust. geol. 40, 277-
282.

— 1966. Zur kenntnis der Odontopleuridae — Trilobiten aus dem bohmischen Altpalaozoikum. Acta
Univ. Carol. Geol. 4, 289-304.

— 1973. Uber hypostoma von Odontopleuriden (Trilobita) und ihrer systematik. Cas. Miner. Geol.

18, 301-307.

— 1975. Revision der Beyrichschen typen bohmischer trilobiten. Sb. geol. red. Paleont. 17, 1-69.

— 1981. Studie zur morphologie und phylogenie der familie Otarionidae R. and E. Richter, 1926
(Trilobita). P alaeontographica , A173, 160-208.

— and horbinger, f. 1985. New taxa of Proetacea (Trilobita) from the Silurian and Devonian of
Bohemia. Cas. Miner. Geol. 30, 237-252.

- 1986. New trilobites of the families Odontopleuridae, Lichidae and Raphiophoridae from the
Silurian and Devonian of central Bohemia. Ibid. 267-278.

— and pek, i. 1985. Phylogeny and taxonomy of family Cheiruridae (Trilobita). Sb. Praci pfir. Fak.
palack. Univ. Olomouci , Geographica-Geologia , 14, 83, 107-193.

ramskold, l. 1984. Silurian odontopleurid trilobites from Gotland. Palaeontology , 27, 239-264.

— 1985. Silurian phacopid and dalmanitid trilobites from Gotland. Stockh. Contr. Geol. 40, 1-62.
Raymond, p. e. 1916. New and old Silurian trilobites from southeastern Wisconsin, with notes on the genera

of the Illaenidae. Bull. Mus. comp. Zool. Harv. 60, 1-41.
reed, F. r. c. 1902. Notes on the genus Lichas. Q. Jl geol. Soc. Lond. 58, 59-82.

rickards, R. b. 1978. Silurian. In moseley, f. (ed.). The geology of the Lake District. Occ. Publ. Yorks. Geol.
Soc. 3, 130-145.

salter, j. w. 1853. A monograph of the British trilobites from the Cambrian, Silurian and Devonian
Formations, part 1. Palaeontogr. Soc. [Monogr.], 1-80.
schrank, e. 1969. Odontopleuriden (Trilobita) aus silurischen Geschieben. Ber. dt. Ges. geol. Wiss. A. 14,
705-726.

1970m Calymeniden (Trilobita) aus Silurischen Geschieben. Ibid. 15, 109-146.

— 19706. Die trilobiten des Silurs der Bohrung Leba 1 (Ostseekusts der VR Polen). Ibid. 373-386.

— 1972. Proetacea, Encrinuridae und Phacopina (Trilobita) aus silurischen Geschieben. Geologie , 76,

1-117.

Shirley, J. 1933. A redescription of the known British Silurian species of Calymene (s.l.). Mem. Proc. Manchr
lit. phil. Soc. 77, 5-67.

— 1936. Some British trilobites of the family Calymenidae. Q. Jl geol. Soc. Lond. 92, 384-422.
siveter, d. j. 1980. Evolution of the Silurian trilobite Tapinocalymene from the Wenlock of the Welsh

Borderlands. Palaeontology, 23, 783-802.

1986 (for 1985). The type species of Calymene (Trilobita) from the Silurian of Dudley, England. Ibid.
28, 783-792.

snajdr, m. 1958. Nekolik novych rodu trilobitu z celede Scutelluidae. Vest, ustfed. Ust. geol. 33, 177-184.
1960. A study of the family Scutelluidae (Trilobitae). Rozpr. ustfed. Ust. geol. 26, I 222. [In Czech with
English summary, 223-265.]

— 1975. New Trilobita from the Llandovery at Hyskov in the Beroun area, central Bohemia. Vest, ustfed.
Ust. geol. 50, 311-316.

— 1977. New genera of Proetidae from the Barrandian, Bohemia. Ibid. 52, 293-298.

— 1978. Llandoverian trilobites from Hyskov (Barrandian area). Sb. geol. ved. Paleont. 21, 7-47.

— 1979. Note on the regenerative ability of injured trilobites. Vest, ustfed. Ust. geol. 54, 171 173.

SIVETER: SILURIAN TRILOBITES FROM IRELAND

161

1980. Bohemian Silurian and Devonian Proetidae (Trilobita). Rozpr. ust fed. Ust. geol. 45, 1 324.

1982. Bohemian Silurian and Devonian Calymenidae (Trilobita). Cas. Mirier. Geol. 27, 371-378.

— 1984a. Ivanopleura and Borkopleura, new odontopleurid genera from the Bohemian Silurian (Trilobita).
Vest, ust fed. Ust. geol. 59, 49-52.

1 984/7. On the genus Odontopleura Emmrich, 1839 (Trilobita). Ibid. 95 104.

1984c. Revision of the trilobite type material of I. Hawle and A. J. C. Corda, 1847. Sb. ndr. Mas. Praze,
39B, 129-212.

— 1984 d. Bohemian Ordovician Odontopleuridae (Trilobita). Sb. geol. ved. Paleont. 26, 47-82.

— 1984c. Remarks to Bohemian Silurian otarionine trilobites. Vest, iistfed. Ust. geol. 59, 283-289.
temple, j. t. 1970. The Lower Llandovery brachiopods and trilobites from Lfridd Mathrafal, near Meifod,

Montgomeryshire. Palaeontogr. Soc. [Monogr.], 1-76.

tomczykowa, e. 1957. Trilobites from the Wenlock and Lower Ludlow graptolitic shales of the Swiety Krzyz
Mountains. Biul. Inst, geol 122, 83-143. [In Polish with Russian and English summaries.]
thomas, a. t. 1978. British Wenlock trilobites, part 1. Palaeontogr. Soc. [Monogr.]. 1-56.

— 1980. Trilobite associations in the British Wenlock. In Harris, a. l., Holland, c. h. and leake, b. e.
(eds.). The Caledonides of the British Isles reviewed. Spec. Publ. geol. Soc. Lond. 8, 447-451.

— 1981. British Wenlock trilobites, part 2. Palaeontogr. Soc. [Monogr.]. 57 99.

— and owens, R. m. 1978. A review of the trilobite family Aulacopleuridae. Palaeontology, 21, 65-81.

— and rushton, a. w. a. 1984. Trilobites in British stratigraphy. Spec. Rep. geol. Soc. Lond. 16,

1-78.

turner, j. s. 1935. Gotlandian volcanicity in western Europe. Geol. Mag. 72, 145-151.

vanek, j. 1959. Celed Lichaidae Hawle et Corda 1847 ze stredoceskeho starsiho paleozoika (Trilobitae).

Bohem. cent. A, 1, 77 168. [In Czech with English summary.]
watkins, R. 1978. Silurian marine communities west of Dingle, Ireland. Palaeogeog. Palaeoclimatol.
Palaeoecol. 23, 79-118.

weber, v. N. 1932. Trilobites of the Turkestan. Trudy-Geol. Korn, ns 178, 3-151. [In Russian with English
summary.]

whittard, w. F. 1938. The Upper Valentian trilobite fauna of Shropshire. Ann. Mag. nat. Hist. (11), 1, 85-
140.

1961 . The Ordovician trilobites of the Shelve Inlier, West Shropshire, part 6. Palaeontogr. Soc. [Monogr.].
197-228.

Whittington, h. b. 1956a. Type and other species of Odontopleuridae (Trilobita). J. Paleont. 30, 504 520.

— 1956/?. Silicified Middle Ordovician trilobites: the Odontopleuridae. Bull. Mm. comp. Zool. Harv. 114,
155-288.

— 1983. Calymene Brongniart, 1822 in Brongniart and Desmarest, 1822 (Trilobita): Proposed conservation.
Z.N. (S.) 637. Bull. zool. Norn. 40, 176-178.

and Campbell, K. s. w. 1967. Silicified Silurian trilobites from Maine. Bull. Mas. comp. Zool. Harv. 135,
447-483.

-and siveter, d. j. 1986. Type species of the genus Calymene Brongniart (Trilobita) in Brongniart and
Desmarest, 1822 and proposed suppression of the name tuberculatus Briinnich, 1781: rider to Z.N. (S) 637.
Bull. zool. Nom. 43, 105- 106.

Typescript received 28 March 1988
Revised typescript 4 July 1988

DEREK J. SIVETER

Department of Geology
The University of Hull
Hull HU6 7RX, UK

Present address:

The University Museum
Parks Road
Oxford OX1 3PW, UK

THE COMPOSITION AND PA LAEOG EOG R A PH I C A L
SIGNIFICANCE OF THE ORDOVICIAN OSTRACODE
FAUNAS OF SOUTHERN BRITAIN,
BALTOSCANDIA, AND IBERO-ARMORICA

by JEAN M. C. VANNIER, DAVID J. SIVETER and ROGER E. L. SCHALLREUTER

Abstract. A review of more than 250 genera has established the taxonomic composition, patterns of
geographical and stratigraphical distribution, and faunal links for the Ordovician ostracodes of the British
Isles, Ibero-Armorica, and Baltoscandia. Compositional and diversity changes of the faunas can be correlated
with tectonic and ecological controls.

Four orders (Beyrichiocopa, Platycopa, Podocopa, and Leperditiocopa), comprising over fifty families and
more than 800 species are represented. In the three regions ostracodes show high diversity at all taxonomic
levels, especially in Baltoscandia. Palaeocopes and binodicopes represent some 85% of the total number of
Ordovician genera in the three regions. Palaeocope dominance increases from Ibero-Armorica to the British
Isles and then to Baltoscandia. The same dominant families of palaeocopes and binodicopes (tetradellids,
ctenonotellids, bolliids, and circulinids) occur in the three domains. Palaeocope and binodicope diversity
increases during the Arenig-Llanvirn and maximum diversity at all taxonomic levels is reached in all three
regions during the Llandeilo-early Caradoc interval. The later Ordovician is marked by a general decline of
palaeocopes.

Faunal links most clearly occur between Baltoscandia and the British Isles, with twenty-seven genera and
some uppermost Ordovician species in common. There is a uniform increase in faunal similarity between
Britain and Baltoscandia throughout the Ordovician, whereas British/Ibero-Armorican and Ibero-Armorican/
Baltoscandian generic contacts show an irreversible decline (for the Ordovician) after Llanvirn/Llandeilo
times.

Faunal composition and diversity vary with environment: binodicope-rich faunas are typically associated
with clastic and unstable environments (e.g. Ibero-Armorica); more diversified, palaeocope-rich faunas are
typically associated with more stable conditions and carbonate sedimentation (e.g. Baltoscandia). The
biological effects of sea-level changes are attested by, for example, a diversity increase during the Llandeilo
early Caradoc (transgression) and a diversity decrease during the later Caradoc- Ashgill (regression).

The evolving pattern of ostracode links support: 1, the northwards movement of the southern part of the
British Isles (microcontinent Avalonia) towards Baltica and the consequent closing of Tornquist’s Sea by the
Caradoc/Ashgill; 2, the development of the Rheic Ocean between Gondwana (including Ibero-Armorica) and
Avalonia by the mid-late Ordovician.

For more than 130 years taxonomic data have been published on Ordovician ostracodes from
Europe, especially on taxa from the erratic boulders of the Baltic regions (see Schallreuter 1964
et seq.). During this period no investigation has specifically addressed the nature, comparative
composition, and possible palaeogeographic significance of the overall spatial and stratigraphic
distribution of the various Ordovician ostracode faunas of Europe. One major problem has been
that, until recently, most of the Ordovician ostracodes from the British Isles and western and
southern Europe were poorly documented as compared to those of Baltoscandia, a neglect
stemming partly from their often relatively unattractive state of preservation (see Siveter 1978).
However, the last 10 years have witnessed substantial research on key British (e.g. Siveter 1978;
Jones 1986, 1987) and Ibero-Armorican (e.g. Vannier 1986u, b) Ordovican ostracodes, providing
much needed monographic and distributional data. Furthermore, such studies coincided with the

IPalaeontology, Vol. 32, Part 1, 1989, pp. 163-222, pis. 24-30.|

© The Palaeontological Association

164

PALAEONTOLOGY, VOLUME 32

text-fig. I . Occurrence of the main Ordovician ostracode faunas from a, Ibero-Armorica; b, British Isles; c,
Baltoscandia; and other localities in central and eastern Europe (excluding Czechoslovakia). Key papers from
which the main data were collected are as follows:

a, Ibero-Armorica (see Vannier 1983a, 6 , c, 1984a, b, c, 1986a, 6; Vannier and Schallreuter 1983; Lethiers
et al. 1985)

1 3. Armorican Massif, western France: 1, Crozon Peninsula; 2, Lavel area and Menez Belair; 3, Domfront
area, Normandy.

4. Bu?aco syncline, Portugal.

5. Toledo Mountains and Guadalupe, central Spain.

b, British Isles

6. Wales and southern England (Siveter 1978, 1982a, 6, 1983; Schallreuter and Jones 1984; Jones and
Siveter 1983; Schallreuter and Siveter 1983; Jones 1984, 1985, 1986, 1987).

7. Ireland (Orr 1985a, b).

c, Baltoscandia

8. Oslo area (Henningsmoen 1948, 1953, 1954a, 6; Dons and Henningsmoen 1949; Qvale 1980).

9. Siljan district (Thorslund 1940; Hessland 1949; Jaanusson 1957, 1966).

10. Jamtland (Thorslund 1940).

I I. South Bothnian area (Jaanusson 1957).

12. Vastergotland (Henningsmoen 1948; Thorslund 1948; Jaanusson 1957; Schallreuter 19846).

13. Ostergotland (Jaanusson 1957).

14. Tvaren area (Thorslund 1940; Jaanusson 1957).

15. Scania (Troedsson 1918; Schallreuter 1980a).

16. Oland (Jaanusson 1957; Schallreuter 1977c).

17. Gotland (Schallreuter 1969a, 19716, 1972, 19846).

18. Gotska Sandon (Jaanusson 1966).

19. Nyland (Martinsson 1956).

20. Leningrad area (Mannil 1963).

21. Pskov district (Neckaja 1973).

22. Estonia (Sarv 1959, 1962, 1963; Meidla 1983, 1986).

[continued opposite

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

165

revision of the major Ordovician ostracode taxa from Baltoscandia (for example, Schallreuter
1976, 1982a, 1983a) and with a major study (Schallreuter 1978(7) focusing on problems of
classification of Ordovician ostracodes. European Ordovician ostracode faunas are now sufficiently
well documented for an assessment of their affinity and palaeogeographical significance to be
undertaken.

This study deals with ostracode faunas from Ibero-Armorica, the British Isles, and Baltoscandia
(text-fig. 1): domains fundamental to an understanding of the Ordovician palaeogeography of
‘Europe’. Faunal and facies distributions (Cocks and Fortey 1982) indicate that in the early
Ordovician Baltoscandia was sited in temperate latitudes, Ibero-Armorica was positioned on high
latitude Gondwana, together with southern Britain which was later in the Ordovician to form part
of a northerly drifting microcontinent, Avalonia (e.g. McKerrow and Cocks 1986; see also text-
fig. 38).

Herein we present the stratigraphic range and geographic distribution for nearly all of the
recorded Ordovician ostracode genera across Europe (excluding Czechoslovakia), representing
over 200 genera belonging to all major lower Palaeozoic ostracode groups. The aim of the paper
is to compare such distributions from Baltoscandia, the British Isles, and Ibero-Armorica; to
determine possible taxonomic links through the Ordovician; and to correlate any changes in the
patterns of ostracode faunal composition and distribution with possible tectonic or other controls.
Similar types of syntheses are already published for benthic macrofossil groups such as trilobites
(for example, Whittington and Hughes 1972) and brachiopods (for example, Williams 1973, 1976).
To what extent do Ordovician ostracode distributions and dynamics confirm or modify the
separation of Gondwana, Baltica, and Avalonia (and North America: see Schallreuter and Siveter
1985) by such bodies as the Rheic and Iapetus oceans and Tornquist’s Sea (text-fig. 38)?

THE STUDY AREAS: DISTRIBUTION OF ORDOVICIAN OSTRACODES IN EUROPE

Most Ordovician ostracodes known from Europe are recorded either from Baltoscandia, the British
Isles, or Ibero-Armorica (text-fig. 1): areas which reflect three Ordovician palaeogeographic
domains (see above). Correlation with the British sequence (Atlantic graptolite zones) is given in
text-fig. 2. The Estonian sequence (text-fig. 2: 22), zoned on conodonts, chitinozoa, and shelly
macrofaunas (e.g. Mannil 1966, 1971; Roomusoks 1970) is often used as a standard for the Baltic
Ordovician succession. For practical reasons, the Ontikan, Viruan, and Harjuan ‘Series’ of the

23. Latvia (GailTte 1971, 1975a, 6; Gaillte in Ulst et al. 1982).

24. Lithuania (Sidaraviciene 1971, 1975; Pranskevicius 1972; Neckaja 1973 and cited papers).

25. North-eastern Poland (Sztejn 1985).

26. Pomerania (Bednarczyk 1974).

27. Bornholm (Poulsen 1978).

Northern Germany— erratic boulders (see Schallreuter 1984a for further references).

28. Hiddensee.

29. Berlin.

30. Schleswig-Holstein.

31. Isle of Sylt (Schallreuter 1984a, b , 1985a, 1986, 1987).

32. Munster area (Schallreuter 19856).

Central and eastern Europe

33. Moscow syneclise (Prokofiev and Kuznetzov 1982).

34. Podolia (Abushik and Sarv 1983; Krandiesky 1969).

35. Volyn (Krandiesky 1975).

36. Thuringia (Blumenstengcl 1965; Kniipfer 1968).

Locality 8 is in Norway; 9-18, 27 in Sweden; 19 in Finland; 20-24, 33-35 in USSR; 25, 26 in Poland; 28, 29
in German Democratic Republic; 30-32, 36 in Federal Republic of Germany.

166

PALAEONTOLOGY, VOLUME 32

text-fig. 2. Correlation of Ordovician stratigraphy in Ibero-Armorica and Baltoscandia with the British
series and stages, mainly based on the distribution of Atlantic graptolites. All selected localities from
Baltoscandia and Ibero-Armorica are numbered as in text-fig. 1. Correlations based on Bruton (1984),
Hammann et al. (1982), Paris (1981), Williams et al. (1972), Bruton and Williams (1982), and further
references cited therein. Lw, Lower; L, Limestone; M, Mudstone; Mb, Member; Fn, Formation; S, Shale.
Atlantic graptolite genera: Dd, Didymograptus; Dl, Dicellograptus; Dp, Diplograptus\ G, Glyptograptus ;

N, Nemagraptus.

Baltic are sometimes employed herein. Correlation of the Ibero-Armorican successions particularly
uses chitinozoan biostratigraphy (Paris 1981).

Ibero-Armorica (text-fig. Ia)

This area comprises the Armorican Massif of north-west France and Portugal and Spain of the
Iberian Peninsula. Faunal data have been collected from most of the Palaeozoic synclines (Vannier
1986a, b) extending east-west throughout the Armorican Massif (Cogne 1971), from the Alcudian
area of central Spain (Hammann et al. 1982), and from the Bu5aco syncline in Portugal (Paris
1979; Vannier 1986a, b ). Ordovician sequences in Armorica and Iberia (text-fig. 2) show strong
similarities. Coeval shallow water deposits occur in both regions: for example, Arenig (Armorican)
quartzites; Llanvirn, Llandeilo, and Caradoc siltstones, mudstones, and sandstones; and late
Ordovician glacio-marine sediments. Furthermore, except for the Ashgill carbonates of the Rosan
Formation (Armorican Massif) and the Porto-do-Santa-Anna Formation (Portugal), there is a
general absence of limestone. The types of facies present, extending southwards to Morocco
(Destombes 1962, 1971) and presumably to Saudi Arabia (Fortey and Morris 1982; Vannier and
Vaslet 1987), correspond to an inner shelf environment probably (in the earlier Ordovician) at
relatively high latitudes and forming the northern edge of the Gondwana continent (Cocks and
Fortey 1982, text-fig. 4).

Striking faunal resemblances, many at specific level, have been documented between Armorica
and Iberia for many Ordovician benthic groups (Henry et al. 1976; Paris and Robardet 1978;
Henry 1980; Vannier 1986a, b). Trilobites and ostracodes both show increased diversity during the
Llandeilo and scarce, low diversity yet widely distributed Caradoc faunas occur in Armorica and

VANNIER ET AL ORDOVICIAN OSTRACODE FAUNAS

167

Iberia. Most Arenig-Caradoc ostracode species from Iberia are also found in Armorica (Vannier
1983a, b, 1986 a, b ).

Trilobite and ostracode associations are not, however, always uniformly distributed over the
areas in question. Some specific and generic level differences between the northern and southern
part of the Armorican massif are possibly related to sedimentary differences such as grain size and
organic content. During the Llandeilo, and to some extent the Llanvirn, Armorican Massif trilobite
and ostracode associations (Henry 1980; Vannier 1986a, b ) seem to follow a north-south deepening
bathymetric gradient; a possible deep water area occurs from Ancenis to the Montagne Noire in
southern France (Dean 1966; Henry 1980; Cocks and Fortey 1982). Comparable coeval differences
are also observed between northern and southern trilobite (Henry 1980) and ostracode faunas
within the Alcudian zone of central Spain.

All these sedimentological and faunal similarities are presumed to result from similar environmen-
tal influences such as clastic and water energy conditions. Together they support the notion that
Iberia and Armorica were two geographically closely related areas (on Gondwana) throughout the
Ordovician.

Baltoscandia (text-fig. lc)

Ordovician ostracode faunas are known from central Baltoscandia (outcrops and boreholes in
Estonia, Latvia, Lithuania, Leningrad area, Sweden, and Norway) and are also abundantly
documented from erratic boulders of Scandinavia and northern central Europe. In contrast with
the thick terrigenous Ordovician deposits in Ibero-Armorica (over 1500 m for the Armorican
Massif), the epicontinental Ordovician of Baltoscandia rarely exceeds 200 m in total thickness.
Jaanusson (1976) distinguished three major facies belts which are geographically fairly constant
throughout the Ordovician. The north ‘Estonian and Lithuanian’ facies belt contains mainly
calcarenites (includes areas 22, 23, 24 respectively in text-figs. 1 and 2). The central Baltoscandian
facies belt (Jamtland, Siljan district, Ostergotland, Oland: text-figs. 1 and 2, areas 9, 10, 13, 16) is
characterized by carbonates, calcilutites, and calcarenites during the middle Ordovician. Graptolite
shales predominate in the Scanian facies belt (Oslo region, Vastergotland, and Scania: text-figs. 1
and 2, areas 8, 12, 15).

General bathymetry can be inferred from these facies belts. Water depth seems to increase from
the east and north-east towards the south-western Wendian Basin through successive Estonian,
central Baltoscandian, and Scandian facies belts respectively. Contrasting markedly with the
epicontinental carbonates are the thick terrigenous deposits of the Wendian Basin running beyond
‘Tornquist’s line’ (Jaanusson 1976, fig. 6) and separating Baltoscandia from other European
palaeogeographic units.

Generic and specific faunal differences mainly occur between the different facies belts and
apparently relate to sedimentary and bathymetric changes. For example, Arenig cyclopygid
trilobites (Poulsen 1965) are closely associated with dark marginal limestones and shales. Except
for some widespread faunal changes associated with physical events such as sea-level changes, in
many cases Baltoscandian benthic faunas (trilobites, ostracodes) seem at times to change
independently in each facies belt. Using ostracode faunal logs Jaanusson (1976, p. 324) suggested
that ‘each major zone was affected by environmental changes that were specific to the zone’. For
example, the dramatic ostracode faunal change at the Kukruse-Idavere boundary (text-fig. 2) in
northern Estonia is unknown in corresponding Swedish sequences (Dalby Formation). This change
may be related to a major crater-like impact structure, the oldest fill of which is of early Idavere
age (Kala et al. 1984; Lindstrom 1987). By contrast, the ostracode faunal change at the base of
the Skagen Limestone (Caradoc) occurs throughout the central Baltoscandian facies belt.

The nature of both the (benthic) trilobite and planktonic faunas and the (‘temperate’) carbonates
indicate an intermediate latitudinal position for Baltoscandia during the Arenig (Cocks and Fortey
1982). By the end of the Ordovician the occurrence of bahamitic limestone places Baltoscandia
within the tropics (Jaanusson 1973).

168

PALAEONTOLOGY, VOLUME 32

The British Isles (text-fig. 1b)

British Ordovician ostracode faunas are known mainly from the Welsh Borderland, Wales, and
northern England (Siveter 1978; Jones 1986, 1987). Some Ashgill ostracodes are known from
southern Ireland (Orr 1985ft, h). During the early Ordovician, the Welsh Basin is often characterized
by possible marginal outer shelf and slope environments, sometimes associated with transgressive
facies (e.g. during the Llanvirn) and volcanism (Williams 1980; Fortey 1984). The deeper
environments of the Welsh Basin generally occur to the west, as exemplified by the graptolite black
mudstones with bathypelagic and blind benthic trilobites in the Arenig of South Wales (Fortey
and Owens 1978). Sandstones, siltstones, and various limestones found along an arcuate belt from
the Welsh Borders through to South Wales reflect more diversified environments and bottom
conditions appearing in the Welsh Basin during the Flandeilo (Williams 1 980). Ostracode abundance
and diversity both decrease towards the supposed deeper parts of the Welsh Basin (Jones 1986,
1987). The early Caradoc transgression is often followed by environmental conditions (dark
graptolite shales) which seem to affect considerably the distribution of the ostracode fauna in
Wales. Shelf limestones and breaks in succession are common features of the Ashgill of southern
Britain (Williams 1980).

Southern Britain, including the Fake District and southern Ireland, formed part of the Avalonian
microcontinent which supposedly drifted northwards, away from its parent plate Gondwana,
through the Ordovician (McKerrow and Cocks 1986; Pickering et al. 1988). Northern Britain (not
considered herein: Ordovician ostracode faunas unrevised) was attached to the low latitude North
American plate on the northern margin of the lapetus Ocean during the Ordovician. In general,
the Cambrian to Silurian evolution of this ocean is well documented using sedimentological,
structural, tectonic, and faunal evidence. McKerrow and Cocks (1976) argued for a late Ordovician
oceanic width of 2000-3000 km. Schallreuter and Siveter (1985) demonstrated mid- and late
Ordovician ostracode faunal links across the lapetus Ocean, between the North American plate
and southern Britain -Baltoscandia, and suggested that relevant plates may have been in closer
proximity— a notion also proposed by Pickering el al. (1988). The position of Avalonia relative to
Gondwana, Baltica, and North America during the Ordovician is addressed herein based on
ostracode evidence.

Other European areas

Data are also available from the Moscow syneclise, Podolia, Volyn, and Thuringia (text-fig. 1: 33-
36 respectively). The ostracode faunas and (limestone) sediments of Podolia (Abushik and Sarv
1983) suggest that Baltoscandian influences extended to the south-west parts of Europe during the
Ordovician. The upper Ordovician ostracodes from Thuringia (Kniipfer 1968) have previously
been assumed to be endemic, but in fact also show some Baltic affinities (unpublished information).
The Ordovician ostracode faunas of Czechoslovakia (e.g. Pribyl 1975, 1979) will be the subject of
another paper.

ORIGIN AND RELATIVE VALUE OF THE OSTRACODE DATA

For faunal comparison it is fundamental to appreciate that documentation of the various European
ostracode faunas is not uniform (text-fig. 3). Baltoscandian ostracode faunas are considerably
more fully documented compared to those of the relatively more recently monographed British
and Ibero-Armorican faunas. Text-figs. 10-17 record some 200 genera from Baltoscandia and
about fifty and thirty from the British Isles and Ibero-Armorica respectively. The availability and
qualitative and quantitative nature of the data on European ostracode faunas also reflects
lithological, preservational, and other factors such as techniques of study.

Most of the Ibero-Armorican ostracode faunas occur in siltstones and mudstones of upper
Arenig to Caradoc (Marshbrookian) age (Paris 1981; Vannier 1986ft, b). Lack of information from
the upper Caradoc and in the Arenig (Armorican sandstones Formation) seems directly related to

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

169

lithological factors, ostracodes being virtually absent in coarse-grained deposits. Ostracodes are
most abundant in the Llandeilo siltstones and mudstones of the Armorican Massif and Portugal.
Ashgill ostracodes are as yet undescribed from Ibero-Armorica but are known to occur within
limestones of the Rosan Formation at Lostmarc’h in the Armorican Massif (see Paris 1981).
Documented Ibero-Armorican ostracodes were obtained from isolated stratigraphic levels and
localities and continuous coastal profiles. They are often strongly distorted tectonically and, except
for a poorly silicified fauna in the upper part of the Andouille Formation (upper Llandeilo and/or
early Caradoc), occur exclusively as moulds.

□

COARSE CLASH CS (SANDSTONE, QUARTZITE ,ETC)

FINE CLASTICS (SILTSTONES, MUDSTONES, ETC)

CARBONATES

PUBLISHED OSTRACODE DATA STUDY IN PROGRESS

• SECTION, ISOLATED OUTCROP OR BOREHOLE MATERIAL

O ERRATIC BOULDER MATERIAL

□ MOULD MATERIAL 89 SILICIFIED MATERIAL

o CALCAREOUS MATERIAL FROM SILICEOUS ROCKS

text-fig. 3. Nature and range of faunal data from the Ordovician of: a, Ibero-Armorica; b, the British Isles;
c, Baltoscandia. The vertical line between the arrows indicates the stratigraphic range of possible faunal

comparisons between the three regions.

Ordovician ostracodes of southern Britain (Siveter 1978; Jones 1986, 1987) were obtained mostly
from frequently silicified calcareous sediments and as mould faunas in clastic deposits. There is a
lack of ostracodes particularly from the Arenig and the early Llanvirn (the oldest published fauna
comes from the upper Llanvirn Fairfach Group) and from within the Caradoc (the only
recorded Actonian-Onnian ostracodes are scarce mid-uppermost Onnian binodicopes). The Ashgill
limestones from England (Jones 1986, 1987) and Ireland (Orr 1985c/, 6 ) contain a promising, rich
fauna.

Ordovician ostracode faunal data of Baltoscandia are available from the Arenig (BII: Sarv 1959)
to the uppermost Ashgill (FII: see Schallreuter 1984/7). Except for faunas described from the
Caradoc of Norway (Qvale 1980) and Sweden (Sularp Shale, Scania: Schallreuter 1976) and from
the uppermost Ordovician of Norway (Troedsson 1918), most of the documented Baltoscandian
ostracodes were extracted as shells from limestones. Numerous Baltoscandian ostracode faunas
are also preserved as calcareous microfossils but within non-calcareous rock matrix: for example,
in middle Ordovician ‘Backsteinkalk’ cherty limestones or upper Ordovician ‘Ojlemyrflint’ cherts
(Schallreuter 19826). In such cases, hydrofluoric acid techniques are needed to recover the
ostracodes. The best-preserved Baltoscandian faunas have been recovered from (Arenig to Ashgill)
erratic boulders. Unfortunately, many species were first described from boulders for which it is
often difficult to determine an exact provenance and age. However, many of the lithologically and
faunally distinctive erratic boulders are dated as exactly as the faunal data from the outcrop allow
(e.g. see Schallreuter 19696, 1971a , 19856) and in many cases the erratic boulder faunas are much
more completely known than those of the correlative beds in the Scandinavian-Baltic region
generally.

Ostracode faunal data common to Baltoscandia, the British Isles, and Ibero-Armorica currently
span an upper Llanvirn to about late Caradoc time interval (text-fig. 3). The present study
concentrates on this period though other comparisons, for example between Baltoscandian and
British Ashgill faunas, are also made.

170

PALAEONTOLOGY, VOLUME 32

So.

Superfamily

Family

Subfamily

A

B

C

1 Ventrygyrinae

2 Bubnof fiopsinae

•

PRI MIT I OPS ACE A

i Primitiopsidae

3 Venzavellinae

o

4 Anisocyaminae

5 Bugariktellinae

•

2 Graviidae

1 Subfam. nov.

3 Oepikellidae

1 Oepikellinae

2 Ampletochilininae

•

•

•

1 Eurychilininae

©

4 Eurychilinidae

2 Piretellinae

•

3 Chilobolbininae

•

5 Oepikiidae

•

•

6 Euprimitiidae

1 Euprimitiinae

2 Gryphiswaldensiinae

•

1 Hithinae

•

7 Tvaerenellidae

7

3 Nodambichilininae

•

4 Martinssonopsinae

•

1 Tallinnellinae

Tu)

©

8 Ctenonotellidae

3 Wehrliinae

•

•

•

HOLLINACEA

4 Ctenonotellinae

5 Guberellinae

•

•

1 Tetradellinae

•

•

•

2 Sigmoopsinae

•

•

•

3 Perspicillinae

•

9 Tetradellidae

4 Glossomorphitinae

5 Dillobellinae

•

•

•

®

6 Sylthinae

7 Gunnaropsinae

•

•

8 Subfam. nov.

•

•

10 Sarvinidae

•

11 Fam.nov.

1 Hollininae

•

12 Hollinidae

2 Triemilomatellinae

3 Tetrasacculinae

4 Nodellinae

•

13 Hollinellidae

14 Cherskiellidae

15 Egorovellidae

16 Soanellidae

text-fig. 4. Systematic position and pres-
ence of Ordovician palaeocope ostracode
families and subfamilies in: A, Ibero-
Armorica; B, British Isles; C, Baltoscandia.
The drawing represents a typical valve of
the suborder Palaeocopa (Order Beyrichi-
ocopa). Systematic classification mainly
based on Schallreuter 1978a.

MORPHOLOGICAL CHARACTERISTICS OF THE MAJOR ORDOVICIAN

OSTRACODE TAXA

The Ordovician ostracode faunas treated herein fall into four orders (Beyrichiocopa, Platycopa,
Podocopa, and Leperditiocopa), thirteen suborders, and over fifty families (text-figs. 4 and 5). An
outline of the morphological differences between major groups is given in order to appreciate the
possible significance that morphology might have in association with particular stratigraphic,
spatial, or palaeoecological patterns of distribution.

Problems of classification of Ordovician ostracodes have been discussed by Schallreuter (1966<r/,
b , 1967, 1968a, b , 1973a, 1975, 1978a, 1979, 1985c). The major taxonomic groups are distinguished
on the basis of the overall lobal, sulcal, and dimorphic characteristics of the shell, the presence of
an inner lamella, and on valve overlap conditions. These features almost certainly reflect vital
aspects of the biology (and ecology) of the ostracode.

Suborder Palaeocopa (text-fig. 6)

Palaeocopes are non-sulcate to quadrilobate, typically 0-5-5-0 mm long (adults), and are the most
diversified and prolific Ordovician ostracode group. Their most important lobal-sulcal sculptures
are a preadductorial node (L2) and an adductorial sulcus (S2). An anterior lobe (LI), posterior
lobe (L4), and posteroventral elevation/lobe (L3) also often occur. Adventral sculptures are closely
related to sexual dimorphism and possible brood care; domiciliar dimorphism is rare. Marginal
sculptures as well as a slight (usually left) valve overlap often occur (for example, see Schallreuter
1982a).

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

171

text-fig. 5. Systematic position and pres-
ence of Ordovician non-palaeocope ostra-
code families and subfamilies in: A, Ibero-
Armorica; B, British Isles; C, Baltoscandia.
The non-palaeocope Ordovician faunas dis-
cussed herein fall into four orders: Order
Beyrichiocopa: suborders Binodicopa (2),
Leiocopa (3), Paraparchitocopa (4), and
Eridostraca (5). Order Platycopa: suborders
Cytherelliformes (6), Punciocopa (7), uncer-
tain suborder (8). Order Podocopa: sub-
orders Metacopa (9), Cypridocopa (10),
Cytherocopa (11), and Parapodocopa (12).
Order Leperditiocopa (13).

Typical forms from all the listed suborders
(except 12) are illustrated for general com-
parison. All drawings are of external views
except those for suborders Metacopa and
Cytherocopa (9 and 1 1), whose most signifi-
cant features are internal (stop-pegs and
inner lamella respectively). O., Order; So.,
Suborder. Numbers given to families follow
on from text-fig. 4.

Palaeocopes may be classified into the infraorders Beyrichimorpha and Primitiopsiomorpha.
The latter (comprising only a single superfamily, Primitiopsacea) are non- to bi-sulcate, have
posterior antral dimorphism and, in contrast with other palaeocopes, normally have a right valve
overlap (text-fig. 9). Beyrichiomorphs (e.g. see Schallreuter 1977/, 1978a, 1987) typically have either
an antrum (Hollinomorpha) or crumina (Cruminata = post-Ordovician) as an egg/brood-care area
in heteromorphs (presumed females).

The Hollinomorpha comprises:

a. Superfamily Hollinacea (text-fig. 9). Highly diversified throughout the Ordovician. Non-
sulcate to quadrilobate. Velar dimorphism sometimes occurs, as does ‘marginal’ dimorphism.
Antral dimorphism is expressed as a dolonal or admarginal concavity either without ( = a botulus)
or with loculi. Another adventral dimorphic sculpture, the histium, is present in some groups (e.g.
Tetradellidae). Domicilar, egorovellid (see Schallreuter 1978a) size and/or proportional dimorphism
can also be closely associated with typical antral dimorphism.

b. Superfamily Eurychilinacea (text-fig. 9). In contrast with Hollinacea the velum of Eury-
chilinacea is, typically, tubulose. In many cases both velar dimorphism and dimorphism
affecting marginal structures occur. A loculate antrum rarely occurs and histial sculptures are
unknown.

172

PALAEONTOLOGY, VOLUME 32

DIMORPHIC FEATURES

OVERLAP

text-fig. 6. Main morphological features of palaeocopes, binodicopes, leiocopes, and paraparchitocopes.
From left to right: ventral view; left lateral view; and a schematic cross-section of the ventral contact margin.
Two types of contact margin occur in binodicopes: 1, no overlap; 2, reverse overlap. The following
abbreviations apply to text-figs. 6-9. AA, admarginal antrum; AP, adductorial pit; AMS, adductorial muscle
scars; AR, adventral ridge; AS, adventral sculpture; CG, contact groove; DS, dorsal spine; ET, eye tubercle;
H, histium; HA, histial antrum; IL, inner lamella; L1-L4, lobes 1, 2, 3, 4; LA, lamella; LP, lappet; LV, left
valve; MB, marginal brim; MS, marginal surface; MSC, marginal sculpture; Nl, anterior node; N3, posterior
node; NA, nauplioconch; OP, overlap platform; OSR, outer stop-ridge; PA, paleola; PAL, preadductorial
lobe; PAN, preadductorial node; PDC, posterior domiciliar concavity; PV, pseudovelum; PVL, posteroventral
lobe; RV, right valve; S, sulcus; S2, adductorial sulcus; SL, sulcament; SP, stop-pit; SR, stop-ridge; SRE,
shell reticulation; STP, stop-peg; UM, umbo; V, velum; VC, ventral concavity; VD, velar dolon; VE,

vestibulum; VSR, ventral stop-ridge.

Suborder Binodicopa (text-fig. 6)

Compared to palaeocopes, binodicopes have a relatively simple morphology with comparatively
few taxonomically useful features. Valve sculpture consists of prominent to obsolescent lateral
nodes (Nl, N3), one of which often has a marked spine or is sometimes divided into three swellings.
Unlike most palaeocopes a velum is absent, though a rounded peripheral ridge (pseudovelum) may
occur between lateral and marginal surfaces. The nodes may be completely or incompletely
connected by ridges (for example, see Vannier 1986«). Possible dimorphism in binodicopes mainly
involves slight differences in domicilial shape or in carapace size and outline (e.g. see Schallreuter
1 980c/; Vannier 1 986c/). Intraspecific reversal of valve overlap occurs (Schallreuter 1980/?).

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

173

text-fig. 7. Main morphological features of eridostracans, cytherelliformes, punciocopes, and lomatopisthiids
(suborder uncertain (8); see text-fig. 5). From left to right: ventral view; external left lateral view; internal
left lateral view (except for lomatopisthiids); and a schematic cross-section of the ventral contact margin. All

abbreviations are listed under text-fig. 6.

Suborder Leiocopa (text-fig. 6)

Most leicopes listed herein (text-fig. 16) belong to the family Aparchitidae (see Schallreuter 1973a).
Most are smooth or punctate, lack any lobal-sulcal sculptures, and are slightly umbonate.
Overlapping right valves have a contact ridge and a contact groove running parallel to the margin
(Schallreuter 19776). Adventral sculpture may occur (only on the left valve) as a ridge or a row
of spines.

Suborder Paraparchitocopa (text-fig. 6)

The paraparchitaceans discussed herein all belong to the Jaanussoniidae (Schallreuter 19716, 1986).
They are small to medium-sized, non-sulcate, and inequivalved. The left valve usually bears an
umbonate sculpture and/or a dorsal spine of varying shape and position. Right valve overlap;
sexual dimorphism may occur (Schallreuter 19716).

Suborder Eridostraca (text-fig. 7)

Compared to palaeocopes, binodicopes, and metacopes, Ordovician eridostracans form a minor
group, but uniquely for ostracodes show molt retention (see Schallreuter 1977a, 1987). Only the
Conchoprimitiidae and Eridoconchidae are represented in the Ordovician of Europe. Adventral
sculpture may occur within these groups (Schallreuter 1988). A well-marked umbo usually occurs,
as do two dorsal nodes on the nauplioconch (first instar). A broad inner lamella (paleola: see

174

PALAEONTOLOGY, VOLUME 32

text-fig. 8. Main morphological features of metacopes, cypridocopes, and leperditiocopes. Overlap conditions
of metacopes are illustrated by transverse sections 1 and 2. From left to right: ventral view; external left
lateral view and (except for leperditiocopes) internal left lateral view; and a schematic cross-section of the
ventral contact margin. All abbreviations are listed under text-fig. 6.

Gramm 1984) and a sulcament occur internally. Valves connect simply, without overlap (Schallreuter
1977«).

Suborder Cytherelliformes (text-fig. 7)

The commonest cytherelliformes are the Monotiopleuridae, which typically have a contact groove,
left or right valve overlap, and slight ventral concavity. Dimorphism of the kloedenellid type
(Guber and Jaanusson 1964) is expressed as a broad posterior domiciliar concavity in heteromorphs.

Suborder Punciocopa (text-fig. 7)

This suborder is essentially represented by the Kirkbyacea (see Schallreuter and Jones 1984), which
are non-dimorphic and which typically have an adductorial (‘kirkbyan’) pit, a ventricular concavity,
and reticulation (sensu Schallreuter 1973 b).

Suborder uncertain , Family Lomatopisthiidae (text-fig. 7)

Lomatopisthiids occur rarely in the Ordovician of Europe (Schallreuter 1978r/). They have a strong,
ridge-like velar sculpture and a special type of posterior domiciliar dimorphism (Guber and
Jaanusson 1964).

Suborder Metacopa (text-fig. 8)

Metacopes are abundantly documented from throughout the Ordovician of Baltoscandia. The
taxonomy of the Podocopa, including metacopes, is partly based on internal features (Adamczak
1976; Schallreuter 1979). Metacopes are inequivalved, normally lack a calcified inner lamella, and

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

175

text-fig. 9. Main morphological features of three major Ordovician palaeocope superfamilies: primitiop-
saceans, eurychilinaceans, and hollinaceans. External lateral and ventral views of a left heteromorphic valve
are figured in each case. All abbreviations are listed under text-fig. 6.

typically have a larger left valve bearing a contact groove or internal stop-ridges (stop-pegs) (see
Schallreuter 1978c/, 1979, 1986). Antero- and posteroventral outer stop-ridges may also occur
(Schallreuter 1979, pi. 1). Dimorphism is unknown.

Suborder Cypridocopa (text-fig. 8)

Typical Cypridocopa (superfamily Bairdiacea) are mostly very small (<0-5 mm) to medium-sized
(1-2 mm), oblong Ordovician podocopes with a broad inner lamella and often distinct vestibula
(Schallreuter 19786, 1979, 1986). Overlap conditions are like those of metacopes: the smaller right
valve often has a continuous ventral stop-ridge.

Suborder Cytherocopa (text-fig. 8)

All known European Ordovician Cytherocopa belong to the Syltheridae. Most have a broad inner
lamella without vestibula and flattened anterior and posterior external surfaces (Schallreuter 1977c/,
19786, 1984c/).

Suborder Parapodocopa

The Conodomyridae and Fam. nov., represented by the single genera Conodomyra and Oejlemyra
respectively, are assigned to this small group (Schallreuter, unpublished). Conodomyra (Schallreuter
1977c/, 1986) has a completely fused inner lamella which, in contrast with most other Ordovician
podocopes, is broadest centroventrally.

Suborder Leperditiocopa (text-fig. 8)

Ordovician leperditiocopes are extremely rare in Europe. They belong to the families Isochilinidae
and Kiaeriidae (Schallreuter 1984c/). All Baltoscandian forms are kiaeriids (Schallreuter 1984//),
having an anterior and posterior (centroventral) marginal brim and one of two stop-pegs internally
on the overlapping right valve.

STEPS USED IN THE ANALYSIS OF THE OSTRACODE FAUNAS

The occurrence and stratigraphic distribution are given for 223 genera from Baltoscandia, the
British Isles, and Ibero-Armorica (text-figs. 10-17). This accounts for some 800 species and
subspecies. Taxa from Podolia (text-fig. I: 34), Thuringia (text-fig. I: 36), and those of uncertain

176

PALAEONTOLOGY, VOLUME 32

PALAEOCOPA

ORDOVICIAN

Aremg | Llanvirn |Llandello| Caradoc I Ashgill

Bi | Bn I B ui |C lal c'lbl CicIC ll ICnllDi 1 D n I D m I E I Fi a I F i bl Fi c | Fn

Platybolbina
R ( Platybolbina }
R ( Rimabolbina )

P.( Reticulobolbina )
P(Abruptobolbina )
Cystomatochilina

Levisulculus

Moekowia

Gellensia

Ampletochilina

Femerensia

Gotula

Swantina

Oepikella

Actinochilina

Piretella

Hesperidella

Chilobolbina

Laccochilina

L.(Laccochitina)

L.[ Prochilina I

Oepikium

Duringia

Bubnoffiopsis

Eurocyamus

Lembitites

Ardennea

Hillmeria

1-2

1-4

Neoschmidtella

MWMiCTt'tWiaRtiW'i BALTOSCANDIA l l BRITISH ISLES I BERO-ARMORI CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION

SUBGENERA

text-fig. 10. Stratigraphic range and occurrence of Ordovician palaeocope ostracode genera and subgenera
of the superfamilies Eurychilinacea and Primitiopsacea (see text-fig. 4) in: A, Ibero-Armorica; B, British Isles;
C, Baltoscandia. Stratigraphy as for the British Series (Arenig to Ashgill) and the main stratigraphic
subdivisions of the Baltic Ordovician (Bi to Fn; see text-fig. 2). N = taxonomic position of genera and
subgenera according to the serial numbers of families and subfamilies as listed in text-fig. 4; — indicates
‘taxonomic position as above’; S = total number of species within each genus or subgenus. Occurrences
mainly based on papers given in text-fig. 1. All symbols apply to text-figs. 10 17.

systematic position have not been plotted. Occurrences of some (especially the palaeocope)
subgenera are included (text-figs. 10-14). The data were drawn from published and (rarely) in
press studies.

Qualitative and quantitative analysis of these data has been undertaken in four steps (text-
fig. 18). First, the taxonomic composition (in % of genera) of the Baltoscandian, British, and Ibero-
Armorican ostracode faunas is compared for the Ordovician as a whole. Comparison is made at
various taxonomic levels, such as subordinal, familial, and (for tetradellids and ctenonotellids)
subfamilial levels. The relative abundance in the Ordovician of palaeocopes and binodicopes is
also compared between the three areas. Secondly (text-fig. 18: 2), the changing taxonomic

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

177

PALAEOCOPA

ORDOVICIAN

Arenlq |Llanvirn|Llandeilo| Caradoc I Ashgill

Bi | Bn | B julc lalc'ib I Cic I C ii |C ml Pi 1 D n I D m I E I Fial Fib I Fic I Fi

Gryphiswaldensia

Dogoriella

Caprabolbina

Steinfurtia

Bolbina

Brevibolbina

Hithis

Cavhithis

Bolbihithis

Bromidella

Uhakiella

Euprimites

Tvaerenetta

Piretia

Lennukella

Bichilina

Nodambichilina

Eoaquapulex

Tallinnella

Tetrada

Pseudorakverella

Neotsitrella

Quadriha

Q.(Quadritia)

Q.(Krutatia)

Homeokiesowia

Brephocharieis

4-5

i

7-3
7- 4

5- 6
2

Homeoceratopsis

H —

« BALTOSCANDIA l l BRITISH ISLES c==. 1 I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION

SUBGENERA

text-fig. 1 1 . Stratigraphic range and occurrence of Ordovician palaeocope ostracode genera and subgenera
of the families Euprimitiidae, Tvaerenellidae, and Ctenonotellidae (in part; see also text-figs. 4 and 12) in:
A, Ibero-Armorica; B, British Isles; C, Baltoscandia. All symbols as for text-fig. 10.

composition and diversity of the ostracode faunas through the Ordovician within each area have
been examined. Thirdly (text-fig. 18: 3), generic links are identified between the total Ordovician
ostracode faunas of Baltoscandia, the British Isles, and Ibero-Armorica (for example, see text-fig.
29). Fourthly (text-fig. 18: 4), the changing pattern of genera common to two or more of the areas
is examined through Ordovician time.

COMPARISON OF THE TAXONOMIC COMPOSITION OF BALTOSCANDI AN,
BRITISH, AND I BE RO- A R M OR I C A N ORDOVICIAN OSTRACODE FAUNAS

Baltoscandia

The higher taxa. Thirteen suborders are represented in the Ordovician ostracode faunas
of Baltoscandia (text-fig. 19). The main groups are the palaeocopes (54% of total genera)
and binodicopes (18%); other groups include the metacopes (9%) and a range of minor groups

178

PALAEONTOLOGY, VOLUME 32

I m m Hill BALTOSCANDIA I I BRITISH ISLES .'.i:— i I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION

SUBGENERA

text-fig. 12. Stratigraphic range and occurrence of Ordovician palaeocope ostracode genera and subgenera
of the family Ctenonotellidae (in part; see also text-figs. 4 and 1 1) in; A, I hero- Armorica; B, British Isles; C,

Baltoscandia. All symbols as for text-fig. 10.

( < 1 %— 4- 5%). The total fauna reflects broad morphological diversity: compare, for example, a
palaeocope, a leiocope, an eridostracan, and a leperditiocope (text-figs. 6-8). The palaeocope:
binodicope ratio is about 3 : 1 (% genera) and almost 5 : 1 (% species) (text-fig. 22).

Composition of the palaeocopes. Although primitiopsacean, eurychilinacean, and hollinacean
palaeocopes are represented in Baltoscandia (text-figs. 4 and 9), about 75% of the palaeocope
genera are hollinaceans (text-fig. 23). Of the six hollinacean families present, tetradellids are
dominant (37% genera; 32% species). Typical tetradellids (for example, Sigmoopsis rostrata, PI. 27,
fig. 10) have a long sigmoidal adductorial sulcus (S2) and a well-developed dimorphic (botulate or
loculate) velum.

Other important palaeocope groups are the hollinacean ctenonotellids (17% genera; 20% species)
and tvaerenellids (13%; 18%), and the eurychilinacean oepikellids (10%; 10%) and eurychilinids
(4%; 7%). Ctenonotellids (for example, Tallinnella: see Opik 1937; Sarv 1959) have a long S2,
botulate velar dimorphism, and in many cases show a lobal reduction and dissolution into cristae.
In tvaerenellids, dimorphism is expressed as a well-marked dolon running up to the dorsal margin
(for example, in Bromidella sarvi : see Schallreuter 19836)- In contrast, oepikellids (for example,
Platybolbina ( Rimabolbina ) rimer, see Schallreuter 1983a, pi. 13, figs. 5 and 6) and eurychilinids

VANNIER ET AL.. ORDOVICIAN OSTRACODE FAUNAS

179

PALAEOCOPA

ORDOVICIAN

Arenig |Llanvirn|Llandeilo| Caradoc I Ashgill

Bi I B II I B ull C lal C Ibl C ic I C 11 1 C III I D I I D ll I D III I E I F.a | Fib I Fic 1 Fll~

Tetradella

Polyceratella

Ogmoopsis

O. (Ogmoopsis)

O- ( Quadridigitalis )

Regiopsis

Pleuroidella

Consonopsis

Ceratopsis

Kiesowia
K. ( Kiesowia )

K. ( Pseudotallinnella '
K. I Carinoboibina)

5-7

6

14-16

2

Sigmoopsis

Sigmoopsoides

Reuentahna

Semikiesowia

Severobolbma
Valdarella _
Braderupia
Concavhithis

Sigmobolbina

Foramenella

Lomatobolbina

Perspicillum

P { Perspicillum)
P. I Eospicillum )

P. ( Podolibolbina 1

Pentagona _

Ceratobolbina

Deefgella

4

1-2

BALTOSCANDIA r i BRITISH ISLES I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION
SUBGENERA

text-fig. 13. Stratigraphic range and occurrence of Ordovician palaeocope ostracode genera and subgenera
of the family Tetradellidae (in part; see also text-figs. 4 and 14) in: A, Ibero-Armorica; B, British Isles; C,

Baltoscandia. All symbols as for text-fig. 10.

(for example, Piretellci triebeli: see Schallreuter 1983a, pi. 13, fig. 1) have a very simple lobal
sculpture, a velum consisting mainly of hollow tubules, and velar dimorphism associated with
marginal dimorphism.

Ordovician tetratellids of Baltoscandia (text-fig. 25) consist of four main subfamilies: Glossomorph-
itinae (33% of tetradel 1 id genera), Sigmoopsinae (21%), Perspicillinae (18%), and Tetradellinae
(15%). The occurrence and morphological variation of a second adventral sculpture (the histium),
in addition to the velum, determines the recognition of these subfamilies (see Schallreuter 1978a).
Sigmoopsines are characterized by both histial and velar dimorphism; in contrast, in tetradellines
the histium (present in tecnomorphs and heteromorphs) is not dimorphic. In most perspicillines
the histium occurs only in heteromorphs and is connected anteriorly to the velum. Glossomorphiti-
nae have a dimorphic velum and heteromorphs exhibit an adventral sculpture (histio-velum) which

180

PALAEONTOLOGY, VOLUME 32

PALAEOCOPA

ORDOVICIAN

Arenig iLlanvirn |Llandeilo| Caradoc I Ashgill

Bi | Bn |B II |C.a I C lb I Cic I C n ICinlDi 1 D n | Dm | E | F,a | F b I Fic I F n

Hippula
H. ( Hippula I

H. (Cetona )

H. ( Pseudocetona )

Aulacopsis

Hesslandella

Glossomorphites

Huckea

Wehrlina

Vittella

Collibolbina

Pseudohippula

I v aria

Octobolbma

Loculibolbina

Gracquma

Jeanlouisella

Pelecybolbina

Sylthis

9 - 5
9 - 6

Scrobisylthis

Disulcinoldes

Airina

Gunnaropsis

Harperopsis

Histina

Cymabolbina

Latebina

Severopsis

Hastatellina

Distobolbina

Sarvina

Grammolomatella

Semibolbina

Eolomatella

Aloculatia

MMMUwiiMiHim BALTOSCANDIA C ) BRITISH ISLES I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION

SUBGENERA

text-fig. 14. Stratigraphic range and occurrence of Ordovician palaeocope ostracode genera and subgenera
of the families Tetradellidae (see also text-fig. 13), Sarvinidae and Hollinidae (see text-fig. 4) in: A, Ibero-
Armorica; B, British Isles; C, Baltoscandia. All symbols as for text-fig. 10.

is thought to represent a coalesced velum and histium. The Baltoscandian ctenonotellid fauna
(text-fig. 26) is dominated by the subfamilies Steusloffiinae (43% of ctenonotellid genera) and
Tallinnellinae (31%). Their distinctive feature is lobal sculpture (Schallreuter 1976, 1978tf).

Composition of the binodicopes. As dimorphic adventral sculptures are typically lacking in
binodicopes, their classification is especially difficult. Lobal and adventral morphology (see text-
fig. 6) distinguish binodicope families, of which four are dominant in Baltoscandia: bolliids (35%
of binodicope genera), circulinids (22%), aechminids (16%), and spinigeritids (13%) (text-fig. 24).

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

181

BINODIC OP A

Arenlg [Llanvirn |Llandeilo[

ORDOVICIAN

I A s h g I l~T

Bi j B II IB II I C ia | C ibl C icl C i 1 Cm I D i I D n |Dm| E I Fia I Fib I Fic I Fn

Duplexibollia

Klimphores

Vaivanovia

Laterophores

Parphores

Sagittovum

Bollia

Bullaeferum

Reginea

Copelandia

Satiellina

Scanipisthia

Ritatia

Lardeuxella

Quadrijugator

CrescentiUa

Pseudulrichia

Antiaechmina

Spinaechmma

Aechmina

Gen. nov. A

Vogdesella

Circulina

Pedomphalella

Easchmidtella

Orechma

Pariconchopri mitla

Bolliaphores

Pseudbolli a

Thibaulina

Conspicillum

R i villi n a

Eocytherel/a

Ojlemyra

Conchoprimitiella

Spinigerites

Pyxion

Kinnekullea

Tsitrella

Byrsolopsina

BALTOSCANDIA L- -- - l BRITISH ISLES I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION
SUBGENERA

text-fig. 15. Stratigraphic range and occurrence of Ordovician binodicope ostracode genera (see text-fig. 5)
in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. All symbols as for text-fig. 10. For gen. nov. A: see

Vannier 1986r/.

182

PALAEONTOLOGY, VOLUME 32

LEIOCOPA, PARAPARCHITOCOPA, ERIDOSTRACA & PLATYCOPA

N

S

ORDOVICIAN

A

B

C

Arenig

Llanvirn

Llandeilol Caradoc

Ashgill

Bi | B n | B m | C la |C lb I C ic | C

ii IC ml Di | D n | D in | E | Fia | Fib I Fic I F„

Brevidorsa

24

6

•

•

Baltonotella

1 - 2

-

— ■

•

Miehlkella

14

—

•

•

Longidorsa

4

•

Hemiaechminoides

27

2

— -

—

•

Kayina

1

- — ■

•

Jaanussonia

3

•

—

•

Hemeaschmidtella

3

— — - —

—

•

Obliquisylthis

1

■■■■—

•

Conchoprimitia

29

19

•

•

Ahlintella

1

_

•

Cryptophyllus

31

1

—

•

Pygoconcha

-

2

—

—

•

Eridoconcha

2

•

Primitiella

32

18

•

•

Unisulcopleura

2

•

Cavopleura

1

•

Spinopleura

3

—

—

•

Foveaprimitiella

1

—

•

Sudon

2

—

—

•

Karinutatia

2

•

Disparigonya

1

—

•

Fallaticella

33

2

•

Leperditella

-

1

—

•

Martinssonozona

34

1

•

Gebeckeria

?

1

•

Ordovizona

35

2

—

—

•

•

Nonsulcozona

1

—

•

Europisthia

36

1

•

BALTOSCANDIA I' I BRITISH ISLES c bb^jbbim I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION
SUBGENERA

text-fig. 16. Stratigraphic range and occurrence of Ordovician leiocope, paraparchitocope, eridostracan, and
platycope ostracode genera (see text-fig. 5) in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. All

symbols as for text-fig. 10.

Characteristically, bolliids have two prominent lateral nodes (as in Klimpores planus : see
Schallreuter 1980/d and an adventral sculpture (pseudovelum) separating lateral and steep marginal
surfaces. Such nodes occur as spine-like structures in most aechminids (as in Antiaechmina
pseudovelata: see Schallreuter 1977c), which also have an acute angle between the ventral surface
and contact plane. Circulinids (for example, Vogdesella subovata : see Schallreuter 1980a) have
rather obsolete nodes, a very convex domicilium, and a more or less complete pseudo-velum; the
marginal surface (mainly flattened) forms an acute angle with the contact plane (as in most
aechminids). Spinigeritids differ from circulinids by their more elongate carapace and slight
swellings anteriorly and posteriorly (for example, Spinigerites spiniger: see Schallreuter 1980a).
Baltic spinigeritids provide a rare example of dimorphism (merely involving valve size) within the
binodicopes.

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

183

BALTOSCANDIA c - . . I BRITISH ISLES cr— I BERO-ARMOR I CA

INFERRED STRATIGRAPHIC DISTRIBUTION

UNCERTAIN OR INCOMPLETELY DOCUMENTED STRATIGRAPHIC DISTRIBUTION

SUBGENERA

text-fig. 17. Stratigraphic range and occurrence of Ordovician podocope and leperditiocope ostracode genera
(see text-fig. 5) in; A, Ibero-Armorica; B, British Isles; C, Baltoscandia. For genus Y: see Schallreuter 1979.

All symbols as for text-fig. 10.

In summary, Baltic Ordovician ostracode faunas are dominated by palaeocopes (especially
tetradellids and ctenonotellids) and binodicopes (especially bolliids).

The British Isles

The higher taxa. Though the database on British Ordovician ostracode faunas is not as complete
as that for Baltoscandia, palaeocopes and binodicopes again represent the dominant components.
Minor groups (also found in Baltoscandia) include metacopes, eridostracans, and cytherelliformes.
The palaeocope : binodicope ratio (text-fig. 22) is about 1-5:1 (% genera) and 1-8:1 (% species).
Palaeocope dominance over binodicopes is somewhat less marked than in Baltoscandia.

184

PALAEONTOLOGY, VOLUME 32

1

A

B

C

0

•

9

•

2

A

B

C

AS

CC

LLO

LLV

AR

*

4

B

C

0

text-fig. 18. Schematic illustration of successive steps followed in the
present study. Faunal data from: A, Ibero-Armorica; B, British Isles; C,
Baltoscandia. I , comparison of the taxonomic composition of the ostracode
faunas of the three regions for the entire Ordovician (O); 2, the changing
taxonomic composition and diversity of ostracode faunas within each area
through the Ordovician from the Arenig (AR) to Llanvirn (LLV), Llandeilo
(LLO), Caradoc (CC), and Ashgill (AS) series; 3, faunal similarities
between Ibero-Armorica, the British Isles, and Baltoscandia for the whole
Ordovician (O); 4, changes in the nature and amount of taxa common
to two regions (AnB = Ibcro-Arnrorica/British Isles; AnC = Ibero-
Armorica/Baltoscandia; BnC = British Isles/Baltoscandia) and between the
three regions (X = AnBnC) from the Arenig to the Ashgill.

Composition of the palaeocopes. The main families are tetradellids (48% genera; 53% species),
ctenonotellids(33%; 36%), tvaerenellids(7-5%;4-5%), oepikellids(7-5%;4-5%), and oepikiids (3-5%;
2%). These are the same five main components and show the same relative abundance as in the
Baltic palaeocope fauna (text-fig. 23).

There are some differences between British and Baltic palaeocope faunas: ctenonotellids are
relatively much more important palaeocope components in the British Isles, and tvaerenellids are
relatively less important than in Baltoscandia. Differences are also apparent in the composition of
tetradellids in the two areas (text-fig. 25): most British tetradellids belong to the endemic subfamily
Gunnaropsinae (46% genera; 52% species), but perspicillines (18% of tetradellid genera in
Baltoscandia) and the Sylthinae do not seem to occur in the British Isles. Tetradellines and
glossomorphitines are almost equally represented in both regions: 15-5% and 23% and 15% and
33% of total tetradellid genera and species in the British Isles and Baltoscandia respectively (text-
fig. 25). Gunnaropsines typically show well-developed quadrilobation occasionally associated with
cristae, an inclined histial ridge uniting anteroventrally with the velum, and a dolonal antrum. As
for the ctenonotellids (text-fig. 26), the tallinnellines and (then) the steusloffines are the two major
components of the second most generically abundant family in both regions.

Composition of the hinodicopes. The British and Baltic binodicope faunas show strong similarities,
consisting of bollids (34% and 35% respectively of total binodicope genera), circulinids (40%; 22%),
aechminids (13%; 14%), and spinigeritids (1 1%; 13%) (text-fig. 24).

Ibero-Armorica

The higher taxa. The overall taxonomic composition of the Ibero-Armorican fauna (Vannier 1986a,
b) is not fundamentally different from Baltoscandia or Britain, mainly consisting of palaeocopes
and binodicopes with minor proportions of metacopes, eridostracans, and leiocopes (text-fig. 19).

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

185

PALAEOCOPA
PARAPARCH1 TOCOPA
PUNC I OCOPA
CYPRIDOCOPA

|_| BINODICOPA

\Z//\ ER I DOSTRACA

H SUBORDER INC,

BPARAPODOCOPA+
CYTHEROCOPA

jgffg] LE I OCOPA

CYTHERELLI FORMES
P"l METACOPA
m LEPERDITIOCOPA

OCCURRENCE IN BRITISH ISLES

OCCURRENCE IN I BERO-ARMOR I CA

text-fig. 19. Generic percentage abundance of ostracode suborders for the Ordovician and its series in
Baltoscandia. Data for the Caradoc is particularly abundant and well dated and is divided into two parts at
the Dii/Dm boundary (see text-fig. 2).

text-fig. 20. Changes in generic abundance of the
four major ostracode suborders (given in % of the
total fauna) known throughout the Ordovician of
Baltoscandia. Lower and upper part of Caradoc as
defined in text-fig. 19.

PALAEOCOPES

30

B I NOD I COPES
METACOPES
LEIOCOPES

186

PALAEONTOLOGY, VOLUME 32

Palaeocopes

text-fig. 21. Changes in the composition of the
ostracode fauna from Baltoscandia through the Ordo-
vician, given in % of genera of palaeocopes, binodi-
copes, and ‘other groups’. 1, Arenig; 2, Llanvirn; 3,
Llandeilo; 4, 5, lower part and upper part of the
Caradoc (as defined in text-fig. 19); 6, Ashgill.

However, a notable difference for Ibero-Armorica is its very low palaeocope : binodicope ratio
(0-7 : 1, based on % of genera). Binodicopes form the dominant ostracode group, thereby contrasting
markedly with the British Isles and Baltoscandia with ratios of 1-5 ; 1 and 3-1:1, respectively (text-
fig. 22).

Composition of the palaeocopes. Only three palaeocope families have been recognized in the
Ordovician of Ibero-Armorica (text-fig. 23): Tetradellidae (58% palaeocope genera; 66% palaeocope
species), Ctenonotellidae (26%; 24%), and Tvaerenellidae (16%; 8%). Thus, the three major
components of the British and Baltic faunas also occur (with a particularly high proportion of
tetradellids) in Ibero-Armorica. Detailed analysis of each family shows additional similarities.
Ibero-Armorican tetradellids (text-fig. 25) consist of glossomorphitines (44% of tetradellid genera
and 55% species), tetradellines (30%; 22-5%), and sigmoopsines (13-5%; 15-5%), groups present in
similar relative proportions within the two other regions. Differences are mainly expressed by the
absence of (‘British’) gunnaropsines and (‘Baltic’) perspicillines and sylthines. The Ibero-Armorican
ctenonotellid fauna (text-fig. 26), in common with those in Baltoscandia and the British Isles,

ORDOVICIAN

% G % S

Ar Llv

% G % G

Llo Cc,

% G % G

Cc2 /4s

% G % G

□ □ PALAEOCOPA

□

BINODICOPA

text-fig. 22. Generic (and specific) percentage abundance to palaeocope of binodicope ostracodes for the
Ordovician and its series in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. The Caradoc is divided
into two, at the Dn/Dm boundary (see text-fig. 2). All data from text-figs. 10-15.

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

1X7

Ar

%G

Llv

% G

Llo

% G

Cc,

% G

AC ABC ABC ABC

| TETRADELLIDAE
I 0EPIKI1DAE

| | CTENONOTELU DAE DlSj TVAERENELLI DAE | | EUPRIMITI IDAE

I I SARVINIDAE V/$i HOLLI N I DAE PR 1 MI TI OPSACEA

RNl OEPI KELLIDAE

EURYCHI L I N I DAE

text-fig. 23. Generic (and specific) percentage abundance of palaeocope families for the Ordovician and its
series in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. The Caradoc is divided as in text-fig. 22. All

data from text-figs. 10 14.

( ,r") ORDOVICIAN

% S % G

ABC ABC

As

% G

□

C I RCULI N I DAE

AECHMI N I DAE

□

SPINIGERITIDAE

\ SUPERFAM . NOV ,

I QUADR I JUGATOR 1 DAE

SKS

DREPANELLI DAE

text-fig. 24. Generic (and specific) percentage abundance of binodicope families for the Ordovician and its
series in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. The Caradoc is divided as in text-fig. 22. All

data from text-fig. 15.

contains a high proportion of wehrliines and tallinnellines (both 33% of total ctenonotellid genera).
In contrast with the other regions, steusloffines are apparently unknown in Ibero-Armorica, while
ctenonotellines reach a comparatively high proportion (33% of ctenonotillid genera).

Composition of the binodicopes. Ordovician binodicopes of Ibero-Armorica are remarkably similar
in general composition to those of the British Isles and Baltoscandia (text-fig. 24). For example,
the relative importance of bolliids (the dominant binodicope group in Ibero-Armorica) is almost
exactly the same in the three regions (32%, 34%, 35%, respectively, of genera); this is also true for
aechminids (19%, 13%, 16%) and to a lesser extent for circulinids (26%, 40%, 22%). The only
significant faunal anomaly of Ibero-Armorican binodicopes is the presence of the Quadrijugatori-
idae, a superfamily otherwise known only from North America.

188

PALAEONTOLOGY, VOLUME 32

ORDOVICIAN

TETRADELL1NAE

SIGMOOPSINAE

GUNNAROPSINAE

SYLTHINAE

GLOSSOMORPHITINAE
SUBFAMILY NOV.
PERSP IC I LLI NAE
DI LOBELL I NAE

ORDOVICIAN

% S % G

ABC ABC

TALLI NNELLI NAE
WEHRLI I NAE

STEUSLOFFI INAE
CTENONOTELLINAE

text-fig. 25 (left). Generic (and specific) percentage abundance of tetradellid (hollinacean) subfamilies for
the Ordovician (Arenig to Ashgill) in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. All data from

text-figs. 13 and 14.

text-fig. 26 (right). Generic (and specific) percentage abundance of ctenonotellid (hollinacean) subfamilies
for the Ordovician (Arenig to Ashgill) in: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. All data from

text-figs. 11 and 12.

In summary, a comparison of the taxonomic composition of the Ordovician ostracod faunas of
Ibero-Armorica, Britain, and Baltoscandia shows:

1. A remarkable diversity at virtually all taxonomic levels of the faunas in all three areas.

2. Higher taxonomic diversity for the Baltoscandian faunas— mainly reflected in the diversity
of suborders and of palaeocope families. This feature undoubtedly partly reflects the greater
amount of taxonomic work undertaken on Ordovician ostracodes from the Baltic compared to
those from the other two areas.

3. Close similarities, between each domain, in the occurrence and relative proportions of all the
dominant taxonomic groups of palaeocopes (tetradellids, ctenonotellids) and binodicopes (bolliids,
circulinids, aechminids).

4. Differences in the palaeocope : binodicope ratio show a gradually increasing dominance of
palaeocopes from Ibero-Armorica to the British Isles and then to Baltoscandia.

5. Other, minor differences occur in the composition of the Tetradellidae and Ctenonotellidae.
At the subfamily level each area may display unique characteristics, such as the occurrence of
gunnaropsines in the British Isles and the absence of steuslofhnes in Ibero-Armorica.

GENERIC ABUNDANCE OF ORDOVICIAN PALAEOCOPES AND BINODICOPES IN
BALTOSCANDIA, BRITAIN, AND IBERO-ARMORICA

Changes in generic abundances of palaeocopes and binodicopes, from Arenig to Ashgill, in the
three regions investigated are given in text-fig. 27.

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

189

text-fig. 27. Total numbers of genera, through the Ordovician, of palaeocope and binodicope ostracodes
and of selected palaeocope families (Tetradellidae, Ctenonotellidae, and Tvaerenellidae) from: a, Ibero-
Armorica; b, British Isles; c, Baltoscandia. Ar = Arenig; Llv = Llanvirn; Llo = Llandeilo; Cc = Caradoc;
As = Ashgill. All data from text-figs. 10-17. Four different scales are used to enable comparison of abundant

and minor groups.

Baltoscandia (text-fig. 27c)

Palaeocopes in general (for example, major families Tetradellidae, Ctenonotellidae), form a relatively
poor late Arenig-early Llanvirn fauna (maximum 23 genera), slightly decrease in abundance during
the Llanvirn and are marked by a steep increase in numbers of genera (from 25 to 50) during the
Llandeilo. Maximum numbers of palaeocope genera (about 60) occur during the early Caradoc;
this is followed by a gradual reduction into the Ashgill (about 30 genera). Interestingly, although
similar variations in numbers of genera are also observed in binodicopes up to the mid-Caradoc,
their maximum abundance is reached during the Ashgill (14 genera) and coincides with a marked
reduction in numbers of palaeocope genera.

British Isles (text-fig. 27b)

Generic abundance in the British Isles shows a similar pattern to that of Baltoscandia. Highest
generic abundance of palaeocopes (for example, tetradellids, ctenonotellids) and binodicopes occurs
during the early Caradoc. However, a later Caradoc reduction in numbers of genera in both groups
is more dramatic than that seen in Baltoscandia and, moreover, the British faunas show a more
marked increase during the late Llanvirn-Llandeilo. As for Baltoscandia, British late Caradoc
ostracode faunas are dominated by binodicope genera. Ashgill ostracode faunas from Britain are
still insufficiently documented to place significance on its numbers of genera plotted in text-fig. 27.

Ibero-Armorica (text-fig. 27a)

Variations in generic abundance of Ibero-Armorican palaeocopes and binodicopes differ slightly
from those of British and Baltic faunas by having a palaeocope (e.g. tetradellid and ctenonotellid)
maximum (lower-middle Llandeilo) and subsequent decrease (late Llandeilo) which is earlier, and
a binodicope pattern whose maximum (late Llandeilo-early Caradoc) does not coincide with that
of palaeocopes and which thereafter (during the later Caradoc) shows a more gradual decrease
and greater numbers of genera compared to palaeocopes.

190

PALAEONTOLOGY, VOLUME 32

VARIATION OF OSTRACODE FAUNAL COMPOSITION, THROUGH THE
ORDOVICIAN, IN B ALTOSCAN DIA, BRITAIN, AND I BE RO- A R M O R IC A

Baltoscandia

Faunal diversity at subordinal level increases notably between the lower and upper Ordovician:
the Arenig, Llandeilo, Caradoc, and Ashgill having six, seven, eleven, and twelve suborders
respectively (text-fig. 19).

During the Arenig to Llandeilo, palaeocopes predominate, comprising 73%, 68%, and 75% of
the Arenig, Llanvirn, and Llandeilo ostracode genera, respectively. Other higher taxa, such as
binodicopes (e.g. 12% in the Llanvirn), metacopes (e.g. 7% in the Arenig), or leiocopes (e.g. 6% in
the Llanvirn), are minor faunal components. No significant changes occur during this period (text-
fig. 19).

During the Caradoc to Ashgill, newly introduced ostracode groups or groups previously forming
minor faunal components become relatively more important; they include metacopes (4% of lower
Caradoc genera; 6% in upper Caradoc; 12% in Ashgill), binodicopes (13% in lower Caradoc; 18%
in upper Caradoc; 21% in Ashgill), and cytherelliformes (7% in lower Caradoc). Palaeocopes
consequently show a marked drop in relative importance from the lower Caradoc through to the
Ashgill, where they comprise some 40% of ostracode genera. A similar, though less important,
trend also affects leiocopes, which form 5% of genera in the Llandeilo and 1% in the Ashgill.
Cypridocopes, parapodocopes, cytherocopes, punciocopes, and leperditiocopes seem to appear first
during the early Caradoc.

To reinforce these points, the generic percentage abundance (of the total ostracode fauna) of
the four major suborders (Palaeocopa, Binodicopa, Metacopa, Leiocopa) known from throughout
the Ordovician of Baltoscandia are given in text-fig. 20. It is again seen that the late Llandeilo-
early Caradoc is a period of significant faunal changes, when both palaeocopes and leiocopes begin
to decrease, and binodicopes and metacopes start to increase in relative abundance (see also text-
fig. 21).

Palaeocopes dominate binodicopes throughout the Ordovician of Baltoscandia, though the
palaeocope : binodicope generic ratio (see text-fig. 22) decreases gradually from about 11-5:1 in
the Arenig to 2 : 1 for the Ashgill. Ratios for the Llanvirn, Llandeilo, and upper and lower Caradoc
are 6, 8-5, 5-5, and 3-5: 1 respectively.

Diversity of palaeocope families (text-fig. 23) reaches a maximum during the Caradoc and
Ashgill (ten families), thus contrasting with the Arenig-Llandeilo interval (five to seven families).
The late Llandeilo-early Caradoc interval is characterized by the appearance of new palaeocope
taxa such as Oepikellidae, Sarvinidae, Hollinidae, and Primitiopsacea. Text-figs. 23 and 28c clearly
show:

a. The relatively stable importance of tvaerenellid and tetradellid palaeocopes throughout the
Ordovician, comprising 18-25% and 31-37% respectively, of the total palaeocope genera.

b. A gradual decrease of the relative abundance of ctenonotellids from the Arenig to the Ashgill
(25% and 5% respectively, of the total palaeocope genera).

c. In contrast to euprimitiids, a maximum relative abundance of eurychilinid genera during the
Llandeilo.

Binodicope higher taxa diversity increases reasonably uniformly from lower to upper Ordovician
(text-fig. 24), four families occurring during the Llanvirn and six in the Ashgill. The Llandeilo to
early Caradoc again appears as a period of significant faunal change (text-fig. 28c). For example,
the percentage of circulinid genera (of the total binodicope fauna) drops from a 40% maximum
during the late Llandeilo-early Caradoc to 8% in the Ashgill. Similar variations are observed
for the Spinigeritidae (text-fig. 28c, graph 5). The zenith of the circulinids apparently coincides
with a period of reduction of bolliids (minimum in the Llandeilo) and of aechminids (early
Caradoc).

VANNIER ET AL .: ORDOVICIAN OSTRACODE FAUNAS

191

text-fig. 28. Generic percentage abundance of
palaeocopes and binodicopes and of some of their
important families through the Ordovician from: a,
Ibero-Armorica; b, British Isles; c, Baltoscandia.

I Percentage of binodicope genera of the total
palaeocope binodicope fauna.

2, 3, 4, 5. Percentage of bolliid, circulinid, aechmi-
nid, and spinigeritid genera, respectively, of the total
binodicope fauna.

6. Percentage of palaeocope genera of the total
palaeocope binodicope fauna.

7,8,9. 10. Percentage of tetradellid, ctenonotellid,
tvaerenellid, and eurychilinid genera, respectively, of
the total palaeocope fauna.

Ar = Arenig; Llv = Llanvirn; Llo = Llandeilo;
Cc = Caradoc; As = Ashgill.

7 BINODICOPA

2 BOLL I I DAE

3 C I RCULI N I DAE

4 AECHMINIDAE

5 SPINIGERITIDAE

6 PALAEOCOPA

7 TETRADELLIDAE

8 CTENONOTELLIDAE

9 TVAERENELLIDAE
10 EURYCHILINIDAE

British Isles

There is a lack of data from the Arenig, where ostracodes are rare, and the Ashgill data are
unpublished. Comparisons will thus be restricted to Llanvirn late Caradoc faunas and will consider
only the (well-documented) palaeocopes and binodicopes.

Palaeocopes are most dominant during the Llandeilo (palaeocope : binodicope generic ratio =
1-6 : 1 ) and early Caradoc (1 -4 : 1 ), but show a substantial reduction in the upper Caradoc (0-25 ; 1 ),
when binodicopes become the most important ostracode group (text-fig. 22). A similar, though
less marked, trend was noted for Baltoscandia.

Analysis of important groups of palaeocopes and binodicopes (text-figs. 23, 24, 28b) show
diametrically opposite trends during the Llandeilo-early Caradoc. This interval witnesses a

192

PALAEONTOLOGY, VOLUME 32

maximum development of ctenonotellids (40% of Llandeilo palaeocopes), circulinids (45% of
Llandeilo binodicopes), and aechminids (26% of early Caradoc binodicopes) and coincides with
an important reduction of tetradellids (55% of Llandeilo palaeocopes), bolliids, and spinigeritids
(18% and 8% respectively, of early Caradoc binodicopes).

The most important differences in trends compared to Baltoscandia (text-fig. 28) affect the
palaeocope faunas, especially tetradellids, whose significance to palaeocope faunas is much more
constant in Baltoscandia than in Britain. The reduced importance of British ctenonotellids from
Llandeilo to late Caradoc contrasts with their gradual decrease throughout the Ordovician of
Baltoscandia. On the other hand, the two major binodicope groups in both Baltoscandia and the
British Isles, circulinids and bolliids, have comparable percentage values (in relation to their
respective binodicope faunas) and trends throughout the Ordovician (text-fig. 28).

Ibero- Armorica

The palaeocope : binodicope generic ratio through the Ordovician in Ibero-Armorica (text-fig. 22)
indicates two major phases: a gradual increase in relative abundance of palaeocopes from the
Arenig (0-3:1) to a Llandeilo maximum (1:1), followed by a marked palaeocope reduction
beginning in the Caradoc (0-2: 1). The same general trend was noted for both Baltoscandia and
the British Isles. The trend in Ibero-Armorica is distinctive in that the faunal changes seem to
occur earlier in the Caradoc than in the British Isles and is much more abruptly expressed than
in Baltoscandia.

The only really significant trend observed within the palaeocope fauna of Ibero-Armorica is
provided by the tetradellids, whose parabolic graph is exactly the same as for the British tetradellids,
with minimum relative generic abundance in the Llandeilo (text-fig. 28).

As with the palaeocopes, the most diversified binodicope faunas in Ibero-Armorica are known
during the Llandeilo and early Caradoc (text-fig. 24), a period marked by an increase in aechminids
(28% of Llandeilo binodicope genera) and spinigeritids (10% of early Caradoc binodicope genera)
as well as by the presence of circulinids (28% and 10% of Llandeilo and early Caradoc binodicope
genera, respectively) and bolliids (14% and 30%, respectively). Apparently, circulinids are not
represented within the late Caradoc faunas.

Ratios of bolliid, circulinid, and aechminid genera with respect to the total binodicope fauna
of Ibero-Armorica are comparable to those for the British Isles and Baltoscandia. Thus, the
bolliid : binodicope ratio reaches a minimum almost simultaneously in all three regions, during the
late Llandeilo-early Caradoc. However, the maximum value of the circulinid : binodicope ratio
occurs during the Llanvirn in Ibero-Armorica compared to the Llandeilo-Caradoc for the British
Isles and the early Caradoc for Baltoscandia.

In summary, variations in generic abundance and faunal composition of Baltic, British, and
Ibero-Armorican ostracode faunas show:

1. General common trends throughout the Ordovician:

a. An increasing importance (% of genera) of palaeocopes over binodicopes during the Arenig-
Llanvirn to a Llandeilo peak followed by a general decrease (marked in Ibero-Armorica and
the British Isles) of palaeocopes during the Caradoc.

b. Except for Baltic binodicopes, palaeocopes and binodicopes reach maximum generic diver-
sity during the Llandeilo-early Caradoc, a trend confirmed at a higher taxonomic level
with a particularly steep increase in numbers of suborders recorded in Baltoscandia in the
Llandeilo.

c. Comparable variations in the relative percentage of major binodicope families (Bolliidae and
Circulinidae), and a peak of binodicope generic abundance during the early Caradoc.

2. Each domain has its own faunal characteristics, which include:

a. Persistent high diversity of families and genera of palaeocopes and binodicopes in Baltoscandia
up to the Ashgill.

b. Similar parabolic graph trends of British and Ibero-Armorican tetradellids through the

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

193

Ordovician (Llandeilo minimum), thus contrasting with only minor variations in generic
abundance of the group in Baltoscandia.

c. A uniform decline of Baltoscandian ctenonotellids, a trend not observed in the British Isles
or Ibero- Armorica.

d. A marked reduction in generic abundance of palaeocopes in Britain and Ibero-Armorica
during the Caradoc, compared to a more gradual decrease in Baltoscandia.

3. Faunal events suggested by variations of generic abundance of ostracode faunas seem to be
diachronous in the three domains. The maximum abundance of palaeocopes occurs during the
early Llandeilo, the late Llandeilo-early Caradoc and the early Caradoc in Ibero-Armorica, the
British Isles and Baltoscandia, respectively. Diachronism is similarly observed for the rise and
decline of this group.

OSTRACODE GENERIC LINKS BETWEEN IBERO-ARMORICA, THE BRITISH
ISLES, AND BALTOSCANDIA FOR THE ORDOVICIAN

The four palaeocopes Platybolbina ( Reticulobolbina ) spongiosoreticulata and Uhakiella magnified
(see Jones 1985, 1986), and Sylthere vonhachti (R. Orr, pers. comm.) and Quadritia (Krutatia)
iunior (see PI. 30, fig. 8), known from the Ashgill of the British Isles and Baltoscandia, represent
the species-level contacts between the Ordovician of the three regions in question. Numerous links
exist at generic and subgeneric level (text-figs. 10-17, 29-35). Text-fig. 29 names genera common
to pairs of areas (E, Ibero-Armorica/British Isles; F, Ibero-Armorica/Baltoscandia; G, British
Isles/Baltoscandia), and to all three regions (X). As for the data on faunal composition, comparisons
will essentially focus on palaeocopes and binodicopes, which represent about 85% of the total
common genera listed (text-fig. 29). The total number of genera of palaeocopes and binodicopes
recorded herein from the Ordovician of each area is 12 and 16, respectively, in Ibero-Armorica;
27 and 15 in the British Isles; and 104 and 34 in Baltoscandia (text-fig. 30).

Generic affinities between Ibero-Armorica and the British Isles

These two regions have fifteen ostracode genera in common in the Ordovician (text-fig. 29): one
metacope ( Medianella ), and seven palaeocopes and seven binodicopes belonging to several

E

F

G

X

CD 1

1

Hastatellina

Ceratopsis

Gracquina

Bichilina

Duringia

Homeokiesowia

Tallinnella

Piretopsis

Schallreuteria

Sigmoopsis

Gotula

Henningsmoenia

Platybolbina

Uhakiella

Hippula

Ogmoopsis

Vittella

Euprimites

Quadritia

CD

2

Copelandia

Satiellina

Aechmina

Eocytherella

Kinnekullea

Byrsolopsina

Bullaeferum

Crescentilla

Easchmidtella

Pedomphalella

Conchoprimitiella

Spinigerites

Klimphores

Laterophores

Gen . nov . A

Pseudulrichia

Vogdesella

Pari conchoprimitia

CO

Miehlkella

Brevidorsa

Pachydomelloides

Conchoprimitia

Primitiella

Ordovizona

Medianella

text-fig. 29. Ordovician (Arenig-Ashgill) genera of Palaeocopa (I), Binodicopa (2), and other groups (3)
occurring in two or more of the following regions: A, Ibero-Armorica; B, British Isles; C, Baltoscandia. E
= genera common between A and B only; F = between A and C only; G = between B and C only; X =
between all three regions (A, B, C). In further text-figs. A, B, C, E, F, G, and X will indicate corresponding

number or percentages of genera.

194

PALAEONTOLOGY, VOLUME 32

A

B

C

E

F

G

X

E+X

F+X

G+X

E+x -ni

A+B- (E+X)

F+x -n-

A+C- (F+X)

G+X -R3

B+C- (G+X)

PAL

12

27

104

3

i

1 1

4

7

5

15

0.22

0.04

0.13

BIN

16

15

34

1

5'

6

6

7

1 1

12

0.29

0.28

0.44

PAL+BIN

28

42

138

4

6

17

10

1 4

1 6

27

0.25

O.IO

0.19

text-fig. 30. Similarity indices lor the Ordovician palaeocope (PAL), binodicope (BIN), and combined
palaeocope-binodicope faunas variously between Ibero-Armorica, British Isles, and Baltoscandia. A, B, C:
total number of genera known from Ibero-Armorica, British Isles, and Baltoscandia, respectively. E, F, G,
X: number of genera common between Ibero-Armorica/British Isles, Ibero-Armorica/Baltoscandia, British
Isles/Baltoscandia, and between all three areas, respectively. Rl, R2, R3: similarity indices for Ibero-
Armorica/British Isles, Ibero-Armorica/Baltoscandia, and British Isles/Baltoscandia, respectively.

families including the Tvaerenellidae, Ctenonotellidae, Tetradellidae, Bolliidae, Aechminidae, and
Circulinidae.

In order to illustrate the relative importance of cosmopolitan palaeocope and binodicope genera,
both graphical representation of results (e.g. text-figs. 29, 31, 32) and a similarity index
(text-fig. 30) are used. The similarity index is defined as the ratio between the total number of
genera in common between two areas (e.g. Ibero-Armorica and the British Isles = E + X: see
text-figs. 29 and 30) and the total ‘endemic' fauna (in this case A+ B— (E + X)) from the two areas
in question. The similarity indices (Rl) for Ibero-Armorican and southern British Ordovician
palaeocopes, binodicopes, and combined palaeocope-binodicope faunas are 0-22, 0-29, and 0-25,
respectively.

Generic affinities between Ibero-Armorica and Baltoscandia

These two regions have a high number of binodicope (11) compared to palaeocope (5) genera in
common (text-figs. 29, 31, 32). The binodicopes include bolliids (3), aechminids (3), circulinids (2),
and spinigeritids (1). The palaeocopes are tvaerenellids (2), tetradellids (2), and ctenonotellids (1).
The relevant similarity index (R2: text-fig. 30) is very low for palaeocopes (0-04) and very high for
binodicopes (0-28). Although similarity indices Rl (Ibero-Armorica/British Isles) and R2 reach
almost identical figures for binodicopes (0-29 and 0-28, respectively), they differ markedly for
palaeocopes (0-22 and 0 04, respectively). Metacope genera (2), leiocopes (2), eridostracans (1),
and cytherelliformes (1) are also common to Ibero-Armorica and Baltoscandia.

Generic affinities between the British Isles and Baltoscandia

A list of twenty-nine ostracode genera in common (text-fig. 29) indicates close ostracode faunal
affinities between the British Isles and Baltoscandia in the Ordovician. Fifteen palaeocope and
twelve binodicope genera occur in common (text-figs. 31 and 32), consisting of Ctenonotellidae
(6), Tetradellidae (4), Oepikellidae (2), Tvaerenellidae (2), Oepikiidae ( 1 ), Circulinidae (4), Bolliidae
(3), Aechminidae (3), and Spinigeritidae (2). Metacope (1) and punciocope (1) genera are also
common to both regions. The British Isles/Baltoscandia similarity index, R3 (text-fig. 30),
consequently reaches high values, especially for binodicopes (0-44). Rl and R2 for binodicopes
are only 0-29 and 0-28, respectively. Nevertheless, despite a high number of palaeocope genera in
common, the R3 palaeocope value (013) is less than the equivalent palaeocope similarly index for
the British Isles and Ibero-Armorica (= 0-22; Rl, text-fig. 30).

Ostracode genera common between the three regions

A substantial proportion (25%) of the genera in common between the regions in fact occur within
all three areas (text-fig. 29); they comprise binodicopes (six genera), palaeocopes (four), and

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

195

PALAEOCOPES

BINODICOPES

o

7

12

II

text-fig. 31. Number of Ordovician (Arenig to Ashgill) palaeocope, binodicope, and palaeocope + binodicope
genera common between Ibero-Armorica, British Isles, and Baltoscandia. The size of each area is proportional
to the number of genera. Number of genera in common are indicated within the shaded areas.

metacopes (one). Moreover, apart from occurring in Baltoscandia, Ibero-Armorica, and the
British Isles, Quadritia and Ogmoopsis are also known from Czechoslovakia (unpublished data).
Interestingly, genera common to the three regions are mainly binodicopes (text-figs. 29 and 32).
This characteristic is also observed in the higher binodicope similarity index for each of the three
faunal comparisons made compared to the equivalent index for palaeocopes (Rl, R2, R3: text-
fig. 30). This cosmopolitan character amongst the binodicopes recorded here can be extended
geographically: for example, Laterophores, Pseudulrichia , Vogdesella , and Pariconchoprimitia are
also present in the Ordovician of North America (see Schallreuter 1980«; Schallreuter and Siveter
1985; Jones 1986, 1987). Klimphores and Pariconchoprimitia also apparently occur in the Ordovician
of Saudia Arabia (Vannier and Vaslet 1987).

VARIATIONS IN OSTRACODE GENERIC LINKS BETWEEN IBERO-ARMORICA,
THE BRITISH ISLES, AND BALTOSCANDIA THROUGH THE ORDOVICIAN

How ostracode faunal similarities between the three areas may have changed through the Ordovician
(text-figs. 33-35) is clearly important because of its possible palaeogeographical implications.

196

PALAEONTOLOGY, VOLUME 32

text-fig. 32. Number of Ordovician (Arenig to Ashgill) palaeocope, binodicope, and palaeocope + binodicope
genera common between Ibero-Armorica, British Isles, and Baltoscandia. The size of each area is proportional
to the number of genera. Number of genera in common are indicated within the shaded areas (E, F, G, X,

as defined in text-fig. 29).

Ibero-Armorican British relationships (text-fig. 35, graphs 1 and 4)

It is during the Llanvirn that British and Ibero-Armorican ostracode faunas show closest affinity,
as noted by the relatively high percentage of binodicope, and particularly palaeocope, genera in
common. Palaeocopes and binodicopes both show a drop to a minimum percentage of genera in
common between Ibero-Armorica and Britain during the Llandeilo, and thereafter show a recovery
of the numbers of genera in common into the early Caradoc. The later Caradoc is characterized
by very low percentages of binodicope, and particularly of palaeocope, genera in common between
the two regions.

Ibero-Armorican- Baltoscandian relationships (text-fig. 35, graph 2)

In contrast to the uniform palaeocope and binodicope links between Ibero-Armorica and Britain,
the palaeocopes and binodicopes common to Ibero-Armorica and Baltoscandia clearly exhibit
mutually opposite trends. The uniform Llanvirn-Caradoc percentage increase of binodicope genera
in common contrasts markedly with an irreversible decrease of palaeocopes in common during the
Llandeilo-Caradoc interval.

British- Baltoscandian relationships (text-fig. 35, graph 3)

The evolving pattern of ostracode links between these two regions shows: first, very similar
variations in the percentage of genera in common for both palaeocope and binodicope faunas,
from the Llanvirn through to the Ashgill. Secondly, ostracode faunal contact steadily enhances
for both groups during this period, the only difference being a slightly lower increase in the
percentage of cosmopolitan binodicope genera during the Llandeilo. Even though faunal data
from the Ashgill of the British Isles are still incompletely known or published, it is important to
note that all documented British Ashgill palaeocope genera also occur in contemporaneous
Baltoscandian sediments. These striking late Ordovician faunal affinities also exist at the subgeneric
level and, to some extent (see above), even at the specific level.

Ibero-Armorican- British- Baltoscandian relationships (text-fig. 35, graphs 5 and 6)

The percentages of binodicopes and palaeocopes which are common to all three areas at given
times through the Ordovician are clearly very different. No subdivision of the Ordovician (as

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

197

IBERO-ARMORICA /
BRITISH ISLES/
BALTOSCANDIA

PALAEOCOPES BINODICOPES

©

text-fig. 33. Percentage change in geographical affinity, through the Ordovician, of cosmopolitan Ibero-
Armorican palaeocope, binodicope, and palaeocope + binodicope genera, respectively. Percentage of Ibcro-
Armorican genera common with: the British Isles; Baltoscandia; and both the British Isles and Baltoscandia.
1, Llanvirn to Caradoc inclusive; 2, Llanvirn; 3, Llandeilo; 4, 5, Caradoc Cel and Cc2, respectively (as
defined in text-fig. 22). The shaded area in each case encompasses the plots for the time periods 2-5 inclusive.

BRITISH ISLES /
IBERO-ARMORICA /
BALTOSCANDIA

o

text-fig. 34. Percentage change in geographical affinity, through the Ordovician, of cosmopolitan British
palaeocope, binodicope, and palaeocope + binodicope genera, respectively. Percentage of British genera
common with: I bero- Armorica; Baltoscandia; and both Ibero-Armorica and Baltoscandia. 1, Llanvirn to
Ashgill inclusive; 2, Llanvirn; 3, Llandeilo; 4, 5, Caradoc Cel and Cc2, respectively (as defined in text-fig.

22). The shaded area in each case encompasses the plots for the time periods 2-5 inclusive.

defined in text-fig. 35) contains palaeocope genera common to the three regions. By contrast,
binodicope generic similarities between the three areas exist in each Ordovician time period (text-
fig. 35) and, moreover, gradually increase overall through the Ordovician.

Affinities of Ibero- Armor ican faunas (text-fig. 33)

The affinities of cosmopolitan Ordovician palaeocope and binodicope genera of Ibero-Armorica
are given in a triangular diagram (text-fig. 33). Taking the data for palaeocopes and binodicopes
together, the cluster points suggest that, in general through the Ordovician, cosmopolitan Ibero-

198

PALAEONTOLOGY, VOLUME 32

text-fig. 35. The percentage of Ibero-Armorican,
British, or Baltoscandian palaeocope or binodi-
cope genera, respectively, common to one or both
of the other areas for each of the given time
periods of the Ordovician.

1, 2, 5. Percentage of Ibero-Armorican genera
common to: the British Isles only (graph 1);
Baltoscandia only (2); and to both the British Isles
and Baltoscandia (5).

3, 4, 6. Percentage of British genera common
to: Baltoscandia only (graph 3); Ibero-Armorica
only (4); and to both Ibero-Armorica and Balto-
scandia (6).

Armorican Ordovician ostracode faunas have a Baltic character and a relatively poorly marked
British influence. However, this signature is less sharp in the early Caradoc (text-fig. 33). Differences
between the geographical affinity of cosmopolitan Ibero-Armorican palaeocopes and binodicopes
are illustrated by the position of the shaded areas within their respective triangular diagrams; the
binodicopes are somewhat more widespread in their distribution.

Affinities of British faunas (text-fig. 34)

The British cosmopolitan palaeocope and binodicope ostracodes together show mainly Baltic and
only minor Ibero-Armorican influences throughout the Ordovician. Again, the cosmopolitan
palaeocopes and binodicopes differ by the same characteristic as previously noted for Ibero-
Armorican faunas (text-fig. 33), binodicopes being somewhat more cosmopolitan in their distribution
(compare respective shaded areas in text-fig. 34).

PALAEOECOLOGICAL IMPLICATIONS OF THE OSTRACODE DISTRIBUTIONAL

PATTERNS

The general composition, diversity, distributional patterns, and faunal dynamics of Baltoscandian,
southern British, and Ibero-Armorican Ordovician ostracodes can be broadly related to palaeoeco-
logical and palaeogeographical factors. In particular, some of their major compositional and
distributional characteristics may be interpreted in the light of two palaeobiological models.
According to the ecological model of Bretsky and Lorenz (1970), species diversity within
communities is correlatable with environmental conditions, especially environmental stability. They
see faunal diversity as a result of speciation processes which probably occur, in situ , preferentially
more often in stable, offshore areas, a notion Eldredge (1974) considered valid and applicable to
benthic Palaeozoic faunas. Fortey (1984) stressed that major physical processes such as eustatic
sea-level changes are partly the trigger for important faunal events, such as speciation and diversity

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

199

shifts, observed in the Ordovician. Both these models can be tested using the patterns of Ordovician
ostracode distribution in our study areas.

1 . The influence of environmental conditions

The Bretsky and Lorenz model (1970; Eldredge 1974) recognizes two types of faunal communities
and related environmental conditions. Communities in unstable environments (and low spatial
heterogeneity) are expected to show low faunal diversities; large numbers of individuals per specific
population; significant morphological variations within species; low rates of evolution and hence
persistence of stable community structures; and are less severely affected by sudden environmental
change. Communities of stable environments (and high spatial heterogeneity) are expected to be
characterized by high faunal diversities; relatively small numbers of individuals per species
population; limited morphological variation; high rates of evolution and hence frequent faunal
reorganizations and changes in community structure; and are normally adversely affected by sudden
environmental changes and are thus subject to widespread extinctions. Unstable environments are
considered to be basically near-shore areas and stable environments are more off-shore areas
(Bretsky and Lorenz 1970).

Within our study areas, two main types of Ordovician ostracode faunas are recognized, based
on the relative abundance of palaeocope and binodicope genera (text-fig. 36). Faunas characterized
by a low palaeocope : binodicope ratio (<1) seem to be related to predominantly fine-gramed
detrital environments (e.g. Ibero-Armorica), whereas palaeocope-dominated faunas (palaeo-
cope : binodicope ratio > 1 ) appear to be related to carbonate-dominated environments (e.g.
Baltoscandia). This bifaunal distinction is supported by the following evidence:

a. Binodicopes are the dominant ostracode group throughout the Ordovician within Ibero-
Armorica (text-fig. 22) where their associated sediments are essentially fine-grained elastics
(siltstones and mudstones) with intercalations of sandstones; there is an absence of limestone
deposition from Arenig to Caradoc inclusive (text-fig. 3). In the Ordovician of the British
Isles, where both clastic and carbonate facies occur, the palaeocope : binodicope ratio is
substantially higher ( > 1) than in Ibero-Armorica for each major interval of Ordovician time.
The highest palaeocope : binodicope values are reached in the palaeocope-rich faunas in
Baltoscandia, in association with carbonate-dominant sedimentation.

h. In a well-documented, rare example of a Baltic Ordovician ostracode fauna from detrital
sediments, from the middle Ordovician Sularp Shale of Scania, Sweden (Schallreuter
1980u), binodicopes are relatively abundant (palaeocope : binodicope ratio = 0-4) and its two
commonest species are binodicopes. Comparisons with Scandinavian faunas from calcareous
sediments are particularly significant. For example, the middle Ordovician Upper Dalby
Limestone (Schallreuter 19846) is clearly dominated by palaeocopes (palaeocope : binodocope
ratio = 3).

c. Similar observations are made in the British Isles, from the Ordovician of the Welsh
Borderland. Although incomplete, the Ordovician (Caradoc) succession in the Welshpool
District (Pen-y-Garnedd Shale and Phosphorite: see Williams el al. 1972) consists exclusively
of shales in which the only ostracodes recorded are binodicopes (Jones 1986, 1987).

d. Ostracode assemblages from the middle Ordovician of Saudi Arabia (Vannier and Vaslet
1987) also seem to confirm a close relationship between elastics (in this case siltstones) and
binodicopes, with palaeocopes apparently being absent.

The two, broad and taxonomically based types of ostracode generic faunas correlated above
with carbonate and detrital-dominated environments, respectively, have many of the distinctive
characteristics (text-fig. 36) of the two types of faunal communities distinguished by Bretsky and
Lorenz (1970):

a. Stability of environmental conditions. During the middle Ordovician the Baltoscandian shelf
environment consisted of a wide carbonate platform (Jaanusson 1976) bordering the Baltic

200

PALAEONTOLOGY, VOLUME 32

text-fig. 36. Schematic representation of the two major types of ostracode
assemblages recognized in the regions studied, as related to their character-
istic environmental conditions and exemplified by Ibero-Armorican and
Baltoscandian faunas respectively. The arrows around the circle indicate
rare (Ibero-Armorica) and frequent (Baltoscandia) faunal reorganization.

continent and extending to both near-shore marginal and deeper water clastic facies. Ostracode
faunas, dominated by palaeocopes, occupied a wide range of platform habitats across a bathymetric
gradient (Jaanusson 1976). Binodicope-rich faunas seem to be restricted to more detrital-influenced
and unstable environments.

Middle Ordovician Ibero-Armorican environments are placed within the context of a widespread
platform bordering the Gondwanan continent, with dominant silts and coarser sedimentation.
Throughout the middle Ordovician of the Armorican Massif, tempestites (Guillocheau 1983)
indicate persistent unstable, storm-related conditions across the bathymetric profile. Distribution
of the Llandeilo ostracode fauna in the Armorican Massif seems to be depth-related (Vannier
19866).

b. Diversity . Considering the Ordovician as a whole, the lowest faunal diversity occurs in Ibero-

VANNIER ET AL.: ORDOVICIAN OSTRACODE FAUNAS

20!

Armorica. Even during the Llandeilo and early Caradoc, when Ibero-Armorican ostracode
faunas seem to be particularly prolific, diversity at all taxonomic levels is much lower than
that of equivalent Baltic faunas and is apparently also slightly lower than that of British
faunas. The fact that during the Ordovician the Baltic was always nearest to, and Ibero-
Armorica always the furthest from, the equator is also an important possible reason to
account for the observed levels of diversity.

c. Number of individuals per population. Assemblages poor in numbers of species and genera,
but very rich in individuals, are most frequently observed throughout all the detrital successions
of Ibero-Armorica. Typical examples are the Llandeilo siltstones and mudstones of the
Armorican medio-syncline (Vannier 1986a, b), or the Armorican Caradoc assemblages
dominated by a few species of the palaeocope Hastatellina and the binodicope Satiellina.
Similar observations have been made about the Baltic ostracode faunas from detrital horizons.
For example, the siliceous Sularp Shales of Sweden (see above: Schallreuter 1980a) contain
a relatively small number of species in which the only two abundantly represented belong to
the binodicope genera Pariconchoprimitia and Spinigerites. This contrasts markedly with the
highly diversified ostracode assemblages of most Baltic carbonate horizons.

d. Rates of speciation. Evolutionary trends amongst the ostracode species of the areas studied
are still too poorly documented to attempt a comparative evaluation of the rates of speciation.
Nevertheless, detailed information on the stratigraphical range and specific diversity of genera
(text-figs. 10-17) would suggest that the rate of speciation amongst Baltic ostracode faunas
during the Ordovician was relatively high compared to those of Ibero-Armorica or the British
Isles. The large number of species within Baltic genera such as Sigmobolbina, Tetrada ,
Tetradella, Laccochilina , Bolbina, or the subgenus Hippula (Hippula) (text-figs. 10 14)
apparently has no equivalent amongst British (if we exclude Ashgill faunas: Jones 1986, 1987,
Orr (unpublished)) or Ibero-Armorican faunas.

e. Persistence of community structure. Some evidence for the nature of this factor can be deduced,
at least for the middle Ordovician of Baltoscandia where thorough documentation of its
palaeocope-dominant ostracode faunas has recognized high faunal turnover (Jaanusson 1976)
and associated frequent rearrangement of species communities. In these stable environments
the persistence of a stable community structure is apparently low.

/. Spatial heterogeneity. Though it is difficult to assess the degree of spatial heterogeneity and
niche-partitioning in the fossil record, the occurrence of high diversity, at all taxonomic levels,
of a wide range of macrofossil (e.g. trilobites, brachiopods, bivalves, echinoderms, bryozoans,
etc.) and microfossil groups suggest that both factors had relatively high values on the
carbonate platform of Baltoscandia. By contrast, the relatively poor benthic faunas of
Ibero-Armorica (e.g. trilobites: see Henry 1980) would indicate that the reverse was the
case there.

In summary, in the middle Ordovician we have a diverse, frequently reorganized, palaeocope-
dominant fauna in relatively stable, carbonate-rich environments in Baltoscandia and a low
diversity but relatively longer-lived series of binodicope-rich faunal communities adjusted to often
unstable conditions in Ibero-Armorica. In Britain, both aspects of this bimodal pattern can be
expected to be represented.

2. The influence of sea-level changes

As emphasized by many authors (see Fortey 1984), major sea-level variations during the Ordovician
may have induced faunal and diversity changes within various fossil groups such as trilobites
(Shaw and Fortey 1977), graptolites, condonts, and shelly faunas (Jaanusson and Bergstrom 1980)
Thus, ‘faunal changes which occur at (Ordovician) Series boundaries are as much a product of
environmental shift as of evolutionary novelty’ (Fortey 1984, p. 39). Fortey has predicted that the
biological effects of a eustatic transgressive event will, depending on local circumstances, encompass:
the promotion of high speciation rates and diversity in shelf areas (because of spatial heterogeneity

202

PALAEONTOLOGY, VOLUME 32

and the species area effect: Eldredge 1974; Ludvigsen 1982); high endemism in shelf areas; the
diachronous, shelfwards displacement of previously extra-cratonic facies and faunas and a possible
associated apparent breakdown of faunal provinciality; an on-shelf migration of tropical 'mound'
faunas and a relative scarcity of island faunas. The faunal signatures from a regressive pulse are
predicted (Fortey 1984) to be: the occurrence on interior cratonic sites of shallow water deposits
whose fossil content is poor and/or which shows taxonomic jumps; a peripheral site location for
‘ancestors’ of later on-shelf faunas; a higher incidence of both island faunas themselves and of the
shelf edge siting of tropical mound faunas and their associated debris slides.

Analysis of the ostracode faunas and facies of Ibero-Armorica, Baltoscandia, and the British
Isles for a time interval of one of the major eustatic cycles recognized for the Ordovician, the
Llandeilo to Ashgill, reveals features consistent with Fortey’s model:

a. The ostracode faunal events recorded from all three areas during the Flandeilo-early Caradoc,
a time of major transgression (Fortey 1984, text-figs. 3-5), include a steep increase in the
generic diversity of both binodicopes and palaeocopes, culminating in the maximum
Ordovician generic diversity for the palaeocopes (text-fig. 27a-c).

By contrast, the later Caradoc is a period of regression (Fortey 1984), an event which
continued into the Ashgill with its well-documented glacial episode. Fate Caradoc times are
marked by the decline of binodicope (except for Baltoscandia) and palaeocope diversity in
all three areas (text-fig. 27a-c).

b. In general, the percentage of ostracode genera in common between the British Isles and Ibero-
Armorica decreases through the early Ordovician, but this trend is somewhat (and temporarily)
reversed for both palaeocopes and binodicopes for the late Llandeilo-early Caradoc (text-
fig. 35, graphs 1 and 4). The latter time interval is a period of transgression and, irrespective
of plate motion, separated faunas might be expected to show increased contact during such
events (Fortey 1984).

c. In the Dalby Formation of Vastergotland, Sweden (text-figs. 1 and 2), a significant change
in the ostracode fauna is associated with a change from calcilutites to dark mudstones
(Jaanusson 1976, p. 323). The appearance of new ostracodes, known previously only from
the deeper water areas of Scania, is the kind of local faunal shift associated with the on-shelf
migration of facies belts, in response to a transgressive pulse, as predicted by Fortey (1984).

EXPLANATION OF PLATE 24

Palaeocope and binodicope ostracodes from the Ordovician of Ibero-Armorica (left) and British Isles (right).

Figs. 1 -6. Palaeocopa. I , Ceratopsis sp. nov., upper part of Caradoc Series, Corral de Calatrava, Ciudad Real
district, Spain; heteromorphic right valve (IGR 32000), x 30. 2, Ceratopsis inflata Jones, 1986, upper part
of Llandeilo Series, Dryslwyn, Dyfed, Wales; tecnomorphic left valve (BM OS 12647), x 56 (Jones 1986,
pi. 11, fig. 13). 3, Hastatellina normandiensis Pribyl, 1975, Caradoc Series, Louredo Formation, near
Cacemes, Buyaco syncline, Portugal; heteromorphic right valve (IGR 30560/1A), x 28 (Vannier 19866, pi.
1, fig. 5). 4, Hastatellina ? sp., Llandeilo Series, Dryslwyn Castell, Dyfed, Wales; tecnomorphic left valve
(BM OS 12663), x 50 (Jones 1986, pi. 14, fig. 12). 5, Gracquina hispanica (Born, 1918), Llanvirn Series,
Alisedas, near Almaden, Ciudad Real district, Spain; heteromorphic right valve (SMF X/E 3716/1),
x 36 (Vannier 19866, pi. 4, fig. 1). 6, Gracquina vannieri Jones, 1986, Llanvirn Series, Huntingdon,

Cambridgeshire, England; heteromorphic right valve (GSM 8589), x 35 (Jones 1986, pi. 14, fig. II).

Figs. 7 10. Binodicopa. 7, Copelandia kerfornei Vannier, 1986a, Gres de Kermeur Formation, Caradoc Series,
Raguenez, near Crozon, Finistere, France; right valve (IGR 7052/2), x 35 (Vannier 1986a, pi. 5, fig. 3). 8,
Copelandia melmerbyensis Jones 1987, Woolstonian Stage, Caradoc Series, Melmerby, Cumbria, England;
right valve (GSM 2546B), x51 (Jones 1987, pi. 2, fig. 11). 9, Thibautina rorei Vannier, 19846, Le Pissot
Formation, Llanvirn Series, Domfront, Orne, France; right valve (IGR 5184C), x71 (Vannier 1986a,
pi. 11, fig. 4). 10, Conspicillum bipunctatum (Jones and Holl, 1869), middle part of the Llandeilo Series,

Builth, Powys, Wales; right valve (NMW 8416G), x 35 (Jones 1987, pi. 4, fig. 3).

All scanning electron micrographs of external lateral views. Figs. I, 3, 5-8, 10 are cast from external moulds.

PLATE 24

VANNIER et al Ordovician ostracodes

204

PALAEONTOLOGY, VOLUME 32

Sandstones

Siltstones

Mudstones

Carbonate

CZT

&

T ransgression

Regression

Palaeocope

Binodicope

text-fig. 37. Generalized interpretation of lithofacies and ostracode faunal changes in Ibero-Armorica and
Baltoscandia related to a transgressive (e.g. late Llandeilo-early Caradoc) and a regressive (e.g. late Caradoc-
Ashgill) cycle. The type of changes listed are those predicted in the model of Fortey (1984). The size of the
ostracodes illustrated is proportional to the diversity of the respective palaeocope- or binodicope-dominated

ostracode fauna.

PALAEOGEOGRAPHICAL SIGNIFICANCE OF THE OSTRACODE
DISTRIBUTIONAL PATTERNS

Cocks and Fortey (1982) have focused on the nature and reliability of employing various types of
faunal evidence for oceanic separation in the Lower Palaeozoic. Similarly, the composition, affinity,
and faunal dynamics of European Ordovician ostracode faunas can be used to test models of the
existence and demise of contemporaneous oceans (e.g. Schallreuter and Siveter 1985).

The typical Lower Palaeozoic ostracode species (e.g. a palaeocope) is benthic, has no known
pelagic larval stage, and a distributional pattern often encoding the developmental history of its
sedimentary basin (Siveter 1984; Schallreuter and Siveter 1985). The distribution of continents is
best defined using shallow water ‘endemics’, as deep water forms of benthic groups potentially

VANNIER ET AL.. ORDOVICIAN OSTRACODE FAUNAS

205

text-fig. 38. Ordovician palaeogeographic evolution of the ‘North Atlantic’ region. Arenig reconstruction
is modified after Cocks and Fortey (1982). Later Ordovician reconstructions are modified in part after
McKerrow and Cocks (1986) (for the Caradoc) and follow Pickering et al. (1988). The reconstructions also
concur with the data on ostracode faunal dynamics described herein.

have a wider dispersal capacity which precludes their effective use as tools in palaeogeographical
reconstruction (Cocks and Fortey 1982). Bathymetric/community analysis has been undertaken
and persuasively argued to account for Ordovician trilobite distributions (e.g. Fortey 1975; Fortey
and Owen 1978), but no equivalent detailed analyses have yet been undertaken of facies and any
associated ostracode communities, and lie outside the scope of the present study. Nevertheless,
virtually all the ostracode genera cited herein can be judged to belong broadly to shelf rather than
deep water environments and their use as general indicators of palaeogeography is considered
valid.

206

PALAEONTOLOGY, VOLUME 32

Faunal and facies distributions indicate the existence of two major early Ordovician continents
(Cocks and Fortey 1982) relevant to the present study (see text-fig. 38): Gondwana, which included
Ibero-Armorica; and Baltica, embracing the Baltic regions and adjacent parts of Scandinavia and
the Russian platform. Southern Britain is thought to have been part of a microcontinent, Avalonia
(e.g. McKerrow and Cocks 1986), which after the early Ordovician broke away from high latitude
Gondwana to drift northwards towards lower latitude Baltica and the North American continent
(Laurentia) which lay astride the equator on the other side of the Iapetus Ocean.

Schallreuter and Siveter (1985) have already addressed the nature and palaeogeographical
significance of ostracode faunal contracts across the Ordovician Iapetus Ocean between North
America and ‘Europe’. Our ostracode data presented herein (e.g. text-fig. 35) endorse the Ordovician
oceanic separation of Baltica, Avalonia, and Gondwana and the nature of the development of the
Rheic and Tornquist tracts (text-fig. 38). Particularly important is the fact that the ostracode faunal
dynamics indicate an Ordovician development for the Rheic Ocean.

1. Iapetus Ocean

An assessment of the ostracode faunal connections across the Iapetus Ocean (Schallreuter and
Siveter 1985), between Europe (southern Britain and Baltoscandia) and North America, dispelled
notions that Ordovician ostracodes showed strict endemicity (thirty middle-late Ordovician genera
are common to both sides) and concluded that by the late Ordovician the two plates may have
been in closer ‘effective’ proximity than previously supposed (e.g. McKerrow and Cocks (1976)
suggested a minimal 2000-3000 km width for Iapetus even by the late Ashgill). Indeed, Pickering
(1987) and Pickering et at. (1988; also Hutton 1987) argue that the available tectonic, stratigraphical,
palaeontological, palaeomagnetic, igneous, and sedimentological data from Newfoundland, the
British Isles, and Scandinavia suggest that by the late Ordovician-early Silurian the Iapetus Ocean
separating Laurentia from eastern Avalonia (southern Britain) and Baltica had closed, at least in
part, with the consumption of intervening oceanic crust, although marine seaways clearly persisted
until the late Silurian.

EXPLANATION OF PLATE 25

Palaeocope and binodicope ostracodes from the Ordovician of Ibero-Armorica {left) and Baltoscandia {right).

Figs. 1-4. Palaeocopa. 1, Bichilina sp., Llanvirn Series, Ligne, Loire-Atlantique, France; tecnomorphic (?)left
valve (IGR 5610), x 39. 2, Bichilina prima Sarv, 1959, middle Ordovician Backsteinkalk erratic boulder.
Northern Germany; heteromorphic left valve (GPIMH 2600), x 56 (Schallreuter 1983a, pi. 12, fig. 3). 3,
Euprimites sp., Llandeilo Series, Guichen, Ille-et-Vilaine, France; left valve (IGR 5611), x 65. 4, Euprimites
minor (Thorslund, 1940), middle Ordovician Backsteinkalk erratic boulder, Northern Germany; heteromor-
phic right valve (GPIMH 2628), x 47 (Schallreuter 1983a, pi. 12, fig. 6).

Figs. 5-10. Binodicopa. 5, Vogdesella ngakoi Yannier 1986a, Andouille Formation, Caradoc Series, Andouille,
Mayenne, France; right valve (holotype, IGR 30415/1), x 37 (Vannier 1986a, pi. 9, fig. 1). 6, Vogdesella
subovata (Thorslund, 1948), Sularp Shale, Caradoc Series, Gislovshammar, Scania, Sweden; right valve
(GPIMH 2291), x 37 (Schallreuter 1980a, pi. 5, fig. 4). 7, Satiellina henningsmoeni (Nion, 1972), Pont-de-
Caen Formation, Caradoc Series, Domfront, Orne, France; left valve (IGR 7086/1), x 40 (Vannier 1986a,
pi. 7, fig. 1). 8, Satiellina biloba (Troedsson, 1918), uppermost part of Ordovician, Rostanga, Scania,

Sweden; left valve (LO 2885), x 37 (Troedsson 1918, pi. 2, fig. 21). 9, Aechmina sp., Andouille Formation,
Caradoc Series, La Touche, Andouille, Mayenne, France; right valve (IGR 30502/1), x 66 (Vannier 1986a,
pi. 8, fig. 5). 10, Spinaechmina keitumensis Schallreuter, 1984a, Ojlemyrflint erratic boulder, upper part of
the Ordovician, Isle of Sylt, North Sea, Germany; right valve (holotype, GPIMH 2915), x 90 (Schallreuter
1984a, pi. 1, fig. 5).

All external lateral views; all scanning electron micrographs except fig. 8. Figs. 1, 3, 5, 7, 9 are casts from
external moulds.

PLATE 25

VANNIER el al Ordovician ostracodes

208

PALAEONTOLOGY, VOLUME 32

2. Tornquisf s Sea

Late Ordovician trilobite and other faunas demonstrate that Tornquist’s Sea, the ocean which had
earlier separated Baltica from Gondwana, had effectively closed at least to the extent that it no
longer provided a barrier to faunal migration (Cocks and Fortey 1982).

As noted in our data, both palaeocope and binodicope ostracodes show an impressive continuous
steep increase in the percentage of common genera between southern Britain (Avalonia) and
Baltoscandia (Baltica) throughout the Ordovician (text-fig. 35, plot 3). This generic faunal similarity
is particularly marked in the Caradoc and is enhanced in the Ashgill with the occurrence of species
in common (see above). Thus, the ostracode evidence concurs with other faunal and geological
evidence (Cocks and Fortey 1982; McKerrow and Cocks 1986; Pickering et al. 1988) that Avalonia
and Baltica were probably in close proximity by the late Ordovician. McKerrow and Cocks (1986)
in fact propose a possible early Ashgill collision between the two plates.

3. Rheic Ocean

By contrast with the lapetus Ocean or Tornquist’s Sea, the timing and nature of the development
of the Rheic Ocean has not been so well documented faunally. It has been recognized to exist
by lower Silurian times based, for example, on brachiopod and phytoplankton evidence (Cocks
and Fortey 1982, fig. 6 and p. 475), when it appears to have separated subtropical northern
Europe (e.g. Baltoscandia-southern Britain) from Gondwana (containing e.g. Ibero-Armorica and
Bohemia) positioned at higher latitudes to the south. Cocks and Fortey (1982, p. 472) also note
that by the Caradoc the trilobite faunas of Scandinavia and Britain were distinctively different
from those of the impoverished (typically high latitude) faunas of Bohemia and Morocco ( = on
Gondwana) and that the palaeogeography of the respective plates probably had a ‘configuration

EXPLANATION OF PLATE 26

Binodicope, leiocope, eridostracan, and nretacope ostracodes from the Ordovician of Ibero-Armorica (left)
and Baltoscandia (right).

Figs. 1-4. Binodicopa. 1, Eocytherella nioni Vannier, 1986a, Pont-de-Caen Formation, Caradoc Series,
Domfront, Orne, France; right valve (holotype, IGR 71036/1), x 69 (Vannier 1986a, pi. 13, fig. 4). 2,
Eocytherella troedssoni Bonnema, 1931, uppermost part of Ordovician, Rostanga, Scania, Sweden; left
valve (LO 2888), x 65. 3, Kinnekullea morzadeci Vannier, 1986a, La Sangsuriere Formation, Caradoc
Series, Saint-Germain-sur-Ay, Manche, France; right valve (holotype, IGR 7052/4), x 35 (Vannier 1986a,
pi. 12, fig. 2). 4, Kinnekullae hesslandi Henningsmoen, 1948, Black Tretaspis Shale, upper part of the
Ordovician, Kinnekulle, Vastergotland, Sweden; left valve (UM no. ar. os. 95), x 59 (Henningsmoen 1948,
pi. 25, fig. 5).

Figs. 5 and 6. Leiocopa. 5, Brevidorsa sp., Le Pissot Formation, Llanvirn Series, Domfront, Orne, France;
right valve (IGR 5613), x 62. 6, Brevidorsa limbata (Sidaraviciene, 1975), Ojlemyrflint erratic boulder,
upper part of the Ordovician, Isle of Sylt, North Sea, Germany; left valve (GPIMH 3412), x 59 (Schallreuter
1986, pi. 5, fig. 9).

Figs. 7 and 8. Eridostraca. 7, Conchoprimitia sp., Cacemes Formation, Llandeilo Series, near Cacemes,
Bu5aco syncline, Portugal; right valve (IGR 5613), x 25. 8, Conchoprimitia leperditioidea Thorslund, 1940,
‘ Ludibundus Limestone’, middle part of the Ordovician (lower Viruan; see text-fig. 2), Bodahamn boring,
Oland, Baltic Sea, Sweden; right valve (UM no. ol. 831), x 22 (Jaanusson 1957, pi. 15, fig. 4).

Figs. 9 and 10. Metacopa. 9, Medianella sp., Le Pissot Formation, Llanvirn Series, Domfront, Orne, France;
left valve (IGR 5612), x 94. 10, Medianella robusta (Kummerow, 1924). Ojlemyrflint erratic boulder,

upper part of the Ordovician, Isle of Gotland, Baltic Sea, Sweden; left valve (GPIMH AG G21/4), x 35.

All scanning electron micrographs except figs. 4 and 8. All external lateral views, except fig. 10 (internal
view). Figs. 1, 3, 5, 7 are casts from external moulds. Fig. 9 is a cast from an internal mould.

PLATE 26

VANNIER et al. , Ordovician ostracodes

210

PALAEONTOLOGY, VOLUME 32

not unlike that given for the early Silurian by the later Ordovician'. Indeed, a corollary of the
general northern drift of Avalonia (southern Britain) to close with Baltica (see above) and a low
latitudinal position for Gondwana in the late Ordovician (accounting for the Ashgill glaciation)
would be a widening Rheic Ocean throughout the Ordovician. McKerrow and Cocks (1986)
speculate that this ocean may have been initiated as early as the Cambrian.

Ostracode evidence suggests the presence and widening of the Rheic Ocean from as early as the
Llanvirn-Llandeilo and throughout the upper Ordovician (e.g. text-fig. 35). For example, the
percentage of palaeocope genera common to the British Isles and Ibero-Armorica shows a
distinctive overall drop during the Llanvirn to upper Caradoc time interval (text-fig. 35, graphs 1
and 4). Punctuating this overall decrease is a slight reversion towards somewhat increased faunal
similarity during the late Llandeilo-early Caradoc, which may be due to the coeval transgressive
event (see above). A decrease is also evident in the percentage of palaeocope genera common to
Ibero-Armorica and Baltoscandia, which shows a sharp uninterrupted drop from the Llandeilo to
early Caradoc (text-fig. 35, graph 2). These trends are consistent with the notion of the middle-
late Ordovician opening of the Rheic Ocean and the coeval closing of the Tornquist’s Sea.

Somewhat different trends are observed for the non-dimorphic, binodicope ostracodes which,
through the Ordovician, show: a picture of fairly consistent generic links between the British Isles
and Ibero-Armorica (text-fig. 35, graphs 1 and 4; the late Llandeilo-early Caradoc enhancement
of generic similarity may, again, be due to the coeval transgression); and a regular increase of the
percentage of genera in common between Ibero-Armorica and Baltoscandia. These binodicope
faunal contacts, apparently inconsistent with the picture obtained from palaeocopes, have no
immediately obvious explanation. Binodicopes may have had wider dispersal capacities than

EXPLANATION OF PLATE 27

Palaeocope ostracodes from the Ordovician of the British Isles (left) and Baltoscandia (right).

Fig. 1. Duringia triformosa Jones, 1984, upper part of the Llandeilo Series, Dryslwyn, Dyfed, Wales;
tecnomorphic left valve (paratype, BM OS 12260), x 63 (Jones 1986, pi. 1, fig. 3).

Fig. 2. Duringia spinosa (Kniipler, 1968), uppermost part of the Caradoc Series, Gebersdorf, Thuringia,
Germany; tecnomorphic right valve (GPIMH 2730), x 87 (Schallreuter 1984a, pi. 11 (10), fig. 1).

Fig. 3. Homeokiesowia epicopa Siveter, 1982a, Llandeilo Series, near Llandeilo, Dyfed, Wales; heteromorphic
left valve (BM OS 6670), x 28 (Siveter 1982a, pi. 9 (90), fig. 4).

Fig. 4. Homeokiesowia frigida (Sarv, 1959), Backsteinkalk erratic boulder, middle part of the Ordovician
(idavere (Cm) to Keila (Dn) stage; see text-fig. 2), Klein-Horst, Pomerania, Poland; heteromorphic right
valve (GPIMH 2023a), x 53 (Schallreuter 1979, pi. 6 (76), fig. 3).

Fig. 5. Schallreuteria (Schallreuteria) superciliata (Reed, 1910), Longvillian Stage, Caradoc Series, Melmerby,
Cumbria, England; heteromorphic right valve (paralectotype, SM A 109836), x 23 (Siveter 19826, pi. 9
(96), fig. 2).

Fig. 6. Schallreuteria (Schallreuteria) lippensis Schallreuter, 1984a, Backsteinkalk erratic boulder, middle part
of the Ordovician (lower upper Viruan; see text-fig. 2), Lippe, Hohwacht Bay, Baltic Sea, Germany;
heteromorphic right valve (GPIMH 2902), x 39 (Schallreuter 1984a, pi. 11 (8), fig. 1).

Fig. 7. Tallinnellal tomacina Jones, 1986, middle part of the Llandeilo Series, Carmarthen, Dyfed, Wales;
tecnomorphic right valve (paratype, BM OS 12595), x 22 (Jones 1986, pi. 3, fig. 10).

Fig. 8. Tallinnella sebyensis Jaanusson, 1957, erratic boulder, Upper grey Orthoceras Limestone, middle part
of the Ordovician (lower Viruan (C,b)), Linauer Moor, near Trittau, East of Hamburg, Germany;
tecnomorphic left valve (GPIMH AG G21/3), x45.

Fig. 9. Sigmoopsis duftonensis (Reed, 1910), Longvillian Stage, Caradoc Series, Melmerby, Cumbria, England;
tecnomorphic left valve (paralectotype, SM A 29914b), x 24 (Jones 1986, pi. 13, fig. 4).

Fig. 10. Sigmoopsis rostrata (Krause, 1892), Backsteinkalk erratic boulder, middle part of the Ordovician,
Northern Germany; heteromorphic left valve (GPIMH 2583), x 43 (Schallreuter 1983a, pi. 10, fig. 1).

All scanning electron micrographs of external lateral views. Figs. 5 and 9 are casts from external moulds.

PLATE 27

VANNIER et al., Ordovician ostracodes

212

PALAEONTOLOGY, VOLUME 32

palaeocopes, or even peculiar substrate preferences. Moreover, it must always be borne in mind
that binodicopes are generally less well documented than palaeocopes and would benefit from a
thorough taxonomic reappraisal.

THE PLATES

Plates 24-30 are designed to demonstrate congeneric ostracode occurrence between the three domains treated
in the text. The figures include scanning electron micrographs of our own material and pertinent illustrations
after previous authors. Restrictions on the availablity of some specimens have prevented their illustration by
scanning electron microscopy.

Abbreviations of repositories used in plate explanations: BM, British Museum (Natural History), London;
GSM, British Geological Survey, Keyworth (ex London); GPIG, Geologisches Institut der Universitat
Greifswald; GPIMH, Geologisch-Palaontologisches Institut und Museum der Universitat, Hamburg (AG =
Archiv fur Geschiebekunde); IGR, Institut de Geologie de l’Universite de Rennes; K, Ulster Museum, Belfast;
LO, Geological and Mineralogical Institute, Lund University; NMW, National Museum of Wales, Cardiff;
SM, Sedgwick Museum, Cambridge University; SMF, Senckenberg Museum, Frankfurt am Main; UM,
Museum of the Palaeontological Institute, University of Uppsala.

Acknowledgements. We thank Drs M. G. Bassett and K. Pickering for discussion. D.J.S. gratefully
acknowledges support from the University of Leicester Research Board and the Royal Society. J. V.
gratefully acknowledges support from the Alexander-von-Humboldt Foundation, the Centre de la Recherche
Scientifique, and the Royal Society, and acknowledges facilities provided whilst a research fellow at the
Geological-Palaeontological Institute, University of Hamburg and the Department of Geology, Leicester
University. We thank Pauline Siveter for typing the manuscript.

EXPLANATION OF PLATE 28

Palaeocope, binodicope, platycope (kirkbyacean), and eridostracan ostracodes from the Ordovician of the
British Isles (left) and Baltoscandia (right).

Figs. 1-4. Palaeocopa. 1 and 2, Gotula gotlandica (Schallreuter, 1967). I . Lower Limestones Member, Portrane
Limestone, Ashgill Series, near Dublin, Eire; incomplete heteromorphic right valve (K 10022), x 35
(Schallreuter and Orr 1985, pi. 12 (152), fig. 2). 2. Ojlemyrflint erratic boulder, upper part of the Ordovician
(upper Harjuan; see text-fig. 2), Vale, Gotland, Sweden; tecnomorphic left valve (GPIMH 3258), x 37
(Schallreuter and Orr 1985, pi. 12 (154), fig. 1). 3, Henningsmoenia costa Orr, 19856, Portrane Limestones,
Cautleyan Stage, Ashgill Series, Dublin, Eire; tecnomorphic left valve (K 10029), x48 (Orr 19856, pi. 12
(64), fig. 1). 4, Henningsmoenia gunnari (Thorslund, 1948), erratic boulder, middle part of the Ordovician
(Idavere (Cm) or Johvi (DO Stage; see text-fig. 2), Northern Germany; heteromorphic right valve (GPIMH
2596), x 51 (Schallreuter 1983a, pi. 11, fig. 7).

Figs. 5 and 6. Binodicopa. 5, Spinigerites hadros Jones, 1987, Onnian Stage, Caradoc Series, Welshpool,
Powys, Wales; left valve (holotype, BM OS 12567), x 36 (Jones 1987, pi. 8, fig. 7). 6, Spinigerites spiniger
(Lindstrom, 1953), Sularp Shale, Caradoc Series, Gislovshammar, Scania, Sweden; right valve (GPIMH
2314), x 58 (Schallreuter 1980a, pi. 9, fig. 2).

Fig. 7 and 8. Platycopa (Kirkbyacea). 7, Gebeckeria dryslwynensis Schallreuter and Jones, 1984, Ashgill
Series, Dryslwyn Castell, near Llandeilo, Dyfed, Wales; left valve (paratype, BM OS 12282), x 68
(Schallreuter and Jones 1984, figs. 1 and 2). 8, Martinssonozona ordoviciana Schallreuter, 1968, Ojlemyr-
flint erratic boulder, upper part of the Ordovician, Isle of Sylt, Northern Germany; left valve (GPIMH
3417), x 105 (Schallreuter 1986, pi. 6, fig. 6).

Figs. 9 and 10. Eridostraca. 9, Eridoconcha plerilamella Jones, 1987, Costonian Stage, Caradoc Series,
Lampeter-Velfrey, Dyfed, Wales; right valve (BM OS 12818), x 60 (Jones 1987, pi. 9, fig. 15). 10,

Cryptophvllus gutta Schallreuter, 19686, Ojlemyrflint erratic boulder, upper part of the Ordovician, Gotland,
Sweden; left valve (GPIMH AG G21/1), x 68.

All scanning electron micrographs of external lateral views.

PLATE 28

8

VANNIER el al., Ordovician ostracodes

214

PALAEONTOLOGY, VOLUME 32

REFERENCES

abushik, a. and sarv, l. i. 1983. Ostrakody molodovskogo gorizonta Podolii [Ostracodes from the Molodova
Stage of Podolia]. Paleontologija drevnego paleozoja Pribaltiki i Podolii , 101 134. Tallin. [In Russian.]

adamczak, F. 1976. Middle Devonian Podocopida (Ostracoda) from Poland; their morphology, systematics
and occurrence. Senckenberg. leth. 57, 265-467.

bednarczyk, w. 1974. The Ordovician in the Koszalin-Chojnice region (Western Pomerania). Acta geoi.
poi. 24, 581-600.

blumenstengel, h. 1965. Zur Ostrakodenfauna eines Kalkgerolls aus dem Thiiringer Lederschiefer (Ordovi-
zium). Freiberger ForschHft. (C), 182, 63-78.

born, a. 1918. Die Calymene Tristam-Stufe (Mittlers Untersilur) bei Almaden. Senckenberg. Naturf. Ges.,
Abb. 36, 309-358.

brenchley, p. j. and cocks, l. r. m. 1982. Ecological associations in a regressive sequence. The latest
Ordovician of the Oslo-Asker District, Norway. Palaeontology, 25, 783-815.

bretsky, p. w. and lorenz, d. m. 1970. Adaptive response to environmental stability: a unifying concept in
paleoecology. Proc. North Amer. Paleont. Convention, Chicago, 1969, 522-550.

bruton, D. L. (ed.). 1984. Aspects of the Ordovician System. Pa/aeont. Contr. LJniv. Oslo, 295, 1-228.

bruton, d. l. and williams, s. h. 1982. Field excursion guide; IV International Symposium on the Ordovician
System, Oslo, Norway. Ibid. 279, 1-217.

cocks, L. R. M. and fortey, r. a. 1982. Faunal evidence for oceanic separations in the Palaeozoic of Britain.
J. geoi. Soc. 139, 465-478.

cogne, j. 1971. Le Massif Armoricain et sa place dans la structure des socles ouest-europeens: Fare hercynien
ibero-armoricain. In: Histoire structural du Golfe de Gascogne. Pnbl. Inst, franqais Petrole (coll. Colloques
et Seminaires), 1, 1-23.

dean, w. t. 1966. The Lower Ordovician stratigraphy and trilobite faunas of the Landeyren Valley and the
neighbouring district of the Montagne Noire, south-western France. Bull. Br. Mas. nat. Hist. (Geoi.), 12,
245-353.

destombes, j. 1962. Stratigraphie et paleogeographie de FOrdovicien de FAnti-Atlas (Maroc). Un essai de
synthese. Bull. Soc. geoi. Fr. 7, 453-460.

EXPLANATION OF PLATE 29

Binodicope ostracodes from the Ordovician of the British Isles (left) and Baltoscandia (right).

Fig. 1. Bullaeferum llandeiloensis Jones, 1987, upper part of the Llanvirn Series, Llandeilo, Dyfed, Wales;
right valve (holotype, BM OS 12743), x 59 (Jones 1987, pi. 1, fig. 1).

Fig. 2. Bullaeferum tapaensis (Sarv, 1959), Ojlemyrflint erratic boulder, upper part of the Ordovician, Isle of
Sylt, North Sea, Germany; left valve (GPIMH 3409), x 78 (Schallreuter 1986, pi. 5, fig. 4).

Fig. 3. Pariconchoprimitia improba Jones, 1987, upper part of the Llanvirn Series, Llandeilo, Dyfed, Wales;
left valve (BM OS 12777), x 65 (Jones 1987, pi. 5, fig. 17).

Fig. 4. Pariconchoprimitia conchoides (Hadding, 1913), Sularp Shale, Caradoc Series, Gislovshammar, Scania,
Sweden; right valve (GPIMFI 2299), x 39 (Schallreuter 1976, pi. 6, fig. 1).

Fig. 5. Easchmidtella elementa Jones, 1987, upper part of the Llandeilo Series, Dryslwyn, Dyfed, Wales; right
valve (holotype, BM OS 12767), x 66 (Jones 1987, pi. 5, fig. 1).

Fig. 6. Easchmidtella fragosa (Neckaja in Abushik et al., 1960), Bachsteinkalk erratic boulder, middle part
of the Ordovician (Idavere (Cm) or Johvi (D,) stage, upper Viruan; see text-fig. 2), Kammin, near Greifswald,
Germany; left(?) valve (GPIMH AG G21/5), x 80.

Fig. 7. Pedomphalella expraeputia Jones, 1987, Costonian Stage, Caradoc Series, Lampeter-Velfrey, Dyfed,
Wales; right valve (BM OS 12786), x 69 ( Jones 1987, pi. 6, fig. 1).

Fig. 8. Pedomphalella jonesii (Krause, 1897), Backsteinkalk erratic boulder, middle part of the Ordovician
(Idavere (Cm) or Johvi (Dfi Stage, upper Viruan; see text-fig. 2), near Stralsund, Pomerania, Germany; left
valve (GPIMH 2722), x 82 (Schallreuter and Siveter 1985, pi. 69, fig. 8).

Fig. 9. Conchoprimitiella dvfedensis Jones, 1987, upper part of the Llandeilo Series, Dryslwyn, Dyfed, Wales;
right valve (BM OS 12805), x 43 (Jones 1987, pi. 7, fig. 1).

Fig. 10. Conchoprimitiella eremita Schallreuter, 1980«, Sularp Shale, Caradoc Series, Gislovshammar, Scania,
Sweden; left valve (holotype, GPIMH 2292), x 65 (Schallreuter 1976, pi. 4, fig. 4).

All scanning electron micrographs of external lateral views.

PLATE 29

VANNIER el a!., Ordovician ostracodes

216 PALAEONTOLOGY, VOLUME 32

destombes, J. 1971. L’Ordovicien au Maroc. Essai de synthese stratigraphique. Mem. Bur. Rech. geol. min.
73, 229-235.

dewey, j. F. 1969. Evolution of the Appalachian/Caledonian Orogen. Nature, Loud. 222, 124 129.
dons, J. a. and henningsmoen, g. 1949. Two New Middle Ordovician Ostracods from Oslo. Norsk geol.
Tisskr. 28, 27-32.

eldredge, n. 1974. Stability, diversity and speciation in Paleozoic epeiric seas. J. Paleont. 48, 540-548.
fortey, r. a. 1975. Early Ordovician trilobite communities. Fossils Strata, 4, 339-360.

— 1984. Global earlier Ordovician transgressions and regressions and their biological implications. In
bruton, d. l. (ed.). Aspects of the Ordovician System. Palaeont. Contr. Univ. Oslo, 295, 37-50.

— 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. Bull Br. Mus. nat. Hist.
(Geol.), 36, 63-75.

EXPLANATION OF PLATE 30

Palaeocope and binodicope ostracodes from the Ordovician of Ibero-Armorica [left), the British Isles (middle),
and Baltoscandia (right).

Figs. 1-9. Palaeocopa. 1, Ogmoopsis arcadelti Vannier, 19866, Llandeilo Series, Traveusot, Guichen, Ille-et-
Vilaine, France; tecnomorphic left valve (holotype, IGR 5307/1), x 23 (Vannier 19866, pi. 8, fig. 1). 2,
Ogmoopsis (Quadridigitalis) siveteri Jones, 1986, Harnagian Stage, Caradoc Series, Cressage, Shropshire,
England; heteromorphic left valve (holotype, BM OS 12654), x 16 (Jones 1986, pi. 12, fig. 1). 3, Ogmoopsis
alata Sarv, 1959, erratic boulder, lower part of the Ordovician (Kunda Stage (Bm); see text-fig. 2), Ahlintel,
near Munster, Germany; tecnomorphic left valve (Westfalisches Museum fur Naturkunde in Miinster,
WMN no. A15), x61 (Schallreuter 19856, pi. 4, fig. 2). 4, Vittella sp., Schistes du Fresne Formation,
Llanvirn Series, Ligne, Loire-Atlantique, France; tecnomorphic left valve (IGR 5290/1), x 32 (Vannier
19866, pi. 3, fig. 1). 5, Vittella fecunda Siveter, 1983, upper part of the Llandeilo Series, Dryslwyn, near
Llandeilo, Dyfed, Wales; heteromorphic left valve (holotype, BM OS 7777), x 32 (Siveter 1983, pi. 10 (14),
fig. 1). 6, Vittella vittensis Schallreuter, 1964, erratic boulder, middle part of the Ordovician (lower upper
Viruan; see text-fig. 2), Isle of Hiddensee, Baltic Sea, Germany; heteromorphic left valve (holotype, GPIG
no. Os 289), x40 (Schallreuter 1964, pi. 11, fig. 3). 7, Quadritia (Krutatia) tromelini Vannier and

Schallreuter, 1983, Cacemes Formation, Llandeilo Series, Cacemes section, near Cacemes, Bu^aco syncline,
Portugal; right valve (paratype, IGR 5101/A1), x 46 (Vannier and Schallreuter 1983, pi. 9, fig. 5). 8,
Quadritia (Krutatia) iunior Schallreuter, 1981, Ashgill Series, Dryslwyn Castell, near Llandeilo, Dyfed,
Wales; right valve (BM OS 13372), x43. 9, Quadritia (Krutatia) krausei Schallreuter, 1976, Backsteinkalk
erratic boulder, middle part of the Ordovician (Idavere (Cm) or Johvi (Dfi Stage; see text-fig. 2), Northern
Germany; left valve (GPIMH 2591), x 47 (Schallreuter 1983a, pi. 11, fig. 1).

Figs. 10-18. Binodicopa. 10, Klimphores vogelweidei Vannier, 1986a, Traveusot Formation, Llanvirn Series,
Laille, Ille-et-Vilaine, France; right valve (holotype, IGR 5255A/8), x61 (Vannier 1986a, pi. 1, fig. 1). 11,
Klimphores morgani (Jones, 1890), Onnian Stage, Caradoc Series, Welshpool, Powys, Wales; left valve
(specimen now broken), x 79 (Jones 1987, pi. 1, fig. 14). 12, Klimphores planus Schallreuter, 1966c, erratic
boulder, middle part of the Ordovician (lower upper Viruan; see text-fig. 2), Isle of Hiddensee, Baltic Sea,
Germany; right valve (GPIMH 2230), x 65 (Schallreuter 19806, pi. 7 (10), fig. 4). 13, Laterophores varesei
Vannier, 1986a, Andouille Formation, Caradoc Series, Andouille, Mayenne, France; left valve (holotype,
IGR 30415/2), x 77 (Vannier 1986a, pi. 2, fig. 1). 14, Laterophores elevatus Jones, 1987, upper part of the
Llanvirn Series, Llandeilo, Dyfed, Wales; right valve (holotype, BM OS 12751), x 64 (Jones 1987, pi. 2,
fig. 1). 15, Laterophores lateris Schallreuter, 1968a, Backsteinkalk erratic boulder, middle part of the

Ordovician (lower upper Viruan; see text-fig. 2), Isle of Hiddensee, Baltic Sea, Germany; right valve
(GPIMH AG G21/2), x 75. 16, Pseudulrichia raguenezensis Vannier, 1986a, Gres de Kermeur Formation,
Caradoc Series, Raguenez, Crozon, Finistere, France; right valve (holotype, IGR 30576/1), x 32 (Vannier
1986a, pi. 5, fig. 2). 17, Pseudulrichia conispina Jones, 1987, upper part of the Llandeilo Series, Dryslwyn,
Dyfed, Wales; right valve (BM OS 12762), x 59 (Jones 1987, pi. 3, fig. 5). 18, Pseudulrichia ullehmanni

Schallreuter, 1981, Ojlemyrflint erratic boulder, upper part of the Ordovician (Harjuan Series; see text-fig.
2), Isle of Sylt, North Sea, Germany; right valve (paratype, GPIMH 2151), x 74 (Schallreuter 1981, fig.
9).

All external lateral views; all scanning electron micrographs except fig. 6. Figs. 1, 2, 4, 7, 10, 11, 13, 16 are
casts from external moulds.

PLATE 30

VANNIER et al Ordovician ostracodes

218

PALAEONTOLOGY, VOLUME 32

fortey, r. a. and Owens, R. m. 1978. Early Ordovician (Arenig) stratigraphy and faunas of the Carmarthen
district. South-west Wales. Ibid. (Geol.), 30, 225-294.

gailIte, l. 1971. Ostrakody semejstva Bolliidae Boucek ordovika Latvii [Ostracoda of the family Bolliidae
Boucek in the Ordovician of Latvia], Paleontologia i Stratigrafiya paleozoya i mezozoya Pribaltiki i
Byelorussii, 45-47. [In Russian.]

— 1975a. Novye vidy ostrakod verchnego ordovika Latvii [New species of Ostracoda from upper Ordovician
of Latvia]. In grigelis, a. a. (ed.). Fauna i Stratigrafiya paleozoya i mezozoya Pribaltiki i Byelorussii , 45
57. [In Russian.]

— 19756. Ostrakody iz pogranicnych sredne-i verchneordovikskich otlozenij Zapadnoj Latvii [Ostracoda
from middle and upper Ordovician boundary in West Latvia], In grigelis, a. a. (ed.). Ibid. 59-67 . [In
Russian.]

gramm, m. n. 1984. Inutrennie struktury rakovie paleozojskich ostrakod [Internal structure of the carapace
of Palaeozoic ostracodes]. Nauka Ross. 72 pp., 32 pis. [In Russian.]
guber, a. l. and jaanusson, v. 1964. Ordovician ostracodes with posterior domiciliar dimorphism. Bull. geol.
Instn Unix. Upsala, 43, 1-43, 6 pis.

guillocheau, F. 1983. Les depots de tempetes. Le modele de l’Ordovicien moyen ouest-armoricain. Thesis
(unpublished). University of Brest, Prance.

hammann, w., romano, m. and robardet, m. 1982. The Ordovician System in Southwestern Europe (Prance,
Spain and Portugal). I.U.G.S. publ. 1, 1-47.

henningsmoen, G. 1948. The Tretaspis Series of the Kullatorp core. Bull. geol. Instn Univ. Upsala , 42, 374-
432.

— 1953. The Middle Ordovician of the Oslo region (Norway). 4. Ostracoda. Norsk geol. Tidsskr. 32, 35-
56.

1954a. Lower Ordovician of the Oslo region (Norway). Ibid. 33, 41-68.

19546. Upper Ordovician of the Oslo region (Norway). Ibid. 69 108.

henry, j. L. 1980. Trilobites ordoviciens du Massif Armoricain. Mem. Soc. geol. miner. Bretagne , 22, 1-250,
48 pis.

— melou, n., nion, j., Paris, f., robardet, m., skevington, d. and thadeu, d. 1976. L’apport de Graptolites
de la Zone a G. teretiusculus dans la datation de faunes benthiques lusitano-armoricaines. Ann Is Soc. geol.
N. 96, 275-281.

hessland, i. 1949. Investigations of the Lower Ordovician of the Siljan district, Sweden. I. Lower Ordovician
ostracods of the Siljan district, Sweden. Bull. geol. Instn Univ. Upsala , 33, 97-408.
hutton, d. h. w. 1987. Strike-slip terranes and a model for the evolution of the British and Irish Caledonides.
Geol. Mag. 124, 405-425.

jaanusson, v. 1957. Middle Ordovician ostracodes of central and southern Sweden. Ibid. 37, 173-442.

— 1966. Ordovician ostracodes with supravelar antra. Ibid. 43, 30 pp., 3 pis.

1973. Aspects of carbonate sedimentation in the Ordovician of Baltoscandia. Lethaia, 6, 1 1 34.

1976. Paunal dynamics in the Middle Ordovician (Viruan) of Balto-scandia. In bassett, m. g. (ed.).

The Ordovician System; proceedings of a Palaeontological Association symposium, Birmingham , 1974,
301-326.

— 1985. Punctional morphology of the shell in platycope ostracodes- a study of arrested evolution.
Lethaia, 18, 73-84.

— and bergstrom, s. m. 1980. Middle Ordovician faunal spatial differentiation in Baltoscandia and the
Appalachians. Alcheringa, 4, 89-1 10.

JONES, C. R. 1984. On Duringia triformosa Jones sp. nov. Stereo-Atlas Ostracod Shells, 11, 13-16.

— 1985. Llandeilo and Caradoc (Ordovician) beyrichiocope Ostracoda from England and Wales. Ph.D.
thesis (unpublished). University of Leicester, Great Britain.

— 1986. Ordovician (Llandeilo and Caradoc) beyrichiocope Ostracoda from England and Wales. Part I.
Palaeontogr. Soc. [Monogr.], 76 pp., 22 pis.

1987. Ordovician (Llandeilo and Caradoc) beyrichiocope Ostracoda from England and Wales. Part II.
Ibid. 37 pp., 8 pis.

jones, t. r. 1890. On some Palaeozoic Ostracoda from North America, Wales and Ireland. Q. Jt geol. Soc.
Lond. 46, 1 31.

— and holl, h. b. 1869. Notes on the Palaeozoic bivalved Entomostraca, No. IX. Some Silurian species.
Ann. Mag. nat. Hist. ser. 4, 3, 211-227.

— and siveter, d. j. 1983. On Harperopsis scripta (Harper). Stereo- Atlas Ostracod Shells, 10, 5-12.
kala, e., puura, v. and suuroja, k. 1984. Glabnye verty stroenia kiardlaskogo pogrevennogo kratera

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

219

[Main features of the Kardia buried crater]. Eesli NSV Tead. Akad. Geol. Inst. Uurim. 33, 17. [In
Russian.]

knupfer, J. 1968. Ostrakoden aus dem Oberen Ordovizium Thiiringens. Freiberger ForschHft. (C), 234,
5-29.

krandievsky, v. s. 1969. Stratigrificne posirennja ostrakod v ordovic' kicli vidkladach Volino-Podillja
[Stratigraphic distribution of Ostracoda in the Ordovician deposits of the Volyn-Podolye], Dopov. Akad.
Nauk ukr. RSR { B ), 10, 870 874. [In Ukrainian.]

1975. Persi znachidki nizn' ordovic' kich ostrakod v Zarchidnij Volini [First occurrences of the lower

Ordovician ostracodes in Western Volyn], Ibid. 8, 691 694. [In Ukrainian.]
lethiers, p., le fevre, j., vannier, j. and weyant, m. 1985. Paleozoique. In OERTLI, H. j. (ed. ). Atlas des
ostracodes de France. Bull. Centres Rech. E.xplor. Prod. Elf- Aquitaine, Mem. 9, 33 87.

LINDSTROM, m. 1987. A report on activities and outlooks in the Baltoscandian Ordovician (carbonate platform
or carbonate basin?). Newsletter (April 1987) of the Working Group on Ordovician Geology of Baltoscandia
(WOGOGOB).

ludvigsen, r. 1982. Upper Cambrian and Lower Ordovician trilobite biostratigraphy of the Rabbit-kettle
Formation, western District of Mackenzie. Fife Sci. Contr. R. Ont. Mus. 134, 1 188.
mckerrow, w. s. and cocks, l. r. m. 1976. Progressive faunal migration across the Iapetus Ocean. Nature,
Lond. 263, 304-306.

— 1986. Oceans, island arcs and olistostromes: the use of fossils in distinguishing sutures, terranes
and environments around the Iapetus Ocean. J. geol. Soc. 143, 185-191.
mannil, r. m. 1963. Voprosy sopostavlenija ordovikskich otlozenij Estonii i Leningradskoj oblasti [On the
correlation of the Ordovician strata of Estonia and Leningrad region], Eesti NSV Tead. Akad. Geol. Inst.
Uurim. 13, 3-40. [In Russian.]

1966. Istoriya razvitiya Baltiyskogo basseyna v ordovike [Evolution of the Baltic Basin during the

Ordovician], Eesti Tead. Akad. Geol. Inst. Tallinn, 200 pp. [In Russian.]

1971. Distribution of selected Ordovician Chitinozoa assemblages and species in Northern Europe and

their stratigraphical evaluation. Mem. Bur. Rech. geol. min. 73, 309 311.
martinsson, a. 1956. Neue Funde kambrischer und ordovizischer Geschiebe inr siidwestlischen Finnland.
Bull. geol. Instn Univ. Upsala, 36, 79-105.

meidla, T. 1983. Ostrakody pogranicnych sloev vormsiskogo i pirguskogo gorizontov v zapadnoj Estonii
[Ostracodes from the Vormsi Pirgu boundary of Estonia]. Eesti NSV Tead. Akad. Geol. Inst. Uurim. 32,
53-59. [In Russian ]

— 1986. Novye ostrakody iz ordovika Pribaltiki [New ostracodes from the Ordovician of the Baltic Area]
Ibid. 35, 10 19. [In Russian.]

mitchell, a. h. G. 1984. The British Caledonides: interpretations from Cenozoic analogues. Geol. Mag. 121,
35-46.

neckaja, a. i. 1973. Ostrakody ordovika i silura SSSR [Ordovician and Silurian ostracodes in the U.S.S.R.].

Trudy vses. neft. nauchno-issled. geol.-razv. Inst. 324, 1 104. [In Russian.]
opik, a. a. 1937. Ostracoda from the Ordovician Uhaku and Kukruse formations of Estonia. Ann Is Soc. Reb.
Nat. invest. Univ. tartu. 43, 65-138.

orr, R. j. 1985a. On Henningsmoenia gunnari (Thorslund). Stereo-Atlas Ostracod Shells, 12, 57 60.

19856. On Henningsmoenia costa Orr sp. nov. Ibid. 61 68.

Paris, F. 1979. Les chitinozoaires de la formation de Louredo, Ordovicien superieur du synclinal de Bu9aco
(Portugal). Palaeontographica A, 164, 24-51.

1981. Les chitinozoaires dans le Paleozoique du sud-ouest de l’Europe (cadre geologique — etude

systematique— biostratigraphie). Mem. Soc. geol. miner Bretagne, 26, 1-412, 26 pis.

— and robardet, m. 1978. Paleogeographie et relations ibero-armoricaines au Paleozoique ante-carbonifere.
Bull. Soc. geol. Fr. 7, 1121-1126.

pickering, k. t. 1987. Deep-marine foreland basin and forearc sedimentation: a comparative study from the
Lower Palaeozoic Northern Appalachians, Quebec and Newfoundland. In leggett, j. k. and zuffa, g. g.
Marine clastic sedimentology, 190 -211.

PICKERING, K. T., bassett, M. G. and siveter, D. J. 1988. Late Ordovician early Silurian destruction of the
Iapetus Ocean: Newfoundland, British Isles and Scandinavia, a discussion. Trans. R. Soc. Edinb. (In press).
poulsen, v. 1965. An early Ordovician trilobite fauna from Bornholm. Meddr dansk. geol. Foren. 16, 49 1 13.

— 1978. Dalmanitina beds (late Ordovician) on Bornholm. Damn. geol. Unders. 1976, 53-87.
pranskevicius, A. A. 1972. Ostrakody silura Juznoy Pribaltiki [South Baltic Silurian ostracodes], Trudy

Litovsk. Nauk. Geol. Inst. 15, 1-280, 42 pis. [In Russian ]

220

PALAEONTOLOGY, VOLUME 32

pribyl, a. 1975. Hastatellina gen. n., eine neue Ostracoden-Gattung und ihre Vertreter aus dem bohmischen
und nichtbohmischen Ordovizium. Paldont. Z. 49, 11-23.

1979. Ostracoden der Sarka-bis Kraluv Dvur-Schichtengruppe dcs bohmischen Ordoviziums. Sb. ndr.

Mus. Praze, B. 33, 53-145.

prokofiev, v. a. and Kuznetzov, a. g. 1982. Fauna i nekotorye voprosy stratigrafii ordovikskich otlozeniy
Moskovskoy sineklisy [Fauna and certain stratigraphic problems in Ordovician deposits of Moscow
syneclise], Bui. MOIP (mojsk. obscestva ispytatelej prirody , Otd. geol. 57, 67-82. [In Russian.]
qvale, G. 1980. New Caradocian ostracodes from the Oslo-Asker district, Norway. Norsk geol. Tidsskr. 60,
93-116.

roomusoks, a. 1970. Stratigrafiya viruskoy i harjuskoy seriy (ordovik) severnoy Estonii [Stratigraphy of
the Viruan and Harjuan Series (Ordovician) in northern Estonia]. Tartu Riiklik Ulik. 1-346, 9 pis. [In
Russian.]

sarv, l. i. 1959. Ostrakody ordovika Estonskoy SSR [Ordovician ostracods in the Estonian SSR]. Eesti NSV
Tead. Akad. Geol. Inst. Uurim. 4, 1-211, 32 pis. [In Russian.]

1962. Ostrakody porkuniskogo gorizonta i llandoveri Estonii [Ostracods from the Pokurni Stage and

Llandovery of Estonia. Ibid. 9, 95 141. [In Russian.]

— 1963. Novye ostrakody ordovika Pribaltiki [New ostracods of the Ordovician of East Baltic], Ibid. 13,
161-188. [In Russian ]

schallreuter, r. e. l. 1964. Neue Ostracoden der Uberfamilie Hollinacea. Ber. geol. Ges. D.D.R. 2, 87-93,
pis. 11-13.

1966 a. Zur Taxonomie und Phylogenie der Ostrakodenfamilie Ctenonotellidae Schmidt, 1941 (Palaeoco-

pina, Hollinacea). Geologie, 15, 197-215.

1966 b. Zur Taxonomie und Phylogenie der Ostrakodenfamilie Tetradellidae Swartz, 1936 (Palaeocopina,

Hollinacea). Ibid. 846-875.

— 1966c. Drepanellacea (Ostracoda, Beyrichiida) aus mittelordovizischen Backsteinkalkgeschieben I.
Klimphores planus g.n.sp.n. und Vaivanovia hiddenseensis g.n.sp.n. Ber. dt. Ges. geol. Wiss. A, 11, 393-402.

1967. Postkriptum zur Taxonomie der Tetradellidae (Ostracoda). Neues Jb. Geol. Paldont. Mh. 7, 431 -

446.

1968a. Zur Taxonomie und Phylogenie der ordovizischen Beyrichicopida (Ostracoda). Ber. dt. geol. Ges.

Wiss. A. Geol. Paldont. 13, 177-183.

19686. Zur Taxonomie und Phylogenie der Eridostraca (Ostracoda). Paldont. Z. 42, 105-119.

1969a. Untergattungen der Ostrakodengattung Platybolbina. Geologie , 18, 877-879.

19696. Alter und Heimat der Backsteinkalkgeschiebe. Hercynia , 6, 285-305.

1971a. Zum Alter der Rollsteinkalkgeschiebe. Neues Jb. Geol. Paldont. Mh. 11, 690-696.

19716. Asymmetrische ordovizische Ostrakoden. Ibid. 4, 249-260.

1971c. Ostrakoden aus Ojlemyrgeschieben (Ordoviz). Ibid. 1971, 423-431.

1972. Vier neue Arten der Ostrakodenfamilie Rectellidae. Zool. Anz. 188, 254-260.

1973a. Die Ostracodengattung Hyper chilarina und das A parchites-prob\em . Geol. For. Stockh. Fork. 95,

37-49.

19736. Tvaerenellidae (Ostracoda, Palaeocopina) aus Backsteinkalk-Geschieben (Mittelordoviz) Nord-

deutschlands. Palaeontographica A , 144, 55-111.

1975. Palaeocopine Ostrakoden aus Backsteinkalk-Geschieben (Mittelordoviz) Norddeutschlands. Ibid.

149, 139-192.

1976. Ctenonotellidae (Ostracoda, Palaeocopina) aus Backsteinkalk-Geschieben (Mittelordoviz) Nord-
deutschlands. Ibid. 153, 161-215.

1977a. On Cryptophyllus gutta Schallreuter. Stereo-Atlas Ostracod Shells , 4, 1-8.

19776. On Miehlkella criboporata Schallreuter gen. et sp. nov. Ibid. 9-16.

1977c. On Antiaechmina pseudovelata Schallreuter sp. nov. Ibid. 5, 29-32.

— 1977<7. Zwei neue ordovizische Podocopida (Ostracoda) und Bemerkungen zur Herkunft der Cytheracea
und Cypridacea. Neues Jb. Geol. Paldont. Mh. 12, 720-734.

1977e. Eine neue Art der Ostrakodengattung Pelecvbolbina aus dem Ordoviz von Oland. Geol. For.

Stockh. Forh. 99, 409-411.

— 1977/ Taxonomie und Phylogenie der palaozoischen Ostrakodengattung Semibolbina Jordon. Paldont.
Z. 51, 32-51.

1978a. Zur Phylogenie und Systematik der ordovizischen Ostrakoden. Unpublished, 173 pp.

— 19786. Zwei weitere ordovizische Cytheracea (Ostracoda, Podocopida). Neues Jb. Geol. Paldont. Mh. 9,
567-576.

VANNIER ET AL.\ ORDOVICIAN OSTRACODE FAUNAS

221

1979. Ordovician podocope ostracodes. In Proceedings of the VII International Symposium on
Ostracodes, Beograd (Taxonomy, biostratigraphy and distribution of ostracodes), 25-28.

— 1980a. Ostrakoden aus dem Sularpschiefer (Mittelordoviz) von Schonen (Schweden). Palaeontographica
A, 169, 1-27.

19806. On Klimphores planus Schallreuter. Stereo- Atlas Ostracod Shells , 7, 9- 16.

1981. Ordovician ostracodes from Baltoscandia. Geol. For. Stockh. Forh. 103, 6171.

— 1982a. Tetradellidae (Ostracoda, Palaeocopa) aus Backsteinkalk-Geschieben (Mittelordoviz) Nord-
deutschlands (mit Ausnahme der Glossomorphitinae). Palaeontographica A , 178, 1 48.

19826. Extraction of ostracodes from siliceous rocks. In bate, r. h., robinson, e. and sheppard, l. m.

(eds.). Fossil and Recent ostracods , 169 176. Ellis Howard, Chichester, for the British Micropalaeontological
Society.

1983a. Glossomorphitinae und Sylthinae (Tetradellidae, Palaeocopa, Ostracoda) aus Backsteinkalk-

Geschieben (Mittelordoviz) Norddeutschlands. Palaeontographica A , 180, 126-191.

19836. On Bromidella sarvi Schallreuter. Stereo- Atlas Ostracod Shells , 10, 25-28.

1984a. Geschiebe-Ostrakoden I [Ostracodes from erratic boulders I]. Neues Jh. Geol. Paldont. Abh. 169,

1-40.

19846. Middle Ordovician ostracodes from Sweden. Geol. For. Stockh. Forh. 106, 93-99.

1984c. Sigmobolbina (Ostracoda) aus mittelordovizischen Sylter Hornstein-Geschieben. Neues Jb. Geol.

Paldont. Mh. 7, 33-38.

1984 d. Neufunde der gehornten Leperditiocopen-Gattung Kiaeria (Ostracoda) in silurischen Geschieben

Westfalens sowie ihre systematische und phylogenetische Stellung. Paldont. Z. 58, 131 142.

1985a. Mikrofossilien von Sylt. In hacht, u. von. (ed.). Fossilien von Sylt, 77 91.

19856. Ein ordovizisches Kalksandstein-Geschiebe aus Westfalen. Geol. Paldont. Westf. 4, 23-52.

1985c. Homeomorphy, phylogeny and natural classification: case studies involving Palaeozoic ostracodes.

9th Internat. Symp. Ostracoda (Evolutionary biology of Ostracoda, its fundamentals and applications).
Programs and Abstracts, Shizuoka, July 1985, 94 95.

1986. Ostrakoden aus Ojlemyrflint-Geschieben von Sylt. Preprint from hacht, v. von. Fossilien von

Sylt //(hacht, i. m. von. (ed.). Hamburg), 31 pp., 3 pis.

1987. Geschiebe-Ostrakoden II [Ostracodes from erratic boulders II], Neues Jb. Geol. Paldont. Abh. 174,

23-53.

1988. Ostrakoden (Muschelkrebse). In press.

— and jones, c. R. 1984. A new Ordovician kirkbyacean ostracode. Neues Jb. Geol. Paldont. Mh. 7, 416

— and ORR, R. J. 1985. On Gotula gotlandica (Schallreuter). Stereo-Atlas Ostracod Shells , 12, 149 156.
and siveter, d. j. 1983. On Tallinnellina dissita Schallreuter & Siveter. Ibid. 10, 1 4.

1985. Ostracodes across the Iapetus Ocean. Palaeontology , 28, 577 598.

shaw, f. c. and fortey, r. a. 1977. Middle Ordovician facies and trilobite faunas in N. America. Geol. Mag.
114, 409-430.

sidaraviciene, N. v. 1971. Novye ostrakody iz srednego i verchnego ordovika Litvy [New Ostracoda from
Middle and Upper Ordovician of Lithuania]. Paleontologiya i stratigrafiya Pribaltiki i Byelorussii
[ Palaeontology and stratigraphy of the Baltic and the Byelorussia ], 3, 23-36. [In Russian.]

— 1975. Novye ostrakody ordovika Juznoy Pribaltiki [New Ordovician ostracodes of South Baltic
Area]. In grigelis, a. a. (ed.). Fauna i Stratigrafiya pa/eozoya i mezozoya Pribaltiki i Byelorussii , 21 43.
[In Russian ]

siveter, d. J. 1978. The Ordovician. Geol. J. Spec. Issue , 8, 41-56.

— 1982a. On Homeokiesowia epicopa Siveter sp. nov. Stereo- Atlas Ostracod Shells , 9, 89-92.

19826. On Schallreuteria super ciliata (Reed). Ibid. 93-100.

1983. On Vittella fecunda Siveter sp. nov. Ibid. 10, 13 16.

— 1984. Habitats and mode of life of Silurian ostracodes. Spec. Pap. Palaeont. 32, 71 85.

sztejn, j. 1985. Malzoraczki ordowiku w polnocno-wschodniej Polsce [Ordovician ostracods in north-eastern
Poland], Biul. Inst. Geol. 350, 54 89. [In Polish. |

thorslund, p. 1940. On the Chasmops Series of Jemtland and Sodermanland (Tvaren). Sver. geol. Unders.
Afh. (Ser. C), 436, I 191, 15 pis.

thorslund, p. 1948. The Chasmops Series of the Kullatorp core. Bull. geol. Instn Univ. Upsala , 32, 343 373,
pis. 20-22.

troedsson, g. t. 1918. Om Skanes Brachiopodskiffer. Acta Univ. land, (nf 2), 15, I 1 10, 2 pis.
ulst, r. z., gailite, l. k. and Jakovleva, v. i. 1982. Ordovik Latvii [ Ordovician of Latvia ], 295 pp., 8 pis.
Riga (Zinatne). [In Russian !

222 PALAEONTOLOGY, VOLUME 32

vannier, j. 1983a. Rivillina, a new ostracode (Bradoriida?) genus of the Armorican Massif, France. Alcheringa,
7, 1-8.

— 1983/7. On Bollia delgadoi Vannier sp. nov. Stereo-Atlas Ostracod Shells , 10, 95 98.

1983c. On Hastatellina normandiensis (Pribyl). Ibid. 99-102.

1984a. Ostracodes ordoviciens du Massif Armoricain. Thesis (unpublished), University of Rennes,
France.

1984/>. On Raimbautina hammanni Vannier gen. et sp. nov. Stereo-Atlas Ostracod Shells, 11, 111-118.

1984c. On Thibautina rorei Vannier gen. et sp. nov. Ibid. 119-122.

1986a. Ostracodes Binodicopa de l’Ordovicien (Arenig-Caradoc) ibero-armoricain. Palaeontographica
A, 193, 77-143.

— 1986/7. Ostracodes Palaeocopa de l’Ordovicien (Arenig-Caradoc) ibero-armoricain. Ibid. 145-218.

— and schallreuter, R. E. L. 1983. Quadritia (Krutatia) tromelini sp. nov., Ostracode du Llandeilo ibero-
armoricain. Interet paleogeographique. Geobios , 16, 583-599.

- and vaslet, d. 1987. Ostracodes from the early Ordovician of central Saudi Arabia. Saudi Arabian
Directorate General of Mineral Resources (open file report). BRGM-OF-07-31, 20 pp., 2 pis.
Whittington, h. b. and hughes, c. p. 1972. Ordovician geography and faunal provinces deduced from
trilobite distribution. Phil. Trans. Roy. Soc. Lond. Ser. B, 263, 235-278.
wilcox, c. J. and lockley, m. g. 1981. A re-assessment of facies and faunas in the type Llandeilo
(Ordovician), Wales. Palaeogeogr. Palaeoclimat. Palaeoecol. 34, 285-314.

WILLIAMS, A., STRACHAN, I., BASSETT, D. A., DEAN, W. T., INGHAM, J. K., WRIGHT, A. D. and WHITTINGTON,
H. b. 1972. A correlation of Ordovician rocks in the British Isles. Geol. Soc. Spec. Rep. 3, 1-74.
williams, h. 1973. Distribution of brachiopod assemblages in relation to Ordovician palaeogeography. Spec.
Pap. Palaeont. 12, 241-269.

1976. Plate tectonics and biofacies evolution as factors in Ordovician correlations. In bassett, m. g.
(ed.). The Ordovician System , 29-66. University of Wales Press and National Museum of Wales, Cardiff.

1980. Structural telescoping across the Appalachian Orogen and the minimum width of the Iapetus
Ocean. Spec. Pap. geol. ,4ss. Can. 20, 422-440.

JEAN M. C. VANNIER

Institut de Geologie
Universite de Rennes
35042 Rennes-Cedex
France

DAVID J. SIVETER

Department of Geology
University of Leicester
Leicester LEI 7RH, UK

ROGER E. L. SCHALLREUTER

Geol.-Palaont. Institut
Universitat Hamburg
Bundesstrasse 55

Typescript received 22 February 1988 D2000 Hamburg 13

Revised typescript received 4 July 1988 West Germany

FULENGIA , A SUPPOSED EARLY LIZARD
REINTERPRETED AS A PROSAUROPOD

DINOSAUR

by s. e. evans and a. r. milner

Abstract. The skull of Fulengia youngi Carroll and Galton, a supposed lizard from the Upper Triassic/Lower
Jurassic of China, is re-examined and compared with contemporary prosauropod dinosaurs. On the basis of
its teeth, and the construction of the maxilla and mandible, the skull of Fulengia is reinterpreted as that of
a juvenile prosauropod dinosaur. It most closely resembles specimens of Gyposaurus sinensis Young, now
generally acknowledged to be juveniles of the common Lufeng anchisaurid Lufengosaurus. Fulengia youngi
is formally proposed to be a junior synonym of Lufengosaurus huenei. The earliest unequivocal fossils of
lizards are of Upper Jurassic age.

In 1977, Carroll and Galton announced the discovery of a ‘modern' type of lizard from the Late
Triassic of China. The specimen formed part of the collections of the Catholic University of Peking
(CUP), which are now housed in the Field Museum of Natural History, Chicago, USA. The
specimen, CUP 2037, was originally catalogued as a juvenile of the prosauropod dinosaur
Yunnanosaurus huangi by Simmons (1965, p. 63) but Carroll and Galton (1977) reinterpreted it as
the skull of a lizard and renamed it Fulengia youngi.

Prior to the description of Fulengia , the earliest true lizards were those of the Upper Jurassic of
Europe, America, and China (see Estes 1983, for a full list). The Upper Triassic kuehneosaurs,
considered by some authors (Robinson 1962, 1967; Colbert 1966, 1970; Carroll 1975) as true
lizards, have more recently been regarded as an independent radiation (Hoffstetter 1962, 1967;
Kluge 1967; Evans 1980, 1984, 1988; Gauthier 1984; Benton 1985). Likewise, the Upper Permian
genera described by Carroll (1975) have subsequently been argued to lack the character-states
diagnostic of lizards (Evans 1980, 1984, 1988; Benton 1985). All of these genera retain primitive
diapsid character-states including: paired median skull-roofing elements; complex palatal dentition;
toothed parasphenoid; large lacrimal; simple subpleurodont teeth; and amphicoelous vertebrae.
The genus Fulengia , by contrast, was described as having several derived lacertilian character-
states including: fused median roofing bones; small lacrimal; serrated pleurodont teeth, procoelous
vertebrae; and, in reconstruction, a temporal region closely resembling that of a modern iguanid
lizard. This combination of derived character-states and early geological age made the specimen
of great potential significance to those studying the evolution of squamates. However, having
examined the holotype and other associated specimens from the same assemblage, we conclude
that the original interpretation (Simmons 1965) was more nearly correct and that Fulengia is the
skull of a juvenile prosauropod dinosaur

LOCALITY AND HORIZON

The specimen was recovered from the Deep Red Sequence of the Lower Lufeng or Fengjiahe
Formation (previously the Lower Lufeng Series), at TaTi in Yunnan Province. This horizon was
originally interpreted as late Upper Triassic (Young 1946, 1951; Simmons 1965) but there is a
recent consensus amongst Chinese workers that it is Lower Jurassic in age (Chen et al. 1982; Sun
et al. 1985; Sun and Cui 1986). Cooper (1982) comes to the same conclusion from faunal evidence.

IPalaeontology, Vol. 32, Part 1, 1989, pp. 223-230.|

© The Palaeontological Association

224

PALAEONTOLOGY, VOLUME 32

1 cm

H

text-fig. 1. ‘ Fulengia young?, CUP 2037, holotype, in a, left dorsolateral view, b, right
ventrolateral view, c, ventral view. Abbreviations: A, articular; An, angular; D, dentary; F,
frontal; H, hyoid; J, jugal; L, left; Mf, mandibular fenestra; Mx, maxilla; N, nasal; P, parietal;
Pm, premaxilla; Po, postorbital; Q, quadrate; R, right; San, surangular; V, vertebra.

HOLOTYPE SPECIMEN (CUP 2037)

The holotype of Fulengia is a small, very mineralized nodule. The specimen comprises a small skull, just
under 4 cm long, with a single associated vertebra. Parts of the skull roof, jaws, antorbital, and temporal
regions are preserved. In many places, bone junctions are very difficult to identify, particularly where already
fragmented bones have been superimposed. No further preparation has been possible. The general outlines
of the specimen, as figured by Carroll and Galton (1977, fig. 1) are correct, but we disagree with some aspects
of their interpretation (ibid. figs. 1 and 2) with respect to the identification of bones and the position of
suture lines (text-fig. 1 ).

Carroll and Galton (1977) gave six characters which support the hypothesis that Fulengia is a lizard:

a , pleurodont teeth with iguanid-like serrations, suggesting an early herbivorous specialization;

b, small lacrimal;

c, absence of a posterior jugal process;

d, configuration of the bones in the temporal region, and the expanded quadrate;

e , median frontal and parietal;

/, association of the above characters with a procoelous vertebra.

These characters bear re-examination:

a. Serrated, pleurodont teeth. The tooth crowns are small, spatulate, and finely serrated, with yellowish
enamel. The bases are long and smoothly rounded, as seen where the lateral wall of the maxilla has been
broken away on the right side (text-fig. 2a), This is unusual for pleurodont teeth. Iguanid teeth detached
from the jaw have a flattened labial surface which is eroded where the tooth contacts the bone (text-fig. 2b,
d; Edmund 1969). The smoothly rounded bases accord better with a thecodont implantation (text-fig. 2c)
and the spatulate crowns of Fulengia resemble most closely those of small contemporary prosauropods (see
below).

b. Small lacrimal. As interpreted by Carroll and Galton (1977, fig. la), Fulengia differs from Kuehneosaurus
and other primitive reptiles in having a reduced lacrimal like that of modern lizards and early sphenodontids.
However, their proposed suture line between the lacrimal and maxilla does not exist. The ‘small lacrimal’ is.

EVANS AND MILNER: LIZARD REINTERPRETED AS DINOSAUR

225

text-fig. 2. Dentition and tooth implantation, a, maxillary teeth of ‘ Fulengia , CUP
2037, in lateral (labial) view; b, maxillary tooth of Iguana in labial view; c, cross-section
of the thecodont tooth of a crocodile, showing a mature tooth being reabsorbed by a
replacement (simplified from Romer 1956, fig. 206 b)\ d, cross-section of the pleurodont
tooth of an iguanid lizard, Ctenosaura (redrawn from Edmund 1969, fig. 10e).
Abbreviations: ca, cavity; co, area of contact with jaw; cr, crown; la, labial wall; re,

replacement tooth.

in fact, the narrow ascending process of the maxilla (text-fig. 1a) which separates the deep anterior region,
shown by the long roots and deep premaxilla (text-fig. 1a), from the shallower posterior section. The shape
of the maxilla thus resembles that of the Lufeng prosauropods (text-fig. 4a, b) more closely than that of
lizards.

c. Absence of a posterior jugal process. The body of the left jugal is separated from the maxilla by a straight
suture (text-fig. 1a). The telescoping of the skull has left the postorbital region unnaturally short and almost
certainly carried the jugal further forward than its original position. Dorsally, the jugal is extended into a
postorbital process, but ventrally it ends abruptly. Carroll and Gabon (1977) interpret this as natural and
reconstruct the jugal with a smooth posterior border, but the edge is broken, not smooth. Immediately behind
the break, there is a bone fragment which we interpret as the base of a quadratojugal process.

d. Configuration of the bones in the temporal region , and the flared quadrate. The elements identified by
Carroll and Galton (1977) as squamosal and supratemporal are bone fragments which, in a telescoped skull,
cannot be identified with any assurance. The most obvious bones are the postorbital and the quadrate. The
quadrate appears to be short and flared in a manner more closely resembling that of a lepidosaur than an
archosaur. However, the dorsal head is obscured by the postorbital and matrix and the proportions may not
be as they appear, particularly as a second specimen, CUP 2038fi (text-fig. 3b), has a quadrate which is
similar at its ventral end but more elongate dorsally.

e. Median frontal and parietal. Carroll and Galton (1977, fig. 1 a, b ) identify two superimposed plates of
bone as the left and right halves of a median frontal bone, but they appear to be separate left and right
ossifications (text-fig. 1a, b). The right has a straight medial edge. The parietal region is too distorted for
accurate interpretation.

f. Procoelous vertebrae. A single elongated vertebral centrum is preserved in association with the holotype
skull (text-fig. 1a-c). One end, to the left, is convex with a small central pit; the other end, to the right, is
slightly concave with a central pit. The ventral surface is lightly keeled. There is no neural arch and no
evidence to determine which end is anterior and which posterior. The vertebra could as easily be opisthocoelous
as procoelous, and, in fact, resembles the axis of Yunnanosaurus robustus as figured by Young (1951, pi. 7)
(text-fig. 4c).

None of the lizard-like features of CUP 2037 can be confirmed by us. There is no reduced
lacrimal, the jugal may possess a posterior process, and the frontals are paired. The temporal
region and the parietal(s) are too crushed for interpretation, and the quadrate is incompletely
exposed. The dentition and the vertebra are equally or more consistent with identity as a tiny
prosauropod dinosaur. This alternative is explored after discussion of CUP 2038.

226

PALAEONTOLOGY, VOLUME 32

text-fig. 3. Lufengosaurus huenei , CUP 2038. a, CUP 2038a; b, c, reversed sides of
CUP 2038 b. (NB, CUP 2038 comprises two distinct specimens, here designated a and
b for ease of references.) Abbreviations: D, dentary (larger individual); Ds, dentary
(smaller individual); eo, possible exoccipital; H, hyoid; Mx, maxilla; Q, quadrate; V,

vertebra.

ADDITIONAL MATERIAL (CUP 2038)

A search through the CUP Lufeng collection yielded two small nodules from TaTi, collected at the same
time as the Fulengia holotype and placed together under the next catalogue number (2038). Like the holotype,
they are catalogued as Yunnanosaurus huangi. The specimens resemble Fulengia in their size and general
preservation (creamish-white and highly mineralized), and may be part of the same accumulation, but the
bones are dissociated and represent more than one individual (text-fig. 3). Simmons (1965) suggests that such
nodules are coprolitic in origin, but this is uncertain. Much of CUP 2038 is very difficult to interpret. CUP
2038# (text-fig. 3a) is a jumble of bone fragments, amongst which only a vertebra and a hyoid element
(?ceratobranchial) can be identified with any assurance. CUP 20386 is better and includes a quadrate, a
possible exoccipital, maxillary fragments, and the dentaries of two different-sized individuals (D and Ds, text-
fig. 3b, c). The teeth are identical to those of Fulengia , but the jaw fragments show them to be thecodont,
as inferred from the holotype. The ventral part of the quadrate is of similar proportions to that of Fulengia ,
but the main body is taller (text-fig. 3b), supporting the interpretation that the quadrate of the holotype is
partially obscured.

SYSTEMATIC POSITION AND RECONSTRUCTION

Several aspects of the structure of CUP 2037 identify it as a prosauropod dinosaur.

a. Possession of an antorbital fenestra close to the naris. As described above, the maxilla of CUP
2037 is a long straight bone with a narrow ascending process separating the deep anterior region
from the shallow posterior ramus (text-fig. 1a). The posterior border of the ascending process is
depressed like that of material referred to Gyposaurus (text-fig. 4a, b). There the depression
continues on to the surface of the posterior ramus (covered by the jugal in CUP 2037) and borders
the antorbital fenestra. The antorbital fenestra is a character of the Archosauria including the
Proterosuchidae (Benton 1985, p. 125), and an antorbital fenestra positioned close to the naris is
a character of the Archosauria excluding the Proterosuchidae (Benton 1985, p. 126).

b. Thecodont teeth. Within the sauropsid amniotes, this is a character of the Archosauria. The
smooth exposed roots of CUP 2037 resemble those of thecodont teeth (text-fig. 2c), not the

EVANS AND MILNER: LIZARD REINTERPRETED AS DINOSAUR

227

text-fig. 4. a, right maxilla of the prosauropod dinosaur Lufengosaurus ( Gyposaurus ) huenei,
CUP 148-4-2006a; b, left maxilla and dentary of Lufengosaurus ( Gyposaurus ) huenei , CUP 148-
4-2006a; c, axis vertebra of the prosauropod dinosaur Yunnanosaurus robustus (redrawn from
Young 1951 , pi. 7). (NB, CUP 148-4-2006 comprises five parts representing at least two individuals
of different age; CUP 148-4-2006a is part of the smaller individual.) Abbreviation: od, odontoid.

attachment faces of the roots of pleurodont teeth (text-fig. 2b, d). CUP 2038 has identical teeth
which are certainly thecodont.

c. Mandibular fenestra. Carroll and Galton (1977) noted that the massive jaws of Fulengia
distinguish it from known lizards; they are more consistent with the deep jaws of prosauropods.
In their figures (1977, lc, e), Carroll and Galton depict an opening in each mandible at the junction
between the dentary, angular and surangular, but do not show such an opening in the reconstruction.
These openings are present on both mandibles (text-fig. lc) and match the structure and position
of the small mandibular fenestrae of Lufengosaurus (Young 1941a). The presence of such mandibular
fenestrae is a character of the Archosauria (excluding the Proterosuchidae) (Benton 1985, p. 126).

d. Dentition. The spatulate tooth crowns with pointed tips and up to eight serrations per side
most closely resemble those of anchisaurid prosauropods (Galton 1985a, fig. 5d-h; text-fig. 4).
The length of the teeth is also characteristic. Among the herbivorous archosaurs, such deep rooted
anterior maxillary teeth are most consistent with the condition found in sauropodomorphs
(sauropods and prosauropods).

e. Hyoid skeleton. The robust hyoid elements of CUP 2037 (text-fig. 1b), identified as
ceratobranchials by Carroll and Galton (1977, fig. 1), are relatively large for a lizard but correspond
to the large ceratobranchials of prosauropods such as Plateosaurus (Galton 19856).

The dominant sauropsids of the Lufeng Formation are a group of prosauropod dinosaurs:
Lufengosaurus huenei (Young 1941a); Gyposaurus sinensis (Young 19416); Yunnanosaurus huangi
(Young 1942); Lufengosaurus magnus (Young 1947); and Yunnanosaurus robustus (Young 1951).
All five have been recorded from TaTi (Simmons 1965). They form a size series from the tiny
Gyposaurus sinensis to the large Lufengosaurus magnus. A detailed analysis of the Lufeng material
led Rozhdestvensky (1965) to conclude that they form an ontogenetic series of a single species, a
view supported by Galton (1976) and Galton and Cluver (1976). The senior name for this species
is Lufengosaurus huenei and it is placed in the family Anchisauridae (Galton 1985a). Cooper (1981)
concluded that the Lufeng anchisaurids could be accommodated within the contemporaneous
genus Massospondylus but has not been followed in this by Galton (1985a) who retained the generic
name Lufengosaurus. Galton (1985a), however, split the Lufeng prosauropods into two groups on
the basis of dental morphology. Lufengosaurus (including Gyposaurus) was placed in the

228

PALAEONTOLOGY, VOLUME 32

Anchisauridae (defined by possession of denticulate teeth without wear facets). Yunnanosaurus
formed the basis of a new family Yunnanosauridae (teeth with wear facets but no denticles).

More recently, Z. Yang (=C. C. Young, 1982a, 6) has described some tiny dinosaur jaw
fragments from the Lufeng. He placed Tawasaurus minor from Heiguopeng (Yang 1982a) in
the Fabrosauridae and Dianchongosaurus lufengensis from Zhangjiawa (Yang 19826) in the
Heterodontosauridae. However, Dong (pers. comm, to Sun et al. 1985) is sceptical about Yang’s
attribution of this material to the Ornithischia and we would agree with him. Reference to Yang’s
figures (1982a, pis. 1 and 2; 19826, fig. 2) suggests that both specimens have small serrated teeth
which bear a close resemblance to those of Lufengosaurus ( Gyposaurus ) and Fulengia. It may be
that both of Yang’s specimens are juvenile prosauropods. The two localities are less than one
kilometre from TaTi. Further discussion of the relationships of these specimens is beyond the
scope of this work and has no direct bearing on the conclusions outlined below.

text-fig. 5. a, reconstruction of the skull of “Fulengia , CUP 2037, in lateral view; b, lateral view
of the skull of the prosauropod Lufengosaurus huenei (from Young 1942, fig. 4, and Galton
1985a, fig. 4c); c, lateral view of the skull of the prosauropod Massospondylus (simplified from

Cooper 1981, fig. la).

There are no unambiguous character-states supporting the lizard status of Fulengia and it may
be noted that no further supposed lizards have been identified in the Lufeng fauna (Sun et al.
1985). The presence of thecodont teeth, in conjunction with mandibular and antorbital fenestrae,
supports the hypothesis that Fulengia is an archosaur. In its dental characteristics and in the
structure of the maxilla, jugal, and mandible, Fulengia most closely resembles the Lufeng
prosauropod material referred to Lufengosaurus and its junior synonym Gyposaurus. Text-fig. 5a
shows a new reconstruction of CUP 2037 as a prosauropod, in comparison with Lufengosaurus
(text-fig. 5b) and Massospondylus (text-fig. 5c). It is easily accommodated in Rozhdestvensky’s
(1965) ontogenetic series and represents a stage slightly younger than that of the smallest specimens
of Gyposaurus sinensis. We therefore formally propose Fulengia youngi to be a junior synonym of
Lufengosaurus huenei. CUP 2038 (in particular CUP 20386), formerly catalogued as Yunnanosaurus
huangi , belongs here also. As noted above, Cooper (1981) has proposed that the genus Lufengosaurus
is a junior synonym of Massospondylus. This is a problem beyond the scope of the present paper
and does not affect our conclusion with respect to Fulengia.

With the reinterpretation of Fulengia as a prosauropod, the earliest described lepidosaurs which
can be referred unequivocally to the Squamata are from the Upper Jurassic of Europe, North
America, and Asia (Evans 1984, 1988; Benton 1985). The earliest known lizard from the People’s
Republic of China is Yabeinosaurus from the Upper Jurassic deposits of Tsaotzushan and Liaoning
(Estes 1983).

Acknowledgements. We express our thanks to the staff of the Field Museum of Natural History, Chicago, in
particular John Bolt and Mary Carman, for access to Lufeng material and to John Attridge for his helpful
comments on the manuscript. This work was partly funded by a grant from the Central Research Fund,
University of London.

EVANS AND MILNER: LIZARD REINTERPRETED AS DINOSAUR

229

REFERENCES

benton, m. j. 1985. Classification and phylogeny of the diapsid reptiles. Zool. J. Linn. Soc. 84, 97 164.
carroll, r. L. 1975. Permo-Triassic ‘lizards' from the Karroo. Palaeont. afr. 18, 71 87.

— and galton, p. M. 1977. ‘Modern’ lizard from the Upper Triassic of China. Nature , Loud. 266, 252-
255.

Chen, p., li, w., chen, J., ye, c., wang, z., shen, y. and sun, s. 1982. Stratigraphical classification of Jurassic
and Cretaceous in China. Scientia sin., B. 25, 1227 1248.
colbert, e. h. 1966. A gliding reptile from the Triassic of New Jersey. Am. Mus. Novit. 2246, 1-23.

1970. The Triassic gliding reptile Icarosaurus. Bull. Am. Mus. not. Hist. 143, 87 142.

cooper, m. r. 1981. The prosauropod dinosaur Massospondylus carinatus Owen from Zimbabwe: its biology,
mode of life and phylogenetic significance. Occas. Pap. natn. Mus. Rhodesia ( B ), nat. Sci. 6, 689-840.

— 1982. A mid-Permian to earliest Jurassic tetrapod biostratigraphy and its significance. Arnoldia Zimbabwe ,
9, 77 104.

edmund, a. G. 1969. Dentition. In gans, c. and parsons, t. s. (eds.). Biology of the Reptilia. 1, 117-200.
Academic Press, London.

estes, R. 1983. Sauria terrestria; Amphisbaenia. Handbuch dev Paldoherpetologie 10A, 1-249. Gustav Fischer
Verlag, Stuttgart.

evans, s. e. 1980. The skull of a new eosuchian reptile from the Lower Jurassic of South Wales. Zool. J.
Linn. Soc. 70, 203-264.

1984. The classification of the Lepidosauria. Ibid. 82, 87-100.

— 1988. The early history and relationships of the Diapsida. In benton, m. j. (ed.). The Phylogeny and
Classification of Tetrapods. Vol. 1, Amphibians, Reptiles and Birds. Syst. Ass. Spec. Vol. 35A, 221 -253.
Clarendon Press, Oxford.

galton, p. m. 1976. Prosauropod dinosaurs (Reptilia: Saurischia) of North America. Postilla , 169, 1 98.

— 1 985a. Diet of prosauropod dinosaurs from the late Triassic and early Jurassic. Lethaia , 18, 105 123.

- 19857). Cranial anatomy of the prosauropod dinosaur Plateosaurus from the Knollenmergel (Middle
Keuper), Upper Triassic of Germany. Geologica Palaeont. 19, 1 19 159.

— and cluver, m. 1976. Anchisaurus capensis (Broom) and a revision of the Anchisauridae (Reptilia:
Saurischia). Ann. S. Afr. Mus. 69, 121-159.

Gauthier, J. A. 1984. A cladistic analysis of the higher systematic categories of the Diapsida. Ph.D. thesis
(unpublished). University of California, Berkeley.

hoffstetter, r. 1962. Revue des recentes acquisitions concernant l’histoire et la systematique des squamates.
Coll, intern. CNRS, 104, 243-285.

— 1967. Coup d’oeil sur les Sauriens ( = Lacertiliens) des couches de Purbeck (Jurassique superieur
d'Angleterre). Ibid. 163, 349-371.

kluge, a. g. 1967. Higher taxonomic categories of gekkonid lizards and their evolution. Bull. Am. Mus. nat.
Hist. 135, 1-60.

robinson, p. l. 1982. Gliding lizards from the Upper Keuper of Great Britain. Proc. geol. Soc. Lond. 1601,
137- 146.

1967. The evolution of the Lacertilia. Coll, intern. CNRS , 163, 395-407.

romer, a. s. 1956. Osteology of the Reptiles. University of Chicago Press, Chicago.

Rozhdestvensky, a. k. 1965. Growth changes in Asian dinosaurs and some problems of their taxonomy.
Int. geol. Rev. 8 (1966), 792-793.

simmons, D. j., 1965. The non-therapsid reptiles of the Lufeng Basin, Yunnan, China. Fieldiana , Geol. 15, 1
93.

sun, a. and cui, K. 1986. A brief introduction to the Lower Lufeng saurischian fauna (Lower Jurassic:
Lufeng, Yunnan, People’s Republic of China). In padian, k. (ed.). The Beginning of the Age of Dinosaurs,
275-278. Cambridge University Press, Cambridge.

— li, y. and wu, x. 1985. A verified list of the Lufeng Saurischian fauna. Vertebr. palasiat. 23,

1-11.

yang, z. 1982a. A new ornithopod from Lufeng, Yunnan. Selected works of Yang, Z., 29-35. Science Press,
Beijing. [In Chinese.]

— 1982 b. On a new genus of dinosaur from Lufeng, Yunnan. Ibid. 38 42. Science Press, Beijing. [In
Chinese.]

young, G. c. 1941a. A complete osteology of Lufengosaurus huenei Young (gen. et sp. nov.) from Lufeng,
Yunnan, China. Paleont. sin. (ns C), 7, I -53.

230

PALAEONTOLOGY, VOLUME 32

young, G. c. 1941/^. Gyposaurus sinensis Young (sp. nov.), a new Prosauropoda from the Upper Triassic
Beds at Lufeng. Bull. geol. Soc. China , 21, 205-252.

— 1942. Yunnanosaurus huangi Young (gen. et sp. nov.), a new Prosauropoda from the Red Beds at Lufeng.
Ibid. 22, 63-104.

1946. The Triassic vertebrate remains of China. Am. Mus. Novit. 1324, 1 14.

- 1947. On Lufengosaurus magnus Young (sp. nov. ) and additional finds of Lufengosaurus huenei Young.
Paleontol. sin. (NS C), 12, 1 53.

1951. The Lufeng saurischian fauna in China. Ibid. 13, 1-96.

SUSAN E. EVANS

Department of Anatomy and Developmental Biology
University College London, Windeyer Building
Cleveland Street, London W1P 6BD

Typescript received 28 March 1988
Revised typescript received 12 May 1988

ANDREW R. MILNER

Department of Biology
Birkbeck College

Malet Street, London WC1E 7HX

A PSILOCERATID AMMONITE FROM THE
SUPPOSED TRIASSIC PENARTH GROUP OF
AVON, ENGLAND

by D. T. DONOVAN, M. T. CURTIS Cind S. A. CURTIS

Abstract. An ammonite from the Penarth Group (Upper Triassic?) at Chipping Sodbury, Avon, England,
is described. This is believed to be the first ammonite to be reported from these beds or from any rocks in
Britain currently regarded as Triassic in age. It is small, but its characters are similar to those of Jurassic
psiloceratids rather than to those of late Triassic families.

Beds of the Penarth Group form the overburden to Carboniferous Limestone in large quarries to
the north of Chipping Sodbury. Sections were described by Reynolds and Vaughan (1904),
Reynolds (1938), and Curtis (1981). Recently, removal of overburden at Hampstead Farm Quarry
exposed good temporary sections. The Penarth Group here comprises two units, the Westbury
Formation below and the Cotham Member of the Lilstock Formation above, the latter extending
to the surface. The section measured was as follows:

(Ground surface)

metres

Cotham Member

Cotham Marble: nodules of algal limestone

Bufif calcareous mudstone with impersistent thin lime-

stone beds

1 86

Westbury Formation

Impersistent limestone with channelled base, passing

into laminated mudstone

008-0-35

Grey and black blocky shales

0-80

Crystalline limestone, impersistent

0-0-23

Black, grey mottled, blocky shales

0-55

Black fissile to blocky shales

0-80

Black laminated shales

0-40

Conglomeratic bone bed

0-0-20

Carboniferous Limestone

The total thickness of the Penarth Group is about 5 m. Thicknesses of individual beds are averages
as some are variable.

The uppermost bed of the Westbury Formation has yielded an abundant and varied vertebrate
fauna, and invertebrates of several groups including well-preserved ostracods and echinoid spines.
The ammonite was found by M.T.C. and S.A.C. in this uppermost bed at ST 726839. It has been
deposited in the City of Bristol Museum and Art Gallery with the registration number BRSMG
Ce 9715.

MODE OF STUDY

On account of its small size, the ammonite was photographed with a scanning electron microscope
using backscattered electrons (Taylor 1986) to avoid the need to give the presently unique specimen
a metallic coating. This did not show up the suture lines well and these were drawn under ordinary
light with a camera lucida and a low power binocular microscope.

|Palaeontology, Vol. 32, Part 1, 1989, pp. 231 -235.|

© The Palaeontological Association

232

PALAEONTOLOGY, VOLUME 32

DESCRIPTION

The fossil is an internal mould in ‘limonite’ which may replace pyrite, suggesting origin in an
anaerobic sediment. The specimen is 3-8 mm in diameter and consists of three and three-quarter
whorls (text-fig. 1a, b). It is an evolute form, each whorl overlapping about one-quarter of the
preceding one. The umbilicus is about 40% of the diameter. The whorl thickness and whorl height
both measure 1-2 mm, so that the whorl section is roughly circular except for the impressed area;
due to the small size of the fossil it was not possible to draw the whorl section accurately. The
internal mould is smooth showing no ornament. The protoconch, whose visible part is about
80 pm in diameter, is seen on the left side (text-fig. lc, d). The body-chamber comprises just
over half a whorl and is probably complete. Several suture lines are visible including the last
(text-fig. 1e). They are asymmetrical, the centre of the median saddle lying to the left of the mid-
line of the venter. The saddles are therefore broader on the right side than on the left.

The presence of the body-chamber shows that the fossil is not the nucleus of a larger ammonite.
There is no clear indication that the shell is mature; it could be the adult of a very small species,
or the young of a larger form.

COMPARISON

In view of its stratigraphical position, in beds currently classed as late Triassic (Warrington et al.
1980), comparisons have been made with latest Triassic and earliest Jurassic ammonites. The term
Rhaetian is used here for the latest Trias without prejudice as to whether it should be used for the
uppermost stage or substage of the Trias (Wiedmann et al. 1979; Tozer 1980, 1981).

The small size of the fossil limits the characters available for identification: these are the suture
line, shell form, and shell ornament (or rather lack of it). The suture line is indistinguishable from
the early sutures of the Hettangian genus Psiloceras, and also from several other early Hettangian
genera placed in the Family Psiloceratidae (Schindewolf 1962). The sutures of most Rhaetian
ammonites (Wiedmann 1972; Tozer 1979) are not similar: in these the external saddle is typically
longer and nearly parallel-sided. Furthermore, members of the Choristoceratidae, the last family
to die out in the Alpine Trias, have only two to two-and-a-half saddles in the external suture line,
compared to three in the present form. Asymmetry of the suture is common in, though not
diagnostic of, the Psiloceratidae, but rare or absent from late Triassic genera.

The Triassic Phylloceratina ( Rhacophyllites , Eopsiloceras ), believed ancestral to Psiloceratidae
(Tozer 1971; Guex 1987), with smooth, relatively evolute shells, are the Triassic ammonites most
nearly similar to the present form. The suture line of Rhacophyllites has subdivided lobes and
phylloid saddles at a whorl height of about 1 mm, as in the form described as Phyllytoceras
zlambachense Wiedmann (1972, p. 584, text-fig. Mb). The genus Phyllytoceras was proposed by
Wiedmann (1970) with the type species, P. intermedium , believed to be from the Karnian (Triassic)
of Iran. Whatever the real identity of the type species, P. zlambachense has been reinterpreted as
Rhacophyllites sp. (Krystyn 1974, p. 142). The suture line of Eopsiloceras is not known at a size
comparable to that of the new ammonite but, at larger sizes, it is more differentiated than in
comparable Rhacophyllites , and is therefore probably so on the inner whorls (Leo Krystyn, pers.
comm.).

Many ammonites are smooth to diameters greater than 4 mm, so that the absence of shell
ornament in the present form is of limited value for comparison. However, in the Choristoceratidae,
which include the youngest Rhaetian ammonites, ornament is already present at a diameter of less
than 4 mm (Tozer 1979). The straight-shelled member of this family, Rhabdoceras , has coiled
innermost whorls, but already shows non-planispiral coiling at a size comparable to that of the
new form (Tozer 1979; Wiedmann 1972, pi. 1, fig. 5).

A new genus and species Primapsiloceras primulum was recognized by Repin (in Polubotko and
Repin 1981) from beds said to lie below the Planorbis Zone of north-east Siberia. It is ribbed from
a diameter of at least 3 mm and is thus not comparable with the new find. Its suture line at this

DONOVAN ET AL PSILOCERATID AMMONITE FROM AVON

233

text-fig. 1. Psiloceratid ammonite from the Westbury
Beds, Hampstead Farm Quarry, Chipping Sodbury,
Avon. Bristol Museum no. Ce 9715. a, b, left and right
sides, x 18. c, umbilical region of left side of same
specimen, showing protoconch, x 75. d, same, x 225. e,
last three suture lines of the specimen shown in a-d,
x 30, at whorl height 0-9 mm.

234

PALAEONTOLOGY, VOLUME 32

size has not been illustrated. Guex (1987, text-fig. 1) shows it as an early derivative of Psiloceras
and believes it to be of early Hettangian age (op. cit., p. 459).

We conclude that, when compared with the stratigraphically nearest forms, the specimen is
indistinguishable from the inner whorls of Psiloceras and is to be placed in Psiloceratidae. On
account of its small size we do not assign it to a genus or a species.

Tozer (1971, 1981) proposed that a single ammonite lineage connected late Triassic and earliest
Jurassic forms. The only possible Triassic ancestors of this lineage are the Discophyllitidae,
which differentiated around the Triassic- Jurassic boundary into Juraphyllitidae+ Phylloceratidae,
retaining many discophyllitid features, and the Psiloceratidae with, in general, simpler suture lines
and more evolute shells (Guex 1987). The present find shows that psiloceratid characters had
already appeared in Britain at the horizon represented by the top of the Westbury Formation.

STRATIGRAPHY

The Westbury Formation has long been correlated with the late Triassic ‘Rhatische Stufe’ of the
eastern Alps (in modern terms the Kossener Schichten) on the basis of the bivalve fauna, especially
Rhaetavicula contorta (Portlock), common to both units. The history of subdivision and correlation
at this level has been related by Pearson (1970). Moore (1861, p. 487), writing about southern
England, recognized a Rhaetic Formation comprising Avicula contorta beds below and White Lias
above. In the Geological Society of London’s Triassic correlation chart Warrington et al. (1980,
pp. 18, 54) were cautious, noting that British late Triassic bivalves are principally those of black
shale facies and ‘of limited use in correlation with the more calcareous sequences of the Standard
[i.e. “the Alpine region’’]’ (p. 18) and again that the macrofossils of the Penarth Group ‘though
generally indicative of a late Triassic age, do not permit direct correlation with the Standard
Sequence’ (p. 54). They preferred to abandon the term ‘Rhaetic’, used in Britain since Moore
(1861) wrote, in favour of the term Penarth Group (op. cit., p. 13) of which the Westbury
Lormation is the lowest component.

The authors of the Triassic correlation chart (Warrington et al. 1980, p. 10) adopted ‘the first
appearance of ammonites of the genus Psiloceras ’ to mark the base of the Planorbis Zone and
therefore of the Jurassic System in Britain. This was in agreement with the authors of the Jurassic
chart, in which Torrens and Getty (in Cope et al. 1980, p. 22) wrote ‘there is no evidence in Britain
for an earlier ammonite fauna than that with Psiloceras planorbis . . At that time (1980) the
first known appearance of Psiloceras was at an apparently constant horizon a few metres above
the base of the Lias, the lithostratigraphic unit above the Penarth beds.

The present find raises two related questions. The first is the correlation of the British sequence
with the Alpine. The latter does not show a continuous record of ammonites, the last Choristoceras
marshi of the Trias being succeeded by a barren gap before the first Psiloceras of the Jurassic in
the area south of the Wolfgangsee, Austria (L. Krystyn, pers. comm.). The new find may indicate
that the top of the Westbury Formation correlates with a higher horizon in the Alpine sequence
than had been thought.

The second question concerns the base of the Jurassic System in Britain. As the present find is
indistinguishable from Psiloceras , it could be claimed that it marks the ‘first appearance of . . .
Psiloceras ’ and therefore places the base of the Jurassic at the top of the Westbury Formation.
We do not do this, but propose discussion of the desirability of defining the base of the Planorbis
Zone in a stratotype section, which had not been done when Torrens and Getty wrote (in Cope
et al. 1980, p. 21) and has not been done since.

Acknowledgements. We thank P. D. Taylor for taking SEM pictures, R. J. G. Savage for the use of the
camera lucida, and E. T. Tozer and L. Krystyn for discussion. We also thank the management of ARC,
Chipping Sodbury Quarry, for access to the site.

DONOVAN ET AL.\ PSILOCERATID AMMONITE FROM AVON

235

REFERENCES

cope, j. c. w., getty, t. a., howarth, m. k., morton, N. and torrens, H. s. 1980. A correlation of Jurassic
rocks in the British Isles. Part One. Introduction and Lower Jurassic. Spec. Rep. Geol. Soc. Loud. 14, 1
73.

Curtis, m. t. 1981. The Rhaetic-Carboniferous Limestone unconformity at Southlields Quarry, Chipping
Sodbury, Avon. Proc. Bristol Nat. Soc. 40, 30 35.

guex, j. 1987. Sur la phylogenese des ammonites du Lias inferieur. Bull. Soc. vaud. Sci. nat. 78, 455-469.
krystyn, l. 1974. Probleme der biostratigraphischen Gliederung der Alpin-Mediterranen Obertrias. SchrReihe
Erdwiss. Komm. Oesterr. Akad. Wiss. 2, 137-144.

moore, c. 1861. On the zones of the Lower Lias and the Avicula contorta Zone. Q. Jl geol. Soc. Lond. 17,
483-516, pis. 15 and 16.

pearson, d. a. b. 1970. Problems of Rhaetian stratigraphy with special reference to the lower boundary of
the stage. Ibid. 126, 125- 150.

polubotko, i. v. and repin, yu. s. 1981. Identification of a new ammonite zone at the base of the Jurassic
System. Dokl. Akad. Nauk SSSR, 261, 1394-1398. [In Russian ]

Reynolds, s. h. 1938. A section of Rhaetic and associated strata at Chipping Sodbury, Glos. Geol. Mag. 75,
97 102, pi. 4.

— and vaughan, a. 1904. The Rhaetic Beds of the South Wales Direct Line. Q. J! geol. Soc. Lond. 60,
194-214, pi. 18.

schindewolf, o. h. 1962. Studien zur Stammesgeschichte der Ammoniten. Lieferung II. Akad. Wiss. Lit.

Mainz, Abh. Math.-nat. Kl. Jhrg. !%2, 11 1-257, pi. 3.
taylor, p. d. 1986. Scanning electron microscopy of uncoated fossils. Palaeontology , 29, 685-690, pi. 52.
tozer, E. t. 1971. One, two or three connecting links between Triassic and Jurassic ammonoids? Nature ,
Lond. 232, 565-566.

— 1979. Latest Triassic ammonoid faunas and biochronology, western Canada. Geol. Surv. Can. Pap.
79-1 B, 127 135, pi. 16.1.

— 1980. Latest Triassic (Upper Norian) ammonoid and Monotis faunas and correlations. Riv. ital.
Paleont. Stratigr. 85, 843-876, pi. 63.

— 1981. Triassic Ammonoidea: classification, evolution and relationship with Permian and Jurassic forms,
66-100. In house, m. r. and senior, j. r. (eds.). The Ammonoidea. Spec. Vol. Syst. Ass. 18, 1 593.

WARRINGTON, G., AUDLEY-CHARLES, M. G., ELLIOTT, R. E., EVANS, W. B., IVIMEY-COOK, H. C., KENT, P., ROBINSON,
p. l., shotton, f. w. and taylor, F. M. 1980. A correlation of Triassic rocks in the British Isles. Spec. Rep.
Geol. Soc. Lond. 13, 1-78.

wiedmann, j. 1970. Uber den Ursprung der Neoammonoideen Das Problem einer Typogenese. Eclog. geol.
Helv. 63, 923 1020, pis. 1 10.

— 1972. Ammoniten-Nuklei aus Schlammproben der nordalpinen Obertrias — ihre stammesgeschichtliche
und stratigraphische Bedeutung. Mitt. Ges. Geol. Bergbaustud. Innsbruck , 21, 561-622, pis. 1-6.

— fabricius, f., krystyn, l., reitner, L and urlichs, M. 1979. Uber Umfang und Stellung des Rhat.
Diskussionsbeitrag zur Sitzung der Internationalen Subkommission fur Trias-Stratigraphie in Munchen,
Juli 1978. New si. Stratigr. 8, 133 152.

D. T. DONOVAN

Department of Geology
University of Bristol
Queens Road
Bristol BS8 1 RJ, UK

M. T. CURTIS and S. A. CURTIS
2 Ribblcsdale, Thornbury
Bristol BS12 2DW, UK

Typescript received 12 May 1988
Revised typescript 12 July 1988

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 1985, 28, pp. 793-800).

Special Papers in Palaeontology is a series of substantial separate works conforming to the style of Palaeontology.

SPECIAL PAPERS IN PALAEONTOLOGY

In addition to publishing Palaeontology the Association also publishes Special Papers in Palaeontology. Members may
subscribe to this by writing to the Membership Treasurer: the subscription rate for 1989 is £37-50 (U.S. $66) for Institutional
Members, and £18-50 (U.S. $32) for Ordinary and Student Members. A single copy of each Special Paper is available to
Ordinary and Student Members only, for their personal use, at a discount of 25% below the listed prices. Special Papers
up to 3 years old are available to members via the appropriate Membership Treasurer. Non-members may obtain these
(at cover price) from Basil Blackwell, 108 Cowley Road, Oxford OX4 1JF, England. Older issues are available to Members
and Non-members from the Marketing Manager.

RECENT PALAEONTOLOGICAL ASSOCIATION

PUBLICATIONS

Special Papers in Palaeontology

Numbers 2-26 are still in print and are available (post free) together with those listed below:

27. (for 1981): Late Devonian Acritarchs from the Carnarvon Basin, Western Australia, by G. playford and r. s. dring.
78 pp., 10 text-figs., 19 plates. Price £15 (U.S. $23).

28. (for 1982): The Mammal Fauna of the Early Middle Pleistocene cavern infill site of Westbury-sub-Mendip, Somerset,
by m. j. bishop. 108 pp., 47 text-figs., 6 plates. Price £25 (U.S. $38).

29. (for 1982): Fossil Cichlid Fish of Africa, by J. A. H. van couvering. 103 pp., 35 text-figs ., 10 plates. Price £30 (U.S.
$45).

30. (for 1983): Trilobites and other early Arthropods. Edited by d. e. g. briggs and p. d. lane. 276 pp., 64 text-figs., 38
plates. Price £40 (U.S. $60).

31. (for 1984): Systematic palaeontology and stratigraphic distribution of ammonite faunas of the French Coniacian, by
w. j. Kennedy. 160 pp., 42 text-figs., 33 plates. Price £25 (U.S. $38).

32. (for 1984): Autecology of Silurian organisms. Edited by M. G. bassett and j. d. lawson. 295 pp., 75 text-figs., 13 plates.
Price £40 (U.S. $60).

33. (for 1985): Evolutionary Case Histories from the Fossil Record. Edited by j. c. w. cope and p. w. skelton. 202 pp.,
80 text-figs., 4 plates. Price £30 (U.S. $45).

34. (for 1985): Review of the upper Silurian and lower Devonian articulate brachiopods of Podolia, by o. i. Nikiforova,
t. l. modzalevskaya and m. g. bassett. 66 pp., 6 text-figs., 16 plates. Price £10 (U.S. $15).

35. (for 1986): Studies in palaeobotany and palynology in honour of N. F. Hughes. Edited by d. j. batten and d. e. g.
briggs. 178 pp., 29 plates. Price £30 (U.S. $50).

36. (for 1986): Campanian and Maastrichtian ammonites from northern Aquitaine, France, by w. j. Kennedy. 145 pp., 43
text-figs., 23 plates. Price £20 (U.S. $35).

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. $35).

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. $50).

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. $50).

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. $50).

Field Guides to Fossils

These are available from the Marketing Manager. Prices do not include postage. Reduced prices are available to Members.

1. (1983): Fossil Plants of the London Clay, by m. e. collinson. 121 pp., 242 text-figs. Price £7-95 (U.S. $12).

2. (1987): Fossils of the Chalk, compiled by E. owen; edited by a. b. smith. 306 pp., 59 plates. Price £1 1 50 (U.S. $18).

3. (1988): Zechstein Reef fossils and their palaeoecology, bv N. hollingworth and t. Pettigrew, iv + 75 pp.. Price £4-95
(U.S. $8).

Other Publications

Please contact the Marketing Manager; do not send payment in advance.

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $30).

1985. Atlas of Invertebrate Macrofossils. Edited by j. w. Murray. Published by Longman in collaboration with the
Palaeontological Association, xiii + 241 pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24.95.

© The Palaeontological Association, 1989

Palaeontology

VOLUME 32 • PART 1

CONTENTS

Amural arachnophyllid corals from the Silurian of the North
Atlantic area

C. T. SCRUTTON 1

Evaluation of a thecideidine brachiopod from the middle
Jurassic of the Cotswolds, England

P. G. BAKER 55

A new mitrate from the Upper Ordovician of Norway, and
a new approach to subdividing a plesion

A. J. CRASKE and R. P. S. JEFFERIES 69

A new camerate crinoid from the Arenig of South Wales

s. k. donovan and j. c. w. cope 101

Silurian trilobites from the Annascaul inlier, Dingle penin-
sula, Ireland

d. j. siveter 109

The composition and palaeogeographical significance of the
Ordovician ostracode faunas of southern Britain, Baltoscan-
dia, and Ibero-Armorica

J. M. C. VANNIER, D. J. SIVETER and R. E. L. SCHALLREUTER 163

Fulengia, a supposed early lizard reinterpreted as a prosauro-
pod dinosaur

s. E. evans and a. r. milner 223

As psiloceratid ammonite from the supposed Triassic Penarth
Group of Avon, England

D. T. DONOVAN, M. T. CURTIS and S. A. CURTIS 231

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

ISSN 0031-0239

H Palaeontology

VOLUME 32 • PART 2 JULY 1 989

Published by

The Palaeontological Association ■ London
Price £25-50

THE PALAEONTOLOGICAL ASSOCIATION

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

COUNCIL 1988 1989

President '. Dr J. D. Hudson, Department of Geology, University of Leicester, Leicester LEI 7RH
Vice-Presidents'. Dr L. B. Halstead, Department of Geology, University of Reading, Reading RG6 2AB
Dr P. W. Skelton, Department of Earth Sciences, The Open University, Milton Keynes MK7 6AA
Treasurer : Dr M. E. Collinson, Department of Biology, King’s College, London W8 7AH
Membership Treasurer. Dr H. A. Armstrong, Department of Geology, University of Newcastle,

Newcastle upon Tyne NE1 7RU

Institutional Membership Treasurer : Dr A. W. Owen, Department of Geology, University of Dundee,

Dundee DD1 4HN

Secretary: Dr P. Wallace. The Croft Barn, Church Street, East Hendred, Oxon 0X12 SLA
Circular Reporter: Dr D. Palmer, Department of Geology, Trinity College, Dublin 2
Marketing Manager : Dr V. P. Wright, Department of Geology, University of Bristol, Bristol BS8 1 RJ
Public Relations Officer : Dr M, J. Benton, Department of Geology, The Queen’s University of Belfast, Belfast BT5 6FB

Editors

Dr M. J. Benton, Department of Geology, The Queen’s University of Belfast, Belfast BT5 6FB
Dr J. E. Dalingwater, Department of Environmental Biology, University of Manchester, Manchester M13 9PL
Dr D. Edwards. Department of Geology, University of Wales College of Cardiff, Cardiff CF1 3YE
Dr C. R. C. Paul, Department of Earth Sciences, University of Liverpool, Liverpool L69 3BX
Dr P, A. Selden, Department of Extra-Mural Studies, University of Manchester, Manchester M13 9PL
Dr P. D. Taylor, Department of Palaeontology, British Museum (Natural History), London SW7 5BD

Other Members

Dr J. A. Crame, Cambridge Dr E. A. Jarzembowski, Brighton

Dr G. B. Curry, Glasgow Dr R. A. Spicer, London

Overseas Representatives

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

MEMBERSHIP

Membership is open to individuals and institutions on payment of the appropriate annual subscription. Rates for 1988 are:
Institutional membership ...... £50-00 (U.S. $79)

Ordinary membership £2L00 (U.S. $38)

Student membership ....... £1 L50 (U.S. $20)

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

There is no admission fee. Correspondence concerned with Institutional Membership should be addressed to Dr A. W. Owen,
Department of Geology, The University, Dundee DD1 4HN. Student members are persons receiving full-time instruction at
educational institutions recognized by the Council. On first applying for membership, an application form should be
obtained from the Membership Treasurer: Dr H. A. Armstrong, Department of Geology, University of Newcastle, Newcastle
upon Tyne NE1 7RU. Subscriptions cover one calendar year and are due each January; they should be sent to the
Membership Treasurer. All members who join for 1989 will receive Palaeontology , Volume 32, Parts 1-4. Enquiries
concerning back numbers should be directed to the appropriate Membership Treasurer.

Non-members may subscribe, and also obtain back issues up to 3 years old, at cover price through Basil Blackwell, Journals
Department, 108 Cowley Road, Oxford OX4 1JF, England. For issues older than 3 years contact the Marketing Manager.

Cover : Ammonite ( Grossouvria ) from the Oxford Clay (Jurassic) of Woodham, Bucks., containing geopetal pyrite stalactites.
Direct print from a thin section ; pyrite white. Diameter of specimen approximately 10 mm. (See Sedimentolog v, 29, 639-667,
1982.)

PALAEOTIDAL CHARACTERISTICS DETERMINED
BY MICRO-GROWTH PATTERNS IN BIVALVES

by TERUFUMI OHNO

Abstract. SEM studies of growth patterns in the ligament groove of late Pleistocene oysters and in
unidentified Miocene shell fragments have enabled the fossils’ position within the intertidal zone and the
tidal regime to be reconstructed. Whereas the present-day tidal regime in Osaka Bay has a strong diurnal
inequality seen especially in the heights of low water, 70 000 years ago it possessed only weak diurnal
inequality, as in the Miocene at another locality, where it was weakly discernible in the heights of high water.
The late Pleistocene borehole material represents a mixed assemblage for it includes individuals that lived at
various intertidal levels. The majority of the Miocene individuals lived at mean tide level. The arrangement
of growth lines of alternating thickness inverts every 2 weeks. Sequences are plotted with successive days on
the ordinate and the time of day on the abscissa. For the Recent this results in a vertical partition of thicker
and thinner growth lines. If the partition is not vertical when fossil growth lines are plotted, a change in the
number of days per month is indicated. The plots for the Miocene material indicate that the length of a
synodic month in terms of the number of synodic days of that time was essentially the same as the present.

In many invertebrate hard parts we can observe traces of macroscopic and microscopic growth
(Neville 1967; Clark 1974; Termier and Terntier 1975; Scrutton 1978; Chave and Erben 1979).
These are commonly called growth lines, increments, bands, rings or ridges. Sequences of these
features (referred to as ‘growth patterns’ in this paper) reflect the growth history of organisms as
well as the environmental factors influencing it.

Wells (1963) called attention to the usefulness of analysis of fine growth features. He counted
the number of the fine growth ridges between the major annulations on coral epilhecas and
assumed the latter formed annually and the former daily. His values were c. 400 per year for the
Devonian Period and c. 390 for the Carboniferous Period. They were in good agreement with the
estimates of days of the year obtained through present-day astronomical observations. Since then,
various palaeontological studies of fine growth patterns have been made.

Molluscan shells, especially bivalves, have well defined and easily discernible growth patterns.
They are also very abundant in aquatic environments, not only in present seas but also in the
geologic past. Therefore they have been intensively studied as suitable material for growth pattern
analysis.

In the early stage of growth pattern study, daily rhythm was thought to be ubiquitous and
responsible for making growth patterns irrespective of the habitats of bivalves, but a few workers
were conscious of the variability in shell growth rhythm. Le Gall ( 1970) suggested that tidal rhythm
is responsible for the formation of fine striations on the shell surface of Mytilus edulis and called
these striations strie maree. Evans (1972, 1975) demonstrated that individuals of the bivalve
Clinocardium nuttallii (Conrad) living in the intertidal zone form their shell with tidal rhythm. In
recent years data about the manner of the shell growth of intertidal bivalves have accumulated
(Richardson, Crisp and Runham 1979, 1980a, b , 1981; Richardson, Crisp, Runham and Gruffydd
1980; Ohno 1983, 1985; Deith 1983). Extensive experiments on living animals in these studies
clearly show that shells of intertidal bivalves grow with tidal rhythms and that the resulting patterns
reflect features of the tides and related phenomena. Tidal growth rhythm is also known in
gastropods (Antoine and Quemerais-Pencreac'h 1980; Ekarante and Crisp 1982; Ohno and
Takenouchi 1984) as well as in barnacles (Bourget and Crisp 1975; Crisp and Richardson 1975;
Bourget 1980).

IPalaeontology, Vol. 32, Part 2, 1989, pp. 237-263, pis. 31-32.|

© The Palaeontological Association

238

PALAEONTOLOGY, VOLUME 32

Apart from Deith's (1983) successful archaeological application of the tidal shell growth of
Cerastoderma edule (Linne) in estimating the seasonal shellfish collecting activity of a Mesolithic
site of southern Scotland, there are few works concerning tidal growth of prehistoric or fossil
bivalve shells. Berry and Barker (1975) interpreted recurring clustering of growth increments on
fossil bivalves as reflecting the change between neap and spring tides. Pannella (1976) showed a
single growth pattern in a Late Cretaceous bivalve shell Limopsis striatus-punctatus Evans and
Schumand, which he interpreted to be of tidal origin. Ohno (1984) made a short report on the
tidal growth rhythm on late Pleistocene bivalve shell fragments.

In this paper I will introduce a method of recognizing tidal shell growth patterns based on recent
experiments with living bivalves. Further I will demonstrate the existence of unequivocal tidal shell
growth patterns in fossil bivalves based on this method. I will also try to reconstruct the features
of tides of the biotopes in which fossil bivalves lived. Finally, an easy method of estimating the
change of the number of lunar days per synodic month based on the tidal growth pattern of fossil
bivalve shells will be proposed.

MATERIAL AND METHODS

Material examined

Fossil bivalve shells from two regions in Japan were examined in the present study.

One group were obtained from two boreholes from the basal part of a late Pleistocene marine clay layer
in the western part of Osaka City. The day is called the Mai 2 Clay by local geologists. (The abbreviation
Ma means marine.) Around Osaka more than ten cycles of marine clay beds and nonmarine clastic deposits
of late Pleistocene age are known. The former represent the transgressive phases and the latter regressive
phases of late Pleistocene eustatic sea-level changes. The basal part of the Mal2 Clay bed may have been
deposited in the early phase of a transgression. This clay bed is correlated with a clay bed at the site of the
off-shore international airport about 30 km south-south-west of the boreholes on the basis of pollen and
Foraminifera (Chiji 1984; Furutani 1984; Nakaseko et al. 1984). The Emiliania huxley acme-zone is found in
the Mai 2 Clay bed at the airport site (Okamura and Yamauchi 1984), and its age is estimated to be 0 07 my
(Gartner 1977).

Bivalve fossils were obtained from the Miocene Mizunami Group, exposed around Mizunami City, about
40 km north-east of Nagoya City. Itoigawa (1981) subdivided this group into four formations; the Toki
Lignite-bearing, the Hongo, the Akeyo, and the Oidawara Formations in upward sequence. The first two
are fresh water and the latter two marine deposits. The fossil shell samples examined in the present study
were collected at a locality called Takenami (loc. 23 of Itoigawa 1974) from an outcrop of a shell bed
considered to belong to the Akeyo Formation, assigned to N.8 of Blow’s zonation (Itoigawa 1981). The
molluscan assemblage of the bed indicates an embayment with a muddy bottom and an intertidal water
depth (Itoigawa et al. 1974).

All the samples illustrated here are stored in the Department of Geology and Mineralogy, Kyoto University,
Kyoto, Japan (registration no. JCTO-OOOl JCTO-0013).

Methods of observation

Growth patterns of fossil bivalve shells were studied using a scanning electron microscope (SEM) mainly
with BEI (back-scattered electron image) mode. This mode enhances surface topography of samples, and
thus is suitable for observation of microgrowth patterns.

Growth patterns in oyster shells are observable on the surface of the ligament groove. Fragments with
the grooves are mounted on the stage after cleaning with water in an ultrasonic bath for about 30 seconds
and then coated with gold. In other samples the growth patterns are observed along the cut sections of the
valves. If the samples are free of matrix they are embedded in plastic, then cut radially between the umbo
and the ventral margin of the shell to obtain the longest growth sequence possible. If they are in a
sedimentary matrix, the cut direction is not predeterminable. Thus rock samples containing shell fragments
are cut in an arbitrary direction and shell fragments suitable for the observation of growth patterns are
selected. The cut shell surfaces are ground with powders up to no. 3000, polished with diamond paste, and
then etched with 0T mol HC1 for about 10 to 30 seconds. Then they are coated with gold and examined
by SEM.

OHNO: BIVALVE MICRO-GROWTH PATTERNS

239

TYPES OF TIDES

To understand tidal shell growth patterns, it is necessary to be familiar with the different types of
tides.

Tide is a periodic rise and fall of water level of the sea caused by the gravitational forces of the
moon and sun on the water mass of the earth. The most familiar 12-4 hourly rise and fall of
the water is caused by the moon’s gravitational force, which is strengthened and weakened by the
sun’s gravitational force causing a 2-week change resulting in spring and neap tides. Obliqueness
of the moon’s orbit to the earth’s equator causes diurnal inequality of the tides. This is most clearly
expressed as the difference between the heights of the two succeeding high tides or of the two low
tides of a day. The periodicities of these oscillations, reflecting the celestial movements of the earth,
moon, and sun, are constant, but their amplitudes and phases are variable from place to place,
depending upon the depth of the ocean, shape of the shore line, etc. Thus the actual tides in the
world oceans varies from place to place.

Tides can be classified into several types according to the intensity of diurnal inequality. In the
present paper, they are divided into four types (text-fig. 1). If the diurnal inequality is weak, 12-4
hourly rise and fall of water occurs resulting in approximately two low tides and two high tides
per day (semidiurnal tide; text-fig. 1a). The increase of the diurnal inequality of the tides is most
clearly expressed in the increasing difference of the heights of two succeeding high tides as well as
of two succeeding lew tides (mixed tide). The difference is emphasized in the heights of low tides
(text-fig. 1b), or that of high tides (text-fig. lc). There are all possible intermediates between these
two extremes and the inequality may appear in heights of both low and high tides in one tidal
curve. Finally, when the diurnal inequality becomes sufficiently large, approximately only one high
tide and one low tide appear per day, with a periodicity of 24-8 hours (diurnal tide; text-fig. Id).

TIDAL GROWTH PATTERNS
Components of growth: growth line and growth increment

Growth of bivalve shells consists of two components, namely growth lines and growth increments.
They are most easily observed in sections vertical to the shell surface. Growth lines are thinner
shell layers, resistant against etching with diluted HC1 or other etching agents. Growth increments
are layers between growth lines. They are usually thicker than the growth lines and form the major
part of the shell. Similar components are also visible on the bivalve shell surface as well as on the
ligament area: the thinner ridges or grooves will be called growth lines and the thicker stripes
between them, growth increments, by analogy with the components seen on the cut surfaces.
Growth of bivalve shells results in the formation of a sequence of growth lines and growth
increments one beside another, which will be called growth patterns.

Intertidal growth patterns: previous studies

Studies of the correlation between bivalve shell growth patterns, tides, and related phenomena,
have been done with Cerastoderma edule (Linne) (Richardson, Crisp and Runham 1979, 1980a, h ,
1981; Richardson, Crisp, Runham and Gruffydd 1980; Ohno 1983, 1985; Deith 1983), Clinocardium
nuttallii (Conrad) (Evans 1972, 1975), and Fragum unedo (Linne) (Ohno 1985). The results of these
studies are summarized here (see also Table 1).

Tidal exposure and growth line formation

Growth line formation in intertidal bivalve shells has been most intensively studied in Cerastoderma
edule. This species lives along European coasts where tides are semidiurnal with weak diurnal
inequality. Individuals near the MTL (mean tide level) are exposed every 12-4 hours at low tide.
Richardson et al. (1979) and Ohno (1985) let their marked individuals grow in natural or simulated
tidal cycles. These shells formed growth lines in almost precise correspondence with the number

240

PALAEONTOLOGY, VOLUME 32

2

1,5

1

.5

0

-.5

-1

-1.5

-2

14. B DAYS

I 1

text-fig. 1. Four typical tidal types and possible arrangement patterns of tidal growth lines corresponding
to them. (The tidal height in arbitrary scale is shown on the left side of each diagram.) a, semidiurnal tide
with faint diurnal inequality, b, mixed tide with diurnal inequality strongly expressed in the heights of the
low tides, c, mixed tide with diurnal inequality strongly expressed in the heights of the high tides, d, diurnal
tide. Hypothetical growth line arrangement patterns are shown at three different levels (a, mean tide level;
b, between MTL and low water; c, near LW) for each tidal type. Three stripes represent parts of shell cut
parallel to the shell growth direction. Patterns are produced assuming that one tidal growth line is formed
at one low tide, but only when the bivalve shell is exposed above water. In the stripe for a specific water
level, one vertical line representing a tidal growth line is drawn under the corresponding low tide, when the
shell is exposed at low tide: ss, single spacing of tidal growth lines; ds, double spacing of tidal growth lines;
lb, irregular bundle (see Table 1 for definitions). (Possible variation in growth line thickness and increment
thickness as well as non-tidal growth lines in irregular bundles are not indicated in these hypothetical patterns,
in order to show clearly the relationship between growth lines and tidal exposure.)

of tidal cycles which they experienced during their growth after the marking. Deith (1983) collected
two sets of specimens of C. edule from a precise locality in an intertidal zone on two different
dates. The number of growth lines from the last winter to the growing edges was counted. The
difference of the means of growth lines between the two sets are well in correspondence with the
number of tides during the interval between the dates of collecting. This study also suggests that
one growth line is formed at each tidal cycle.

To determine when during each tidal cycle a growth line is formed, Richardson et al. (1981)
collected individuals of the species at approximately 1 hour intervals at a fixed point near the MTL
of an intertidal zone for two tidal cycles and compared the amount of shell growth beyond the
last growth line with the phase of the tide at the time of collecting. Their work demonstrates that

OHNO: BIVALVE MICRO-GROWTH PATTERNS

241

one growth line is formed at the end of each tidal exposure (growth line of tidal exposure origin
will be called tidal growth line; Table 1). Consequently, a growth increment which separates two
tidal growth lines is formed during the time between two tidal exposures, i.e. during inundation
at high tide.

Growth lines are also formed subtidally (these will be called non-tidal growth lines), when
bivalves are submerged continuously for more than one semidiurnal tidal cycle. Richardson el al.
(19806) claimed that these lines in C. edule are formed with endogenous tidal rhythms, although
the number of such growth lines formed during each of a series of their experiments varied
significantly from specimen to specimen. Ohno (1985) found that the number of the subtidally
formed growth lines in his experimental samples of C. edule did not correspond to that of the tidal
cycles during the duration of the experiment. He concluded that the subtidal growth line formation
was not rhythmical.

Despite this debate, it is easy to distinguish non-tidal growth lines from tidal ones, because the
former are more weakly defined and more irregularly spaced than the latter (Richardson et al.
19806; Ohno 1983, 1985). The spacing of subtidal growth lines sometimes becomes abnormally

table I Tidal growth patterns, their definitions and causes.

I. Growth lines

Tidal growth lines: growth lines formed during tidal exposure at low tides.

Non-tidal growth lines: growth lines formed subtidally.

II. Patterns in growth line arrangement

Regular bundles: bundles consisting of well defined and more or less regularly spaced tidal growth lines.
Spacing of growth lines in regular bundles can be classified as follows:

(а) Single spacing of tidal growth lines: spacing of tidal growth lines formed during tidal exposure at
low tides occurring approximately 12-4 hours apart. This spacing is very often accompanied by
alternating thicker and thinner growth lines (see III of this Table).

(б) Double spacing of tidal growth lines: spacing of tidal growth lines formed during tidal exposure at
low tides occurring approximately 24-8 hours apart.

If growth lines of 12-4 hourly tidal exposure origin and those of 24-8 hourly tidal exposure origin
occur in one shell growth pattern, the spacing of the latter lines are about twice as wide as that of
the former. Thus the latter ones are referred to as double spacing and the former as single spacing.
Spacing of the tidal growth lines of 24-8 hourly exposure origin are also referred to as double spacing,
even when they solely occur in a growth sequence.

Irregular bundles: bundles consisting of weak and irregularly spaced non-tidal growth lines formed
subtidally.

Alternating regular and irregular bundles: sequence caused by the change of periodic tidal exposures
during spring tides and continuous submergence during neap tides.

III. Pattern in the growth line thickness

Alternating thicker and thinner growth lines: caused by the interference of semidiurnal tidal exposure
and 24 hour change of day and night; thicker growth lines formed at daytime low tide, thinner ones
at night-time low tide. Thus the pattern is diagnostic to semidiurnal tidal exposure. Order of the
arrangement of thicker and thinner growth lines inverts approximately every 28-5 growth lines, which
corresponds to the number of 12-4 hour tidal cycles per one fortnight.

IV. Pattern in the growth increment thickness

Alternating thicker and thinner increment: caused by alternating shorter and longer duration of
submergence at successive high tides because of the diurnal inequality of tides; thicker increment
formed during the longer submergence and thinner one during the shorter submergence. The order
of arrangement of thicker and thinner growth increments inverts approximately every 26-4 increments,
which corresponds to the number of 12-4 hour tidal cycles per half a tropical month.

242

PALAEONTOLOGY, VOLUME 32

wide, more than twice or three times wider than that of the more or less uniformly spaced tidal
growth lines (Ohno 1983, 1985), if they are present in the same individual. Wide spacing of growth
lines documented in a sample of C. edule in plate 8c of Farrow (1972) was also probably caused
by the same mechanism, as its location in the lower intertidal zone suggests.

Patterns in arrangement of growth lines

The arrangement of tidal growth lines alone, or together with non-tidal growth lines, in intertidal
bivalve shells reflects the kind of tidal exposures determined by the type of tides in their habitat
and the level at which the bivalves lived. For example, C. edule that live low in the intertidal zones
are semidiurnally exposed only during spring tides, and continuously submerged during neap tides.
This change is recorded as alternating bundles of more or less regularly spaced tidal growth lines
(which will be called regular bundles) and of weak and irregularly spaced non-tidal growth lines
(irregular bundles). The proportion of the regular bundles to the irregular bundles decreases in the
lower intertidal zone (Ohno 1983).

Patterns in growth line arrangements are also found in Fragum unedo (Linne) and Clinocardium
nuttallii (Conrad). Evans (1972, 1975) studied the growth patterns of C. nuttallii from California,
USA. He did not make field experiments with living individuals, yet correlation between tidal
exposure and growth line formation is fairly obvious. Because of the change of magnitude of the
diurnal inequality of the local tide, his specimens were exposed semidiurnally during a specific
period of half a tropical month, but only once a day during the rest of the time. The resulting
growth pattern consisted of bundles of more or less constantly spaced growth lines alternating
with those of regularly spaced growth lines in which spacing is about twice as wide as those of the
former bundles. The former bundles were interpreted to be composed of growth lines of semidiurnal
tidal exposures and the latter of those of diurnal tidal exposures (Evans 1972, 1975). In this case,
the growth lines are all of tidal origin, but the frequency of their formation alternates periodically
from semidiurnal to diurnal.

The spacing of tidal growth lines formed through semidiurnal tidal exposures will be referred
to as single spacing and that of those formed during diurnal tidal exposures as double spacing.

F. unedo (Ohno 1985) lives in the lower part of the intertidal zone of Ishigaki Island, Japan.
Because of the strong difference in the height of the low tides, due to the strong diurnal inequality
of the local tide, they were exposed once a day during spring tides, but continuously submerged
for several days during neap tides. As a result, regular bundles with double spacing of tidal growth
lines formed during spring tides alternate with those of irregular bundles formed during neap tides.
This bundle alternation pattern is quite similar to that of lower intertidal individuals of Cerastoderma
edule. The only difference is that the tidal growth lines were formed diurnally in F. unedo rather
than semidiurnally.

These patterns of tidal growth lines together represent only those formed at a few levels in the
intertidal zone with two of the four general tidal types. To complement the real arrangement
patterns, hypothetical ones are shown in text-fig. 1 .

Alternating growth line thickness

Several growth patterns occur exclusively in the bivalves of intertidal zones. One of these, perhaps
the most conspicuous one, is the alternation of thicker and thinner lines (text-fig. 2). It was first
discussed by Dolman (1975) based on C. edule. Richardson et al. (1980a) experimentally found
that this pattern was the result of the difference of air or substrate temperature between daytime
and night-time tidal exposures: thicker lines are formed during daytime exposure and thinner ones
during night-time exposure. Ohno (1985) came to the same conclusion and furthermore showed
that, in the summer season at least, the formation of this alternation pattern near the MTL is not
affected by the duration of exposure at low tides, which varies to a certain extent due to the weak
diurnal inequality of the tides in the habitat of his samples.

Formation of the pattern of thicker and thinner growth lines is schematically shown in text-fig. 3.
Because low tide occurs about 50 minutes later than on each preceding day, the ‘daytime’ low

OHNO: BIVALVE MICRO-GROWTH PATTERNS

243

30 20 kszh 10 0

text-fig. 2. Alternating thicker and thinner growth lines, a, radial section of the shell of a recent intertidal
specimen of Cerastoderma edule (Linne) collected near the MTL from Vogelkoje in the vicinity of List/Sylt,
German North Sea. Sample no. P361, registration no. JCTO-13, x 315; growth from right to left. Note that
the order of arrangement of thicker and thinner growth lines inverts from right to left, so that in arbitrary
numbering the lines with odd numbers are thicker near the right corner, whereas lines with even numbers
are thicker near the left corner. Where the inversion occurs, there is a zone of a few lines of similar thickness

(SZ, switch zone).

14.8 days

text-fig. 3. Schematic explanation of the formation of the alternation of thicker and thinner growth lines.
For simplicity, it is presumed that the thicker daytime growth lines are formed between 6 and 18 o’clock and
the thinner night time ones between 18 and 6 o’clock, a, semidiurnal tide for fifty tidal cycles (tidal height
in arbitrary scale); b, sequence of 12-hour day (white segments) and 12-hour night (black segments); c,
schematic presentation of resulting alternation of growth line thickness in a bivalve living near the mean tide
level (MTL). Note the inversion of the order of arrangement of thicker and thinner growth lines every
fortnight. The inversion occurs in switch zones (SZ).

244

PALAEONTOLOGY, VOLUME 32

tide will change into the ‘night-time’ low tide after a fortnight. Consequently the order of appearance
of thicker and thinner lines inverts after every 2 weeks (28-54 semidiurnal tidal cycles). This
inversion is also seen in the middle of the growth sequence of the recent C. edule shown in text-
fig. 2. Where the inversion of growth line thickness occurs, there is a zone with a few lines of more
or less similar thickness. This zone is called ‘switch zone’ in this paper.

This pattern of line thickness alternation and inversion is also recognized in the gastropod
Monodonta labio (Linne) (Ohno and Takenouchi 1984).

text-fig. 4. Alternating increment thickness through strong diurnal inequality in the heights of high tides
(tidal height in arbitrary scale), a , if the inequality is not strong, duration of submergence at each high tide
(here indicated with hi to h7) does not vary significantly from one to another and the corresponding
increments (il to i7) have more or less similar thickness, fi, the inequality in the height of the high tides
causes alternation in duration of submergence at each high tide (hi to h7), which in turn results in the

alternation of increment thickness (il to i7).

Alternation of increment thickness

The characteristics of tides are also reflected in increment thickness. Alternating thicker and thinner
increments were found in C. edule which were experimentally grown near the mean sea level in
List/Sylt, West Germany, by Ohno (1985). Diurnal inequality of the local tide caused alternation
in the duration of submergence at every 12-4 hourly high tide. Ohno (1985) compared the duration
of submergence and the thickness of growth increment formed during each high tide and found
that thicker growth increments were formed during longer submergence and thinner growth
increments during shorter submergence. The formation of such alternation in increment thickness
is schematically shown in text-fig. 4. The order of longer and shorter submergence inverts with a
periodicity of 26-4 tidal cycles (half a tropical month); the order of thicker and thinner increments
inverts after every 26-4 tidal cycles.

Tidal v.y. daily growth rhythms in intertidal bivalves

Claims have been made that intertidal bivalves form their growth patterns with 24-hour solar daily
rhythm. If that is true, the tidal formation of growth patterns observed in the above three species

OHNO: BIVALVE MICRO-GROWTH PATTERNS

245

would lose general applicability in interpreting growth patterns of fossil intertidal bivalves. These
claims are examined below.

The claim of House and Farrow (1968) for daily growth rhythm in the intertidal C. edule was
not accompanied by any field experiments. As stated above, a massive amount of experimental
studies on this species (Richardson, Crisp and Runham 1979, 1980a, b\ Richardson, Crisp, Runham
and Gruffydd 1980; Ohno 1983, 1985) has shown clearly the tidal growth rhythm of intertidal
specimens of C. edule.

The growth rhythm of intertidal individuals of Mercenaria mercenaria (Linne) has been variously
interpreted as: a 24-hourly solar daily growth rhythm (Pannella and MacClintock 1968; Rhoads
and Pannella 1970); a solar daily growth rhythm which is interrupted by tidal exposure at low
tide (MacClintock and Pannella 1969; Pannella 1972, 1975); a purely tidal rhythm (Pannella
1976).

Ohno (1985) carried out experiments with intertidal individuals of M. mercenaria in tidal creeks
in South Carolina. In one group (MC of Ohno 1985) the average of increments (55-7) was near
that of the semidiurnal tidal cycles during the experiment (60). In another group, from a different
tidal creek (MO of Ohno 1985), with occasional interruptions of growth caused by the severe
environmental conditions of the marsh region, the number of increments formed was variable
among individuals. The average number of increments (43-8) exceeded that of the days of experiment
(33) significantly.

Further, Ohno’s specimens formed alternating thicker and thinner growth lines. Examination
of photomicrographs published by Pannella and MacClintock (1968) also show that the alternation
of thicker and thinner growth lines is very common. Their ‘complex increment’ (e.g. their pi. 1,
fig. 5) also reveals itself as nothing more than a pair of increments bounded by growth lines of
alternating thickness. Such alternation of growth line thickness should be, as mentioned already,
interpreted as the result of the interference of the semidiurnal tidal exposures and the 24-hour
change between day and night. Thus the growth pattern is quite probably formed tidally, although
the growth may be occasionally and locally disturbed by environmental stress. The agreement of
the number of the increments formed with the number of days of the experiments in the work of
Pannella et al. (1968) may be coincidental, as Pannella (1975) suggested.

Koike (1973) studied growth line formation in Merterix lusoria (Roeding) from the intertidal
zone of Kyushu, Japan. She divided growth lines arbitrarily into five different types. Based on the
correspondence of the sum of lines of her type A and type B with the number of days of the
experiment, she concluded that the lines of these two types were formed diurnally. However, as
stated by Ekarante and Crisp (1982) the single photograph provided to illustrate the supposed
sixteen daily growth bands clearly contains many more. Koike’s belief in daily growth line
formation is not acceptable without further experimental evidence.

Hall et al. (1974) studied shell growth of intertidal Tivela stultorum (Mave) assuming daily shell
growth rhythm, but they did not confirm their assumption.

As discussed above, there is no substantial evidence of daily growth rhythms among the intertidal
bivalve species.

RECONSTRUCTING ANCIENT TIDAL TYPES
Recognizing fossil tidal growth patterns

All the tidal shell growth patterns observed among living intertidal bivalves discussed earlier in
this paper are summarized in Table 1. The formation of several important growth patterns are
schematically explained in text-figs. 3 and 4. The possible arrangement patterns of tidal growth
lines at three different levels for each of four general tidal types are shown in text-fig. 1 .

Referring to these figures and tables, we can begin to interpret tidally formed growth patterns
among fossil bivalve shells. The first step is to find a fossil bivalve assemblage from one locality,
where individuals show growth patterns which are similar or identical to the present-day tidal

246

PALAEONTOLOGY, VOLUME 32

growth patterns. The frequent occurrence of tidal growth patterns cannot be attributed to chance
and allows us to conclude that we are dealing with fossils from the intertidal zone.

Procedure for reconstructing ancient tidal types

Based on the peculiar combinations of tidal growth patterns characteristic of a specific tidal type,
it is possible to reconstruct the tidal type and habitat of fossil bivalves.

There are several approaches to reconstructing tidal types and the habitat of the fossils based
on their tidal growth. One approach is the flow chart in Table 2. Here, patterns which may appear
in a wide variety of tidal types are considered first. Moving down the flow chart, characters more
and more specific to particular tidal types are taken into consideration.

The growth line thickness alternation accompanied by the inversion of the order of arrangement
of thicker and thinner growth lines is diagnostic of semidiurnal tidal exposure, although the
inversion does not always occur if the preserved growth sequence is too short or if the bivalves
are submerged continuously during the time of the 2 weeks favourable for its formation. Semidiurnal
tidal exposure can be seen in semidiurnal tides and two types of mixed tides (text-fig. 1a-c). The
common occurrence of double spacing of tidal growth lines together with the alternation and the
inversion in growth line thickness is indicative of mixed tides with diurnal inequality expressed in
the heights of low tides. The common occurrence of alternation of growth increment thickness
accompanied with the alternation and the inversion of growth line thickness indicates mixed tides
with diurnal inequality expressed in the heights of high tides. The common occurrence of the
alternation and the inversion in growth line thickness without the diagnostic features of diurnal
inequality described above is characteristic of semidiurnal tides with insignificant diurnal inequality.

Under the lack of growth line thickness alternation the existence of double spacing and alternation
of regular and irregular bundles is characteristic of the diurnal type of tides. In practice, the
recognition of double spacing as such is not possible, if it is not accompanied by the alternating
regular and irregular bundles. Double spacing alone appears to be nothing more than regular
spacing of growth lines and therefore may be confused with regularly spaced growth lines with a
different periodicity. The co-occurrence of the alternating regular and irregular bundles, which is
the reflection of spring and neap tides, ensures that the accompanying regular spacing of growth
lines is formed through tidal exposures. To summarize, in the absence of alternating growth line
thickness, these two growth patterns indicate the existence of diurnal tides.

Once the tidal type of the habitat of a fossil assemblage is reconstructed, the intertidal level for
the fossil individuals can be easily deduced by comparing the arrangement patterns of tidal growth
lines with possible arrangement patterns for reconstructed tidal curves (text-fig. 1). For example,
in an intertidal zone with a semidiurnal tide with insignificant diurnal inequality of tide, the
individuals with growth patterns consisting exclusively of single spacing live near the mean tide level
(MTL), whereas those with the alternation of the regular and irregular bundles live in the lower
part of the intertidal zone. The increasing proportion of irregular bundles in growth patterns
indicates increasingly lower living levels of bivalves.

ESTIMATE OF THE CHANGE IN THE NUMBER OF SYNODIC DAYS PER
SYNODIC MONTH IN THE GEOLOGICAL PAST: A GRAPHIC METHOD

It is inferred that tidal friction caused the change in the velocity of the earth’s rotation and the
moon’s orbital motion throughout the geological past. This resulted in the change of the periods
between two succeeding low tides and the length of the solar day, and consequently in the change
of the length of the synodic month in terms of the number of synodic days.

As mentioned earlier, the order of the arrangement of growth lines of alternating thickness
inverts every 2 weeks. The number of growth lines between two inversions (Ng) is equal to the
number of semidiurnal tidal cycles per fortnight (Nf). Because the duration of a semidiurnal tidal
cycle is half a synodic day and that of a fortnight half as long as a synodic month, the number Nf

OHNO: BIVALVE MICRO-GROWTH PATTERNS

247

equals the number of synodic days per a synodic month (Ns). Therefore the growth line thickness
alternation, if found in fossil bivalves, will provide information on the number of synodic days
per synodic month (Ns) in the past. The simplest method is to count the total number of growth
lines between several switch zones, then divide it by the number of intervals between the switch
zones. Because switch zones usually contain several lines of similar thickness, the resulting value
is accompanied by a certain amount of ambiguity. If a long sequence can be obtained the ambiguity
will become negligible. But in fossil bivalves, and even in living ones, it is very difficult to obtain

248

PALAEONTOLOGY, VOLUME 32

long and undisturbed growth sequences. However, it is easy to tell qualitatively whether the Ns
was larger or smaller than the present value with the help of a simple method described below.

The thickness of each growth line in a sequence of alternating growth line thickness is classified
as follows:

thicker = a line thicker than the preceding and succeeding lines.

thinner = a line thinner than the preceding and succeeding lines.

undifferentiated = a line equal in thickness to the preceding and the succeeding lines.

This sequence is then plotted with successive days on the ordinate and the time of the day on
the abscissa. Because tidal growth lines are formed at each low tide, the thickness of each growth
line is plotted with the interval of semidiurnal low tides: a thicker growth line with a large solid
rectangle; a thinner one with a small quadrate; a point for an undifferentiated line.

For the Recent sequence of alternating growth line thickness the result of such a plot is a vertical
partition of the thicker and thinner growth lines. The thicker ones of daytime tidal exposures are
distributed in a vertical zone in the middle of the co-ordinate, at both sides of this zone thinner
ones of night-time tidal exposures are distributed (text-fig. 5a).

If the two parameters, the period of semidiurnal tides and the duration of the solar day, are
different from the present ones, the number of the growth lines of alternating thickness between
two succeeding inversions, which corresponds to the number of synodic days per one synodic
month, may also vary. If the amount of change in both parameters is known, an appropriate co-
ordinate with correct low tide intervals can be prepared and the sequence of growth line thickness
alternation formed under the changed condition can be plotted on it. For example, text-fig. 5M is
a plot of a hypothetical alternation sequence on the appropriate co-ordinate for the time when the
period of the semidiurnal tides is slightly longer than the present-day value. Text-fig. 5c 1 is another
plot for the slightly shorter period of the semidiurnal tides on the appropriate co-ordinate. The
result is the vertical partition of the thicker and thinner growth lines in both cases, although the
number of tidal cycles per fortnight is different from the present-day value.

In practice, for fossil growth line thickness alternation patterns the amount of change in the
earth's rotation or that of moon’s orbital motion is not known. However, any fossil sequence can
be plotted on the co-ordinate with the present-day intervals between succeeding low tides. If the
zones with the same symbol on such a plot run parallel to the abscissa, it means that the number
of synodic days per synodic month (Ns) of the period, when the fossils lived, is equal to the
present-day one. If the zone with the same symbol shifts from the vertical row it indicates that the
number of the synodic days per synodic month (Ns) is different from the present-day value. For
example, the plot of hypothetical growth line alternation sequences for the time when the periods
of two succeeding tides is slightly longer than the present-day value yields a shift of growth lines
of different thickness from top right to bottom left (text-fig. 5^2) and for the time of slightly shorter
semidiurnal tidal cycles a shift from top left to bottom right (text-fig. 5c2). The degree of the shift
depends on the amount of change in the motions of the earth and moon compared to present-day
conditions.

Thus the zonation pattern on the co-ordinate with the present-day low tide intervals will tell us
whether or not the number of the synodic days per synodic month was different in the geological
past.

FOSSIL TIDAL PATTERNS
Samples from the late Pleistocene Osaka Group

Twenty-nine shell fragments from two bore holes at the same horizon were examined. More
than half (16) are oyster shells, on which growth patterns are most clearly visible on the surface
of the ligament groove that is coated with a thin layer of acicular crystallites (PI. 31, figs. 5 and
6). Shell material including this layer, which was examined by electron diffraction method using a

MIDNIGHT NOON MIDNIGHT

3 0 0.5 1.0

. MIDNIGHT NOON
Dl 0 05

MIDNIGHT MIDNIGHT NOON MIDNIGHT

1.0 D2 0 0.5 1.0

20-

cn

>-

CC

□

40-

20-

cn

>-

=£

□

40

V

V

&

MIDNIGHT NOON MIDNIGHT MIDNIGHT NOON MIDNIGHT
Cl o 0.5 1.0 C2 0 0.5

1

- V-

—

—

\

If —

w

1+- —

». mm

- -

\ ■

-t —

— 1 —

— ^

— t —

— \

\

=-t —

■ — ^

% \

B

\-m~

\

"i

1 —

1

i k-

~e \ -

~ a |

:

- •"

\

\

•| —

V

—

TEXT-FIG. 5. Plots of alternating growth line thickness (simulation). For simplicity, it is presumed that the
growth lines formed between 6 and 18 o’clock are thicker than those formed between 18 and 6 o’clock.
(Ordinate of the co-ordinates = successive days; abscissa = time of day expressed as a fraction of the period
of day, so that 0, 0-5, and 10 mean 0 o’clock, 12 o’clock, and 24 o’clock, respectively.) a , plot of a sequence
ol the present-day growth line thickness alternation on the co-ordinate with the present-day low tide intervals
(Ns = 28-5). b , pair ol plots for the hypothetical slight lengthening of low tide intervals (0-5% decrease in
angular velocity of the moon s rotation around the earth; rotation velocity of the earth unchanged; Ns =
24 8). bl, plot on the co-ordinate with changed low tide intervals; b2 , plot on the co-ordinate with present-
day low tide intervals. Distribution boundaries of thicker and thinner growth lines are marked with broken
lines, c, pair ol plots lor the hypothetical slight shortening of low tide intervals in comparison with the length
of the solar day (0-5% increase of the angular velocity of the moon’s rotation around the earth; the rotation
velocity of the earth unchanged; Ns = 33-4): cl, plot on the co-ordinate with changed low tide intervals; c2,
plot on the coordinate with present-day low tide intervals. Distribution boundaries of thicker and thinner

growth lines are marked with broken lines.

250

PALAEONTOLOGY, VOLUME 32

transmission electron microscope, consists of calcite with a very minute amount of aragonite.
Aragonite might be a remnant of oyster resilium, as observed in several oyster species (Taylor et
al. 1969). Therefore, the thin layer lining the ligament groove can be considered to be composed
of calcite. On the floor of the ligament groove the broad and flat growth increments are bordered
by growth lines as thin ridges (PI. 31, figs. 1 and 5). On several specimens the thin layer of acicular
crystallites is lacking, probably through dissolution or wear. In such cases the growth lines are
narrow grooves (PI. 3 1 , fig. 7); yet the growth patterns are well developed as the negative impression
of the original topography.

Occurrences of various growth patterns are summarized in Table 3. On all of the specimens,
alternating thicker and thinner growth lines are visible (PI. 31, figs. 1, 2, 4, 5). The ubiquitous
occurrence of the alternating growth line thickness is diagnostic of the semidiurnal tidal exposure.
On several of them the inversion of the order of arrangement of thicker and thinner growth lines
also can be seen (PI. 31, figs. 1 and 4), which strengthens the diagnosis. Consequently the specimens
must have experienced frequent semidiurnal tidal exposure.

How strong was the diurnal inequality of the tides in the fossil habitat? The rare occurrence of
both double spacing of growth lines and the alternation of increment thickness (Table 3) suggest
the weakness of the diurnal inequality expressed in the heights of the low tides and of the high
tides, respectively. It is therefore concluded that the tidal type of the habitat of the present
specimens was a semidiurnal one with weak diurnal inequality. The schematically reconstructed
tidal curve is shown in text-fig. 6b, and deviates strongly from the mixed type of tide with strong
diurnal inequality expressed in the heights of low tides of the present-day Osaka Bay (text-fig. 6a).

The examined shells seem to contain individuals from various levels within the intertidal zone.
Several specimens show a continuous sequence of growth line thickness alternation of more than
28-5 growth lines (PI. 31, fig. 1), which is larger than the semidiurnal tidal cycles per 2 weeks at
present. Therefore, they must have lived near the mean tide level, where they were exposed at each
low tide, even during neap tides. The majority of specimens showing alternation of regular and
irregular bundles (PI. 31, figs. 2 and 3) must have lived lower in the intertidal zone, where they
formed regular bundles during spring tides and irregular bundles during neap tides. The number
of well-defined growth lines within a single regular bundle (nine to twenty-two) is well within the

EXPLANATION OF PLATE 31

Figs. 1-7. SEMs of material from the late Pleistocene ‘Mai T Clay from Osaka: 1-4 in back scattering
electron image (BEI) mode; 5-7 in secondary electron image (SEI) mode. 1, ligament surface of oyster
shell with more than thirty growth lines (marked with bars and numbers). Growth from left to right.
Alternation of growth line thickness is visible: near the left corner (between the lines numbered 0 to 19)
lines with odd numbers are thicker, whereas near the right corner (between the lines with numbers 24 to
34) lines with even numbers are thicker. Sample no. pi. 1-1, registration no. JCTO-0012; from borehole B,
x 270. 2, ligament surface of oyster shell with alternating regular bundles (marked with solid lines) and
irregular bundles (marked with broken lines). Alphabets correspond to the occurrence of abnormally thick
increments shown in text-fig. Id. Growth from right to left. Sample no. P294, registration no. JCTO-0009;
borehole A, x 44. 3, radial section of bivalve shell fragment showing alternating regular bundles (marked
with solid lines) and an irregular bundle (marked with broken line). Growth from right to left. Sample no.
P103, registration no. JCTO-OOOl; borehole A, x 125. 4, ligament surface of oyster, with alternating
thicker and thinner growth lines as well as the inversion of the order of their arrangement. Lines with even
numbers are thicker near the left corner, whereas those with odd numbers are thicker near the right corner.
Growth from left to right. Sample no. P354, registration no. JCTO-OOll; borehole A, x 250. 5, part of
ligament surface of oyster with alternating thicker growth lines (indicated by thick downward arrows) and
thinner ones (indicated by thin downward arrows). Growth from right to left. Sample no. P294, registration
no. JCTO-0009; borehole A, x 260. 6, enlargement of the area marked with white frame in fig. 5, x 1600.

7, part of ligament surface of oyster with growth lines as grooves. Sample no. P293, registration no. JCTO-

0008, x 280.

PLATE 31

OHNO, Bivalve microgrowth patterns

252

PALAEONTOLOGY, VOLUME 32

table 3. Tidal growth patterns in fossil shell fragments from the ‘Mai 2’ Clay of the late Pleistocene Osaka
Group, p = present; f = faintly expressed; . = not present; — = observation or counting not carried out.
Shell structure: gr. = granular structure; cl. = crossed lamellar structure; ac. = acicular cristallites. Number
of growth lines: if irregular bundles are prevalent in growth sequence, number of growth lines are not counted.
Alternation of thicker and thinner growth lines: a. = line thickness alternation; i. = inversion of order of
occurrence of thicker and thinner growth lines. Alternation of regular and irregular bundles: the number of
pairs is given in parentheses if the alternation is well developed; if growth sequence is without irregular

bundles the number is not given.

Growth line

Alternation
of regular
and

irregular

bundles

Alternation

of

growth

increment

thickness

Sample

no.

Shell

structure

Number

of

growth

lines

Thickness

alternation

Double

spacing

a.

i.

Borehole A

Ligaments of Ostrea sp.
P293 ac.

95 >

P

P294

ac.

160 >

P

P

P295

ac.

—

P

P(5)

P296

ac.

—

P

P(3)

P297

ac.

85 >

P

p(2)

P298

ac.

75 >

P

P

P(3)

P299

ac.

55 >

P

p(3)

P330

ac.

66 >

P

p

P332

ac.

140 >

P

P

P334

ac.

—

P

P

P(2)

P350

ac.

100 >

P

p

p(4)

P351

ac.

50 >

P

P

P

P352

ac.

—

P

P

P(3)

P353

ac.

100 >

P

P354

ac.

70 >

P

p

p(4)

Other bivalve shell fragments
P335 cl. 21

P

P339

—

220 >

P

P342

cl.

200 >

P

P

P343

cl.

—

P

p(5)

P344

cl.

21

P

P

P346

cl.

40 >

P

P

p(2)

Borehole B

Ligament of Ostrea sp.
PI 1-1 ac.

36

P

Other bivalve shell fragments
P095 cl. 50 >

P

f

P097

cl.

30

P

P

P

P099

gr-

28

P

PI 02

cl.

100 >

P

P

PI 03

gr-

34

P

P

PI 2-1

cl.

—

P

P

P12-3

cl.

—

P

f

P

OHNO: BIVALVE MICRO-GROWTH PATTERNS

253

a OSAKA RECENT

Tidal cycle

b OSAKA 70,000 y. B.P.

Tidal cycle

text-fig. 6. Tidal curves along the Osaka Bay in the present day and in late Pleistocene time (c. 70 000 years
bp), a, the present-day tide along the Osaka Bay, drawn from the prediction for July and August, 1980, from
the tide tables published by the Maritime Safety Agency, Japan (ordinate = lidal height; abscissa =
semidiurnal tidal cycles), b, the reconstructed tidal curve for late Pleistocene time on the basis of the fossil
tidal shell growth patterns (tidal height in ordinate is arbitrary; abscissa = semidiurnal tidal cycles). Note
that the late Pleistocene curve has relatively weak diurnal inequality in comparison with the present-day

curve.

maximum growth line number formed through semidiurnal tidal exposures per 2 weeks at present
(28-5), and it decreases inversely with the increase of the relative width of the irregular bundles.

Many oyster specimens show extremely wide growth line spacings (text-fig. 7), which occur
periodically and exclusively in irregular bundles (for example, compare text-fig. Id and PI. 31, fig.
2). This makes it easy to detect the change of spring and neap tides. However, the wide spacings
must be the result of the break of an otherwise constant semidiurnal rhythm of growth line

254

PALAEONTOLOGY, VOLUME 32

formation, as known in the recent intertidal Cerastoderma edule (Ohno 1983, 1985). Thus the
estimate of the length of a fortnight in terms of semidiurnal tidal cycles on the basis of the
periodically swinging curves in text-fig. 7 is inevitably accompanied with a large error and has not
been carried out in the present work.

Samples from the Miocene Mizunami Group

Bivalve shell fragments from the Mizunami Group are all preserved in a consolidated sedimentary
matrix. Therefore specific identification could not be made. Original shell microstructures such as
granular structure (PI. 32, fig. 3) or crossed lamellar structure are very well preserved in most
specimens. Sometimes the outer surface of the shell (PI. 32, fig. 5) or the growth lines (PI. 32,
fig. 4) are lacking because of post depositional dissolution. Only specimens with well-preserved
growth patterns were studied.

Growth patterns are usually easily observed. Sometimes, depending on the shell microstructure
and the amount of etching, the relief of the lines is fairly weak, but most such cases shell growth
patterns are obvious, if photographs of the specimens are observed in oblique position (for example,
PI. 32, figs. 2 and 6).

Forty-six shell fragments were studied (Table 4) and on each shell fragment there are fifteen to
1 1 1 growth lines. All of the specimens show the alternation of growth line thickness (PI. 32, figs.
1, 2, 6, 7; text-fig. 8). About half of them also show the inversion of the order of arrangement of
thicker and thinner lines (PI. 32, figs. 1, 2, 6; text-fig. 8). These observations show that samples
were very frequently exposed at semidiurnal low tides.

The double spacing is seen only in five specimens and its proportion in the growth patterns is
very small. This indicates that the diurnal inequality of tides expressed in the height of low tides
was weak. The alternation of thicker and thinner growth increments occurs in eighteen samples.
It is fairly obvious (PI. 32, fig. 1) only in three of them and is weak or very faint in the others.
Therefore, the diurnal inequality was expressed in the height of the high tides, but it was probably
not strong. These observations together indicate that the type of the tide when and where the
examined shells lived was semidiurnal with a moderate amount of diurnal inequality expressed
in the height of high tides. This tidal type was intermediate between the tidal types shown in

EXPLANATION OF PLATE 32

Figs. 1-7. All figures are SEMs of shells from the Miocene Mizunami Group: 3 and 4 in SEI mode; others
in BEI mode. 1, radial section of shell fragment. Growth lines are seen as narrow white stripes indicated
with arrows and numbers. Alternation of increment thickness is visible: increments between growth lines
numbers 5 and 6, 7 and 8, 9 and 10, 1 1 and 12, 13 and 14 are thicker than adjoining increments. Alternating
growth line thickness as well as inversion of its order are also observable; lines with even numbers are
thicker near the left corner, whereas those with odd numbers are thicker near the right corner. Growth
from left to right. Sample no. P270, registration no. JCTO-0005, x 185. 2, radial section of shell fragment.
Growth lines are seen as narrow white stripes marked with arrows and numbers. Because of the coarse
grain size of shell material the resolution of the lines is not good. Yet lines can be fairly well recognized
if the pictures are observed obliquely from the side. Alternating growth line thickness as well as inversion
of its order are visible: lines with even numbers are thicker near the left corner, whereas those with odd
numbers are thicker near the right corner. Growth from left to right. Sample no. Villa registration no.
JCTO-0006, x 250. 3, part of a shell showing very well-preserved crystallites. Sample no. PI 95, registration
no. JCTO-0002, x 2850. 4, shell fragment in which growth line material (running horizontally) is dissolved
away. Sample no. P20I, registration no. JCTO-0003, x 760. 5, shell fragment in which a part of the outer
layer is not preserved. Well-defined growth lines are still visible in the inner part of the shell. Sample no.
P277fi, registration no. JCTO-0007, x 108. 6, alternating thicker and thinner growth lines. Because of the
coarse grain size the resolution of the lines is not good, yet lines can be fairly well recognized if the pictures
are observed obliquely from the side. Growth from top to bottom. Sample no. P258, registration no.
JCTO-0004, x 370. 7, alternating thicker and thinner growth lines. Each growth line is marked with an
arrow. Growth from top to bottom. Sample no. PI 95, registration no. JCTO-0002 same as illustrated in
Plate 32, fig. 3, x415.

PLATE 32

OHNO, Shell microgrowth patterns

increment thickness

256

PALAEONTOLOGY, VOLUME 32

Id T T f T

0 20 40 60 80 ioo 120 140 160

100-

f

20

40

60

80

O-l T x X X X

0 20 40 60 80 100

20 40 60

text-fig. 7. Measurements of increment thickness on the surface of the oyster ligaments. Note the periodic
change of the increment thickness with recurring appearance of extremely thick increments. Alphabets in d
correspond to irregular bundles in Plate 31, fig. 2 ( a , sample P293; b , P299; c, P332; d , P294: e, P297; f, P298;

g, P350; h, P330; i, P353; j, P354).

OHNO: BIVALVE MICRO-GROWTH PATTERNS

257

table 4. Tidal growth patterns in fossil bivalve shell fragments from the Miocene Mizunami Group, p =
present; f = faintly expressed; s = strongly expressed; . = not present; — = observation or counting not
carried out. Shell structure: gr. = granular structure; cl. = crossed lamellar structure. Alternation of thicker
and thinner growth lines: a. = line thickness alternation; i. = inversion of order of occurrence of thicker and

thinner growth lines.

Sample

no.

Shell

structure

Growth line

Numbe

of

growth

lines

r Thickness

alternation

of regular of

and growth

Double irregular increment

spacing bundles thickness

a.

i.

P195

gr-

43

P

p

P

P206«

cl.

42

P

p

P206 b

cl.

58

P

p

P

P207

gr-

32

P

p

P

P208

gr-

47

P

p

P

P209

gr-

26

P

p

P253

gr-

69

P

P

P254

gr-

36

P

p

P

P256

gr-

75

P

p

P257

—

30

P

p

P

P258

gr-

40

P

p

P265

gr-

55

f

P265 b

gr-

54

P

P266

gr-

43

P

p

s

P267a

gr-

17

P

P2676

gr-

20

P

f

P268

gr-

60

P

P P

P269a

cl.

30

P

P P

P269 b

—

30

P

P

P270

gr-

30

P

p

s

P2706

gr-

16

P

P271

gr-

15

P

f

P276

gr.

59

P

p

Villa

gr-

48

P

p

s

Vlllb

gr-

60

P

P301

gr-

11 1

P

p

p

P304

gr-

26

P

f

P3 1 On

gr-

43

P

p

p

P3 1 Ob

gr-

35

P

p

P31 1

gr-

29

P

p

f

P313

gr-

46 >

P

p

p

P314

gr-

49

P

p

f

P315

cl.

30

P

p

P317

gr-

50

P

P318

gr-

34

P

f

P319

gr-

14

P

P320

gr-

64

P

p

P321

gr-

50

P

p

P322a

gr-

68

P

P322c

gr-

30

P

P323

cl.

49

P

p

P324

gr-

18

P

P325

gr-

58

P

p

f

P326

gr-

23

P

P327

gr-

32

P

p

P328

gr-

17

P

p

p

258

PALAEONTOLOGY, VOLUME 32

text-fig. 8 a-d. Radial cross-section of a bivalve shell from the Miocene Mizunami Group with continuous
sequence of growth lines of alternating thickness (SEM micrograph in BEI mode). The numbering corresponds
to that of the growth lines plotted on the co-ordinate of text-fig. 9cl (sample no. P301, registration no. JCTO-

0010, x 138).

text-fig. 1a and c. With regard to the magnitude of the diurnal inequality, it is most similar to
the one in text-fig. 1a. The less frequent occurrence of irregular bundles in only three specimens
shows that few of the samples lived in the lower part of the intertidal zone. The majority of
the specimens lived near the mean tide level, where they formed one growth line at each
semidiurnal tidal exposure.

NUMBER OF SYNODIC DAYS PER SYNODIC MONTH IN THE MIOCENE EPOCH

Eight of the samples from the Miocene Mizunami Group have a relatively long and continuous
sequence of alternating thinner and thicker growth lines without any sign of disturbance. For
example, one sample (text-figs. 8 and 9 d) has the longest growth sequence with 1 1 1 successive
growth lines. The sequence of alternating growth line thickness of these eight samples is thus
plotted on the co-ordinates with the present-day low tide intervals, in order to see whether the
number of the synodic days per synodic month was different in the Miocene than it is today.

A clear partition of thicker and thinner growth lines is visible in all of the eight plots (text-fig.
9). In plots b, d, and e the left boundary of the thicker growth lines slightly shifts from right at
the top to left at the bottom. But the distribution of the thinner growth lines, as well as the right
boundary of thicker growth lines, is almost parallel to the ordinate. Thus there is no need to
consider the partition of the zones with thicker and thinner lines inclined to the ordinate. In the

OHNO: BIVALVE MICRO-GROWTH PATTERNS

259

other five samples the situation is similar. While certain boundaries of the zone are slightly inclined
from the right to left, other boundaries are almost vertical to the ordinate.

Generally the partition of thicker and thinner growth lines can be regarded as parallel to the
ordinate. Thus it is concluded that the number of the synodic days per a synodic month in the
Miocene epoch was almost the same as that of today.

text-fig. 9. Sequence of alternating growth line thickness obtained from fossil bivalve shells of the Miocene
Mizunami Group plotted on the co-ordinate with present-day low tide intervals: ordinate = successive days;
abscissa = time of day expressed as a fraction of the period of day, so that 0, 0-5, 10 mean 0 o’clock, 12
o’clock, and 24 o’clock, respectively. The time of the first low tide of each plot is arbitrarily determined, a ,
sample PI 95; b , P208; c, P274; d, P301; e, P314 ;/, P322; g, P325; h, P353.

DISCUSSION AND CONCLUSIONS

The two examples described in this paper demonstrate that tidal growth patterns of fossils can be
used to recognize habitats of the ancient intertidal zone, reconstruct tidal types of these habitats,
and infer the living level of bivalves within the intertidal zone. In addition, information can be
obtained on geophysical aspects related to tidal phenomena, e.g. the change of the earth-moon
system through tidal dissipation.

Despite such potential applicability, there have been very few attempts to identify the fossil tidal
growth patterns of bivalves. In addition, the photomicrographs of fortnightly clustering of growth
increments in a Palaeocene venerid shell (Berry and Barker 1968, 1975) and tidal growth patterns
of a Late Cretaceous Limopsis striatus-punctatus (Pannella 1976) do not show any of the tidal

260 PALAEONTOLOGY, VOLUME 32

growth features summarized in Table 1. More material needs to be examined to verify these
studies.

Because of the fragmentary nature of the shells and their preservation in hard matrix, specific
identification of the bivalve fossil shells examined could not be carried out. However, the ubiquitous
occurrence of the various growth patterns among them, which are identical to those seen in living
intertidal bivalves, justifies the interpretation that the two fossil bivalve shell assemblages also grew
intertidally with a tidal growth rhythm. The Osaka samples come from the lower part of the Mal2
Clay bed, which very probably was deposited during an early phase of a transgression. The
examined samples of Mizunami come from the locality where Itoigawa el al. (1974) found many
molluscan species inferred to have lived intertidally. This is in agreement with the present
interpretation that the two examined bivalve assemblages once lived in the intertidal zone.

As mentioned above, the tidal growth patterns are clearly recognizable even in fragmentary
shells. This makes almost any type of shell-bearing sediments suitable for growth pattern analysis.
For example, a vast number of cores stored in various institutions might be examined successfully,
because even in a borehole core of small volume, there is often a large number of shell fragments.
Indeed, the Osaka material obtained from boreholes provided a good sample with successful
results.

The tidal types of the late Pleistocene, as well as the Miocene, reconstructed from the fossil tidal
growth patterns of the shells from the two Japanese localities, is semidiurnal with a weak diurnal
inequality. In comparison to these the tidal type of the present-day Japanese Pacific coasts is
characterized by strong diurnal inequality. This is the first proof of a change of tidal types through
geological time.

Tidal type varies from place to place depending on factors such as configuration of shore lines
and the depth of the sea. Because 70 000 years seems not long enough for the basic geography of
the Osaka Bay to have changed tectonically, the eustatic sea-level change may be the cause of the
drastic change of the tidal type in Osaka Bay since the late Pleistocene. Little can be said at the
moment about the reason for the Miocene tidal type. Intensive collecting of data on fossil tidal
patterns from different localities and geologic ages may provide a much more vivid picture of the
change of tidal type as well as its causes.

Tidal patterns are related to another interesting topic: the history of the earth-moon system.
Through tidal friction, mainly in shallow oceans, the length of the earth day increases (at present
by 2 milliseconds per century) and the moon spirals away from the earth by a few centimetres a
year (Runcorn 1967; Brosche 1971; Goldreich 1972). Thus astronomical cycles such as the length
of the day, days per year, days per month, and synodic months per year are thought to have been
different in the geological past.

Estimates of the length of ‘fortnights’ (2 weeks) or synodic months in the geological past have
been carried out previously by several authors using fossil growth patterns (Berry and Barker
1968, 1975; Pannella and MacClintock 1968; Pannella et al. 1968; Pannella 1972). Clusters
consisting of groups of thin increments with relatively thick ones were interpreted either as formed
fortnightly or monthly and the increment numbers per cluster were estimated. The reasons for
assigning such clustering to fortnightly or monthly periodicity, as well as whether the basic rhythm
is semidiurnal, solar daily, or lunar daily, were not clearly given, as Scrutton (1978) pointed out.
Further, as already mentioned in the observation of the fossil material from Osaka, such a periodic
clustering could result from the break of rhythm of growth line formation in some part of the
growth sequence.

In attempting to estimate astronomical values concerning earth-moon history, two points are
critical for the accuracy of such estimates. One is whether the patterns concerned really reflect the
inferred periodicity. The other is how precisely the periodicity is recorded in the growth pattern.

The alternation of thicker and thinner growth lines together with periodic inversion of their
order of occurrence, described earlier in this paper, is so peculiar and complicated a pattern that
it could not be formed by any other means than those which cause it to form in recent bivalve
shells; namely, the interference of the semidiurnal tidal exposure and the 24-hourly change of

OHNO: BIVALVE MICRO-GROWTH PATTERNS

261

temperature. This pattern, also very easily recognizable in fossil shells, can be used to estimate
whether the number of synodic days per synodic month was shorter or longer in the past than it
is in the present.

How exactly are growth lines formed at each low tide? Ohno (1985) experimentally showed that
near the MTL (mean tide level) of a region with a semidiurnal type of tide with insignificant
diurnal inequality, individual specimens of Cerastoderma edule formed exactly one growth line and
increment at each tidal exposure and immersion respectively. Thus they faithfully recorded the
tidal cycles.

Further, the alternation of thicker and thinner growth lines can be used to check the completeness
of the growth record. Assume, for example, that one growth line fails to be formed at a certain
low tide. In the plot of such a growth sequence on the co-ordinate used in text-figs. 5 a and 9a-/?,
there will be a dislocation in the vertical partition of thicker and thinner growth lines at the point
which corresponds to the missing growth line. With the increase of missing lines, dislocation occurs
much more often and the resulting pattern may be something like that on a chequer-board. Thus
we can exclude plots with dislocation and thus minimize misinterpretation.

Selected sequences of growth lines of alternating thickness from the Miocene Mizunami Group
provide plots that show no detectable inclination in partition. This means that the number of the
synodic days per synodic month about 15 ma ago was almost the same as the present value.

Efforts should be made in the future to determine the alternating pattern of growth line thickness
in specimens from older geological epochs in order to obtain a clear picture of the change in the
number of the synodic days per synodic month throughout earth’s history. The present finding of
tidal growth patterns in oyster shells is encouraging, because oyster shell material is composed of
calcite and resistant against diagenetic change. Oysters also have a good fossil record which begins
in the Late Triassic. Tidal growth patterns most likely exist in fossils other than bivalves, because
tidal growth patterns or rhythms also can be seen in the Recent gastropods (Antoine and Quemerais-
Pencreac’h 1980; Ekarante and Crisp 1982; Ohno and Takenouchi 1984) and barnacles (Bourget
and Crisp 1975; Crisp and Richardson 1975; Bourget 1980).

Tidal growth patterns may also apply to a much wider range of problems than those treated in
this paper. For example, estimates of fossil shell growth rate based on periods of semidiurnal or
diurnal tidal cycles can be easily made and surely other astronomical periodicities can also be
documented in shell growth patterns. For other patterns with periodicities, we have not yet as safe
criteria as those for tidally formed patterns.

Further investigations with intensive experimental studies on recent organisms should be carried
out in order to find out criteria for such periodicities. If we find such criteria in the future, we will
be able to retrieve quite new and valuable information from growth patterns, which cover a
spectrum of problems from the ontogeny of an individual fossil organism to the history of the
earth-moon system.

Acknowledgements. The present work was initiated under the generous supervision of Professor Dr H. K.
Erben of the Institute of Palaeontology, Bonn University. Professor J. Itoigawa, Nagoya University, gave
me information on the fossil material from the Mizunami Group and Professor K. Nakaseko, Osaka
University, provided the late Pleistocene material from Osaka. Dr A. Tsuchiyama of my department kindly
determined the mineralogy of the ligament area of the Pleistocene oyster fossils. Professor M. Kumazawa,
University of Tokyo, Dr D. Walossek, Bonn University, and Professor Y. Nogami, Primate Research
Institute, Kyoto University, provided not only valuable discussions, but also continuous encouragement.
Professor K. Chinzei kindly read the manuscript and made valuable suggestions for improvement, while Dr
E. J. Moore, Corvallis, Oregon, corrected and improved the English of the manuscript. My thanks are due
to all these people. A part of this work was supported by a post-doctoral scholarship from the Japan Society
for the Promotion of Science.

262

PALAEONTOLOGY, VOLUME 32

REFERENCES

antoine, l. and quemerais-pencreac’h, d. 1980. Stries et rythmes de croissance chez la Patclle Patella vulgata
L. C. r. hebd. Seanc. Acad. Sci ., Paris, Serie D, 290, 1127 1130.
berry, w. b. n. and barker, r. m. 1968. Fossil bivalve shells indicate longer month and year in Cretaceous
than present. Nature, Load. 217, 938-939.

1975. Growth increments in fossil and modern bivalves. In rosenberg, g. d. and runcorn, s. k.
(eds.). Growth rhythms and the history of the earth's rotation, 9-25. Wiley, London.
bourget, e. 1980. Barnacle shell growth and its relationship to environmental factors. In rhoads, d. c. and
lutz, R. a. (eds.). Skeletal growth of aquatic organisms, 469-491. Plenum, New York, London.

— and crisp, d. j. 1975. An analysis of the growth bands and ridges of barnacle shell plates. J. mar. biol.
Ass. UK, 55, 439-461.

brosche, p. 1971. Die Bremsung der Erdrotation. Sterne and Weltraum, 1971 (2), 38-40.
chave, k. e. and erben, h. k. 1979. Biomineralization. In fairbridge, r. w. and jablonski, d. (eds.). The
encyclopedia of paleontology, 88-94. Dowden, Hutchinson and Ross, Stroudsburg/Pennsylvania.
chiji, m. 1984. Detailed survey on fossil Foraminifera of the submarine strata at the Kansai International
Airport in Osaka Bay, Central Japan. In nakaseko, k. (ed.). Geological survey of the submarine strata at
the Kansai International Airport in Osaka Bay, Central Japan, 29-36. Calamity Science Institute, Osaka.
[In Japanese.]

Clark, G. R. ii 1974. Growth lines in invertebrate skeletons. A. Rev. Earth Planet. Sci. 2, 77-99.
crisp, d. j. and Richardson, c. a. 1975. Tidally-produced internal bands in the shell of Elminus modestus.
Mar. Biol. 33, 155-160.

deith, m. a. 1983. Molluscan calendars: the use of growth-line analysis to establish seasonality of shellfish
collection at the Mesolithic site of Morton, Fife. J. Archaeol. Sc. 10, 423-440.
dolman, j. 1975. A technique for the extraction of environmental and geophysical information from growth
records in invertebrates and stromatolites. In rosenberg, g. d. and runcorn, s. k. (eds.). Growth rhythms
and the history of the earth's rotation, 191-222. Wiley, London.
ekarante, s. u. k. and crisp, d. j. 1982. Tidal micro-growth bands in intertidal gastropod shells, with an
evaluation of band-dating techniques. Proc. R. Soc. B214, 305-323.
evans, J. w. 1972. Tidal growth increments in the cockle Clinocardium nuttallii. Science, Washington, 176,
416-417.

1975. Growth and micromorphology of two bivalves exhibiting non daily growth lines. In rosenberg,
G. d. and runcorn, s. K. (eds.). Growth rhythms and the history of the earth's rotation, 119-134. Wiley,
London.

farrow, G. e. 1972. Periodicity structures in the bivalve shell: analysis of stunting in Cerastoderma edule
from the Burry Inlet (South Wales). Palaeontology, 15, 61-72.
furutani, m. 1984. Pollen stratigraphy of the submarine strata at the Kansai International Airport in Osaka
Bay. In nakaseko, k. (ed.). Geological survey of the submarine strata at the Kansai International Airport in
Osaka Bay, Central Japan, 91 116. Calamity Science Institute, Osaka. [In Japanese.]
gartner, s. 1977. Calcareous nannofossil biostratigraphy and revised zonation of the Pleistocene. Mar.
Micropaleont. 2, 1-25.

goldreich, p. 1972. Tides and the earth-moon system. Scient. Am. 226, 43-52.

hall, c. a. jr., dollase, w. a. and corbato, c. e. 1974. Shell growth in Tivela stultorum (Mawe, 1823) and
Ccdlista chione (Linnaeus, 1758, Bivalvia): Annual periodicity, latitudinal differences, and diminution with
age. Palaeogeogr. Palaeoclimat. Palaeoecol. 15, 33-61.
house, m. r. and farrow, g. e 1968. Daily growth banding in the shell of the cockle, Cardium edule. Nature,
Lond. 219, 1384 1386.

itoigawa, j. 1974. Geology of Mizunami Group. Bull, of the Mizunami Fossil Mus. no. 1, 9-42. [In Japanese.]

— 1981. Mizunami Area. In tsuchi, r. (ed.). Neogene of Japan— its biostratigraphy and chronology, 62-64.
IGCP-114 National Working Group of Japan, Shizuoka.

shibata, H. and nishimoto, H. 1974. Molluscan fossils from the Mizunami Group. Bull, of the Mizunami
Fossil Mus. no. 1, 43-203. [In Japanese.]

koike, h. 1973. Daily growth lines of the clam Meretrix lusoria—a basic study for the estimation of prehistoric
seasonal gathering. J. anthrop. Soc. Nippon, 81, 122-138.
le gall, m. p. 1970. Methode d’etude des stries de croissance de Mytilus edulis L. mise en evidence du
rhythme et des modalites de leur formation. C. r. hebd. Seanc. Acad. Sci., Paris, Serie D, 270, 509-51 1.
MaccLiNTOCK, c. and pannella, g. 1969. Time of calcification in the bivalve mollusk M. mercenaria (L.)

OHNO: BIVALVE MICRO-GROWTH PATTERNS

263

during the 24 hour period. Abstr. Ann. Mtg. Geol. Soc. Ain. 140.

NAKASEKO, K., TAKEMURA, K., NISHIWAKI, N., NAKAGAWA, Y., FURUTANI, M. and YAMAUCHI, M. 1984. Stratigraphy
of the submarine strata at the Kansai International Airport in Osaka Bay off Senshu, Central Japan. In
nakaseko, K. (ed.). Geological survey of the submarine strata at the Kansai International Airport in Osaka
Bay, Central Japan, 191-198. Calamity Science Institute, Osaka. [In Japanese.]
neville, a. c. 1967. Daily growth layers in animals and plants. Biol. Rev. 42, 421 439.
ohno, t. 1983. A note on the variability of growth increment formation in the shell of the common cockle
Cerastoderma edule. In brosche, p. and suendermann, j. (eds. ). Tided friction and the earth's rotation II,
222-228. Springer, Berlin, Heidelberg.

— 1984. Tidal growth patterns in late Pleistocene bivalves in the boring cores around Osaka Bay, Central
Japan. News of Osaka Micropaleontologists , no. 12, 41-49. [In Japanese with English abstract.]

— 1985. Experimentelle Analysen zur Rhythmik des Schalenwachstums einiger Bivalven und ihre palaobio-
logische Bedeutung. Paldontographica. Abt. A, 189, 63 1 23.

— and takenouchi, k. 1984. Tidal growth patterns in recent Monodonta labio (Linnaeus, 1758) (Gastropoda,
Trochidae). News of Osaka Micropaleontologists , no. 12, 51-56.

okamura, m. and yamauchi, m. 1984. Detailed survey on nannofossils at the Kansai International Airport
in Osaka Bay, Central Japan. In nakaseko, k. (ed.). Geological survey of the submarine strata at the Kansai
International Airport in Osaka Bay, Centred Jeipein, 19-28. Calamity Science Institute, Osaka. [In Japanese ]
pannella, G. 1972. Paleontological evidence on the Earth’s rotational history since Early Precambrian.
Astrophys. Space Sci. 16, 212-237.

— 1975. Paleontological clocks and the history of the Earth’s rotation. In rosenberg, g. d. and runcorn,
s. K. (eds.). Growth rhythms and the history of the earth's rotation , 253-284. Wiley, London.

— 1976. Tidal growth patterns in recent and fossil mollusc bivalve shells: a tool for the reconstruction of
paleotides. Naturwissenschaften , 63, 539-543.

— and MaccuNTOCK, c. 1968. Biological and environmental rhythms reflected in molluscan shell growth.
J. Paleontol. 42 (Mem. 2), 64 80.

- — and Thompson, m. n. 1968. Paleontological evidence of variations in length of synodic month since
Late Cambrian. Science, Washington, 162, 792-796.

rhoads, d. c. and pannella, g. 1970. The use of molluscan shell growth patterns in ecology and paleoecology.
Lethaia, 3, 143-161.

Richardson, c. A., crisp, d. j. and runham, n. w. 1979. Tidally deposited growth bands in the shell of the
common cockle, Cerastoderma edule ( L . ) . Malacologia, 18, 277-290.

— 1980u. Factors influencing shell growth in Cerastoderma edule. Phil. Trans. R. Soc. B210,

513-531.

19806. An endogenous rhythm in shell deposition in Cerastoderma edule. J. mar. biol. Ass.
UK, 60, 991 1004.

— 1981. Factors influencing shell deposition during a tidal cycle in the intertidal bivalve
Cerastoderma edule. Ibid. 61, 465-476.

— and gruffydd, ll. d. 1980. The use of tidal growth bands in the shell of Cerastoderma edule
to measure seasonal growth rates under cool temperate and sub-arctic conditions. Ibid. 60, 977-989.

runcorn, s. k. 1967. Corals and the history of the earth’s rotation. Sea Front. 13, 4-12.
scrutton, c. t. 1978. Periodic growth features in fossil organisms and the length of the day and month. In
brosche, p. and suendermann, j. (eds.). Tided friction and the earth's rotation, 154 196. Springer, Berlin,
Heidelberg, New York.

taylor, j. d., Kennedy, w. j. and hall, a. 1969. The shell structure and mineralogy of the Bivalvia.

Introduction. Nuculacea-Trigonacea. Bull. Br. Mus. nat. Hist. (Zool.) Suppl. 3, 1-125.
termier, h. and termier, g. 1975. Sedimentary behaviour and skeletal textures available in growth cycle
analysis. In rosenberg, g. d. and runcorn, s. k. (eds.). Growth rhythms and the history of the earth's
rotation , 89-102. Wiley, London.

wells, j. w. 1963. Coral growth and geochronometry. Nature, Loud. 197, 948-950.

TERUFUMI OHNO

Department of Geology and Mineralogy
Faculty of Science

Typescript received 16 February 1988 Kyoto University

Revised typescript received 1 August 1988 606 Kyoto, Japan

A LATE PERMIAN FRESHWATER SHARK FROM
EASTERN AUSTRALIA

by MICHAEL R. LEU

Abstract. A new genus and species of elasmobranch, Surcaudalus rostratus , is described from the Late
Permian Rangal Coal Measures, Blackwater, central Queensland. Surcaudalus is characterized by a
palatoquadrate with a well-developed ethmoidal articulation, cladodont (phoebodontiform) dentition, absence
of ribs, dorsal fin-spines with an anterior keel and a flat to concave posterior wall whose posterolateral
margins bear three rows of barb-like denticles, and a non-lunate caudal fin with a well-developed epicaudal
lobe. The phylogenetic significance of fin-spine characteristics, caudal fin morphology, and broad, expanded
occipital segments is discussed.

The Rangal Coal Measures in the Utah Development Company’s open-cut coal mine, 20 km
south-south-west of Blackwater, central Queensland (text-fig. 1), contain several lacustrine mass-
mortality horizons that have yielded a genus of the Bobasatraniformes (Campbell and Duy Phuoc
1983), at least twelve new genera of Palaeonisciformes, and two new genera of Elasmobranchii.

|Palaeontology, Vol. 32, Part 2, 1989, pp. 265-286, pis. 33-34.|

© The Palaeontological Association

266

PALAEONTOLOGY, VOLUME 32

o

O

o

AGE

LLI

Z

o

N

SPORE-

POLLEN

SPORE-

POLLEN

AGE
n Ma

0 25
< <-

cc

o

1 CC
LU

ZONES

ZONES

LU

LU

Cl

East & South

~D
( / )

West Australia

Australia

_

cn

>-

r- 243 —

CO

T.PLAYFORDII

A.TENUISPINOSUS

“

<

GC

cc

<

SCYTHIAN

UJ

P.SAMOILOVICHII

P.SAMOILOVICHII

f—

LU

9/|«

cr

o

CL

CO

O

_l

<

L.PELLUCIDUS.

L.PELLUCIDUS

“

250-

2

LU

TATAR IAN

P. MICROCORPUS

P. MICROCORPUS

- 250

-

cr

UJ

t—

<

°53

KAZANIAN

UPPER STAGE 5

UPPER STAGE 5

text-fig. 2. Correlation of the Late Permian to Early Triassic palynological zones of west, east, and south

Australia (modified after Helby et al. 1987).

The Rangal Coal Measures, the uppermost formation in the Late Permian Blackwater Group,
comprises coal, labile sandstone, carbonaceous shale, siltstone, and mudstone, and is between
107 m and 137 m thick near the Utah Development Company’s mine (Staines 1972; Burgis 1975).
Mortality horizons occur frequently above the Argo Seam; the latter averages 6-1 m in thickness,
dips gently eastwards, and occurs towards the base of the Rangal Coal Measures.

Truswell (in Campbell and Duy Phuoc 1983) identified spores and pollens from a mortality
horizon 6 m above the Argo Seam. She regarded the assemblage as being high in the Upper Stage
5 of Kemp et al. (1977); this is equivalent to a Late Permian (Late Kazanian) age (text-fig. 2).

Three closely spaced mortality horizons occur at a site known as Clinker Hill. The new
elasmobranchs were collected from the middle horizon; it is separated vertically by 300 mm and
400 mm respectively from the horizons immediately above and below. The uppermost mortality
horizon is covered by between 100 mm and 500 mm of overburden, enabling extensive lateral
excavation. The underlying coal seam has ignited, and the original light-grey shales and mudstones
have been oxidized red and baked ‘brick’ hard, thus faithfully preserving the fossils from the
rigours of surface weathering.

PALAEOENVIRONMENTAL INTERPRETATION

Burgis (1975) viewed the Rangal Coal Measures as having accumulated in an environment
characterized by laterally migrating river channels and associated floodplain and lacustrine deposits.
She envisaged peat accumulating in extensive swamps above point-bar sands and muddy overbank
sediments. Differential compaction was thought to have resulted in shallow lakes forming in
depressions over former peat swamps. Burgis (1975, p. 56) interpreted the sequence of events in
the development of the lake basins as one in which ‘compactional growth of mounds . . . disrupted
the system of ephemeral channels and small lakes’.

All three horizons have been excavated, each over an area of 8 m x 4 m. The two lower horizons
are about 5 mm thick; the uppermost horizon is variable and up to 50 mm in thickness. All contain
ontogenetically young fishes (chondrosteans and elasmobranchs); much larger specimens of the
same species have been found in drag-line dumps. The fish are densely packed within the mortality
horizons. Those in the uppermost horizon were obviously buried rapidly in a turbid mass flow, as
evidenced by rip-up clasts and by fish contorted and buried in all orientations; this may have been
caused by mass slumping in a lake or by rapid burial in overbank splay deposits.

LEU: PERMIAN FRESHWATER SHARK

267

Abbreviations and symbols used in text and text-figures

Anatomical: AP, articular process; BCDF, basal cartilage of dorsal fin; CB, ceratobranchials; CT, cusps of
teeth; DF, dorsal fin; DFS, dorsal tin-spine; EART, ethmoidal articulation; ECL, epicaudal lobe; FECL,
fragmentary epicaudal lobe; FS, fleshy snout; FVFIL, fragmentary ventral hypochordal lobe; LHL,
longitudinal hypochordal lobe; LSL, lateral sensory line; MC, Meckel’s cartilage; MSP, mesopterygium; MTP,
metapterygium; NCR, neurocranium; NE, neural element; NM, notochordal mass; OCCR, occipital region;
OR, orbit; PF, pectoral fin; POP, postorbital process; PQ, palatoquadrate; PRP, propterygium; SC,
scapulocoracoid; SCD, sensory canal denticles; STL, subterminal lobe; TBR, tooth bearing ramus; TBRMC,
tooth bearing ramus of the Meckel’s cartilage; TBRPQ, tooth bearing ramus of the palatoquadrate; VHL,
ventral hypochordal lobe; W, white unoxidized region; WCRPQ, weakly calcified region of the palatoquadrate.

Institutional: AMF, Australian Museum Fossil; CMNH, Cleveland Museum of Natural History; QMF,
Queensland Museum Fossil.

Other: PDH, plugged drill hole.

Symbols: regular heavy stipple indicates calcified cartilage; diagonal cross hatching indicates damaged
cartilage surfaces; lines comprising alternating dashes and dots indicate incomplete outline of body.

SYSTEMATIC PALAEONTOLOGY

Class chondrichthyes Huxley 1880
Subclass elasmobranchii Bonaparte 1838
Order euselachii Hay 1902
Superfamily incertae sedis
Genus surcaudalus n. gen.

Etymology. From the Latin sur (above, on) and cauda (tail); m reference to the well-developed epicaudal
lobe.

Type species. Surcaudalus rostratus n. sp.; monotypy.

Diagnosis. Small elasmobranch with a blunt fleshy snout extending beyond the mandibular
symphysis; palatoquadrate with a well-developed ethmoidal articulation; Meckel’s cartilage
relatively deep and robust; dentition cladodont (phoebodontiform); ribs absent; dorsal fin-spines
with an anterior keel and a flat to concave posterior wall whose posterolateral margins bear,
apically, three pairs of small closely spaced barb-like denticles; lateral surfaces of fin-spines having
three major longitudinal costae; caudal fin comprising 26-27% of the total body length and
possessing a well-developed epicaudal lobe.

Surcaudalus rostratus n. sp.

Plates 33 and 34; text-figs. 3 1 I

Diagnosis. As for genus.

Etymology. From the Latin rostratus (rostrum), referring to the elongate fleshy snout extending beyond the
mandibular symphysis.

Holotype. QMF14470A and QMF14470B, part and counterpart respectively, of an articulated skeleton of a
shark, minus the posterior portion of the caudal fin; exposed in lateral view.

Paratypes. AMF72559A and AMF72559B, part and counterpart respectively, of an articulated skeleton of
a shark exposed in lateral view. QMF14471A and QMF14471B, part and counterpart respectively, of an
articulated skeleton of a shark, minus the head region, exposed in lateral view.

Horizon and locality. Rangal Coal Measures of the Blackwater Group (Late Permian [Late Kazanian], high
in the Upper Stage 5 of Kemp et al. 1977), Blackwater, central Queensland, Australia.

268

PALAEONTOLOGY, VOLUME 32

text-fig. 3. Surcaudalus rostratus n. gen. n. sp. paratype (AMF72559A). a, head and pectoral girdle in lateral

view, x2-3. b, drawing of the same specimen.

DESCRIPTION

Body form. All specimens of Surcaudalus rostratus n. sp. are ontogenetically young individuals with slender
fusiform bodies. AMF72559 is 193 mm long and 19 mm deep at its widest point along the trunk. It has a
subterminal mouth and a caudal fin which comprises 26-27% of the total body length. It has a tapering,
blunt, fleshy snout whose length is at least 24-26% that of the neurocranium.

Neurocranium The neurocrania of both QMF14470 (text-figs. 5 and 6) and AMF72559 (text-figs. 3 and 4)
are preserved in lateral view; both are extensively crushed and distorted. The rostrum does not extend
beyond the anterior margin of the palatoquadrates. The postorbital process is very prominent and long
(anteroposterior), with a gently concave anterior margin. The posterior margin of the orbit is situated at 57
67% of the length of the neurocranium (measured from the posterior). The otico-occipital region is moderately
long and is tentatively estimated to comprise 38 -43% of the total length of the neurocranium.

EXPLANATION OF PLATE 33

Figs. 1 -3. Surcaudalus rostratus n. gen. n. sp. 1, an almost complete specimen of the new genus (AMF72559A)
in lateral view, x 1. The circular feature is a plugged drill hole. 2 and 3, part and counterpart of the
holotype (QMF14470A and QMFI4470B, respectively) in lateral view, x 1-5.

PLATE 33

LEU, Australian late Permian freshwater shark

270

PALAEONTOLOGY, VOLUME 32

text-fig. 4. Surcaudalus rostratus n. gen. n. sp. paratype (AMF72559B). a, head and pectoral girdle in lateral

view, x2-3. b, drawing of the same specimen.

A robust, extensively calcified band of cartilage (TBR of text-figs. 3B 6B) extends posterodorsally from
the anterior extremity of the palatoquadrates of both QMF14470 and AMF72559. This cartilage passes over
the ethmoidal articulation and through the dorsolateral portion of the orbit. It exhibits numerous cusps in
all orientations, particularly along its anterior half. Its interpretation remains enigmatic; it may represent the
tooth-bearing ramus of the palatine portion of a palatoquadrate that was displaced dorsally during
compression, ultimately breaking through the weakly calcified ethmoidal and orbital regions.

Visceral skeleton

Palatoquadrate. QMF 14470 and AMF72559 have their palatoquadrates preserved in lateral view. The palatine
portion has a distinct zonal differentiation in the degree of chondrification. The tooth-bearing ramus is
narrow, elongate, and extensively calcified. In AMF 72559 the ventral margin of the ramus is convex ventrally,
as opposed to QMF14470 where it is inclined gently posteroventrally and is weakly dorsally convex. The
condition in AMF72559 has resulted from post-mortem distortion. Portions of the cusps of many teeth are
visible along the anterior half of the ramus. In AMF72559 the majority belong to non-functional teeth as
their apices are dorsally oriented. The palatine is sparsely calcified immediately dorsal to the ramus. This
region comprises 24-29% of the palatine’s dorsoventral height along a line oriented vertically beneath the
posterior half of the ethmoidal articulation; anteriorly its area diminishes. The remainder of the palatine
(which constitutes the majority of its area) is extensively calcified. The posteroventral margin of this region

LEU: PERMIAN FRESHWATER SHARK

271

text-fig. 5. Surcaudalus rostratus n. gen. n. sp. holotype (QMF14470A). a, head and pectoral girdle in lateral

view, x2-5. b, drawing of the same specimen.

slopes posteroventrally towards the mandibular articulation. This margin is a robust, narrow, gently convex
ridge (best preserved in AMF72559B) and appears to be a raised portion of the palatoquadrate rather than
a labial cartilage (text-figs. 3 and 4).

Along the anterodorsal margin of the palatoquadrate a well-developed ethmoidal articulation rises (from
the anterior edge of the palatoquadrate and extending to what appears to be the anteroventral edge of the
orbit) at 23-25° to the notochordal axis. In QMF 14470 the ethmoidal articulation is reinforced by extensive
chondrification. Heading backwards from the ethmoidal articulation the dorsal surface of the palatoquadrate
rises into the floor of the orbit. Its slope steepens as it passes along the anteroventral portion of the postorbital
process. Unfortunately, the dimensions of the otic portion of the quadrate region and the presence or absence
of a postorbital articulation cannot be accurately determined. Traces of cartilage in this region are irregular;
there is uncertainty in differentiating neurocranial from palatoquadrate cartilage. In AMF72559A a lateral
expansion of the ventral quadrate region is present but no trace of the otic portion is preserved. In
AMF72559B extensive calcified cartilage is present but its origin is uncertain. In QMF14470 the dorsal
portion of the quadrate region terminates abruptly and is posteriorly concave. In QMF14470B a roughly
arcuate, posteriorly convex band of cartilage extends dorsally from the posteroventral half of the quadrate
region.

272

PALAEONTOLOGY, VOLUME 32

text-fig. 6. Surcaudalus rostratus n. gen. n. sp. holotype (QMF14470B). a, head and pectoral girdle in lateral

view, x2-5. b, drawing of the same specimen.

Meckel's cartilage. The lower jaw of QMF14470 deepens rapidly posteriorly and is 84% of the height of the
palatoquadrate directly beneath the anterior edge of the postorbital process. Meckel’s cartilage is laterally
convex beneath the tooth-bearing ramus. The ventral margin (heading posteriorly) is inclined posteroventrally;
it turns sharply posterodorsally beneath the poorly preserved articular region. The tooth-bearing ramus is
extensively calcified and robust relative to the majority of Meckel’s cartilage. It is weakly convex dorsally
and tapers posteriorly. It exhibits numerous central cusps of teeth of disparate preservation, particularly
along its anterior half. Meckel’s cartilage of AMF72559 is relatively distorted with the tooth-bearing ramus
poorly preserved along its anterior half. Its posteroventral border is crushed and attenuated.

Axial skeleton

Notochord. The notochord is unconslricted, persistent, and uncalcified. QMF14470 and AMF72559 contain
calcified neural cartilages of disparate preservation (PI. 33, figs. 1-3; PI. 34, fig. 7; text-figs. 3a (arrowed), b
and 7). In the cervical region of AMF72559, sparse traces of cartilage dorsal to the notochord indicate the
presence of weakly calcified neural elements of uncertain form. In its thoracic region, extending backwards

LEU: PERMIAN FRESHWATER SHARK

273

text-fig. 7. Surcaudalus rostratus n. gen. n. sp. paratype (AMF72559A); articulated specimen in lateral view.

from the ventral tip of the anterior fin-spine to the posterior margin of the basal plate of the second fin-
spine, there is a series of between thirty-four and thirty-seven neural cartilages whose apices extend dorsally
beyond the most ventral point of the fin-spines. The first six (approximately) neural cartilages occur ventral
to the anterior basal plate and are preserved as irregular patches of calcified cartilage of indeterminable
morphology. Immediately posterior to these, the series is better preserved and the vertebral cartilages consist
of calcified, wide-based neural arches bearing tapering neural spines. They are 3 mm long, closely spaced,
presumably paired and unfused, and are inclined at 50-55° to the notochordal axis. The anteroventral corners
of the neural arches are anteriorly elongated. There are six neural cartilages beneath the posterior basal plate;
they are shorter and less robust than those anterior to them. Posteriorly only minor traces of neural
calcifications are visible, though presumably they extended to the tip of the dorsal lobe of the caudal fin but
were not, or only weakly, calcified. In QMF14470 a series of approximately thirty-two neural cartilages is
present between the inserted portions of the fin-spines.

No haemal calcifications are present in the cervical and thoracic regions. In QM FI 4471 A there are at least
twelve slender, tapering rocls extending backwards for a short distance from the anterior margin of the anal
fin. These haemal cartilages are 3-5 mm long and are inclined posteroventrally at 40° to the notochordal axis.

Dorsal fins. The anterior and posterior dorsal fins (PI. 33, figs. 1-3; text-fig. 7) are situated above, and extend
posteriorly beyond, the posterior half of the pectoral and pelvic fins, respectively. Both fins extend dorsally
above the apices of the fin-spines, with the anterior fin being the largest of the two. The anterior dorsal fin
of AMF72559 (text-fig. 7) is detached from the posterior edge of the anterior fin-spine at a point just ventral
to the origin of the most ventral barb-like denticle. Both fins are mildly folded and contorted indicating they
were relatively flexible during life.

Most of each fin is preserved, enabling accurate estimation of their dimensions; nevertheless the precise
lateral outline of the fins is unknown because their dorsal tips and posterior borders are not preserved.

Both fins are supported by a ‘triangular’ cartilaginous basal plate (PI. 34, fig. 7; text-fig. 10), whose anterior
portion is inserted in the posteroventrally situated basal notch of the fin-spine. Each basal plate consists of
an expanded dorsal portion and a tapering, anteroventrally directed process extending to near the dorsal
margin of the notochord. The expanded dorsal portion of the anterior basal plate occupies a smaller area
than that of the posterior basal plate. Conversely, its anteroventrally directed process is wider and
approximately one and a half limes longer than the latter.

Radial cartilages, if present, were not calcified.

Anal fin. The anterodorsal portion of the anal fin is preserved on QMF14471; its origin is indicated by a
slight widening of the body in this region. The most anterior point of the fin is located immediately ventral
to the ventral tip of the most anterior haemal element. The dimensions and configuration of the anal fin are
unknown.

274

PALAEONTOLOGY, VOLUME 32

text-fig. 8. Surcaudalus rostratus n. gen. n. sp. para-
type (AMF72559B); notochordal mass and portions
of appended fin lobes.

Caudal fin. In AMF72559 and QM FI 4471 the entire notochordal mass and portions of the appended fin
lobes are present (PI. 33, fig. 1; PI. 34, fig. 4; text-figs. 7 and 8).

The following description is based on characters (text-fig. 12) defined by Thomson (1976, p. 19). It should
be noted that many authors when referring to the epicaudal lobe are in fact describing the notochordal mass,
rather than a distinct dorsally appended fin lobe.

The caudal fin length of Surcaudalus is estimated to comprise 26-27% of the total body length. The
notochordal mass is inclined at 17-25° (heterocercal angle) to the main anterior body axis. The tip of the
caudal fin is characterized by a subterminal lobe (PI. 34, fig. 4). Portions of the epicaudal lobe are preserved
in AMF72559 (PI. 34, fig. 4; text-figs. 7 and 8). It extends to the posterior perimeter of the plugged drill hole;
its height (as preserved) is equivalent to that of the notochordal mass, indicating that Surcaudalus had an
elongate, well-developed epicaudal lobe.

The preserved portions of the ventral hypochordal lobe are insufficient to enable an accurate reconstruction
of its shape and dimensions. However, the functional analyses of Thomson and Simanek (1977) suggest that
Surcaudalus possessed at least a moderately well-developed (in terms of the ventral hypochordal lobe
development in squaloids) ventral hypochordal lobe (text-fig. 1 1). They determined that the presence of an
epicaudal lobe requires a larger angle of rotation for stability and that it significantly reduces the magnitude
of the upward force components in the tail. They concluded that sharks (e.g. Surcaudalus) which possess low
to moderate dorsal thrust angles and a well-developed epicaudal lobe are characterized by a moderately well-
developed ventral hypochordal lobe, because it is essential in order to maintain a balanced thrust that passes
through the centre of balance. Such sharks are usually characterized by a small longitudinal hypochordal
lobe and a large subterminal lobe. Thomson and Simanek (1977, p. 347) also established that a subterminal
lobe is crucial in sharks with a low dorsal thrust angle because it ‘acts somewhat like a kite, passively helping
the fish maintain the correct angle of rotation along the whole of the caudal fin and protecting the tip from
fluttering’.

Appendicular skeleton

Pectoral girdle. This consists of a pair of dorsoventrally elongate scapulocoracoid cartilages (PI. 33, figs. 1
3; text-figs. 3-6) that were presumably not fused along the ventral mid-line and which extend dorsally to just
beyond the most ventral point of the anterior dorsal fin-spine. The long scapular portion is inclined
posterodorsally at an angle of 9-10° to the notochordal axis. It tapers moderately dorsally until 70-73% of
its length, whence the posterior margin curves sharply anterodorsally and the scapular narrows rapidly to a
rounded, blunt apex. Its anterior and posterior margins are weakly concave and convex, respectively. In
AMF72559 the anteroventral flank of the scapular process was shattered during compression. This region is
crushed and attenuated in QM FI 4470. Traces of the cartilage of the articular process and the coracoid
portion are best preserved in AMF72559B (text-fig. 4), preservation is poor and the following reconstruction
must be regarded as tenuous.

The dorsal edge of the articular process extends posteroventrally; its posterior margin is bluntly rounded
and projects only slightly beyond the posterior margin of the coracoid; its ventral edge is distinguished from
the latter by a distinct change in slope.

The coracoid portion is widest (antero-posterior) immediately ventral to the articular process, is inclined
and narrows moderately anteroventrally before extending and tapering anterodorsally. Its dorsal and ventral
margins are concave and convex, respectively. The posteroventral section of the coracoid is extremely poorly
calcified relative to the portion immediately dorsal to it (text-fig. 4). The ventral portion of the coracoid

LEU: PERMIAN FRESHWATER SHARK

275

extends beyond the ventral margin of the body and indicates that, in life, this region was inclined towards
the mid-line.

Pectoral fins. A pectoral fin lobe is present in QMF14470 and AMF72559 (text-fig. 7). In both specimens it
is distorted with the dorsal half preserved against the ventral portion of the flank, its precise shape and
dimensions are indiscernible.

Weakly calcified cartilages of the basipterygial region (text-fig. 4) are present in AMF72559B. The
metapterygium has been dislocated dorsally and lies immediately above the articular process with its proximal
margin against the posterior border of the scapular. It is 5-4 mm long and 1 -6 mm wide; approximately
rectangular in outline with a concave, anteroventrally inclined posterior border. Two irregular patches of
cartilage, which may be traces of the mesopterygium and the propterygium (text-fig. 4), are located just
posterior to the articular process.

Pelvic girdle and fins. In QMF 14470 a calcified pelvic cartilage of indeterminable morphology is located at
the ventral margin of the body, immediately anterior to the most ventral point of insertion of the posterior
fin-spine. Portions of the right fin lobe are present in QMF 14471 (PI. 34, figs. 5 and 10) but they are
insufficient to enable an accurate reconstruction of its outline and dimensions.

Dermal skeleton

Denticles. The head, trunk, and notochordal mass exhibit a shagreen of stellate (Petrodus- like) denticles (PI.
34, figs. 5, 8, 10) whose crowns consist of smooth, rounded ridges radiating out from a central pointed axis.
The radial ridges vary in number from eight to twelve, are generally unevenly spaced, rarely bifurcate, and
distally taper. The perimeter of the denticles is usually circular to rhombic in outline; the denticles, along the
trunk, are 0-42 mm-0-76 mm in diameter. Occasionally two denticles appear to have fused or coalesced. All
denticles are preserved as external moulds in crown view and consequently the morphology of their bases is
unknown.

The lateral flanks of the trunk are densely covered in denticles which ventrally tend to diminish somewhat
in diameter. Posterior to the posterior dorsal fin they gradually diminish in size and become very sparse
along the terminal portion of the notochordal mass. The denticles are sparse in the ventral half of the lateral
flank adjacent to the anal fin; a dense covering is maintained in the dorsal portion. The dorsal half of the
branchial area and head are densely covered in denticles; they appear to be absent along the ventral portion
of the branchial region and are scattered along the rostrum, although this may be an artifact of preservation.

Minute denticles of indeterminable morphology (inferior preservation) are present on all fin webs. The
dorsal fins are extensively covered in denticles that appear rectangular in outline and are aligned in
posteroventrally inclined rows. The pelvic fins exhibit a sparse covering of denticles anteriorly; posteriorly
they become fewer and finally absent. All other fin lobes have denticles but their density and distribution is
uncertain. The notochordal mass exhibits closely packed stellate denticles for most of its length; they become
very sparse towards the origin of the subterminal lobe.

Lateral line. The lateral sensory line is indicated by dual, parallel rows of closely spaced denticles. They are
well preserved along the trunk (text-figs. 3 and 7) where the denticles of the dorsal row tend to be elongated
anteroposteriorly and are approximately horizontally aligned. Those of the ventral row are triangular in
outline and 0-37 mm-0-45 mm in length from their apices to the mid-line of their bases. The vertical spacing
between the apices of the ventral denticles and the ventral edge of the dorsal denticles is 0-45 mm 0-52 mm
along the trunk; this gap decreases (tapers) posteriorwards.

The lateral line (text-fig. 3), extending posteriorly from the posterior margin of the neurocranium, arches
gently dorsally across the scapulocoracoid adjacent to the region where the scapular portion begins to rapidly
narrow dorsally, beyond which it levels out and continues posteriorly, immediately ventral and more or less
parallel to the notochordal axis. It appears to terminate close to the ventral margin of the posterior half of
the notochordal mass of the caudal fin.

Teeth. Both QMF14470 and AMF72559 contain numerous teeth that are, with few exceptions, very poorly
preserved. QMF14470B exhibits three disarticulated teeth (PI. 34, fig. 1) along the dorsal margin of the
palatine portion of the palatoquadrate. These teeth are phoebodontiform; the crown is tricuspid, comprising
a central cusp and two lateral cusps that diverge at 18-5-27-50 from the central cusp. The central cusp appears
to be approximately the same length as the lateral cusps, although the apices of most cusps are damaged.
The teeth are preserved in labial view; their bases are 0-35 mm wide, extremely narrow and gently arched
apically. The teeth are 015 mm— 0 18 mm long from the ventral margin of the base to the dorsal tip of the

276

PALAEONTOLOGY, VOLUME 32

text-fig. 9. Surcaudalus rostratus n. gen. n. sp. paratype (QMF14471B). c, dorsal half of the anterior dorsal
hn-spine in lateral view; note most of the anterior keel (right-hand side) is missing, apart from vestigial
remnants, x 12. b, e, g, transverse sections through the same fin-spine, all figures x 20. The specimen in text-
fig. 9g has been distorted by lateral compression, so that the posterior border is obliquely aligned to the
sagittal axis of the fin-spine, a, d, lateral and posterior views, respectively, of the specimen shown in text-fig.
9e, both figures x 20. f, posterior view of the specimen shown in text-fig. 9h, x 20. Note costae along the

posterior surface.

central cusp. QMF14470A has a tooth (PI. 34, fig. 2) preserved in labial view on the Meckel’s cartilage, close
to the mandibular symphysis; it is 0-42 mm wide and 0-35 mm long. The lateral cusps diverge at 19-21° from
the central cusp.

Several poorly exposed teeth, the majority of which have only the central cusp visible, are present on

EXPLANATION OF PLATE 34

Figs. 111. Surcaudalus rostratus n. gen. n. sp. 1, disarticulated teeth of QMF14470B, x42. 2, tooth of
Meckel’s cartilage of QMF14470A, in labial view, x73. 3, teeth of the anterior portion of Meckel’s
cartilage of AMF72559A, in labial view. Note only the central cusps are exposed, x45. 4, caudal fin of
AMF72559A showing notochordal mass, lateral line, fragmentary traces of the epicaudal lobe (arrowed),
and the subterminal lobe, x 1-7. 5, portion of the pelvic fin of QMF14471A (arrowed), note sparser
density of denticles relative to the ventral flank of the trunk, x3-5. 6, external mould of the posterior
dorsal fin-spine of AMF72559B, x 3. 7, posterior dorsal fin-spine, basal plate, and neural elements of
AMF72559A, x2-5. 8, details of the dermal denticles of the anteroventral flank of QMF14471A, x 7.
9, apical view of the posterior dorsal fin-spine of AMF72559A, note the posterolaterally placed denticles,
x 7. 10, QMF14471 A showing dense shagreen of dermal denticles and the basal portion of the posterior

dorsal fin-spine, x 1. The dorsal margin of the trunk is overlain by a Glossopteris sp. 11, posterior dorsal
fin-spine of AMF72559A, note anterior keel, x 4-7.

PLATE 34

LEU, Australian late Permian freshwater shark

278

PALAEONTOLOGY, VOLUME 32

text-fig. 10. Surcaudalus rostratus n. gen. n. sp. paratype (AMF72559A). a, anterior dorsal fin-spine, basal
cartilage of dorsal fin, neural elements and dorsal portion of scapular region of the scapulocoracoid. b,
posterior dorsal fin-spine, basal cartilage of the dorsal fin and neural elements.

Meckel’s cartilage and the palatoquadrate of AMF72559. A single tooth (PI. 34, fig. 3) on Meckel’s cartilage
of AMF72559A (located beneath the anterior end of the orbit) has the medial portion of the labial side
uncovered. The tooth is 029 mm long from the dorsal tip of the central cusp to the ventral margin of the
base and, at least (distance exposed) 0-25 mm wide along the base. The central cusp is gently inclined
posterodorsally, its maximum basal width being 0 075 mm. The flanks lateral to the principal cusp are
obscured by matrix and too fragile to be prepared, thus the number of lateral cusps is uncertain. A small
inflexion, inclined posterodorsally at 74° to the ventral margin of the base, occurs approximately midway
along the base. Laterally, the flanks on either side of this ridge are gently directed lingually. The spacing
between the mid-points of the principal cusps of the teeth in this region is 0-38 mm; this would also represent
the maximum basal width of a single tooth. A broken tooth (situated two teeth posteriorly from the tooth
described above) exposing the inner surface of the lingual side demonstrates that the base is lingually
expanded. No trace of ornament is preserved on any tooth.

Fin-spines. A fin-spine is present along the anterior margin of both dorsal fins (PI. 33, figs. 1 -3). In AMF72559
the anterior and posterior fin-spines are both 2-6 cm in length and are inclined to the notochordal axis at
56-58° and 69-71°, respectively; they curve gently posteriorly towards their apices. In QMF14470 the anterior
and posterior fin-spines are inclined, respectively, at 53° and 64° to the notochordal axis.

The lateral surfaces of both fin-spines bear three major longitudinal costae (PI. 34, figs. 6 and 11; text-figs.
9 and 10). The costae fade approximately adjacent to the dorsal limit of the triangular basal plate. Dorsally
they tend to merge and become indistinct adjacent to the barb-like denticles. Fine longitudinal striations
cover the entire surface of the fin-spine; they are more or less parallel but do coalesce in places. The anterior
and posterior costae bear five and four striae, respectively; the middle costa has four striae increasing to five

LEU: PERMIAN FRESHWATER SHARK

279

text-fig. 11. A reconstruction of Surcaudalus rostratus n. gen. n. sp. The outline of all fins is hypothetical,
although the dimensions (see text) of the dorsal and caudal fins are regarded as accurate.

ventrally. The posterior surface of the fin-spines is flat to shallowly concave (text-fig. 9b, f, h) and possesses
at least three longitudinal costae (text-fig. 9e, g). It bears along its posterolateral margins three pairs of
closely spaced, small, barb-like, posteriorly directed denticles that increase in size ventrally (PI. 34, fig. 9;
text-fig. 10a, b). The ventral and dorsal limits of the denticle series occur at, respectively, 76% and 88% of
the distance from the most ventral point of the fin-spine to the apex. Their posterior and ventral margins
are inclined, respectively, at 17° and 45° to the posterolateral surface of the spine. A large median costa,
resembling a carina, is present on the anterior surface of the fin-spine. It narrows dorsally and becomes
indistinct close to the apex; it tapers anteriorly in transverse section. The most posterior striae on the anterior
costal ridge are irregularly disjunct in places but remain more or less aligned with the longitudinal axis of
the fin-spine. The most anterior series (comprising at least three rows of striations) are often obliquely inclined
anteroventrally and disjunct.

A basal opening for the triangular supportive cartilage is present along the posterior margin of both fin-
spines. It extends dorsally to a point of closure located 34% of the way towards the apex.

The widest (antero-posterior) point of the fin-spines is proportional to 12% of their length and is located
adjacent to the dorsal closure of the basal notch. Dorsally their widths decrease evenly until about the
position of the most ventral barb-like denticle, after which it decreases rapidly towards their pointed apices.
It rapidly decreases ventrally from the point of insertion.

FUNCTIONAL MORPHOLOGY

The following interpretation is based on studies (Thomson 1976; Thomson and Simanek 1977) of
the body form and locomotion in sharks, specifically the mechanical action of the heterocercal
tail.

Thomson and Simanek demonstrated that for a shark to cruise efficiently, the line of net thrust
from the caudal fin must pass through its centre of balance. To attack prey or evade predation, a
shark can turn suddenly in any direction by altering the balance of forces acting in the tail, thus
causing powerful turning moments about the centre of balance. The head will revolve dorsally and
ventrally, respectively, if the principal forward thrust is directed either posterior or anterior to the
centre of balance. It can then employ its pectoral fins to manoeuvre swiftly at any angle.

The caudal fin of Surcaudalus is characterized by a heterocercal angle of between 17 (estimated,
QM FI 4470) and 25° (AMF72559), a dorsal thrust angle (Thomson and Simanek 1977, p. 346)
estimated to be between 7-5-10°, a large epicaudal lobe, a subterminal lobe, and a ventral
hypochordal lobe.

The moderate heterocercal angle of Surcaudalus indicates that it would have been capable of
producing relatively strong turning moments around the centre of balance, enabling it to rapidly
and efficiently change direction.

Thomson (1976) determined that the epicaudal lobe produces a thrust that is antagonistic to the

280

PALAEONTOLOGY, VOLUME 32

text-fig. 12. Characteristics of the heterocercal tail,
after Thomson (1976), showing the notochordal mass
and various appended fin lobes. Based on the caudal
fin of Centrophorus molluccensis Bleeker, 1860.

hypobatic effect of the notochordal mass and the longitudinal hypochordal lobe, resulting in a
lowering of the centre of effort and restricting the dimensions of the heterocercal and dorsal thrust
angles. He concluded that sharks (specifically squaloids) possessing a well-developed epicaudal
lobe and low to intermediate dorsal thrust angles (intermediate angles range from 10° to 25°) are
characterized by slow cruising speeds. At high speeds, such sharks would not be capable of
maintaining in balance the various thrusts developed by the respective fin lobes. The net thrust
would no longer be directed through the centre of balance and the shark would be unable to
maintain an even keel in a horizontal plane.

In summary, Surcaudalus , when active, would have been capable of high manoeuverability, slow
cruising speeds, and incapable of sustaining high speeds.

In many Recent sharks, particularly the pelagic species, the posterior dorsal fin is typically small
compared to the anterior dorsal fin. This highly adaptive condition results in thrust enhancement
(Lighthill 1975; Webb and Keyes 1982) through hydrodynamic interactions between the first dorsal
fin and the caudal fin. This interaction improves cruising efficiency but limits transient swimming
(fast starts and powered turns) performance because of the reduced body area (Webb and Keyes
1982). The large second dorsal fin of many Palaeozoic sharks ( Ctenacanthus costellatus , Goodrichthys
eskdalensis, Tristychius arcuatus , Dabasacanthus inaskasi, and Surcaudalus rostratus) suggests a
greater reliance on transient swimming relative to most Recent sharks. The major variable affecting
thrust in fast-start kinematics (Webb and Keyes 1982) is the distribution of body depth (to
maximize the mass of water accelerated), especially in the caudal area. Webb and Keyes (1982)
noted that large increments in fast-start performance are developed with relatively small increases
in fin and body depth. They recognized that among actinopterygians, at each successive adaptive
level, a recurring morphology is the fusiform carnivore with a design favouring transient swimming.
They concluded this could be because of the importance of fast starts and powered turns in critical
activities (particularly for juveniles) such as catching prey and avoiding predators. The more
generalized morphology of those Palaeozoic sharks with enhanced transient swimming capabilities
would be at the expense of steady swimming (cruising), because the large posterior dorsal fin would
increase frictional drag (the fins of sharks are not collapsible) and significantly impair thrust
enhancement between the median fins.

DISCUSSION

Several fin-spine characteristics of euselachians— concave posterior wall, posterolaterally situated
denticles, and posteriorly placed central cavity — are widely shared among groups (Rieppel 1982)
such as xenacanths, ctenacanths, hybodonts, and neoselachians (Table 1). I concur with Dick
(1978, p. 107) that the similarities between ctenacanth and neoselachian fin-spines are symplesio-
morphies, and with Young (1982, p. 838) that the same applies to the fin-spines of Antarctilamna. If

table 1. Some fin-spine characteristics of euselachians.

LEU: PERMIAN FRESHWATER SHARK

281

to

3

4®o-5

> r ™ o

5

<D

03 E c 03 Q)

C

C

03

c

o

c c ? o -c

CO

<v

o

o

o

T) D ^ m _

£

£

A longi
rounde
tubercl
the po:
edges

to

-c

c

-O -D

c $

S 3

03 §

m g 5

03 2

fO -*"• (/)

r 15 >- d)

E >..2 o

CTJ

o

03

c

?o»

8 a“

nj “ >>

oj <d £

m « 0)

- £ O T3

03

03

c

03

-O

0)

-c

fco®

lIsS

CL

<

Q.

</)

§2^ l

- O 0) o

55 SS E 8

S If .®

CL o O T3

c

03

03

Q.

03

-it

>

C

c

03

O

0)

03

o

£

c

o

o

1

03

£

o

03

I

= = n

to

0) •
> X

to a>

-= o 2? 03
2 Q. E JD

0) (U w -n

<5

03

03

c c oJ

Q-

■o

03

E o ®

2 $ “ So O
a> a e n- ~

>

-Q

03

0)

O

3

03 > co

n oV.

m w ® Q)

O C O-^-O

E

13

£

CO

$

® w 2 o ^

03

LL w £

TE£-oT

£

to

-C

0)

o

03

O

c

C

03

03

■Q

o

o

O

-Q

<

03

£

_Q

<

o

C5

to

To

“ c . Io?

tr oj J5 to

ft .1

-c

* t> * ® 2 « a>

F° S 2

c

03

o

03

c

03

(D £ ^ OJ ^

° C « w 03 O

™gg!la-§

!§s :!*.

CL O 2

m — _D

2^°>-
i5 c ^ n w

o CD O .2 03

Present

Absent

o

IlflW

w ? w c

O w >> o 0)
Q_ O -Q O-TD

x: w

03

C

'*! c

E

03 _

> 03 ^

■2

o a) £

*-

o

o « s

03

03

n

0)

03

>, 0)

<

<

C

CO 0) 03 \_

t n

>

FINSPINE

CHARACTERISTIC

Posterior Wal

Posterolateral

Situated

Denticles

Posteriorly
Placed Centr;
Cavity

Anterior Kee

282

PALAEONTOLOGY, VOLUME 32

so, this would require that the common ancestor of the xenacanths, ctenacanths, and neoselachians
had already differentiated by Middle Devonian time. Young (1982, p. 840) believes that, if the
similar fin-spines of ctenacanths and neoselachians are symplesiomorphous, then the aplesodic fin
evolved independently in the common ancestor of the neoselachians, and in the common ancestor
of Hybodus and Tristychius. Maisey (1984, p. 366) disputes the supposition that the plesodic
condition is primitive and points out that plesodic pectoral fins occur in Recent euselachians and
are associated with specific modes of life. It is conceivable that a similar distribution occurred in
both ctenacanths and hybodonts and their respective common ancestors. This is clearly the case
among the few ‘hybodonts’ where the pectoral fin is known in detail; Onychoselache is plesodic
(Dick and Maisey 1980), whereas Tristychius and Hybodus are aplesodic. Locomotory adaptations
could be the selective pressure that governs the length of pectoral fin radials and this character
could frequently be derived by convergence in several lineages.

Schaeffer (1981, fig. 26) regards the xenacanths as representing a sister-group to the Cleveland
‘ Ctenacanthus ’ (CMNH6219) on the basis that they both possess multilayered prismatic calcite
and pronounced lateral otic processes. Maisey (1984, fig. 1) noted that both these characters are
shared with other taxa and considers xenacanths to be a specialized group of ctenacanthiform
sharks on the basis of the following synapomorphies (characters 18 and 19, p. 365): dorsal fin-
spines with a pectinate ornament (implies that the ctenacanths and xenacanths [ Antarctilamna ] had
already separated during or prior to the Middle Devonian); and a broad, expanded occipital
segment (the Cleveland ''Ctenacanthus'').

Pectinate ornament (Maisey 1984, character 18) of the ctenacanthiform variety may be a
plesiomorphic euselachian character or convergently derived. The fin-spines of ‘ Ctenacanthus ’
vetustus have a pectinate ornament, be it only on the anterior ribs, which Maisey (1981, p. 15)
notes are coarser and more irregular than the pectinations of C. major. Maisey (1981, p. 20) regards
the fin-spines of ‘C.’ vetustus as ‘notably like those of Mesozoic hybodonts’. Maisey (1982a, p. 2)
notes that Recent squaloid and heterodontid fin-spine morphology and ornamentation does not
vary above generic level, and that the same apparently applies to hybodont fin-spines. The fin-
spines of Antarctilamna differ from those of C. major in being relatively short and broad with little
curvature. They lack posterolaterally placed denticles and typically have a very narrow zone of
insertion, although in two specimens (Young 1982, p. 828) the inserted portion was much more
extensive. The fin-spines of some acanthodians possess pectinate ornament, e.g. Nodacosta Gross,
1940. Lund (1985, p. 16) regards the superficially inserted small and thorn-like fin-spines of
Antarctilamna and Heteropetalus as representing the primitive condition of the dorsal spine. He
also considers the deep, sagittally inserted fin-spines of the ctenacanths, hybodonts, and many
neoselachians to be an apomorphic state. Maisey (pers. comm.) disputes this and regards the spines
of Cladoselache and stethacanthids as deeply inserted.

Schaeffer (1981, p. 60) regards ‘the strong projection of the occipital segment behind the auditory
capsules in Xenacanthus and Tamiobatis ... as a derived condition that relates these taxa’. He
also stated that ‘the considerable projection of the occipital segment behind the capsules in
Xenacanthus , Tamiobatis and Tristychius (Dick 1978) is correlated with several pairs of well spaced
occipitospinal nerve foramina’. Maisey (1984, p. 365) considers the broad, expanded occipital
segments of Xenacanthus , Tamiobatis , and the Cleveland ‘ Ctenacanthus ’ to be synapomorphic. The
terms ‘broad and expanded’ and ‘strong or considerable’ projection are subjective and need to be
quantified to enable more precise comparisons. Various dimensional relationships of the occipital
segments of shark neurocrania are presented in Table 2. Expressed as percentages, the length of
the occipital segment extending behind the otic capsules vs. both the length of the occipital segment
and the total length of the neurocranium for Tamiobatis , Hybodus basanus , and Squalus are 76 :
71:33 and 16:20:5, respectively. The maximum and minimum widths of the occipital segment
behind the otic processes vs. the postorbital width for Xenacanthus, Tamiobatis, and Hybodus are
45:44:57 and 45:30:24, respectively. The minimum width of the occipital segment behind the
otic capsules (width of occipital cotylus) vs. the maximum width for the latter three genera and
‘ Cladodus ’, Cladoselache, and Squalus are 100:68:42:44-51:76:57, respectively. The

LEU: PERMIAN FRESHWATER SHARK

283

table 2. Dimensions of the occipital segments of shark neurocrania, expressed as percentages.

Tam iobatus

Xenacanthus

' C ladodus"

Hybodus

Cladoselache

Cobe lodus

Squalus

( Schaeffer
1981, figs.
19 & 21)

(Schaeffer
1981, fig. 6)

( Schaeffer
1981, fig. 25)

(Maisey 1982,
fig .2)

(Maisey 1983,
fig .15D)

(Maisey 1983,
fig .15F)

(Maisey 1983,
fig.15A)

LENGTH OF OCCIPITAL
SEGMENT

TOTAL LENGTH OF
NEUROCRANIUM

19

29

29

17

16

16

LENGTH OF OCCIPITAL
SEGMENT BEHIND OTIC
CAPSULES

TOTAL LENGTH OF
NEUROCRANIUM

16

20

8

8

5

LENGTH OF OCCIPITAL
SEGMENT BEHIND OTIC
CAPSULES

LENGTH OF OCCIPITAL
SEGMENT

76

71

49

50

33

MAXIMUM WIDTH OF
OCCIPITAL SEGMENT
BEHIND OTIC CAPSULES
WIDTH ACROSS
POSTORBITAL PROCESSES

44

45

43-48

57

24

28

54

MINIMUM WIDTH OF
OCCIPITAL SEGMENT BEHIND
OTIC CAPSULES

WIDTH ACROSS
POSTORBITAL PROCESSES

30

45

20-22

24

20

30

MINIMUM WIDTH OF
OCCIPITAL SEGMENT BEHIND
OTIC CAPSULES

MAXIMUM WIDTH OF
OCCIPITAL SEGMENT
BEHIND OTIC CAPSULES

68

100

44-51

42

76

43

57

MAXIMUM WIDTH OF
OCCIPITAL SEGMENT BEHIND
OTIC CAPSULES
WIDTH ACROSS LATERAL
OTIC PROCESSES

57

43

66

51

50

54

only significant dimensional discrepancy between the occipital segments of Hybodus and Xenacanthus
(regardless of the differences in the relative positions of their postorbital processes) is the degree
to which the lateral hanks of the occipital segment taper posteriorly behind the otic capsules. The
range of variability of this character within the xenacanths and amongst distantly related sharks,
e.g. Cladoselache and Squalus, clearly demonstrates that it is not phylogenetically significant.

It is apparent from the above that the dimensions of the occipital segment of Hybodus resemble
closely the xenacanth/Cleveland ‘ Ctenacanthus ’ condition, and differ significantly from most
neoselachians. In fact, Maisey (1983, pp. 30-31) considers that the otico-occipital region of
Hybodus , although foreshortened, resembles the condition of Xenacanthus and Tamiobatis more
than the neoselachian condition, and he states (1982 b, p. 7) that ‘the otico-occipital region [of
Hybodus ] is short, although the deeply concave articular cotylus of the occiput forms a prominent
posterior extension bounded laterally by triangular extensions’. He (1983, p. 58) also notes that
there are no ‘appreciable differences in the extent to which the occipital arch extends between the
otic capsules in Hybodus and Xenacanthus'. Some extant sharks, such as the hexanchoids (sensu
Compagno 1977), Chlamydoselachus , Hexanchus, and particularly Notorynchus , have occipital
segments that extend much farther posteriorly beyond the otic capsules relative to the majority of
Recent sharks. It is more parsimonious to regard a broad expanded occipital segment as a primitive
character shared by xenacanths, ctenacanths, and hybodonts. In the absence of other shared
characters, it is too tenuous to demonstrate confidently that xenacanths are a specialized group of
ctenacanthiform sharks. I intuitively agree, from a phenetic viewpoint, with Schaeffer’s (1981,
p. 61) conclusion that the Cleveland ‘ Ctenacanthus ’ represents a sister-group to Xenacanthus ,
Tamiobatis , and ‘ Cladodus ’. Maisey (1985, p. 15) notes that the different configurations of the

284

PALAEONTOLOGY, VOLUME 32

shortened posterior portion of the braincases of Hybodus and neoselachians could readily be
derived from a primitive morphotype in which the parachordals were much longer.

Comparisons with the placoderms and acanthodians suggest that a broad expanded occipital
region may be a primitive gnathostome character. Amongst the arthrodires the phyctaeniniids
(Kujdanowiaspis) and the brachythoracids ( Pholidosteus and Tapineosteus) possess extremely
long and broad occipital segments. The petalichthyids (Epipetalichthys, Macropetalichthys, and
Ellopetalichthys) have elongate narrow occipital segments that extend considerable distances beyond
the otic region. The occipital region of the rhenanid Jagorina extends beyond the otic section,
albeit to a much lesser extent (16% of the total length of the neurocranium) in comparison to the
above arthrodires. Acanthodes has a broad, expanded occipital segment that extends beyond the
otic region for 20-5% of the total length of the neurocranium.

During the ontogeny of Recent sharks, the occipital region becomes much shortened (Holmgren
1940). Holmgren (1942) and Jarvik (1980), influenced by their belief in a placoderm ancestry for
sharks, both interpreted this as phyletic recapitulation.

Surcaudalus possesses the following ‘ctenacanthiform’ characteristics: absence of calcified ribs,
‘cladodont’ dentition, fin-spines which have a concave posterior wall, and apically situated
posterolateral denticles. These characters (apart from the posterolaterally placed denticles) are
shared equally with the neoselachian Palaeospinax and, additionally, they both have a non-lunate
caudal fin. The well-developed epicaudal lobe of the caudal fin of Surcaudalus is a character which,
according to Thomson (1976, p. 20), is present only in squalomorph selachians.

Surcaudalus differs from Ctenacanthus and Goodrichthys in possessing an elongate fleshy snout,
a non-lunate caudal fin, basal plates with a well-developed anteroventrally directed process, and
fin-spines which bear lateral longitudinal costae covered by fine longitudinal striations and have
an anterior keel. Tristychius and Surcaudalus both share the latter three characters. Surcaudalus
and Wodnika both lack calcified ribs and have fin-spines with fairly smooth, broad ribs and a
concave posterior wall. Wodnika differs, however, in possessing a more weakly calcified skeleton,
fin-spines of unequal size, and teeth that are low rounded and tumid; although the dentition may
be a function of dietary specialisation.

The non-lunate caudal fin of Surcaudalus is a character that Compagno (1977) and Young (1982,
character 10) regard as a synapomorphy uniting Tristychius , Onychoselache , Hybodus , Palaeospinax ,
and Recent euselachians. Using multivariate analyses, Thomson and Simanek (1977) distinguished
four distinct patterns of shark body shape, each intrinsically related to one of four discrete types
of caudal fin ranging from nearly symmetrical to straight. Thomson (1976, fig. 14) also plotted
various dimensional relationships of elasmobranch tails for both living and fossil genera. He
demonstrated that all fossil genera plot within the same dimensional limits as the Recent forms.

Thomson and Simanek (1977) noted that the morphology of neoselachian caudal fins (whether
lunate or non-lunate) do not equate with current shark systematics ( sensu Compagno 1977). They
concluded (1977, p. 352) that the various tail patterns have been ‘convergently developed in sharks
of different major groups’ and that ‘the four morphological groupings accord better with differences
in mode of life’.

Only three ctenacanthiform sharks with well-preserved caudal fins have been described to date.
Of these, Ctenacanthus and Goodrichthys have lunate caudal fins, whereas Bandringa is non-lunate.
In view of the small sample size, and considering the distribution of caudal fin morphology within
the various neoselachian groups, it is probable that ctenacanthiform sharks possessed a variety of
caudal fin architecture in response to adaptations for specific life habits. This diversity would have
been maintained at successive adaptive levels. Because of possible convergence, the non-lunate
caudal fin of ctenacanths ( Bandringa ), hybodonts, and neoselachians cannot be safely assumed to
be synapomorphic, regardless of whether the morphotypic condition was deeply forked and almost
equilobate. Maisey’s amendment of this character (Maisey 1984, character 35, hypaxial endoskeleton
of tail reduced) is consistent with the record. Bearing in mind the case developed for convergent
derivation of plesodic pectoral fins connected with specialized modes of life, further comparative
anatomy of the caudal endoskeleton of Recent sharks is required to ascertain if discrete caudal fin

LEU: PERMIAN FRESHWATER SHARK

285

shapes are congruent with classifications based on other characters. ‘Advanced’ ctenacanths may
have independently achieved this condition.

It is apparent that, in the absence of well-preserved neurocranial and pectoral fin cartilages, it
would be extremely subjective to differentiate between a ctenacanth and a hybodont. Surcaudalus
has many ‘ctenacanthiform’ characters that are present in other groups, such as fin-spines with a
concave posterior wall, and the absence of calcified ribs, shared with Wodnika , Sphenacanthus,
Palaeospinax , and some Recent euselachians. The scales of Surcaudalus are preserved as external
moulds of the crowns, whilst the fin-spines and teeth have been replaced by a clay mineral from
the illite group during thermal metamorphism of the fish-bearing shales by combustion of the
underlying coal seam. Therefore, it is not possible to determine if Surcaudalus had non-growing
placoid scales, a feature which Maisey (1984, character 34) believes unites Hybodus, Acronemus ,
Tristychius , Onychoselache, Palaeospinax , and Recent euselachians.

Acknowledgements. This work forms part of a Ph.D. at Macquarie University enthusiastically supervised by
Dr John Talent. Dr John Maisey (American Museum of Natural History) read and commented helpfully on
the manuscript. 1 have also benefited from discussion with Dr Alex Ritchie (Australian Museum), Dr John
Long (University of Western Australia), and Dr Gavin Young (Bureau of Mineral Resources). Dr Peter
Forey (British Museum, Natural History), Dr Michael Benton (Queen’s University of Belfast), and an
anonymous referee all made valuable comments on the manuscript. I am also indebted to the staff of the
Utah Development Company, in particular Mr David Watson (Senior Mining Engineer), Mr Alex Wilkins
(Manager), and Mr Peter Isles (Manager) for providing accommodation, excavating equipment, storage, and
hospitality. I am grateful to many of the miners (too numerous to mention) who assisted me in many ways.
Ross Talent, Vladimir Maricic, James Henderson, and Paul Hart all contributed time and labour towards
the arduous task of excavating and packing. The Utah Development Company, C.B.S. Explosives, Comet
Transport, Mr Norm Laever, and Mr John Johnstone (Deputy Director of the Mining Museum, Sydney) all
contributed towards defraying the immense cost of transporting the excavated material to Sydney. Ron
Oldfield, John Cleasby, and Malcolm Ricketts all provided assistance and advice on photography.

REFERENCES

burgis, w. a. 1975. Environmental significance of folds in the Rangal Coal Measures at Blackwater,
Queensland. Rep. Bur. miner. Resour. Geol. Geophys. Aust. 171, v + 64 pp.

Campbell, k. s. w. and duy phuoc, l. 1983. A Late Permian Actinopterygian Fish from Australia.
Palaeontology , 26, 33-70.

compagno, l. j. v. 1977. Phyletic relationships of living sharks and rays. Am. Zool. 17, 303-322.
dick, J. r. F. 1978. On the Carboniferous shark Tristychius arcuatus Agassiz from Scotland. Trans. R. Soc.
Edinb. 70, 63-109.

— 1981. Diplodoselache woodi gen. et sp. nov., an early Carboniferous shark from the Midland Valley of
Scotland. Ibid. 72, 99-113.

— and maisey, j. g. 1980. The Scottish Lower Carboniferous shark Onychoselache traquairi. Palaeontology,
23, 363-374.

gross, w. 1940. Acanthodier und Placodermen aus //elm«rh/s-Schichten Estlands und Lettlands. Ann. Soc.
Reb. natur. Invest. Univ. Tartuensis Const. 46, I 89.

helby, r., Morgan, r. and partridge, a. d. 1987. A palynological zonation of the Australian Mesozoic.
Mem. Ass. australas. Palaeontol. 4, 1 94.

holmgren, N. 1940. Studies on the head in fishes. Part 1. Development of the skull in sharks and rays. Acta
Zoologica , 21, 51-257.

— 1942. Studies on the head in fishes. An embryological, morphological and phylogenetical study. Part 3.
The phylogeny of elasmobranch fishes. Ibid. 23, 129-161.

jarvik, e. 1980. Basic Structure and Evolution of Vertebrates , vol. 1: xvi + 575 pp., vol. 2: xiii + 337 pp.
Academic Press, London and New York.

kemp, e. m., balme, b. e., helby, r. j., kyle, r. a., playford, G. and price, p. l. 1977. Carboniferous and
Permian palynostratigraphy in Australia and Antarctica: a review. J. Bur. miner. Resour. Geol. Geophys.
2, 177-208.

lighthill, m. j., 1975. Mathematical biofluiddynamics , 281 pp. SIAM, Philadelphia.

286

PALAEONTOLOGY, VOLUME 32

lund, r. 1985. Stethacanthid elasmobranch remains from the Bear Gulch Limestone (Namurian E2b) of
Montana. Am. Mus. Novitates , 2828, 1 24.

maisey, j. g., 1981. Studies on the Paleozoic Selachian Genus Ctenacanthus Agassiz: No I Historical Review
and Revised Diagnosis of Ctenacanthus, with a List of Referred Taxa. Ibid. 2718, 1-22.

“1982a. Studies on the Paleozoic Selachian Genus Ctenacanthus Agassiz: No. 2. Bythiacanthus St. John
and Worthen, Amelacanthus, new genus, Eunemacanthus St. John and Worthen, Sphenacanthus Agassiz,
and Wodnika Munster. Ibid. 2722, 1-24.

— 19826. The anatomy and interrelationships of Mesozoic hybodont sharks. Ibid. 2724, 1 48.

— 1983. Cranial anatomy of Hybodus basanus Egerton from the Lower Cretaceous of England. Ibid. 2758,
1-64.

— 1984. Chondrichthyan Phylogeny: A Look at the Evidence. J. vertebr. Paleont. 4, 359-371.

— 1985. Cranial Morphology of the Fossil Elasmobranch Synechodus dubrisiensis. Am. Mus. Novitates,

2804, I -28.

rieppel, o., 1982. A new genus of shark from the Middle Triassic of Monte San Giorgio, Switzerland.
Palaeontology, 25, 399-412.

Schaeffer, B. 1981. The xenacanth shark neurocranium, with comments on elasmobranch monophyly. Bud.
Am. Mus. nat. Hist. 169, 1 66.

staines, h. r. e. 1972. Blackwater Coalfield: correlation of seams in the Rangal Coal Measures, Mackenzie
River to Sirius Creek. Rep. geol. Surv. Queensland , 70, 1-11.

Thomson, k. s. 1976. On the heterocercal tail in sharks. Paleobiology, 2, 19-38.

and simanek, d. e., 1977. Body form and locomotion in sharks. Am. Zool. 17, 343-354.
webb, p. w. and keyes, r. s. 1982. Swimming kinematics of sharks. Fish. Bull. 80, 803-811.
young, G. c., 1982. Devonian sharks from south-eastern Australia and Antarctica. Palaeontology, 25, 817
843.

MICHAEL R. LEU

School of Earth Sciences
Macquarie University

Typescript received 20 March 1988 Sydney, New South Wales

Revised typescript received 3 June 1988 Australia, 2109

A NEW GENUS OF OSMUNDACEOUS STEM
FROM THE UPPER TRIASSIC OF TASMANIA

by R. S. HILL, S. M. FORSYTH and F. GREEN

Abstract. Petrified osmundaceous trunks from the Late Triassic east of Woodbury in central Tasmania are
assigned to a new genus and species, Australosmunda indentata. This species possesses an ectophloic
siphonostele with a parenchymatous pith, but in other respects is similar to Millerocaulis, Osmundacaulis , or
Osmunda. Although leaf gaps are absent the stele is deeply indented where leaf traces arise. A. indentata is
the first osmundaceous species described which has a stele lacking leaf gaps but a parenchymatous pith, and
offers convincing support for the hypothesis that the evolution of the parenchymatous pith and the evolution
of leaf gaps in the xylem were independent transitions. Because of the relatively advanced nature of the leaf
traces, and the presence of species with leaf gaps earlier in the record (Palaeosmunda from the Upper Permian
of Queensland), A. indentata is unlikely to have been an intermediate stage in the development of leaf gaps,
but probably represents a relatively advanced species which has maintained a primitive stelar feature.

The Osmundaceae contains three living genera with about twenty-one species (Hewitson 1962),
but importantly has a long and impressive fossil record (e.g. Miller 1967, 1971; Gould 1970)
beginning in the Upper Permian. Of particular interest in the family is the development of the
modern stelar types from the earliest fossil types. Since petrified stems are very common a great
deal of literature has developed concerning stelar development. Although the evolutionary pathway
is now generally beyond debate, newly found fossil species are constantly adding information in
this area.

The Osmundaceae are relatively common in Permian to Jurassic strata in Australia (Gould 1970,
1973; Tidwell 1987; Tidwell and Jones 1987, and references therein) and some of the species are
of particular evolutionary significance. The discovery of specimens from the Upper Triassic of
central Tasmania which belong to a new genus adds substantially to the record of the family in
the southern hemisphere and provides evidence of a previously undescribed stelar type.

LOCALITIES AND GEOLOGY

Petrified tree-fern stems were found at three separate localities east of Woodbury during mapping of the
Interlaken Quadrangle of the Tasmanian Geological Survey Geological 1 : 50000 atlas series (Forsyth 1986).
The geology of the area consists essentially of gently dipping volcanic lithic sandstone and coal measures of
the Upper Parmeener Supergroup intruded by small and large-scale dolerite sheets and dykes elsewhere
radiometrically dated as being of mid-Jurassic age (Schmidt and McDougall 1977). The strata and some
dolerite bodies were faulted and eroded prior to the extrusion of tholeiitic lava flows that occur a few
kilometres north of the fossil localities. The basalt flows are inferred to be of Early to Middle Miocene (Late
Oligocene?) age (Sutherland and Wellman 1986) and probably previously extended closer to the fossil
localities. The basalt overlies thin veneers of Tertiary rocks that include silcrete and groundwater ferricrete.
Quaternary deposits include veneers consisting largely of dolerite clasts but also containing notable components
derived from Parmeener strata. They occur as talus, alluvial fans, higher level alluvial terraces above the
modern flood plains, and as lag deposits.

Two of the fossil localities can be directly related to Upper Parmeener strata, the third locality consists of
loose material inferred to be derived from Upper Parmeener Supergroup strata and found in a Quaternary
high-level terrace deposit. The similarity of the tree-fern fossils at all localities suggests that the third locality
is not related to Tertiary silicification.

IPalaeontology, Vol. 32, Part 2, 1989, pp. 287 296, pis. 35 36.1

© The Palaeontological Association

288

PALAEONTOLOGY, VOLUME 32

Locality 1. The best stratigraphic control is provided by this locality which occurs high on the northern side
of Brents Sugarloaf and is estimated to be 30 m topographically below the summit (Co-ordinates 55 GEP
412296 (Universal Grid Reference), 147° 29' 53" E., 42° 11' 04" S.). Here a conspicuous block of silicified
material, probably transported peat, contains much plant material including tree-ferns and Dicroidium, and
occurs probably almost in situ in volcanic lithic sandstone of the Upper Parmeener Supergroup. Further
loose pieces of the silicified material occur at the same elevation nearby, suggesting that a lens containing
the material is present. Volcanic lithic sandstone beds at a similar horizon on the eastern side of Brents
Sugarloaf are well exposed. They include breccia beds with intra-basinal lulite clasts up to 1 m in diameter,
beds with silicified stems or tree stumps from a few mm to over I m in diameter, and beds with thin layers
or isolated pebbles, cobbles, and boulders of rounded extra-basinal clasts. Exotic clasts include cleaved and
veined quartzite and acid igneous porphyries. The highest beds at Brents Sugarloaf include black lutite and
several tuff layers just below a capping Jurassic dolerite sill.

Locality 2. This locality (55 GEP 401284, 147° 29' 10" E., 42° 1 1' 43" S.) occurs 1-5 km south-west of locality
1 and is probably separated from it by one or more faults. Tree-fern fossils and the long-ranging late
Palaeozoic to early Mesozoic pollen Falcisporites australis occur in several loose blocks of silicified material
to which volcanic lithic sandstone matrix adheres. Two hundred metres further south and higher the fine-
grained base of a minor Jurassic dolerite sheet is exposed and is probably faulted against an extensive area
of coarse-grained dolerite. Extra-basinal clasts of exotic lithologies similar to those at Brents Sugarloaf but
including Lower Parmeener Supergroup lithologies with Permian shelly fauna occur in lag deposits, and in
talus shed from the fine-grained dolerite sheet near locality 2. From the rock distribution it is concluded that
the extra-basinal clasts and tree-fern fossils are shed from the approximately 40 m of strata underlying the
fine-grained dolerite intrusion.

Locality 3. Abundant silicified wood and a solitary silicified tree-fern (holotype of Australosmunda indentata)
were found as loose clasts in a higher level Quaternary alluvial terrace deposit (55 GEP 398326, 147° 29' 15"
E., 42° 09' 26" S.). The deposit straddles a Jurassic fine-grained dolerite dyke that probably marks the
structural boundary between Upper Parmeener volcanic lithic sandstone upstream to the south and older
quartz sandstone to the north. Dolerite clasts are the most common constituent of the deposit and the
relatively coarse grain-size of many clasts indicates that they have not been derived locally, but have undergone
a minimum transport of 2 5 km. Other clasts include well-rounded quartz porphyry, fossiliferous Lower
Parmeener rocks and other exotic rocks for which no other source exists within the catchment other than
the extra-basinal clast-bearing beds of the volcanic lithic sandstone sequence.

Although the majority of silicified wood clasts appear indistinguishable from silicified wood in the volcanic
lithic sandstone sequence, some clasts could be derived from other (?Tertiary) sources.

Age. The tree-fern fossils from localities 1 and 2 are clearly derived from the Upper Parmeener Supergroup
volcanic lithic sandstone sequence. This sequence is probably best known in eastern and north-eastern
Tasmania where it has been penetrated by numerous fully cored coal exploration bores and has been recently
mapped (Turner et al. 1984). In particular, beneath Fingal Tier the main coal-bearing interval of the sequence
is about 220 m thick and contains coal seams known informally from top to bottom as seams A H (Threader
and Bacon 1983). Exotic cobbles and boulders like those found at Woodbury are not known below about
seam E and tuff beds appear to be confined to above B seam (Calver, in Turner and Calver 1987). Although
the volcanic lithic sandstone and coal measures sequence is widely distributed in Tasmania, neither exotic
clasts nor tuff beds have been commonly reported away from the north-eastern to eastern area except for
the occurrence of one or both of the features at a few localities in the Tasmanian Midlands (Forsyth 1984,
in press a). Where indications are available the exotic clasts and tuff appear to occur in the uppermost beds
of the sequence in the Midlands area.

The similarity of the tree-fern fossil from locality 3 with those from localities I and 2 plus the occurrence
of exotic clasts at all localities suggests the tree-fern fossils may all be derived from a restricted interval of
the volcanic lithic sandstone sequence above the equivalent horizon of seam E and possibly in proximity to
seam B.

Quartz sandstone occurs about 45 m below seam H at Fingal Tier and nearby at Nicholas Range the
interval with quartz sandstone hosts two partly extrusive basalt ‘flows’. Interbedded sediments contain a
microflora probably best compared with the less distinctive microflora of the upper part of the Bowen Basin
Moolayember Formation (de Jersey and Hamilton 1967) and the overlying basalt has been radiometrically
dated at 233 + 5 Ma (Calver and Castleden 1981). At Fingal Tier the interval from seam A to seam G and
possibly to seam H can be correlated with the Craterisporites rotundus Zone (de Jersey 1975; Forsyth, in

HILL ET AL TRIASSIC OSMUNDACEOUS STEM

289

press b ) although from the incomplete palynology carried out to date the nominate zone fossil and
Polycingulatisporites densatus have not been recorded below seam B.

Thermal effects of an overlying dolerite intrusion may prevent the top of the C. rotundus Zone from being
recognized at Fingal Tier, but further south a tuff bed associated with coal near the top of the sequence has
been radiometrically dated at 214+ I Ma (Bacon and Green 1984). Rocks immediately below the tuff and a
lithocorrelate of the coal contain C. rotundus Zone microfloras, whereas 20 m above the coal correlate the
microflora from an overlying dominantly grey lutite sequence has been referred to the lower (Assemblage A)
Polycingulatisporites crenulatus Zone (de Jersey 1975; Forsyth, in press b).

An approximate Carnian age has been indicated for the C. rotundus Zone (de Jersey 1975; Helby et a/.
1987) and this is supported by the New Zealand range of Annulispora follicularis and A. microannulata
(N. J. de Jersey, pers. comm.). Assemblage A (de Jersey 1976) of the P. crenulatus Zone is probably of Norian
age (Stevens 1981; Flelby et al. 1987; Tozer 1984). The coal seam sequence seam A to seam G at Fingal Tier
is therefore considered to be of Late Triassic (Carnian) age and good agreement is shown between the
radiometric dates and the time scale of Webb (1981); the date of 214+1 Ma for tuff near the top of the
sequence comparing favourably with the Norian/Carnian boundary 2 1 5 ± 5 Ma, and the date of 233 ±5 Ma
for basalt underlying the sequence being probably of Middle Triassic age, 225 + 5 Ma to 240 + 5 Ma.

Although the volcanic lithic sandstone sequence has yielded C. rotundus Zone microfloras or slightly older
microfloras with A. folliculosa from several localities in Tasmania (Forsyth, in press a), palynological data
from Woodbury are lacking. Within 15 km of Woodbury, A. folliculosa is present either in the volcanic lithic
sandstone sequence or in underlying rocks, and 35 km south-west from Woodbury at Spring Hill, a microflora
from a sequence either below the volcanic lithic sandstone sequence or alternatively interpretable as a lutite
dominated basal facies of the volcanic lithic sandstone sequence, indicates a C. rotundus Zone age (Forsyth
1984, in press u, b).

Macrofloras at Woodbury immediately underlying a tuff overlain by a sandstone bed with exotic cobbles,
includes Johnstonia coriacea indicating a Late Anisian to Norian age (Retallack 1977). The stratigraphic
position of other beds with Dicroidium odontopteroides and Heidiphyllum elongation cannot be determined
with respect to the interval with fossil tree ferns.

Based on the lithological correlation of the interval with tree-fern fossils at Woodbury with the upper part
of the coal measures at Fingal Tier outlined above, the tree-fern fossils are considered to be Late Triassic
(probably Carnian or early Norian) in age.

MATERIAL AND METHODS

One fossil specimen represents an isolated section of a trunk (PI. 35, fig. I). Transverse sections of this
specimen were taken from the apex and base, and a longitudinal section from the apex. Several other
specimens were found embedded with other plant remains in larger pieces of rock, and transverse sections
of several of these were also cut. However, the best anatomical preservation was found in the isolated
specimen, and the description is based largely on it.

The sections were produced by the lapidary section of the Tasmanian Department of Mines. First, blocks
were cut from the fossils using a diamond saw. The appropriate block faces were then polished on a Logitech
LP30 Production Lapping and Optical Polishing Machine before bonding to ground microscope slides using
the bonding/mounting medium epo-tek 301. Excess material was then removed using a Micro-Trim saw, a
lapping machine, and the Logitech LP30, adjusted to produce sections of 30 ^ m thickness automatically.
Coverslips were applied using epo-tek 301. Material removed from the trunk apex of the holotype to produce
a transverse section was utilized to make a longitudinal section, which cut the central stele at about 5°. The
remaining offeuts are stored with the type specimen.

DESCRIPTION OF SPECIMENS

The fossils (PI. 35, figs. 1 4) are clearly osmundaceous in affinities, with the characteristic stelar structure
and leaf trace arrangement described by Miller (1967, 1971). The stele is very small (about 3-5 mm in dia-
meter), and despite some deep indentations where the leaf traces arise, a leaf gap was not observed (PI. 35,
fig. 4). Inside the xylem ring only a single cell type was observed, and longitudinal section confirmed that
this was parenchyma (PI. 35, figs. 4 7). On this basis the fossil can be said to have a simple siphonostele,
with a parenchymatous pith. The metaxylem elements have conspicuous scalariform pitting (PI. 35, fig. 8).
Each departing leaf trace has only one protoxylem bundle, in an endarch arrangement (PI. 36, figs. I and

290

PALAEONTOLOGY, VOLUME 32

2). The protoxylem divides into two usually at the outer edge of the outer cortex (PI. 36, fig. 3), and continues
dividing beyond there until there are eight or more protoxylem groups present (PI. 36, fig. 4). The metaxylem
in the leaf trace increases in cell number and in overall size as it departs from the stem, and develops into
the C-shape characteristic of many ferns including the Osmundaceae. As the leaf trace moves through the
inner and then the outer cortex, it becomes surrounded by a ring of cells from these two areas, first,
parenchyma and then sclerenchyma. Beyond the outer cortex there is clear development of stipular expansions,
but the sclerenchyma ring maintains a rounded shape and does not extend into these extensions (PI. 36,
fig. 5). Hewitson (1962) noted the taxonomic importance of the presence and positioning of sclerenchyma in
stipular expansions. In these fossils there is one large, rounded bundle in each stipular expansion, as well as
several smaller, scattered bundles (text-fig. 1; PI. 36, fig. 5). The sclerenchyma ring can also have a characteristic
distribution of thick-walled fibres, and in the fossil they occur in an abaxial arch and as two lateral bundles
(text-fig. 1). There are also two large sclerenchyma bundles within the concavity of the C-ring of metaxylem
(text-fig. 1; PI. 36, fig. 6).

One root with a diarch xylem strand arises from each departing leaf trace (PI. 36, figs. 2, 3, 7), probably
before it enters the inner cortex. There is no sign of a mat of external roots, which is common in many fossil
and living osnrundaceous species, in any of our fossil specimens. The roots consistently run parallel to the
stem, suggesting that the species had an upright habit.

The phloem is completely degenerated, but occasional signs of what is probably the endodermis can be
seen external to the central xylem (PI. 36, fig. 5). There is no sign of either phloem or endodermis on the
inside of the xylem ring, and since the parenchymatous pith almost fills the central area in some sections
(e.g. PI. 35, fig. 5) it is assumed that phloem was absent there. The position of the inner cortex is quite clear

text-fig. I . Drawing of a transverse section of a petiole base of Australosmunda indentata gen. et sp. nov.
Note that the thick-walled fibres (F) in the sclerenchyma ring occur in an abaxial arch and as two lateral
bundles. Sclerenchyma (S) occurs in each concavity of the C-ring of metaxylem (C) and as one large, rounded
bundle in each stipular expansion, surrounded by numerous smaller, scattered bundles.

EXPLANATION OF PLATE 35

Figs. 1 7. Australosmunda indentata sp. nov. 1, holotype of A. indentata (GST 10001), showing upright habit
and petiole scars, x 0-5. 2, transverse section of the basal end (GST 10001 A) showing the small stele (S),
the inner cortex (I), the outer cortex (O), the leaf traces in the cortex, and the petiole bases towards the
edge of the section (P), x 1-6. 3, stele (GST 1 0001 A), from the basal end of the stem, showing a continuous
metaxylem cylinder which is deeply indented, but not pierced, by departing leaf traces, x 15. 4, stele (GST
10001B), showing the parenchymatous pith which fills the cavity within the metaxylem cylinder. Traces of
the endodermis (E) can be seen, x 14. 5, transverse section of the pith parenchyma (GST 10001 A), x 225.
6, longitudinal section of the pith parenchyma (GST 1000 1C), x225. 7, longitudinal section of the

metaxylem elements of the stele (GST 10001C). Scalariform pitting can be seen in some areas, x 150.

PLATE 35

HILL et al., Aslralosmunda indent at a

292

PALAEONTOLOGY, VOLUME 32

(PI. 35, fig. 2), but again cell detail is absent. It is probable that this area was parenchymatous. The outer
cortex is well preserved and is sclerenchymatous (PI. 35, fig. 2; PI. 36, fig. 8).

These fossils do not, in our opinion, fall within the range of any described genera of the Osmundaceae.
All extant species have an ectophloic, dictyoxylic siphonostele, or something more advanced, and the same
is true for all fossil species which have been assigned to extant genera or to the form genus Osmundacaulis
(Miller 1971), which has recently been separated into two genera, Osmundacaulis and Millerocaulis (Tidwell
1986). Most other fossil genera have a protostele (e.g. Zalesskya, Bathypteris , Chasmatopteris, Iegosigopteris ,
Petcheropteris, and Thamnopteris). Palaeosmunda , from the Upper Permian of Queensland, usually has an
ectophloic-dictyoxylic siphonostele, but can sometimes be simply siphonostelic (Gould 1970). However,
although this is the closest genus to the fossils in terms of stele structure, Palaeosmunda differs in several
other important ways, especially in the extension of the sclerotic ring in the leaf traces out into the stipular
expansions. Therefore, a new genus is required to accommodate these fossils.

SYSTEMATIC PALAEONTOLOGY

Division pterophyta
Order filicales
Family osmundaceae
Genus australosmunda gen. nov.

Type species. Australosmunda indentata sp. nov.

Derivation of name. Named for the southern occurrence of these osmundaceous fossils.

Diagnosis. Arborescent osmundaceous trunk, with a stem surrounded by a mantle of leaf bases
and adventitious roots; unbranched. Stele an ectophloic siphonostele; pith parenchymatous; xylem
ring consisting of approximately twenty contiguous radial strands, up to eighteen tracheids thick;
leaf gaps incomplete, extending up to three-quarters through the metaxylem ring; phloem, pericycle
unknown, endodermis external only. Cortex differentiated into inner zone where cells are not
preserved, and an outer sclerotic fibrous layer, with short, wide, sclerenchyma cells lining leaf
traces and inner cortex; inner cortex about as wide as outer cortex; leaf traces arise at about 30°
to stele, initially with one endarch protoxylem group; about forty traces in a transverse section of
cortex. Petiole bases stipulate, containing an adaxially curved, C-shaped vascular strand, inner
cortex, and sclerotic ring; sclerotic rings remain rounded or elliptical in transverse section, not
extended into stipules. Roots with diarch xylem strand, arising singly from each departing leaf
trace usually before it enters inner cortex.

Australosmunda indentata sp. nov.

Plate 35, figs. 1-8; Plate 36, figs. 1-8

Holotype. GST 10001; figured in Plate 35, fig. 1, housed at the Tasmanian Department of Mines, Hobart.

explanation of plate 36

Figs. I 8. Australosmunda indentata gen. et sp. nov. 1, departing leaf trace from the central stele (GST
10001 A). Note the single protoxylem group (P) and the lack of a complete leaf gap, x 70. 2, leaf trace in
the outer cortex (GST 10001A), still with one protoxylem group (P). Note the root to the left of the leaf
trace which has arisen from the leaf trace soon after its formation, x 40. 3, leaf traces in the outer cortex
(bottom right) (GST 10001 A) and beyond. Note that the protoxylem bundle (P) divides into two at about
the outer edge of the outer cortex and then continues to divide, x 15. 4, metaxylem bundle in a petiole
base (GST 10001 A) containing a large number of protoxylem groups (e.g. P), x 30. 5, petiole bases (GST
1 0001 A) with stipular expansions. Note the sclerenchyma bundles in the stipular expansions and in the
concavity of the C-ring of metaxylem, x 6. 6, C-ring of metaxylem in a petiole base (GST 10001 A). Note
the two sclerenchyma bundles (S) in the concavity of the C-ring, x 12. 7, diarch root trace (GST 10001A)
soon after its formation from the leaf trace. A protoxylem bundle can be observed at the top and bottom
of the root trace, x40. 8, longitudinal section of the sclerenchyma in the outer cortex (GST 10001C),
x 150.

PLATE 36

HILL et al., Astralosmunda indentata

294

PALAEONTOLOGY, VOLUME 32

Type locality. Locality 3 east of Woodbury, Tasmania.

Derivation of name. Named for the indented nature of the stele.

Diagnosis. Trunks up to 12 cm high, 7 cm diameter; stem 14 mm diameter. Stele 3-5 mm diameter,
pith 2-2 mm diameter; metaxylem tracheids 36 (range 15-55 ^m) in diameter, with scalariform
pitting. Inner cortex includes about thirteen leaf traces in transverse section, fibrous outer cortex
includes about twenty-seven leaf traces. Leaf traces arise with one endarch protoxylem group
which bifurcates at the end of the outer cortex. Leaf bases with C-shaped vascular strand containing
eight or more protoxylem groups, with sclerenchyma bundle in each concavity; sclerenchyma ring
with an abaxial arch and two lateral bundles of thick-walled fibres; stipular expansions each with
one large mass and numerous smaller masses of thick-walled fibres.

Discussion. Australosmunda indentata is the first osmundaceous species described with a simple
siphonostele and a parenchymatous pith. Osmundaceous fossils exhibit a range of stelar types with
the simplest being protosteles composed of central short tracheids and peripheral long tracheids
(e.g. Zalesskya, Thamnopteris ), although in T. kidstoni parenchyma cells occur singly or in clusters
near the periphery of the central xylem tissue (Miller 1971). In all but four of the species with a
protostele there is a zone of decay in the centre of the stem, and Miller (1971) notes that although
typical central xylem tracheids border the vacant zone, other cell types may have occurred within.
Chasmatopteris principalis exhibits an early indication of the formation of leaf gaps, with the
metaxylem cylinder being conspicuously indented (but never pierced) opposite certain leaf traces
(Miller 1971). There is a large morphological distinction between these stelar types and those of
other osmundaceous stems described to date which have conspicuous leaf gaps and usually a
distinct pith of parenchyma and/or sclerenchyma. A. indentata fills part of this morphological
hiatus. There is no trace of short tracheids in the pith of this species, and the formation of leaf
gaps is almost complete (text-fig. 2). The metaxylem cylinder is heavily and frequently indented
opposite the point of origin of leaf traces, but in several sections viewed (including both the base
and apex of the holotype), a leaf gap was not observed.

text-fig. 2. Stylized series of ascending transverse sections showing the separation of a leaf trace from the
stelar xylem of Australosmunda indentata gen. et sp. nov. The black dot represents the protoxylem.

It is notable that this Late Triassic species was growing long after the Upper Permian
Palaeosmunda species from Queensland, which show clear evidence of well-developed leaf gaps
(Gould 1970). However, as has been noted earlier, the arrangement of the sclerenchyma in the leaf
traces of Palaeosmunda was more similar to the primitive protostelic species, whereas A. indentata
has a more advanced sclerenchymatous arrangement.

1mm

HILL ET AL.: TRIASSIC OSMUNDACEOUS STEM

295

Miller (1971) nominated six characters on a primitive to advanced scale which he used for
numerical analyses of osmundaceous stems. The first of these characters is stelar type. The stelar
type of A. indent ata is not recorded in Miller’s list, and does not fit easily into his sequence, since
it is relatively advanced in having a parenchymatous pith and relatively primitive in not having
fully developed leaf gaps. The second character concerns the cell types in the cortex, and in this
A. indentata is intermediate on the scale in having parenchymatous and sclerotic layers of about
equal breadth. The third character concerns the petiole bases, which are primitive in A. indentata
in being closely adhering, although this condition is generalized in the family (Miller 1971). The
fourth character refers to the xylem arrangement in the leaf trace at its point of divergence, and
A. indentata , with its endarch arrangement, is considered to be advanced. The fifth character refers
to the number of leaf traces visible in one cortical cross-section. A. indentata , with about forty, is
considered to have a medium number between the high primitive number of 100-150 and the
advanced condition of only about five to fifteen. The final character deals with the position in
which the first protoxylem bifurcation in the leaf trace takes place, and in this A. indentata exhibits
the primitive condition, with the bifurcation taking place near the outer edge of the outer cortex.

Therefore, on this basis A. indentata has a mixture of primitive, intermediate, and advanced
characteristics. Although A. indentata exhibits an important intermediate type of stelar development,
it must be assumed that this type occurred much earlier, since Palaeosmunda exhibits a more
advanced form in the Permian. The development of leaf gaps in Australosmunda would give rise
to a stem type which would be typical of many fossil Millerocaulis and Osmundacaulis species, and
it is possible that Australosmunda is a precursor of at least some of these species. Several
Osmundacaulis species have recently been described from Australia (Edwards 1933; Gould 1973;
Tidwell 1987; Tidwell and Jones 1987), but none appear to be closely related to Australosmunda
indentata.

The occurrence of a wholly parenchymatous pith in a stele without leaf gaps in A. indentata
offers convincing support for the hypothesis that the evolution of the parenchymatous pith and
the evolution of leaf gaps in the xylem were independent transitions. Previously, the occurrence of
‘delayed’ and ‘incomplete’ leaf gaps in species that also have a mixed pith (e.g. Millerocaulis
dunlopi , M. kolbei (Miller 1971) and M. beardmorensis (Schopf 1978)) left support for the hypothesis
quite ambiguous.

Acknowledgements. Our thanks to Miss J. L. Mackey and Miss J. H. Triffett of the Lapidary Section at the
Tasmanian Department of Mines for preparing the thin sections of the fossils. We are grateful to Professor
C. N. Miller for constructive comments on the manuscript.

REFERENCES

bacon, c. a. and green, d. c. 1984. A radiometric age for a Triassic tuff from eastern Tasmania. Unpubl.
Rep. Dept. Mines Tasm.

calver, c. R. and castleden, r. h. 1981. Triassic basalt from Tasmania. Search , 12, 40-41.
de jersey, n. j. 1975. Miospore Zones in the Lower Mesozoic of southeastern Queensland. In Campbell,
k. s. w. (ed. ). Gondwana Geology, 159-172. A.N.U. Press, Canberra.

— 1976. Palynology and time relationships in the Lower Bundamba Group (Moreton Basin). Qd. Govt
Min. J. 77, 461-465.

— and Hamilton, m. 1967. Triassic spores and pollen grains from the Moolayember Formation. Pubis
geol. Surv. Qd. 336 (Palaeont. Pap. 10).

Edwards, w. n. 1933. Osmundites from central Australia. Ann. Mag. nat. Hist., ser. 10, 11, 661-663.
forsyth, s. m. 1984. Geological atlas 1 : 50,000 series. Sheet 68 (8313S). Oatlands. Explan. Rep. geol. Surv.
Tasm.

— 1986. Geological atlas 1 : 50,000 series. Sheet 61 (8313N). Interlaken. Dept, of Mines, Tasm.

— In press a. Geological atlas 1 : 50,000 series. Sheet 61 (8313N). Interlaken. Explan. Rep. geol. Surv. Tasm.

— In press b. Upper Parmeener Supergroup. In burrett, c. f. and martin, e. l. (eds.). The Geology and
Mineral Deposits of Tasmania.

296

PALAEONTOLOGY, VOLUME 32

gould, r. e. 1970. Palaeosmunda , a new genus of siphonostelic osmundaceous trunks from the Upper Permian
of Queensland. Palaeontology , 13, 1 0 28.

— 1973. A new species of Osmundacaulis from the Jurassic of Queensland. Proc. Linn. Soc. NSW, 98, 86
94.

helby, r., Morgan, r. and partridge, a. d. 1987. A palynological zonation of the Australian Mesozoic. In
jell, p. A. (ed.). Studies in Australian Mesozoic Palynology. Mem. Assoc. Australas. Palaeontol. 4, 1 94.
hewitson, w. 1962. Comparative morphology of the Osmundaceae. Ann. Mo. Bot. Gdn. 49, 57-93.
miller, c. N. 1967. Evolution of the fern genus Osmunda. Contr. Mus. Paleont. Univ. Midi. 21, 139-203.

— 1971. Evolution of the fern family Osmundaceae based on anatomical studies. Ibid. 28, 105-169.
retallack, g. j. 1977. Reconstructing Triassic vegetation of eastern Australasia: a new approach for the

biostratigraphy of Gondwanaland. Alcheringa , 1, 247-278.

Schmidt, p. w. and mcdougall, i. 1977. Palaeomagnetic and Potassium Argon dating studies of the
Tasmanian dolerites. J. geol. Soc. Aust. 24, 321-328.
schopf, j. m. 1978. An unusual osmundaceous specimen from Antarctica. Can. J. Bot. 56, 3083-3095.
stevens, J. 1981. Palynology of the Callide Basin, east-central Queensland. Pap. Dept. Geol. Univ. Qd. 9,
1-35.

Sutherland, f. l. and wellman, p. 1986. Potassium-Argon ages of Tertiary volcanic rocks, Tasmania. Pap.
Proc. R. Soc. Tasm. 120, 77-86.

threader, v. m. and bacon, c. a. 1983. The Department of Mines coal exploration programme, Fingal Tier.
Unpubl. Rep. Dept. Mines Tasm. 1983/46.

tidwell, w. d. 1986. Miller ocaulis , a new genus with species formerly in Osmundacaulis Miller (fossils:
Osmundaceae). Sida , 11, 401 405.

1987. A new species of Osmundacaulis (O. jonesii sp. nov.) from Tasmania, Australia. Rev. Palaeobot.
Palynol. 52, 205-216.

— and jones, r. 1987. Osmundacaulis nerii , a new osmundaceous species from Tasmania, Australia.
Palaeontographica Abt. B. 204, 181-191.

tozer, E. T. 1984. The Trias and its ammonoids: The evolution of a time scale. Geol. Surv. Can. Misc. Rep.
35, 1-171.

turner, n. j. and calver, c. R. 1987. Geological atlas 1 : 50,000 series. Sheet 49 (8514N). St Marys. Explan.
Rep. geol. Surv. Tasm., 1 159.

— castleden, r. h. and baillie, p. w. 1984. Geological atlas 1 : 50,000 series. Sheet 49 (8514N). St Marys.
Dept, of Mines, Tasm.

webb, j. a. 1981. A radiometric time scale of the Triassic. J. geol. Soc. Aust. 28, 107-121.

r. s. hill and f. green

Botany Department
University of Tasmania
GPO Box 252C
Hobart

Tasmania, Australia 7001

S. M. FORSYTH

Typescript received 14 December 1987
Revised typescript received 29 July 1988

Geological Survey of Tasmania
PO Box 56
Rosny Park

Tasmania, Australia 7018

ORIGINAL MINERALOGY OF TRILOBITE
EXOSKELETONS

by n. v. wilmot and a. e. fallick

Abstract. The mineralized exoskeletons of well-preserved trilobites are now composed of low-magnesian
calcite. However, as this is the only form of calcium carbonate to survive in Lower Palaeozoic rocks, such
a mineralogy may be a function of diagenetic processes rather than reflecting primary cuticle composition.
Ferroan calcite replacement has previously been used to infer an original high-magnesian calcite mineralogy
for trilobite exoskeletons. By contrast, petrographic data involving over seventy trilobite species, ranging in
age from Cambrian to Devonian, together with carbon and oxygen stable isotope analyses of specimens from
the Much Wenlock Limestone Formation, England (Wenlock), are here used to infer that trilobites constructed
low-magnesian calcite exoskeletons. Petrographically, the trilobite cuticles share the same preservational
characteristics as low-magnesian calcite organisms such as articulate brachiopods. They also have very similar
isotopic signatures to those of brachiopods, yet differ from crinoids which secreted high-magnesian calcite
ossicles and now commonly contain microdolomite inclusions and secondary voids. Together, these separate
lines of evidence strongly suggest that trilobite exoskeletons originally had a low-magnesian calcite mineralogy.

Most trilobites had heavily mineralized exoskeletons, a characteristic shared by some other marine
arthropods such as decapod crustaceans, cirripedes, and ostracodes. Three trilobite species with
entirely organic cuticles have been described (Whittington 1977, 1985; Dzik and Lendzion 1988),
but this paper is concerned only with those trilobites that had mineralized exoskeletons. In such
trilobites the dorsal exoskeleton and hypostome were predominantly composed of calcium carbonate
with a small proportion of organic matter. If the exoskeleton is decalcified in EDTA, the remains
of the organic framework become apparent as a delicate brown residue of unknown composition
still retaining the general structure of the cuticle (Dalingwater 1973; Teigler and Towe 1975; Miller
1976; Dalingwater and Miller 1977). The ultrastructure of the exoskeleton comprises an outer
prismatic layer with a much thicker principal layer below (text-fig. Id). Several types of cuticular
microstructure may also occur, including laminations, tubercles, and canals (Dalingwater 1973;
Miller 1976; Stormer 1980; Mutvei 1981). Despite increasing understanding of the structure of the
exoskeleton, its original mineralogy has not previously been determined.

Marine invertebrates construct their exoskeletons from a variety of forms of calcium carbonate.
Scleractinian corals use aragonite (A), echinoderms have high-magnesian calcite (HMC) hard parts
( > 5% MgC03), and articulate brachiopods construct low-magnesian calcite (LMC) valves ( < 5%
MgC03); sometimes combinations of these may be used, as in certain molluscs (Milliman 1974;
Wolf et al. 1976; Morrison and Brand 1987). Among modern arthropods, decapod crustaceans
utilize HMC and small amounts of calcium phosphate, whereas most ostracodes use LMC, and
cirripedes form their lateral plates from LMC with some species having A basal plates (Richards
1951; Milliman 1974; Morrison and Brand 1987).

When trilobite exoskeletons are well preserved and still exhibit primary microstructures,
examination of stained thin-sections and X-ray diffraction show they are composed of calcite
(Dalingwater 1973; Teigler and Towe 1975). Analyses by microprobe (Teigler and Towe 1975;
Miller and Clarkson 1980) indicate an LMC mineralogy.

Isolated reports of phosphate within trilobite cuticle (Dalingwater 1973; Teigler and Towe 1975)
probably represent secondary deposits. An LMC mineralogy is to be expected for Palaeozoic
fossils, for although the primary composition of extant organisms or well-preserved Tertiary fossils
can be determined directly, the older the fossil the more likely is its composition to have been

[Palaeontology, Vol. 32, Part 2, 1989, pp. 297-304.|

© The Palaeontological Association

298

PALAEONTOLOGY, VOLUME 32

altered by diagenetic processes. The original composition of Lower Palaeozoic fossils may be
inferred by comparisons with living taxa of the same class, from petrographic evidence (Lohmann
and Meyers 1977; Richter and Fiichtbauer 1978), and from stable isotope data. These principles
are used for trilobites, which have no close modern relatives and, in the following sections, data
are presented which indicate an original LMC mineralogy.

MATERIAL AND METHODS

As part of a wider study on trilobite cuticles, examples of over seventy species of trilobite, ranging in age
from Cambrian to Devonian, were embedded in Araldite (epoxy resin) and made into uncovered thin-
sections. They were then examined by cathodoluminescence, using a Technosyn cold cathode luminoscope,
model 8200 Mkll, at a gun current of 15 18 kV and 400-600 mA. The thin-sections were later stained
(Dickson 1966) and protected by coverslips before examination in transmitted light.

Ten samples of five trilobite species from the Much Wenlock Limestone Formation, England (Wenlock)
were prepared for atomic absorption analyses. Unweathered specimens were scraped with a scalpel to obtain
approximately 50 mg of exoskeleton in powder form. Each sample was then dissolved in 10 ml of concentrated
HC1 diluted by 50%, and boiled on a hot plate for 2 min until the solution became clear, adding more distilled
water as necessary. The solutions were allowed to cool and made up to 100 ml volume with distilled water,
before running on a Perkin-Elmer atomic absorption spectrophotometer, model 460.

Pieces of trilobite cuticle, brachiopod valves, and individual crinoid ossicles from the Much Wenlock
Limestone Formation were prepared for carbon and oxygen stable isotope analyses. Unweathered specimens
were scraped with a scalpel to obtain 5 mg of powder, great care being taken to avoid contamination from
the surrounding matrix. Isotopic ratios were then obtained following established methods (McCrea 1950).
All isotopic values are in parts per million with respect to PDB (Pedee Belemnite). Precision is around
01% (1 <j) or better for both S13C and §lsO.

PETROGRAPHY

Trilobite exoskeletons are best preserved in dark, fine-grained limestones and calcareous mudstones.
Fine primary microstructures, such as canals and laminations, are often retained in such cases and
eye lenses may remain capable of focussing sharp images (Towe 1973). Loss of most primary
structure is inevitable during the inversion of aragonite to calcite. Occasionally, traces of organic
inclusions outline former structures and produce pseudopleochroic calcite in neomorphically altered
aragonite (Hudson 1962). As the preservation of trilobite exoskeletons is generally much better
than this, a primary A mineralogy can be discounted.

Richter and Fiichtbauer (1978) suggested that replacement by ferroan calcite is indicative of a
former HMC composition, as LMC is more stable and therefore not replaced. Trilobite cuticles
are very occasionally preserved in this way. The former existence of HMC can also be indicated
by the presence of microdolomites that formed in a closed system during the conversion to LMC
(Lohmann and Meyers 1977). During this process the excess magnesium ions which are not
incorporated into the LMC lattice form microdolomite inclusions 1-10 /an in diameter, which are
readily identified with cathodoluminescence. Microdolomites have never been found within trilobite
exoskeletons, despite the latter commonly occurring in close association with inclusion-rich
components, suggesting that they may have had an original LMC mineralogy. These differences
are well illustrated by petrological examination of fossils from the Much Wenlock Limestone
Formation, which show clear preservational variations between taxa (text-fig. 1). Gastropods which
had an original A exoskeleton have been replaced by a void-filling cement (Bathurst 1975)
with total loss of primary microstructure. Crinoid ossicles, originally HMC, have abundant
microdolomite inclusions and often contain secondary voids. Both brachiopods (primary LMC)
and trilobites show good preservation, with laminae and canals still visible. Neither taxon ever has
any microdolomite inclusions or shows evidence of secondary dissolution— suggesting that they
were both composed of LMC.

WILMOT AND FALLICK: TRILOBITE EXOSKELETONS

299

text-fig. 1. Composite photomicrographs of fossils from the Much Wenlock Limestone Formation. In each
case, the left-hand side is taken under cathodoluminescence (CL) and the right-hand side under plane
polarized light (PPL), a , crinoid ossicle, NMW 88.22G.1, originally HMC, now with syntaxial overgrowth.
Internally the ossicle has become coarsely crystalline, and under CL, small bright points mark the positions
of microdolomites which formed during the conversion to LMC. x 25. b, gastropod, NMW 88.22G.2,
originally A which has now been totally replaced by a coarse, clear, void-filling cement, x 30. c, strongly
ribbed brachiopod, NMW 88.22G.2, primary LMC retaining its original structure of fine laminations (1).
Under CL the brachiopod is non-luminescent and contains no microdolomite inclusions unlike the surrounding
matrix, x 30. d , trilobite cuticle, NMW 88.22G.2, with thin, outer prismatic layer (p), and fine canals (c)
within small tubercles, still visible under PPL. Under CL the exoskeleton is weakly luminescent, a function
of its trace element content, but no microdolomites are present. x40.

300

PALAEONTOLOGY, VOLUME 32

ATOMIC ABSORPTION ANALYSES

Average composition of ten trilobite specimens from the Much Wenlock Limestone Formation,
England — localities 43 and 64 of Thomas (1978)— were as follows:

%Ca2+

%CaC03

%Mg2+

%MgC03

Total CaMgC03

%Fe2+

35-18

87-94

0-81

2-82

89-89

0-50

±2-73

±6-81

±0-57

±1-98

±6-83

±0-19

(Impurities such as organic matrix, ferroan calcite, pyrite, and water account for

incomplete percentage totals.)

These data indicate an LMC mineralogy.

CARBON AND OXYGEN STABLE ISOTOPE ANALYSES

The principles involved in carbon and oxygen isotope analyses are outlined below; for a more
detailed introduction to stable isotopes and their geological uses, see Hoefs (1980) and Arthur et
al. ( 1 983). The isotopic fractionation of oxygen between calcium carbonate and water is temperature-
dependent (McCrea 1950). Although the isotopic composition of meteoric waters (as measured
from atmospheric precipitation) varies considerably with temperature, altitude, and latitude, the
isotopic composition of sea-water remains relatively constant at a given temperature because of
its much larger mass. It would be expected that marine invertebrates secreting a calcium carbonate
exoskeleton would do so in isotopic equilibrium with the surrounding sea-water. This is the case
in certain organisms, such as articulate brachiopods and Recent cirripedes, though other groups
such as echinoderms exhibit a ‘vital effect’ and secrete exoskeletons that are not in isotopic
equilibrium. This non-equilibrium is thought to result from variations in their physiological
processes, such as use of metabolic carbon dioxide (Milliman 1974; Morrison and Brand 1987).
Hence different marine invertebrate groups possess distinctive ‘isotopic signatures’, related both
to the temperature of the sea-water they inhabit and to their physiology. In general, forms with
LMC exoskeletons tend to be in isotopic equilibrium, whereas those secreting A or HMC hard
parts show a vital effect.

Isotopic signatures can be determined for fossils and coexisting inorganically precipitated
carbonate, including cements. For any locality the relative differences in carbon and oxygen isotopic
values between organisms and the sea-water at the time of their formation should still be apparent,
even with diagenetic effects superimposed, provided the components did not reach isotopic
equilibrium with pore fluids.

On this basis the composition of trilobite exoskeletons was determined using specimens from
the Much Wenlock Limestone Formation, which contains some of the best-preserved trilobites in
Britain. Conodont elements from this formation have suffered very minor thermal maturation, 50-
90 °C (Aldridge 1986), suggesting burial to only 1 — 1-5 km. Previous carbon and oxygen isotope
studies from this formation (Ratcliffe 1987) revealed a clear distinction between the isotopic values
of brachiopods and crinoid ossicles. These data from primary and secondary LMC are then used
here for comparison with the isotopic signature obtained from trilobite cuticles.

Discussion

The isotope data (text-fig. 2) show no relationship between the trilobite species used, or the locality
from which they came. The trilobite data plot much more closely to the brachiopod values than
to the other limestone components, and the spread of points can be explained when viewed in
relation with the diagenetic history of the formation (Ratcliffe 1987).

All originally unstable components now have elevated (313C signatures relative to Silurian
marine carbonates. Similarly, early diagenetic primary LMC cements have relatively high (513C

WILMOT AND FALLICK: TRILOBITE EXOSKELETONS

301

key

o

micrites

early diagenetic

cements

c, 9

burial cements

•

crinoid ossicle

O

brachiopod valve

•

crinoid field

©

brachiopod field

X

trilobite cuticle

-2

-1

X

-4

~T

-3

n r

-2 -1

0

— 1

<5 13 c

PDB

text-fig. 2. Carbon and oxygen stable isotope compositions for allochems and cements from the Much
Wenlock Limestone Formation. Based partly on unpublished data from K. T. RatclifFe.

values. These elevated values reflect the development of early diagenetic pore fluids, so the more
porous the constituent analysed, the more cement it will have incorporated, influencing the net
isotopic value obtained. The micrites all have relatively high r)13C values compared with the
brachiopods, which were in isotopic equilibrium with the sea-water (Lowenstain 1961; Popp et al.
1986). This is probably because they include small amounts of early diagenetic cements with a
high <5 1 3C value. They also show a north-east/south-west trend similar to that of the burial cements,
again indicating incorporation of burial cement in the micrite micro-pores. Crinoid ossicles also
have high <513C values, but with a relatively wide distribution. The composition of these originally
HMC skeletal parts is a reflection of their original isotopic signature, and partial equilibration
with the pore fluids that generated the early diagenetic primary LMC cements.

The brachiopods, being primary LMC and hence very stable, have low <513C values. They also
show very little spread in data due to their low porosity. These factors give brachiopods great
stability, making them suitable for palaeotemperature determinations. The trilobite cuticles also
have low <513C values and are distinct from both the micritic and cement phases, suggesting that
they, like the brachiopods, were constructed from LMC. Trilobite exoskeletons contain numerous
canals which were liable to being infilled by either early cements or burial cement, the degree to
which this happened being reflected in the variation of their isotopic values. When these effects
are taken into account, it can be seen that the isotopic signature of trilobites was originally very
similar to that of the brachiopods: close to isotopic equilibrium with sea-water.

302

PALAEONTOLOGY, VOLUME 32

The degree of porosity within trilobite exoskeletons makes them unsuitable for palaeotemperature
determinations despite their primary LMC mineralogy; eye lenses are a possible exception. Similar
problems arise in using belemnite rostra for this purpose (Veizer 1974). It is notable that the
isotopic composition of Silurian sea-water obtained from these data is lower in <513C than previously
published results from the Silurian of Gotland (Frykman 1986). However, as the <$lsO values are
in the same range, implying the same temperatures, this is probably due to differing productivity
levels between England and Gotland.

The present study emphasizes that fossils should not be studied in isolation, since the isotopic
values of the trilobite specimens can only be interpreted when viewed in conjunction with data
from other limestone allochems and cements. Only in this way can diagenetic effects be identified
and their influences evaluated.

CONCLUDING REMARKS

Comparison of the petrographic and chemical characteristics of taxa of known mineralogy with
those of trilobites occurring in the same samples strongly suggests that mineralized trilobite
exoskeletons were constructed from LMC. This contrasts with Richter and Fuchtbauer’s (1978)
inferred HMC composition based on replacement by ferroan calcite. However, this type of
replacement in trilobites is very rare and probably results merely from local diagenetic conditions.
Where ferroan calcite trilobites have been found in the Much Wenlock Limestone Formation, they
occur in a ferruginous crinoidal grainstone lithofacies which had major diagenetic interaction with
nearby iron-rich mudstones (Ratcliffe 1987), so that the vast majority of the other limestone
components have been affected. Trilobite cuticles were basically stable, being composed of LMC,
but nevertheless subject to alteration by highly reactive pore fluids. Microstructure (Walter 1985)
and the amount and composition of the organic matter within the exoskeleton all influence the
rate of ionic exchange. Even brachiopods, which are recognized as having had a LMC composition,
may occasionally be replaced by ferroan calcite (A. M. Searl, pers. comm.). The total absence of
microdolomites from trilobite cuticles, the absence of secondary voids, and the similarity in type
of preservation and isotopic composition between trilobites and brachiopods, indicate that trilobites
have a primary LMC mineralogy. Given the early appearance of trilobites in the Phanerozoic
record, this is compatible with the general evolutionary trend of biomineralization. This began
with a brief use of calcium phosphate in the Tommotian followed by a change to carbonate— first
LMC, then HMC— until today the majority of marine organisms use A (Wilkinson 1979;
Lowenstam 1981; Lowenstam and Weiner 1983).

Acknowledgements. This work was carried out as part of a Ph.D. project funded by an Aston University
Research Studentship Award. The Isotope Geology Unit is supported by NERC and the Scottish Universities.
N.V.W. thanks Dr A. T. Thomas for all his help and supervision, and Dr A. M. Searl for useful discussion.
We particularly thank Dr K. T. Ratcliffe who allowed free use of his data and discussed the results, as well
as commenting on an earlier version of the manuscript. Dr E. N. K. Clarkson also provided encouragement.
All figured material is deposited at the National Museum of Wales (NMW), Cardiff.

REFERENCES

aldridge, r. j. 1986. Conodont palaeobiogeography and thermal maturation in the Caledonides. J. geol.
Soc. Lond. 143, 177-184.

Arthur, m. a., anderson, t. f., kaplan, i. r., veizer, j. and land, l. s. (eds.). 1983. Stable isotopes in
sedimentary geology , 435 pp. Soc. econ. Paleont. Miner. Short Course 10, Dallas.
bathurst, R. G. c. 1975. Carbonate sediments and their diagenesis (2nd edn.), xix + 658 pp. Elsevier, Amsterdam.
dalingwater, j. e. 1973. Trilobite cuticle microstructure and composition. Palaeontology , 16, 827-839,
pis. 107-109.

WILMOT AND FALLICK: TRILOBITE EXOSKELETONS

303

— and miller, J. 1977. The laminae and cuticular organisation of the trilobite Asaphus raniceps.
Palaeontology , 20, 21-32, pis. 9 10.

dickson, j. a. d. 1966. Carbonate identification and genesis as revealed by staining. ./. sedim. Petrol. 12, 133
149.

dzik, j. and lendzion, k. 1988. The oldest arthropods of the East European Platform. Lethaia, 21,
29-38.

frykman, p. 1986. Diagenesis of Silurian bioherms in the Klinteberg Formation. Gotland, Sweden. In
shroeder, J. a. and purser, b. h. (eds.). Reef diagenesis, 399-423. Springer-Verlag, Berlin, Heidelberg, New
York.

hoefs, J. 1980. Stable isotope geochemistry (2nd edn.), xii + 208 pp. Springer-Verlag, Berlin, Heidelberg, New
York.

Hudson, j. d. 1962. Pseudo-pleochroic calcite in recrystallised shell-limestones. Geol. Mag. 99, 492-
500.

lohmann, k. c. and meyers, w. j. 1977. Microdolomite inclusions in cloudy prismatic calcites: a proposed
criterion for former high-magnesian calcites. J. sedim. Petrol. 47, 1078 1088.
lowenstam, H. a. 1961. Mineralogy, 180/160 ratios, and strontium and magnesium contents of Recent and
fossil brachiopods and their bearing on the history of the oceans. J. Geol. 69, 241 -260.

1981. Minerals formed by organisms. Science , N.Y., 211, 1126 1131.

and weiner, s. 1983. Mineralization by organisms and the evolution of biomineralization. In westbroek,
p. and de jong, E. w. (eds.). Biomineralization and biological metal accumulation , 191-302. D. Reidel,
Dordrecht, Holland.

mccrea, j. m. 1950. On the isotopic chemistry of carbonates and a paleotemperature scale. J. chem. Phys. 18,
849-857.

miller, j. 1976. The sensory fields and mode of life of Phacops rana (Green 1832) (Trilobita). Trans. R. Soc.
Edinb. 69, 337 367, pis. 1-4.

— and clarkson, e. n. k. 1980. The post-ecdysial development of the cuticle and the eye of the Devonian
trilobite Phacops rana milleri Stewart, 1927. Phil. Trans. R. Soc. B288, 461 480, pis. 1-7.

milliman, j. d. 1974. Recent sedimentary carbonates , Part I: marine carbonates , xv + 375 pp. Springer-Verlag,
Berlin, Heidelberg, New York.

morrison, J. o. and brand, u. 1987. Geochemistry of Recent marine invertebrates. Geoscience , Canada , 13,
237-254.

mutvei, h. 1981. Exoskeletal structure in the Ordovician trilobite Flexicalymene. Lethaia , 14, 225-
234.

popp, b. n., anderson, t. f. and sandberg, p. a. 1986. Brachiopods as indicators of original isotopic
compositions in some Palaeozoic limestones. Bull. geol. Soc. Am. 97, 1262-1269.
ratcliffe, k. t. 1987. Sedimentology, palaeontology and diagenesis of the Much Wenlock Limestone
Formation. Ph.D. thesis (unpublished), Aston University.

Richards, A. G. 1951. The integument of arthropods, xvi + 41 1 pp. University of Minnesota Press, Minneapolis.
richter, d. k. and fuchtbauer, h. 1978. Ferroan calcite replacement indicates former high-magnesian calcite
skeletons. Sedimentology , 25, 843 860.

stgrmer, l. 1980. Sculpture and microstructure of the exoskeleton in chasmopinid and phacopid trilobites.
Palaeontology , 23, 237-271, pis. 25-34.

teigler, d. j. and towe, k. m. 1975. Microstructure and composition of the trilobite exoskeleton. Fossils
Strata, 4, 137 149, pis. 1 9.

thomas, a. T. 1978. British Wenlock trilobites (Part 1). Palaeontogr. Soc. [Monogr.], I, 1 56, pis. 1 14.
towe, k. m. 1973. Trilobite eyes: calcified lenses in vivo. Science, N.Y., 179, 1007-1009.
veizer, j. 1974. Chemical diagenesis of belemnite shells and possible consequences for paleotemperature
determinations. Neues Jb. Geol. Paldont. Abh. 147, 91-111.

Walter, l. m. 1985. Relative reactivity of skeletal carbonates during dissolution: implications for diagenesis.
In schneidermann, n. and Harris, p. m. (eds.). Carbonate cements, 3-16. Soc. econ. Paleont. Miner. Sp.
Pub. 36, Tulsa, Oklahoma.

Whittington, h. b. 1977. The Middle Cambrian trilobite Naraoia, Burgess Shale, British Columbia. Phil.
Trans. R. Soc. B280, 409 443.

— 1985. Tegopelte gigas, a second soft-bodied trilobite from the Burgess Shale, Middle Cambrian, British
Columbia. J. Paleont. 59, 1251-1274.

wilkinson, b. 1979. Biomineralization, paleo-oceanography and the evolution of calcareous marine organisms.
Geology, 7, 524-527.

304

PALAEONTOLOGY, VOLUME 32

wolf, k. H., chillingar, G. v. and beales, F. w. 1976. Elemental composition of carbonate skeletons, minerals
and sediments. In chillingar, g. v., bissell, h. j. and fairbridge, r. w. (eds.). Carbonate rocks: Physical
and chemical aspects , 23-149. Elsevier, Amsterdam.

n. v. WILMOT

Department of Geological Sciences
Aston University, Aston Triangle
Birmingham B4 7ET, UK

Present address:

School of Earth Sciences
University of Birmingham
PO Box 363, Birmingham B15 2TT, UK

A. E. FALLICK
Isotope Geology Unit

Typescript received 1 May 1988 Scottish Universities Research and Reactor Centre

Revised typescript received 15 August 1988 East Kilbride, Glasgow G75 OQU, UK

THE MAMMAL-LIKE REPTILE RECHNISAURUS
FROM THE TRIASSIC OF INDIA

by S. BANDYOPADHYAY

Abstract. Rechnisaurus cristarhynchus from the Yerrapalli Formation (Middle Triassic) of the Pranhita
Godavari valley of India was described as a stahleckeriid dicynodont on the basis of its blunt snout and lack
of a high parietal crest. Another large Triassic dicynodont, but with a pointed snout, from the N'tawere
Formation of Zambia was also designated as Rechnisaurus cristarhynchus , while a skull from the Omingonde
Formation of Namibia was named Kannemeyeria simocephala. Keyser and Cruickshank considered all these
three species to be examples of K. cristarhynchus. A re-examination of the Indian R. cristarhynchus shows
that this species is quite distinct from the other kannemeyeriid and stahleckeriid genera. Because of its
incomplete nature the cranial measurements used for the classification of large Triassic dicynodonts cannot
be applied to it. Until complete material of this species is found, R. cristarhynchus from India should be
considered as incertae sedis.

The dicynodonts are a group of mammal-like reptiles which attained a world-wide distribution
during the Late Permian and the Triassic. These terrestrial herbivorous animals were quite
successful in the Permian and a large number of genera are known, but the number of both genera
and species declined in the Triassic.

Roy-Chowdhury (1970) briefly described the skulls of two large Triassic dicynodonts from the
Yerrapalli Formation of the Pranhita-Godavari valley, Deccan, India. Wadiasaurus indicus was
identified as a kannemeyeriid dicynodont while Rechnisaurus cristarhynchus was designated as a
stahleckeriid. However, much confusion has been created regarding the nomenclature as well as
the status of the genus Rechnisaurus (Keyser 1974; Keyser and Cruickshank 1979; Cox and Li
1983; Cruickshank 1986). An attempt is made here to re-examine the status of R. cristarhynchus
in the light of the recent family diagnosis of the Triassic dicynodonts (Cox and Li 1983).

The family Stahleckeriidae was established by Cox (1965) who first classified the Triassic
dicynodonts into four families: Kannemeyeriidae, Stahleckeriidae, Shansiodontidae, and Lystro-
sauridae. The family Stahleckeriidae was distinguished by its blunt snout, wide and low occiput,
short temporal opening, and lack of a parietal crest, and included Stahleckeria and Dinodontosaurus.
Cox and Li (1983), while reviewing anew all the Triassic dicynodonts known so far, proposed
modified and enlarged family diagnoses, but basically adhered to the family arrangement of Cox
(1965). According to the Cox and Li (1983) family diagnosis, the family Stahleckeriidae includes
medium to large dicynodonts characterized by the following features: the snout is wide, blunt, and
pronouncedly elongated, nearly 37-56% of the skull length, and bent in some genera; the jaw
articulation lies posteriorly; the occiput is almost vertical; the ratio of the skull length in the palatal
view to the dorsal skull length is usually more than 100%. The genera included in this family
by Cox and Li (1983) are Dinodontosaurus , Parakannemeyeria , Dolichuranus , Rhinodicynodon ,
Stahleckeria , Sinokannemeyeria , and Zambiasaurus. The present author follows their classification
except for a few alterations (Table 1).

THE STATUS OF RECHNISAURUS CRISTARHYNCHUS ROY-CHOWDHURY

R. cristarhynchus from the Yerrapalli Formation was designated as a stahleckeriid by Roy-
Chowdhury (1970) who came to this decision following Cox’s (1965) family diagnosis. The sole

| Palaeontology, Vol. 32, Part 2, 1989, pp. 305-312.]

© The Palaeontological Association

306

PALAEONTOLOGY, VOLUME 32

table 1. The revised classification of Triassic dicynodonts (modified after Cox and Li 1983).

Family Kannemeyeriidae

Kannemeyeria, Uralokannemeyeria, Shaanbeikannemeyeria , Rhadiodromus, Rabidosaurus , Ischigual-
astia, Placerias , Moghrebeeria , Wadiasaurus.

Family Stahlcckeriidae

Dinodontosaurus, Parakannemeyeria , Dolichuranus , Rhinodicynodon , Stahleckeria, Sinokannemeyeria.
Family Shansiodontidae

Shansiodon, Tetragonias, Angonisaurus, Vinceria( ?)

Incertae sedis

Bar y soma, Elephantosaurus , Jachalaria , Sangusaurus, Zambiasaurus, Rechnisaurus.

text-fig. 1. Rechnisaurus cristarhynchus. ISI R37 (after Roy-Chowdhury 1970). Restoration of the skull in
a, dorsal, 6, ventral, and c, side views. Scale bar 100 mm.

David Stanford

3

rt

c/3*

nr

o

3

nr

rt

3

rt

nr c/5

co c

^ o

rrt ft

Sf °

t; c/3
ft C/5

3 O

CL

ft

3 ~

’ < ns
-• 3
3 CL

T3 3“

n °
O T3
Cl. ft)

ft ^

3' O

TO C

- 5=

3 3

C/3* ’“

_. 3
c/) 3

£ CL

3 ft

ft n,

C/3

S'

3

CL

TO

ns

3.

n’

S'

TO

'<

O

3

cr

ft

o

3

3

n>

-o

3

cn 3
nr 3
ns q_

3

3 3-

TO ^
3- O

3 rt>

>-3 3>
^ n
O ft

^3 '"t

C/3

O *

3 £

o'

3

3

3

rt £

O ^

rt-, Q

o O-

5 §

nr 3

i-h TO
ns

• •<
o

3

3

ft

ft

X!

3

ns

3

r—F

ft

3

ft

nr

3

3

TO

ft

s ^

3 ft
CL

ft 3“
3 3

3. <
ft ft

3

C/)

O
2. £
ft 3
rt".

ft ^

C/3 _

nr

ft

nT

C/3

C/3

C/3

3

ft

ft

&

3

ft

'"a

C/>

nr

D

3

3

<

ft

C-

"d

C/3

3

3

ft

S'

C3S

0

13

o

CL

O

3

r—F

n

O

’-■f

5’

f— F

C/3

3

3

S'

5'

CL

ft

3

ft

3*

3

'<

3

C/3

"a

ft

0.

§

c—F

O

TO

X

O

CL

cr

3

C/3

"0

S3

S“

r—t-

3

'<

sS

3

ft

no

3

ft

H

ft)

O

ft

O

3;

o

3

3

CL

C/3 ’

ns

3

r—F

3

rt

S~

>

C/3

C/3

O

o'

3

C/3

3

O

ft

CL

nr

C/3*

r— F

sr

0

o.

"S'

ft

o

3'

ft

CD

>-h

O

cr

o'

*-*

fD

o

3

nr

3 =r

f ft

8 D

ft ft__
rt fT

3 TO
ft £2

_ „ rt

C n>

3 ^

3‘ O

< 3h

3 _►

nr

ft

o

3

CL

c

3

<’

ft

nr

3

<

ft

CL

ft

n

CL

ft

CL

To members of the Palaeontological Association from the
Printer to the University of Oxford.

BANDYOPADHYAY: RECHNISAURUS FROM INDIAN TRIASSIC

307

holotype skull (text-fig. 1) (ISIR37 in the collection of the Geological Museum, I.S.I., Calcutta)
has a wide and blunt snout which bears a strong midnasal ridge running from the anterior part
of the premaxilla, gradually broadening backwards, and dying out behind the nasofrontal suture.
The ridge is flanked on each side by a deep depression which widens posteriorly as the snout
broadens and terminates where the nasal meets the frontal and the prefrontal. The skull also
possesses a pair of powerful canines curving slightly inwards and placed quite posteriorly in the
maxillae. The interorbital area is quite wide and the temporal openings are apparently broad and
short, evidenced by the short and narrow intertemporal bar formed mostly by the paired parietals.
Presence of a boss immediately behind the pineal foramen and lack of a parietal crest were also
considered as important characters of the species. Roy-Chowdhury (1970) compared the genus
with other genera of the stahleckeriids, known at that time, and found that Rechnisaurus was closer
to Dinodontosaurus ‘but differs in having a high median nasal ridge and a boss behind the pineal
foramen’ (Roy-Chowdhury 1970, p. 137).

In the same year another dicynodont skull from the N’tawere Formation of Zambia was also
designated as R. crist arhynchus (text-fig. 2) by Crozier (1970), who identified it on the basis of
the presence of a strong midnasal ridge flanked by depressions and short and broad temporal
openings. The Zambian skull (no. 421, Bernard Price Institute for Palaeontological Research, also

text-fig. 2. Rechnisaurus cristarhynchus. BPI 3638 (after Crozier 1970) in a, dorsal, b , ventral, and c,
side views of the skull as preserved. (Dots represent matrix; hatchings represent broken bones). Scale bar

100 mm.

308

PALAEONTOLOGY, VOLUME 32

mentioned as BPI 3638 after its field number) is rather incomplete as the intertemporal bar, the
right orbital region, and a good part of the zygomatic arches are missing. While noting the presence
of a pointed snout in the Zambian skull, Crozier (1970) stated that the blunt snout of the holotype
skull (ISIR37) was due to ‘a fracture or erosion, notwithstanding a definite statement to the
contrary of Dr. P. L. Robinson’ (Crozier 1970, p. 39). Crozier (1970) further amplified her statement
by mentioning that ‘the palatal ridges of the type are not bounded anteriorly by any marked rim
as they are in the specimen here . . . which is the more normal condition’. It must be reiterated
here that the snout region of the holotype skull from India is devoid of any fracture or erosion
whatsoever (text-fig. 1) and consequently the basis of assigning the Zambian specimen to R.
cristarhynchus was founded on inadequate characterization and erroneous assumption which later
created confusion in the identification of other material.

Keyser (1973) described a kannemeyeriid skull (text-fig. 3) from the Omingonde Formation of
Namibia as Kannemeyeria simocephala (no. R313 in the collection of the Geological Survey, RSA).
He described the form as having a medium to large-sized skull with tusks in both sexes, zygomatic
arches parallel or subparallel in dorsal view, high and narrow parietal crest with no extensive

text-fig. 3. Kannemeyeria cristarhynchus. R313 (after Keyser and Cruickshank 1979) in a, dorsal, b , ventral,
and c, side views of the skull. Scale bar 100 mm approximately.

BANDYOPADHYAY: RECHNISAURUS FROM INDIAN TRIASSIC

309

exposure of interparietal on the dorsal surface. Subsequently, Keyser and Cruickshank (1979)
compared the skulls of K. simocephala from the Omingonde Formation of Namibia and the
supposed R. crist arhynchus from Zambia and found a great resemblance between the two (Table
2). They observed that both the forms (R313 and BPI 3638) has strong midnasal ridges flanked
by depressions, strong caniniform processes, and short temporal openings but Their parietal crests
not being as high as might be expected in a typical K. simocephala. From this comparative study
they made two conclusions. First, K. simocephala of Namibia (R313) was specifically distinct from
K. simocephala Weithofer. They renamed the Namibian specimen K. crist arhynchus. Secondly, R.
cristarhynchus of Zambia (BPI 3638) not only belonged to the genus Kannemeyeria, but was also
conspecific with the Namibian form. Both forms, therefore, were included in K. cristarhynchus.

table 2. Comparison of skull measurements (in mm) of Kannemeyeria cristarhynchus
from the Omingonde Formation (R 313) and Rechnisauus cristarhynchus from the
N’tawerc Formation (BPI 3638) (after Keyser and Cruickshank 1979).

R313

BPI 3638

Length: a, palatal midline

355

365

b, dorsal midline

409

450*

c, over squamosal wings

444

465*

Width over squamosal

406

454*

Interorbital distance

140

150

Internasal distance

150

160

Width of parietal crest at level of pineal

59

53*

Length behind postorbital on dorsal mid-line

130

140*

Length in front of postorbital on dorsal mid-line

279

310*

Length of internal nares

82

105

Length of fenestra mediopalatinalis

29

18

Diameter of tusks

40x29

36-5x31

Horizontal diameter: orbit

68

95

Horizontal diameter: nares

55

55

Depth: caniniform process

150

145

Interpterygoid
Internal nares

35%

17-6%

Preorbital length

x 100

Total mid-line length

69%

69%

* Estimate on damaged or distorted region.

Keyser and Cruickshank ( 1 979), following an earlier suggestion by Keyser ( 1 974), also concluded
that the generic status of Rechnisaurus was untenable and relegated it to a junior synonym of
Kannemeyeria. This conclusion cannot be accepted as the analysis is based on the characters of
the Zambian skull (BPI 3638) only. The holotype Rechnisaurus (ISI R37) was not taken into
account by Keyser and Cruickshank (1979) and the real difference between Kannemeyeria and
Rechnisaurus remains unexplored.

Unfortunately this erroneous conclusion (which started originally from a misconception) has
been followed by other workers such as Cox and Li (1983). Cruickshank (1986) described a
kannemeyeriid from the Manda Formation of Tanzania and named it as Sangusaurus parringtoni.
The holotype S. edentatus, collected from the N’tawere Formation of Zambia, was recognized on
the basis of some fragmentary skull material by Cox (1969), but later Cox and Li (1983) considered
it as incertae sedis probably because of its indefinite characters. Nevertheless, Cruickshank (1986)

310

PALAEONTOLOGY, VOLUME 32

related S. parringtoni to ‘K. cristarhynchus (Chowdhury)’ on the basis of ‘broad open groove on
parietal mid-line, and a boss immediately behind the pineal opening’. The species might be similar
to S. edentatus but definitely differs from the Indian R. cristarhynchus in having a sharply pointed
snout and lacking a deep depression beside the midnasal ridge. Cruickshank (1986), because of
the incomplete nature of the skull, kept his decision open and stated ‘when more material is known
this decision may have to be reversed’. However, in the same paper, in the discussion of the
evolution of the kannemeyeriid dicynodonts, he used K. cristarhynchus (Roy-Chowdhury) as a key
species. It is interesting to note that the figures he used to illustrate K. cristarhynchus show lobe-
like bars in the intertemporal region (Cruickshank 1986, fig. 4a). However, the Indian specimen
of R. cristarhynchus , although having a somewhat incomplete parietal crest, does not show any
indication of forming any long bars behind (text-fig. 1), nor do the Zambian or Namibian
specimens, so whether the structure of the intertemporal region is a good guide to taxonomic
affinity remains doubtful.

An examination of the holotype skull of R. cristarhynchus (ISI R37) reveals that this form is
quite distinct from other dicynodont genera belonging to the Family Kannemeyeriidae. Its wide
and blunt snout indicate an affinity with stahleckeriids. However, the only skull available for study
is incomplete; most of the occiput, zygomatic arches, and interparietal are missing. In the description
of the material, Roy-Chowdhury (1970) mentioned ‘The zygomatic arches are broken a little
behind the maxillae, but the well preserved post-orbital bar helps in restoring the continuation of
the suborbital bar up to the orbit and also indicates the position of the more posterior extension
of the zygomatic bar. ... In the occiput, only the condyle, the foramen magnum and the median
part of the supraoccipital are preserved, with a minor break above the condyle. The squamosals
are missing save for an isolated piece near the dorsal part of the lateral wing of the right squamosal.’
Because of this incomplete nature of the skull many of the measurements used by Cox and Li
(1983) for taxonomic characterization are not available. Moreover, to ascertain the definite familial
status of R. cristarhynchus , better material will have to be obtained. Until then Rechnisaurus should
be considered as incertae sedis showing some affinity to the Family Stahleckeriidae Cox 1965.

SYSTEMATIC PALAEONTOLOGY

In the light of the above discussion a revised systematics of the three specimens under consideration is given
below.

Family kannemeyeriidae
Genus kannemeyeria Weithofer 1888
Kannemeyeria cristarhynchus (Crozier 1970; Keyser and Cruickshank 1979)

Synonyms. ? Rechnisaurus cristarhynchus Crozier 1970, N’tawere Formation, Zambia; Kannemeyeria simo-
cephala Keyser 1973, Omingonde Formation, Namibia.

Type specimen. R421/BPI 3638, a partial skull 620 mm long and complete lower jaw 320 mm long, in the
collection of the Bernard Price Institute of Palaeontological Research.

Locality and horizon. Locality no. 16 of the Lower Fossiliferous horizon in the N’tawere Formation, Zambia.
Referred specimen. R313 in the collection of the Geological Survey, RSA.

Locality and horizon. Between the lower and middle arenaceous horizons of the Lower Etjo Beds, Omingonde
Formation, Namibia.

Diagnosis. Skull dorsally triangular in outline with very large canine tusks; maxillary process with
exceedingly wide lateral flanges. Wide interorbital region. Midnasal ridge on the anterior and
dorsal surface and shallow depression on either side extending from tip of the snout to the
interorbital region. No postfrontal. Preparietal with low boss in front of pineal foramen. Short and
broad temporal opening. Jugal occupies most of the length of the zygomatic arch. Premaxilla

BANDYOPADHYAY: RECHNISAURUS FROM INDIAN TRIASSIC

31 1

short. Septomaxilla forms posterior wall and floor of the nostril. Distinct ectopterygoid. Behind
the pituitary foramen a small boss consisting of part of the epipterygoid fused to pterygoid.
Secondary palate, with three parallel grooves, the central one running into the vomerine ridge.
The two anterior ridges meet the ridge around the rim of the premaxilla. Moderate-sized labial
fossae at the junction of the maxilla, pterygoid, and the jugal. Maxilla enters the internal narial
passage. Palatine extends the entire length of the pterygoid ramus to meet the maxilla. Low, broad
occiput. Deep quadrate fossa on anterior face of occiput. Lower jaw with massive dentary. Deep
central and shallow lateral grooves on dorsal surface on the dentary. Surangular with short lateral
face. Long S-shaped Meckel’s fossa. Reflected lamina of the angular meets the horizontal flange
of the lateral condyle leaving an oval opening dorsally between the angular and the reflected
lamina. Long and broad shallow condyle allowing longitudinal and lateral movement of the jaws
(after Crozier 1970).

Taxon tentatively assigned to Family stahleckeriidae
Genus rechnisaurus Roy-Chowdhury 1970
Rechnisaurus crist arhynchus Roy-Chowdhury 1970

Type specimen. ISI R37, an incomplete skull about 380 mm long, in the collection of the Geological Museum,
Indian Statistical Institute, Calcutta.

Locality and horizon. 1 km south of Rechni village in the Yerrapalli Formation of the Pranhita-Godavari
valley, Andhra Pradesh, India.

Diagnosis. Moderately large-sized skull, about 380 mm long. Large canine teeth. Wide interorbital
region. Blunt snout. Strong median ridge on anterior and dorsal surfaces of premaxilla continuing
up to the dorsal side of the nasal and flanked by a pair of deep depressions. Powerful anteroventrally
directed caniniform process bearing rugose rounded flange on its posteroventral edge. Short
postorbital region. Short and wide temporal opening. Fairly narrow intertemporal bar, dorsally
concave in cross-section. Parietal crest not high. Low boss immediately behind pineal foramen.
Parietal forms most of the intertemporal bar. Sharp transition between dorsal and occipital surface
(after Roy-Chowdhury 1970).

CONCLUSIONS

The case of the identification and naming of Rechnisaurus highlights problems in several areas:

1. Classification of the Triassic dicynodonts has proved particularly refractory mainly because
the type specimens are distributed in several continents; the majority of Permian types are in South
Africa. Personal observation of specimens is essential in order to produce a consistent and coherent
classification, which makes it unlikely that any one worker will be able to see all specimens. This
makes accurate description in the literature of paramount importance. The present study is a
contribution towards this.

2. Several studies of the functional morphology of Triassic dicynodonts have appeared recently
(Walter 1986; Bandyopadhyay 1988). Snout morphology in particular has often been used as a
pointer to skull function in these animals, so it would be interesting to have the state of the snout
in Rechnisaurus confirmed.

3. The Indian fauna is comparable with dicynodont faunas from other continents from broadly
the same time span and the accurate classification of Rechnisaurus will eventually add to the
knowledge of this fauna.

Acknowledgements. I thank Professor T. K. Roy-Chowdhury of the Indian Statistical Institute for discussion.
I am grateful to Dr G. M. King of the University Museum of Oxford for her suggestions and comments and
critical appraisal of the manuscript.

312

PALAEONTOLOGY, VOLUME 32

REFERENCES

bandyopadhyay, s. 1988. A kannemeyeriid dicynodont from the Middle Triassic Yerrapalli Formation. Phil.
Trans. R. Soc. Loud. B320, 185-233.

cox, c. b. 1965. New Triassic dicynodonts from South America, their origin and relationship. Ibid. B248,
457-516.

- 1969. Two new dicynodonts from the Triassic N’tawere formation, Zambia. Bull. Br. Mus. nat Hist.
(Geol.), 17, 255-294.

and li, J. 1983. A new genus of Triassic dicynodont from East Africa and its classification. Palaeontology ,
26, 389-406.

crozier, E. a. 1970. Preliminary reports on two Triassic dicynodonts from Zambia. Palaeont. afr. 13, 39-45.
cruickshank, a. r. i. 1986. Biostratigraphy and classification of a new Triassic dicynodont from East Africa.
Mod. Geol. 10, 121-131.

keyser, a. w. 1973. A new Triassic vertebrate fauna from South West Africa. Palaeont. afr. 16, 1-15.

1974. Evolutionary trends in Triassic Dicynodontia. Palaeont. afr. 17, 57-68.

and cruickshank, a. r. i. 1979. Origins and classifications of Triassic dicynodonts. Trans, geol. Soc. S.
Afr. 82, 81-108.

roy-chowdhury, t. 1970. Two new dicynodonts from the Yerrapalli Formation of Central India.
Palaeontology, 13, 132-144.

Walter, l. 1986. The limb posture of kannemeyeriid dicynodonts: functional and ecological considerations.
In padian, k. (ed. ). The beginning of the age of dinosaurs, 89-97. Cambridge University Press, Cambridge.

S. BANDYOPADHYAY

University Museum and Department of Zoology
Parks Road, Oxford OX1 3PW

Current address:

Manuscript received 11 February 1988
Revised manuscript received 17 June 1988

Geological Studies Unit
Indian Statistical Institute
203 Barrackpore Trunk Road
Calcutta 700 035, India

ABBREVIATIONS USED IN THE TEXT FIGURES

BO

Basioccipital

PAL

Palatine

bo.t.

basioccipital tubera

p.f.

pineal foramen

car. for

carotid foramen

PMX

Premaxilla

ECT

Ectopterygoid

PO

Postorbital

EO

Exoccipital

PP

Preparietal

EPT

Epipterygoid

PRF

Prefrontal

F

Frontal

PRO

Prootic

f.m.

foramen magnum

PSP

Parasphenoid-basisphenoid complex

f.o.

fenestra ovalis

PT

Pterygoid

IP

Interparietal

pit.f.

Pituitary foramen

J

Jugal

Q

Quadrate

L

Lacrimal

QJ

Quadratojugal

l.f.

labial fossa

SMX

Septomaxilla

MX

Maxilla

SQ

Squamosal

N

Nasal

ST

Stapes

OP

Opisthotic

u.c.

upper canine

P

Parietal

V

Vomer

IOCRINUS IN THE ORDOVICIAN OF
ENGLAND AND WALES

by STEPHEN K. DONOVAN and ANDREW S. GALE

Abstract. Iocrinus contains more described species than any other crinoid genus known from the Ordovician
of England and Wales. British species of this taxon generally have a smooth, conical, dorsal cup and a
proximal stem which is pentagonal in transverse section. A new species, I. pauli, from the Llanvirn of the
Builth Wells area, is unusual in having a ribbed dorsal cup and a proxistele of pentastellate transverse section.
These features have hitherto been noted only in Iocrinus from North America. Iocrinus sp. cf. I. pauli is
recognized from the Lower Llandeilo of Dyfed on the basis of disarticulated columnals and a poorly preserved
crown. I. whitteryi Ramsbottom, from the Caradoc of Shropshire, is refigured to illustrate such diagnostic
features as the anal sac, the brachial articulum, and the stem. The family Iocrinidae Moore and Laudon is
now known to include six genera. With one exception of Lower Llandovery age, all known iocrinids are of
Ordovician age, and can be divided into two groups depending upon the complexity of the anal series.

The disparid crinoid genus Iocrinus Hall is known from the Ordovician of North America and
the United Kingdom. The first British species were not recognized until 1961, when Ramsbottom
described I. shelvensis and I. whitteryi. He also tentatively suggested that Dendrocrinus cambriensis
Hicks may be an Iocrinus. Bates (1965) described a further species, I. brithdirensis , and later (1968)
showed that D. cambriensis was not an Iocrinus , but a member of a new disparid genus
Ramsey ocrinus.

Iocrinus is the most diverse crinoid genus found in the British Ordovician south of the Iapetus
Suture. Hitherto, British species have been differentiated from those of North America in having
a smooth, conical, dorsal cup that lacks ribbing and a proximal stem that is pentagonal, rather
than pentastellate, in transverse section. A new species from the mid-Ordovician of Wales, described
below, shows those features which hitherto were thought to be limited to North American species
of Iocrinus.

Of the other British species of Iocrinus , /. shelvensis (Ramsbottom 1961, pp. 3-4, pi. 1, figs. 3-
8; Donovan 1986, p. 27, pi. 1, figs. 3, 4, 8, 9, text-fig. 12a-j) and I. brithdirensis (Bates
1965; Donovan 1986, p. 25, text-fig. 12k) are both adequately described elsewhere. However,
Ramsbottom’s original illustration of I. whitteryi (1961, pi. 1, fig. 9) did not show the diagnostic
features of the genus such as the anal sac. This species is reillustrated below.

Crinoid terminology used herein follows Moore et al. (1968), Ubaghs (1978), and Webster
(1974). The synonymy of I. whitteryi was compiled in the style advocated by Matthews (1973).

SYSTEMATIC PALAEONTOLOGY

Class crinoidea J. S. Miller, 1821
Order disparida Moore and Laudon, 1943
Family iocrinidae Moore and Laudon, 1943
Genus iocrinus Hall, 1866

Type species. By monotypy; Heterocrinus ( Iocrinus ) polyxo Hall 1866 ( = Heterocrinus suberassus Meek and
Worthen 1865).

Other species. I. brithdirensis Bates 1965; I. crassus (Meek and Worthen 1865); I. pauli sp. nov.; /. shelvensis
Ramsbottom 1961; I. similis (E. Billings 1857); I. trentonensis Walcott 1883; I. whitteryi Ramsbottom 1961.

[Palaeontology, Vol. 32, Part 2, 1989, pp. 313-323.|

© The Palaeontological Association

314

PALAEONTOLOGY, VOLUME 32

Diagnosis. (Modified after Moore et al. 1978, p. T552; Kelly 1978, p. 54.) Dorsal cup conical to
slightly bowl-shaped, composed of five prominent basal plates and five larger radial ossicles. An
anibrachial, supported by the C ray radial, in turn supports the anal series on the left side and a
free arm on the right. Anal sac complex. Arms with at least four isotomous branches. Column
transversely pentagonal or pentastellate proximally, pentagonal in the mesistele and circular
distally. Permanent attachment by distal, non-planar, spiral coil.

Remarks. We agree with Ubaghs (1978, p. T 1 1 8) that the plate supported by the radial in the C
ray is an anibrachial, rather than a superradial (Moore et al. 1978, p. T552).

locrinus pauli sp. nov.

Text-figs. 1-3

Derivation of name. For Dr Christopher R. C. Paul.

Type material. The holotype and three paratype anal sacs occur together on a single slab, British Museum
(Natural History) (BMNH) E71413 (text-figs. I and 2). An external mould without counterpart. Probably
from near Hundred House (see below).

Other material and locality. A single, well-preserved specimen in the private collection of Mr J. J. Savill (text-
fig. 3). An external mould without counterpart. From National Grid Reference SO 0970 5590 (Mr C. Moore,
written comm.), on the road from Hundred House to Llandrindod Wells, Powys, mid-Wales. Didvmograptus
bifidus Beds, although there is some doubt as to the precise age of this locality (see below). A cast is deposited
in the BMNH E71424.

text-fig. 1. locrinus pauli sp. nov., type specimens, BMNH E71413. All specimens preserved as external
moulds, apart from some portions of anal sac which appear to have been infilled. Holotype (H) and paratypes

(P) indicated, x I

DONOVAN AND GALE: IOCRINUS IN ENGLAND AND WALES

315

Horizon. The fauna associated with the type material suggests that this species may be either late Lower
Llanvirn or early Upper Llanvirn. A dissociated trilobite pygidium has been identified as Flexicalymene sp.
cf. F. aurora Hughes (BMNH It 19109) by Drs R. A. Fortey and S. F. Morris. Common brachiopods
(BMNH Be 10656) are assigned to Tissintia plana (Williams) (Dr D. A. T. Harper, written comm.). Thomas
et al. (1984, fig. 10) suggested that F. aurora ranges from the Lower into the earliest Upper Llanvirn, although
Hughes (1969, p. 82) considered the species to be limited to the uppermost Lower Llanvirn. Lockley and
Williams (1981, table, p. 5) noted that T. plana occurs in the Upper Llanvirn and Lower Llandeilo of the
Llandeilo area.

text-fig. 2. Iocrinus pauli sp. nov. holotype, BMNH E71413. Camera lucida
drawing from a latex cast. A, C, D, and E rays indicated. * = anibrachial
plate.

Diagnosis. A species of Iocrinus with a proxistele of pentastellate section and a ribbed, conical
dorsal cup. Basals about as high as wide, radial plates higher than wide. Five primibrachials per arm,
with at least five secundibrachials, five to nine(?) tertibrachials and about thirteen quartibrachials per
branch.

Description. Form of attachment uncertain, but the dististele of the BMNH E71424 (text-fig. 3) is curved
through about 90° (although slightly disarticulated), so that it is reminiscent of the distal stem in I. brithdirensis
(cf. Bates 1965, pi. 45, fig. 5; Donovan 1986, text-fig. 12k). I. pauli was, therefore, probably attached by a
distal, non-planar, spiral coil (Brett 1981), as were other species of Iocrinus (Kelly 1978).

The holotype retains the proximal 13-5 mm of stem (text-figs. 1 and 2) and the column of BMNH E71424
is over 20 mm long (text-fig. 3), but neither is complete. The column is pentastellate in transverse section
proximally, changing to pentagonal shortly below the cup. Latera strongly convex, except in tertinternodals
and quartinternodals, which have weakly convex to planar latera. Articular facet divided into five equal
areola petals surrounding a central, pentagonal lumen. Articulation uncertain, either symplectial or synostosial.
Column heteromorphic, Nl(?) proximally to N4342434 14342434 distally (notation after Webster 1974).
Nodals highest, quartinternodals lowest. Columnal angles swollen, except in quartinternodals.

Dorsal cup monocyclic, wider than high, with an angular, conical outline. Basals low, pentagonal, concave,
presumably five in number, with longitudinal grooves corresponding to the folds of the proximal column.
Radials as high as wide (about twice as high as basals), presumably five in number, with prominent, central,
longitudinal ribs radially positioned and corresponding to the angles of the column. Regions between ribs

316

PALAEONTOLOGY, VOLUME 32

text-fig. 3. Iocrinus pcndi sp. nov., BMNH E71424. Camera lucida drawing from a latex cast. C ray (and
some dissociated brachials from the B ray?) occupies the upper right and centre of the figure. Note the

slightly disarticulated anal pyramid.

DONOVAN AND GALE: IOCRINUS IN ENGLAND AND WALES

317

depressed, but a further pair of ridges, parallel to the adoral surface of the cup, link the main ribs of adjacent
radials. All radials observed are of equal size.

The complex anal sac arises from the anibrachial plate in the C ray and is about 30 mm long by 4 mm
wide in BMNH E71424, reaching 35+ mm in the longest paratype (text-fig. I). The proximal anal series is
best seen in the holotype (text-fig. 2) where it comprises a series of elongate plates, slightly longer than wide,
with central, convex latera flanked by planar flanges. The anal series supports lateral sac covering plates.
The anal sac is terminated by an anal pyramid (text-fig. 3) composed of solid, conical ossicles.

The arms are uniserial, non-pinnulate, isotomously branching at least four times and presumably five in
number. The C ray radial supports an anibrachial (text-figs. 2 and 3). Where apparent, there are five
primibrachials per ray, including those above the anibrachial. Following each successive bifurcation, there
are five secundibrachials, five to nine(?) tertibrachials, and about thirteen (minimum) quartibrachials per
branch. Brachials are stout proximally, slender distally. Latera are planar and unsculptured. Adoral groove
V-shaped, with brachials having a U-shaped section. Brachials are about as high as wide proximally, but
higher than wide distally.

Remarks. I. pauli sp. nov. is easily distinguished from the other British species of Iocrinus , I.
brithdirensis Bates, I. shelvensis Ramsbottom, and I. whitteryi Ramsbottom, all of which have
smooth, conical, dorsal cups and, where known, a proxistele of pentagonal section. I. pauli most
closely resembles the North American species I. subcrassus (Meek and Worthen). However, I.
subcrassus has basals and radials that are wider than high, whereas in I. pauli basals are about as
wide as high, while radials are slightly higher than wide. The base of the dorsal cup in I. subcrassus
appears broader (cf. Moore 1962, pi. 2, figs. 1 and 2a: Kelly 1978, pi. 1, figs. 12 and 13), with a
lower anibrachial plate (cf. Kelly 1978, text-fig. 5), than in I. pauli. Further, it may be relevant
that I. pauli is significantly older than the first I. subcrassus (text-fig. 6) and no other crinoid species
has yet been recognized from both sides of the Iapetus Ocean in the Ordovician.

Of the other species of Iocrinus from North America, I. crassus (Meek and Worthen) has an
unusually large dorsal cup, with a radial : basal height ratio of about 3 : 2, compared with 2 : 1 in
I. pauli. I. similis E. Billings is probably aberrant (Kelly 1978, p. 61) and is only known from a
unique holotype. The proxistele is pentagonal and an additional ossicle is present between the C
ray radial and the anibrachial. I. trentonensis Walcott is very similar to I. subcrassus. A new species
of Iocrinus from the lower Ordovician of North America is awaiting description by Kelly.

Iocrinus sp. cf. I. pauli sp. nov.

Text-fig. 4

Material, locality , and horizon. A partially disarticulated crown, BMNH E69597, an external mould without
counterpart, plus a brachial, BMNH E71418, and five columnals, BMNH E71414 E71416, E71417a, b
(counterparts), E71419u-c (counterparts), all of which are external moulds. All from the Lower Llandeilo,
Llandeilo Flags, Ffairfach, Dyfed, South Wales, NGR SN 628 212 (Bassett 1982, p. 284, fig. 3).

Description. Disarticulated, pentastellate to pentagonal columnals with a small, central, pentagonal lumen
(text-fig. 4a, b, d). Lumen angles correspond to the angles of the columnal. Five separate areola petals, which
vary from rounded-pentagonal to elongate-triangular in outline, correspond to the angles of the lumen.
Symplectial articulation, with parallel crenulae extending between areola petals in more stellate columnals
(text-fig. 4b, d). No pentastellate columnal preserves a crenularium at the extremities of the rays. Columnals
low, with slightly convex latera.

Dorsal cup unknown apart from a single, disarticulated, ?radial plate (text-fig. 4c). The anal sac is
incomplete, but the proximal 17 mm is preserved, along with a more distal fragment. Two parallel series of
lateral sac covering plates are apparent, with a suture between them that possibly conceals the pseudo-anal
series (cf. Kelly 1978, text-fig. 4e).

Arms long, apparently at least 75 mm minimum (text-fig. 4c), uniserial, non-pinnulate, and branching
isotomously. The number of branching events is unknown, but one or ?two are apparent distally. Proximal
brachials are about as high as wide. Distal brachials are smaller and higher than wide. Brachials have a U-
shaped section with a deep, broad, V-shaped adoral groove. Depressed, crescentic ligament pit aboral to the
adoral groove.

318

PALAEONTOLOGY, VOLUME 32

text-fig. 4. Iocrinus sp. cf. I. pauli sp. nov. a, BMNH E714176,
lumen, b, BMNH E71415, articular facet, c, BMNH E69597,
partially disarticulated crown. AS = anal sac; R = radial; ? =
plate of uncertain affinity; other structures are fragments of
arms or disarticulated brachials. d, restoration of articular
facet. Apart from d, all figures are camera lucida drawings
from external moulds (a, b) or a latex cast (c).

Remarks. Features such as the pattern of arm branching, the brachial facet architecture, the large
anal sac, the outline of columnals, and the sculpture of their facets, all indicate that this species is
a typical Iocrinus. It is tentatively recognized as an Iocrinus similar to I. pauli , the only British
species with a pentastellate proxistele and ribbed radials. However, the Ffairfach species differs
from the available material of I. pauli in having a larger crown.

DONOVAN AND GALE: IOCRINUS IN ENGLAND AND WALES

319

Iocrinus whitteryi Ramsbottom, 1961

Text-fig. 5

1869 Cyathocrinus sp.; Morton, p. 19.
v* 1961 Iocrinus whitteryi sp. nov., Ramsbottom, p. 5, pi. 1, fig. 9.

1973 Iocrinus whitteryi Ramsbottom; Webster, p. 157.

1978 Iocrinus whitteryi Ramsbottom; Kelly, pp. 63, 64.

1986 Iocrinus whitteryi Ramsbottom; Donovan, p. 21.

Material , locality, and horizon. Holotype, BMNH E49603, an external mould without counterpart. Paratypes,
BMNH E1365, E49604, E49605 (all external moulds without counterparts); British Geological Survey (BGS
GSM) 85720, 85720u (counterpart external moulds), 85721 (wax squeeze of holotype). Ail specimens are
from Whittery Quarry, east of Marrington Dingle, near Chirbury, Shropshire, NGR SO 274 982. Chirbury
Formation, Whittery Shale Member(?), Caradoc, Soudleyan (Whittard 1979, locality 299, p. 56, fig. 33).

Remarks. Ramsbottom (1961, p. 5) gave an adequate description of I. whittery , but the associated
illustration (pi. 1, fig. 9) is undiagnostic of Iocrinus. Text-fig. 5 is therefore included herein to show
those features of I. whitteryi that are typical of the genus.

text-fig. 5. Iocrinus whitteryi Ramsbottom, 1961. All paratypes apart from d, e. a,
b, BMNH E49604. A pair of anal sacs, juxtaposed as preserved, and showing the
morphology of both the internal and external surfaces. Both are in a similar
orientation, with the plates of the anal series facing into the page. A short
pluricolumnal is illustrated to the left of A, which is also associated with an arm
fragment that branches twice, c, BMNH El 365. A further fragment of anal sac. d,
e, BMNH E49603, holotype. d, dorsal cup showing three rays and some dissociated
brachials. e, articular facet of brachial arrowed in d. All camera lucida drawings
from latex casts. Not all fragments from each specimen have been illustrated.

320

PALAEONTOLOGY, VOLUME 32

The holotype, sadly lacking a counterpart, is the only known dorsal cup (text-fig. 5d; Ramsbottom
1961, pi. 1, fig. 9). This shows three rays, none of which includes an anibrachial and associated
anal series. This specimen must therefore represent either the D-E-A or the E-A B rays. None
of the arms of the holotype appears to branch (but note branching at the ?primaxilliary and
?secundaxilliary levels in text-fig. 5a), so the primaxilliary must be at the level of lBr12 or above.
Dissociated brachial ossicles on the same slab, apparently derived from the holotype, show an
articular facet architecture similar to that of /. shelvensis Ramsbottom (text-fig. 5e).

The anal sacs attributed to I. whitteryi (text-fig. 5a-c) are comparable to those known from
other species of Iocrinus (cf. Ramsbottom 1961, pi. 1, fig. 6; Moore 1962, pi. 2, fig. 2 a-e; Bates
1965, pi. 45, figs. 1, 2, 4-6; Kelly 1978, pp. 10-25, text-fig. 4, pi. 1, figs. 1-3, 5, 6, 14). This complex
structure is diagnostic of the genus. Additionally, a short pluricolumnal, previously undetected,
has been recognized on one of the paratype slabs of I. whitteryi (text-fig. 5a). This appears to be
pentagonal in section, with rounded angles, and is heteromorphic. It is not dissimilar to some of
the less nodose, pentagonal columns of other species of Iocrinus.

PHYLOGENY OF THE IOCRINIDAE

The disparid crinoid family Iocrinidae is now known to have contained at least six genera, spanning
the interval lower Ordovician to Lower Llandovery (the only known Lower Ordovician species,
Iocrinus sp. nov., is awaiting description by Kelly and is not included in this analysis). Iocrinid
genera are differentiated on the basis of column structure, form of attachment, shape of the dorsal
cup, basals exposed or cryptic, interbrachial plates present or absent, first primibrachials free or
attached, and the structure of the anal series. On the basis of these characteristics, it is suggested
that the iocrinids appear to be divided into two natural groups. These groups can easily
be recognized, comprising those species which have either a complex ( Iocrinus ) or a simple
(Caleidocrinus + Ristnacrinus + gen. nov.) anal series (text-fig. 6).

Iocrinus itself can be divided into two lineages, based on whether the proxistele is pentagonal
or pentastellate in section and whether the dorsal cup is unsculptured or ribbed, respectively. Kelly
(1978) considered that the ribbed species probably evolved from an ancestor with a smooth cup,
but the occurrence of I. pauli suggests that the time of divergence of these two lineages was much
earlier than he originally postulated. In text-fig. 6 the suggested relationships of the later species
with a ribbed dorsal cup essentially follows Kelly (1978, text-fig. 15), while those for species with
an unsculptured dorsal cup follows Donovan (1985«, fig. 8).

The features of the two species of Caleidocrinus Waagen and Jahn, and the relationship of
Caleidocrinus to Ristnacrinus Opik, have also been discussed in detail elsewhere (Donovan 1985u).
Ristnacrinus is particularly well recorded on the basis of dissociated columnals, although at least
some of these may be derived from Caleidocrinus (C.) multiramus Barrande. Only the type species
of Ristnacrinus , R. marinus Opik, is known from relatively complete material. 1R. altohasalis
Brower and Veinus is poorly known and its generic assignment is, at best, dubious. It is perhaps
significant that the distinctive columnals of Ristnacrinus have not been reported from North
America. R. cirrifer Le Menn is based only upon cirriferous columnals. As some columnals of the
type species are cirriferous and indistinguishable from Le Menu’s species, this suggests that R.
cirrifer is invalid. Four further species based on disarticulated columnals (Stukalina 1980) appear
to be morphologically distinct. However, because of the general uncertainty regarding the
subdivision of Ristnacrinus, this genus has been left undivided in text-fig. 6.

A new genus (Donovan, in press) differs from Iocrinus in lacking a large and complex anal sac.
The basal plates of Caleidocrinus are either low or cryptic (Donovan 1985u) and interbrachial
plates are present, whereas in the new genus basals are prominent and interbrachial plates are
absent. Ristnacrinus has a slender dorsal cup with cryptic basals and a circlet of five radials
supporting five fixed primibrachs (including the anibrachial in the C ray). The dorsal cup in the
new genus is broad, tapering towards the base, and all brachials are free. The new genus is most
similar to Caleidocrinus (H.) turgidulus Ramsbottom.

DONOVAN AND GALE: IOCRINUS IN ENGLAND AND WALES

321

KEY TO FAUNAL PROVINCES

CD

LU

0

• SOUTH EUROPEAN

CD

LU

LU

0

Q_

If)

LU

■ NORTH AMERICAN

LU

CD

g

CO

C£

>

LU

CD

IS

CD

* BALTIC

* KAZAKHSTAN

<

<

5

3

3

Z

X

© A

HIR

® A

GAN

—

_ nGEN.

® *

—

RAN

/. CRASSUS 1 NOV.

H m I

• *

?#A

RIC

440

CAU

■ i

*# A

* 1

_l

_J

0

i

*
* 1

<

Z

<

X

X

□

* 1

<

CD

<

<

■ i

* |
* 1
* 1

z

£

_i

0

CD

/. SUBCRASSUS-

B 1

■ |

o

X

0

_2_

3

Q_

■

* 1

IB

* 1

z

450-

ONN

/. TRENTONENS/S

w

H 1

■ 1. S/M/LIS 1

* 1
* 1
* 1

<

z

LU

Q

-IAN

,■ ■ i

|

LU

ACT

7 ✓ 1

* 1

-

MAR

-

RISTNACRINUS SPP.-

* 1
* 1

z

<

LON

1

*•

*•

<

1

* x

cn

460-

o

fl WHITTERYI

1

**

**

z

X

CD

o

Q

<

cn

%

LU

\

1

!

**

**

**

<

%

KIR

<

Q

1

* A

<

u

3

o

1

X

o

ROC

CD

1

* A

:>

470-

arna;

■GIAN \

1

1

1

;

I i ,

**

**-.2
** *

**

BLACK

RIVERAK

X 1

1

-

cos

1

i

▲

* A

is

" u "
'1ST

1

1

1 C.(C.I MULTIRAMUS -
1 |

▲ 1
▲

? A

/■-I

<C uj

1 A

-=V-

L

• / SR CF. /. PAULI N N

'-1

z

z

u

ft. SHELVENSIS

1 /. PAUL 1

<

480-

cn

>

; z :

—

- 1

f C. (H.) TURGIDULUS ®

o

O

cn

<

L

1/ BRITHD/R-

1

1

LU

ENSIS |

J

x

l___

^ *

§

I-'

text-fig. 6. A probable phylogeny of the known iocrinids, showing
the distribution of species in both time and space. Faunal data
derived from Bates (1965), Briskeby (1981), Brower and Veinus ( 1974),
Chauvel and Le Menn (1973), Chauvel et al. (1975), Donovan (1984,
1985a, b , 1986, in press), Kelly (1978), Opik (1934), Ramsbottom
(1961), and Stukalina (1980). Stratigraphic data principally from
Ross et al. (1982), but also from Williams et al. (1972). Key: 1 =
Iocrinus; C = Caleidocrinus\ H = Huxleyocrinus; 1 = ? Ristnacrinus
altobasalis Brower and Veinus; 2 = Ristnacrinus marinus Opik.

The precise relationships of two further species of iocrinid are unknown and they are not
included in text-fig. 6. Both belong to monospecific genera and both retain a pentameric column.
This is presumably a primitive feature, yet one of these species, Pariocrinus heterodactylus Eckert,
1984, from the Lower Silurian of Ontario, is the youngest known iocrinid. This species has an
elongate anal tube, rather than a sac as in Iocrinus , and arm branching above the primaxillary is
heterotomous, rather than isotomous.

322

PALAEONTOLOGY, VOLUME 32

The second species with a pentameric column is Peltacrinus sculptatus Warn, 1982, from the
Black Riveran, Bromide Formation of Oklahoma. This also supports a tube-like anal series. The
dorsal cup is ribbed, with a flared base. Arm branching appears to be isotomous throughout.

The true diversity of the genus Iocrinus is probably underestimated in text-fig. 6. Numerous
Iocrinus- like columnals have been recognized from the Middle and Upper Ordovician of Britain
(Donovan 1983) and have been particularly well reported in the Russian literature. However, until
more complete specimens are known, it would be premature to assign there dissociated plates to
new species of Iocrinus.

Acknowledgements. We thank Mr Charles Buist, who found the original specimens of /. pauli , and Mr Jeremy
J. Savill for supplying us with a cast of his specimen. Mr David N. Lewis (BMNH) kindly supplied casts of
I. whitteryi and arranged for photography of BMNH E71413. Dr H. Ivimey-Cook (BGS) sent SKD the
paratypes of I. whitteryi. We give special thanks to Drs R. A. Fortey, S. F. Morris (both BMNH), and
D. A. T. Harper (University College, Galway) for identifying our trilobite and brachiopod material. Dr Peter
R. Sheldon (Trinity College, Dublin) gave invaluable help in suggesting the probable stratigraphic position
of I. pauli.

REFERENCES

bassett, m. g. 1982. Ordovician and Silurian sections in the Llangadog-Llandilo area. In bassett, m. g. (ed. ),
Geological excursions in Dyfed , south-west Wales , 271-287. National Museum of Wales, Cardiff.
bates, d. e. b. 1965. A new Ordovician crinoid from Dolgellau, north Wales. Palaeontology , 8, 355-
357.

— 1968. On ‘ Dendrocrinus' cambriensis Hicks, the earliest known crinoid. Ibid. 11, 406-409.

billings, e. 1857. New species of fossils from Silurian rocks of Canada. Rep. geol. Surv. Can. 1853-1856,
245-345.

brett, c. E. 1981. Terminology and functional morphology of attachment structures in pelmatozoan
echinoderms. Lethaia , 14, 343-370.

briskeby, p. I. 1981. Klassifikasjon av krinoidstilker fra den over-ordoviciske Kalsjoformasjonen pa Hadeland.
Cand. Real thesis (unpublished). University of Oslo.

brower, j. c. and veinus, j. 1974. Middle Ordovician crinoids from southwestern Virginia and eastern
Tennessee. Bull. Am. Palaeont. 66, no. 283, 125 pp.

chauvel, j. and le menn, j. 1973 (for 1972). Echinodermes de l’Ordovicien Superieur de Coat-Carrec, Argol
(Finistere). Bull. Soc. geol. miner. Bretagne , ser. C, 4, 39-61.

— melendez, b. and le menn, j. 1975. Les echinodermes (cystoi'des et crinoi'des) de l’Ordovicien Superieur
de Luesma (Sud de l’Aragon, Espagne). Estudios geol. Inst. Invest, geol. Lucas Mallada 31, 351 364.

DONOVAN, s. K. 1983. Evolution and biostratigraphy of pelmatozoan columnals from the Cambrian and Ordovician
of Britain. Ph.D thesis (unpublished), University of Liverpool.

- 1984. Ramseyocrinus and Ristnacrinus from the Ordovician of Britain. Palaeontology , 27, 623-634.
1985a. The Ordovician crinoid genus Caleidocrinus Waagen and Jahn, 1899. Geol. J. 20, 109 121.
19856 (for 1984). Ristnacrinus and the earliest myelodactylid from the Ashgillian Boda Limestone of
Sweden. Geol. For. Stockholm Forh. 106, 347 356.

— 1986 (for 1984). Pelmatozoan columnals from the Ordovician of the British Isles, part 1. Monogr.
paleontogr. Soc. 1-68.

— In press. Pelmatozoan columnals from the Ordovician of the British Isles, part 2. Ibid.

eckert j. d. 1984. Early Llandovery crinoids and stelleroids from the Cataract Group (Lower Silurian) in
southern Ontario, Canada. Life Sci. Contr. R. Ontario Mus. 137, iv + 83 pp.
hall, J. 1866. Descriptions of new species of Crinoidea and other fossils from the Lower Silurian strata of the
age of the Hudson- River Group and Trenton Limestones , 17 pp. Privately published, Albany, NY.
hughes, c. p. 1969. The Ordovician trilobite faunas of the Builth-Llandrindod Inlier, central Wales, part 1.
Bull. By. Mus. not. Hist. (Geol.), 18, 39-103.

kelly, s. m. 1978. Functional morphology and evolution of Iocrinus, an Ordovician disparid inadunate crinoid.
MS thesis (unpublished), Indiana University.

lockley, m. g. and williams, A. 1981. Lower Ordovician Brachiopoda from mid and southwest Wales. Bull.
Br. Mus. nat. Hist. (Geol.), 34, 1-78.

Matthews, s. c. 1973. Notes on open nomenclature and on synonymy lists. Palaeontology, 16, 713-719.

DONOVAN AND GALE: IOCRINUS IN ENGLAND AND WALES

323

meek, f. b. and worthen, a. h. 1865. Description of new species of Crinoidea, &c., from the Palaeozoic rocks
of Illinois and some of the adjoining states. Proc. Acad. nat. Sci. Philad. 17, 143 155.
miller, I. S. 1821. A natural history of the Crinoidea or lily-shaped animals, with observations on the genera
Asteria, Euryale, Comatula and Marsupites, 150 pp. Bryan and Co., Bristol.

MOORE, R. c. 1962. Ray structures of some inadunate crinoids. Paieont. Contr. Univ. Kans., Echinodermata
Art. 5, 1 -47.

— jeffords, r. m. and miller, T. h. 1968. Morphological features of crinoid columns. Ibid. 8, I 30.

— lane, n. g., strimple, h. l. and sprinkle, j. 1978. Order Disparida Moore & Laudon 1943. In moore,
r. c. and teichert, c. (eds. ). Treatise on invertebrate paleontology. Part T, Echinodermata 2(2), T520 T564.
Geological Society of America and University of Kansas Press, Lawrence, Kansas.

— and laudon, l. r. 1943. Evolution and classification of Paleozoic crinoids. Spec. Pap. geol. Soc. Am.
46, 153 pp.

morton, G. e. 1869. The geology of the country around Shelve. Proc. Lpool geol. Ass. 1, 3-22.
opik, A. a. 1934. Ristncicrinus , a new Ordovician crinoid from Estonia. Tartu Ulik. Geol. -Inst. Toim. 40,
7 pp.

ramsbottom, w. h. c. 1961. The British Ordovician Crinoidea. Monogr. palaeontogr. Soc. 1-36.

ROSS, R. J., jr. et al. 1982. The Ordovician System in the United States. Pub. Int. Un. geol. Sci. 12, 73 pp.
stukalina, G. a. 1980. Novae dannae ob Ordovikskich krinoideyach tsentralnogo Kazachstan. Ezheg. vses.
paieont. Obshch. 23, 216-249. [In Russian.]

thomas, a. t., Owens, R. m. and rushton, a. w. a 1984. Trilobites in British stratigraphy. Spec. Rep. geol.
Soc. Lond. 16, 78 pp.

ubaghs, G. 1978. Skeletal morphology of fossil crinoids. In moore, r. c. and teichert, c. (eds.). Treatise on
invertebrate paleontology. Part T, Echinodermata 2(1), T58 T216. Geological Society of America and
University of Kansas Press, Lawrence, Kansas.

walcott, c. d. 1883. Descriptions of new species of fossils from the Trenton Group of New York. Rep. NY
St. Mas. nat. Hist. (Adv. publ.), 35, 7 pp.

warn, j. m. 1982. Long-armed disparid inadunates. In sprinkle, j. (ed.). Echinoderm faunas from the
Bromide Formation (Middle Ordovician) of Oklahoma. Paieont. Contr. Univ. Kans. Monogr. 1, 77-89.
webster, g. d. 1973. Bibliography and index of Paleozoic crinoids, 1942 -1968. Mem. geol. Soc. Am. 137,
xi + 341 pp.

1974. Crinoid pluricolumnal noditaxis patterns. J. Paieont. 48, 1283 1288.
whittard, w. f. 1979. An account of the Ordovician rocks of the Shelve Inlier in west Salop and part of
north Powys. Bull. Br. Mus. nat. Hist. (Geol.), 33, 1-69. (Compiled by w. t. dean.)

WILLIAMS, A., STRACHAN, I., BASSETT, D. A., DEAN, W. T., INGHAM, J. K., WRIGHT, A. D. and WHITTINGTON,

h. b. 1972. A correlation of Ordovician rocks in the British Isles. Spec. Rep. geol. Soc. Lond. 3, 74 pp.

S. K. DONOVAN

Department of Geology
University of The West Indies
Mona, Kingston 7
Jamaica

a. s. GALE

School of Earth Sciences
Thames Polytechnic
Walburgh House

Typescript received 28 March 1988 Bigland Street

Revised typescript received 1 1 July 1988 London e1 2ng

FIRST RECORDS OF CONODONTS FROM THE
LATE TRIASSIC OF BRITAIN

by ANDREW SWIFT

Abstract. Conodonl elements have been recovered for the first time from British Triassic deposits.
Most specimens are from the Langport Member (Penarth Group, late Triassic) at Normanton Hills,
Nottinghamshire, but elements also occur in the same member at Lavernock Point, South Glamorgan, and
in the succeeding Pr e-planorbis Beds at Barnstone and Blue Hill, Nottinghamshire. Specimens from the Pre-
planorbis Beds may be the youngest conodont elements known from north-west Europe. The occurrence of
Misikella posthernsteini (Kozur and Mock) in the Pr e-planorbis Beds at Barnstone allows correlation with
part of the upper Rhabdoceras suessi and Choristoceras marshi zones of the late Triassic of southern Europe,
and with beds of Rhaetian age in Japan and Papua New Guinea. Other elements have little stratigraphic
value, and may represent the apparatuses of one or two species only.

Recent studies have shown conodonts to be derived from extinct marine jawless craniates whose
grasping/food-processing apparatus consisted of a suite of phosphatic elements (Briggs et al. 1983;
Aldridge et al. 1986). These elements form an important part of the microfossil record and are
extensively used for biozonation in Ordovician to Triassic sediments. In Britain the previous
youngest record of conodont elements was from late Permian, Zechstein 1, sediments (Swift 1986;
Swift and Aldridge 1982, 1985). The stratigraphically youngest conodont elements known are from
Tethyan deposits of late Rhaetian age in southern Europe (Mostler et al. 1978). The later part of
the Rhaetian Stage ( sensu Wiedmann et al. 1979) is represented in complete British sequences by
the Penarth Group and the lowest beds of the Lias, the Pve-planorbis Beds (Warrington et al.
1980). Sampling of the Penarth Group and Pr e-planorbis Beds has produced the first collections
of conodont elements from British Triassic rocks. Amongst the elements recovered is one from
the apparatus of the zonally important species Misikella posthernsteini (Kozur and Mock). Since
these beds are very close to the Triassic-Jurassic boundary, the specimens recovered may be the
youngest known from north-west Europe.

LOCALITIES AND STRATIGRAPHIC SETTING

Collections of conodont elements have been recovered from the Langport Member of the Penarth
Group at two localities: Normanton Hills railway cutting in south Nottinghamshire and Lavernock
Point in South Glamorgan. The succeeding Pr e-planorbis Beds have yielded conodont elements
from two localities in south Nottinghamshire: Blue Hill canal cutting and Barnstone railway
cutting. The distribution of these localities is shown on text-fig. 1, and the stratigraphical context
in text-fig. 2. The base of the Jurassic is placed at the lowest occurrence of the ammonite Psiloceras
in facies of the Lower Lias (Warrington et al. 1980).

Langport Member

Normanton Hills. The most numerous and diverse collections have been recovered from the
Normanton Hills railway cutting (SK 537 246), 5 km north of Loughborough (Browne 1895;
Lox-Strangways 1905; Trueman 1918; Kent 1937). The Langport Member is here represented
by a single bed of hard, homogeneous, pale grey, flinty micrite which occurs between soft cal-
careous mudstones of the Cotham Member and ‘paper’ shales and argillaceous limestones of

(Palaeontology, Vol. 32, Part 2, 1989, pp. 325-333, pi. 37.|

© The Palaeontological Association

326

PALAEONTOLOGY, VOLUME 32

B

text-fig. I. Sample localities in a— south Nottinghamshire and b— -South Glamorgan (based on geological
map of the UK, South. 3rd edn., 1979, British Geological Survey).

SWIFT: LATE TRIASSIC CONODONTS FROM BRITAIN

327

text-fig. 2. Stratigraphic context of horizons yielding conodonts in the British late Triassic. Stratigraphy
based on Warrington et al. (1980). Additional section detail: Normanton Hills (Kent 1937), Lavernock Point
(Iviniey-Cook 1974; Waters and Lawrence 1987), Blue Hill (Ivimey-Cook and Elliott 1969), Barnstone (Sykes

et al. 1970).

the Pr e-planorbis Beds. The bed is very similar lithologically to limestones of the Langport Member
exposed at the classic Rhaetian localities in north Somerset and South Glamorgan. The base of
the Jurassic at Normanton Hills, indicated by the incoming of Psiloceras , is 2-92 m above the top
of the micrite bed (Kent 1937, p. 168).

The occurrence of the Langport Member at Normanton Hills confirms previous intimations of
its presence north of Leicester (Johnson 1950; Kent 1953, 1968, 1970) and it seems likely that the
‘splintery limestone’ or ‘sun-bed’ which sporadically occurs at the top of the ‘Rhaetic’ elsewhere
in Nottinghamshire and Leicestershire (Lamplugh et al. 1908, 1909) also represents the Langport
Member. The patchy distribution may be related to lateral facies variations or breaks in
sedimentation dictated by a basin and swell topography coupled with sea-level oscillations. The
results of such phenomena have been noted in the Langport Member on the Devon coast (Hallam
1960). The top of the micrite bed at Normanton Hills is hummocky, with pockets of winnowed
shell debris and evidence of burrowing organisms, indicative of a shallow-water environment.

Lavernock Point. In South Glamorgan, a few conodont elements have been recovered from the
Langport Member at Lavernock Point (ST 187 682) (Richardson 1905; Ivimey-Cook 1974; Waters
and Lawrence 1987). Here the Member consists of a basal development of hard limestones, ranging
from homogeneous micrites to shelly calcarenites, overlain by calcareous mudstones. The limestones
show evidence of subaerial weathering, with hardground features such as uneven iron-rich upper
surfaces, pyrite infilled burrows, and weathered crusts. The topmost limestone yielded conodont

328 PALAEONTOLOGY, VOLUME 32

elements. The first Psiloceras appear 4-7 m above the top of the Langport Member (Waters and
Lawrence 1987).

Pre-planorbis Beds

Blue Hill , Owthorpe. In the almost overgrown section on the west bank of the old canal at Blue
Hill (SK 682 343) (Lamplugh et al. 1909; Kent 1937), soft calcareous mudstones with impersistent
nodular horizons (base not seen) of the Cotham Member, are overlain by brown shales with very
thin fine calcarenites succeeded by a somewhat fissile bed of fine calcarenite containing Liostreal
sp. and rare conodont elements. Above this the sequence is obscured by topsoil and rubble. No
Psiloceras have been found at this locality, but logs of sections nearby record its appearance
approximately 3 0 m above the conodont-bearing bed (Trueman 1915; Kent 1937). The Owthorpe
Fox Holes boreholes (see text-fig. 1), sunk a short distance away by the National Coal Board,
offer the most recent comparative stratigraphy (Ivimey-Cook and Elliott 1969), and showed
mudstones of the Cotham Member separated by a non-sequence from a succeeding laminated
‘paper’ shale (28-33 cm), which is overlain by greyish-white or pale grey, partially laminated
limestone. The non-sequence may explain the absence of lithologies typical of the Langport
Member at Blue Hill, with the conodont-bearing bed, which is of Liassic aspect, representing part
of a Pre-planorbis Beds sequence.

Barnstone. The sections around Barnstone, now largely obscured, have been the subject of several
papers, with the bone bed at the base of the Westbury Formation attracting particular attention
(Sykes 1971, 1979). The most recent study in the railway cutting (SK 739 358) involved trenched
sections (Sykes, Cargill and Fryer 1970), and a sample collected during a re-examination of one
of these produced a single conodont element. This comes from the apparatus of M. posthernsteini
(Kozur and Mock). The sample was taken from a horizon comparable with that of the conodont-
bearing bed at Blue Hill, i.e. the first limestone development succeeding mudstones with nodules
of the Cotham Member, and before Jurassic beds with Psiloceras. This limestone is at the base of
a sequence of shales and limestones of Liassic aspect which are assigned to the Pre-planorbis Beds
(Sykes et al. 1970). The presence of a non-sequence at the top of the Cotham Member here has
been suggested by Sykes, Cargill, and Fryer (1970), but later questioned on palynological grounds
(Fisher 1972). No physical evidence for this can be observed at the exposure due to its overgrown
nature. Nevertheless, since no biostratigraphical data available for the upper Rhaetian of Britain
offer the degree of refinement necessary to detect such a fine break in the sequence, it is possible
that certain strata of the Langport Member equivalent to those at Normanton Hills, are missing
at Barnstone.

EXPLANATION OF PLATE 37

Fig. 1. Misikella posthernsteini (Kozur and Mock). Inner lateral view of dextral Pa element 6246/9, x 300;
Pre-planorbis Beds, Barnstone.

Figs. 2-14. Unassigned elements. 2-4, M element. 2, inner lateral view of sinistral element 6235/4, x 200;
Langport Member, Normanton Hills. 3, inner lateral view of dextral element 6245/4, x 300; Pre-planorbis
Beds, Blue Hill. 4, inner lateral view of dextral element 6245/2, x 300; Langport Member, Lavernock
Point. 5-8, Sc element. 5, inner lateral view of dextral element 6246/7, x 300; Langport Member, Normanton
Hills. 6, inner lateral view of sinistral element 6247/14, x200; Langport Member, Normanton Hills. 7,
inner lateral view of dextral element 6245/1, x 300; Langport Member, Lavernock Point. 8, inner lateral
view of dextral element 6245/5, x 300; Pre-planorbis Beds, Blue Hill. 9, Pa element. Inner lateral view of
dextral element 6246/6, x 300; Langport Member, Normanton Hills. 10 and 11, Pb element. Inner lateral
views of dextral elements 6235/1, 6236/7, x 300; Langport Member, Normanton Hills. 12, Sa element.
Lateral view of 6246/11, x 300; Langport Member, Normanton Hills. 13, incomplete element A. Inner
lateral view of 6247/12, x 200; Langport Member, Normanton Hills. 14, incomplete element B. Inner
lateral view of 6234/15, x 200; Langport Member, Normanton Hills.

All scanning electron micrographs. Repository of specimens on this plate: Conodont Reference Collection,
Department of Geology, University of Nottingham.

PLATE 37

SWIFT, late Triassic conodonts

330

PALAEONTOLOGY, VOLUME 32

No specimens of Psiloceras have been found in the cutting, where 1-6 m of Pr e-planorbis Beds
are recorded (Sykes et al. 1970).

THE CONODONT FAUNAS

All conodont elements isolated from the British late Triassic are extremely small, transparent, and
delicate. Very little colouration is apparent, indicating a minimal CAI (Colour Alteration Index)
value. The distribution of element types is shown in Table 1. Conodont element designation as
Pa, Pb, M, Sa, or Sc follows Sweet (1981) and is a nomenclature relating to the location of each
element in the feeding/grasping apparatus of conodont animals. No Sb elements have been
recognized in the study material. An integrated functional system for conodont elements based on
complete fossils from the Carboniferous of Scotland (Aldridge et al. 1987) envisaged an anterior
set of grasping S and M elements, followed posteriorly by a shearing pair of Pb elements and a
grinding pair of Pa elements.

table 1. Frequency and distribution of conodont element types in British late Triassic samples

Frequency of element types

Incomplete elements

Locality

Pa

Pb

M

Sa Sc

Type A

Type B

Normanton Hills

1

2

12

1 137

1

1

Lavernock Point

1

5

Blue Hill
Barnstone

1

1

4

Misikella posthernsteini (Kozur and Mock, 1974 ) from Barnstone

The recovery from a 2-25 kg sample of a Pa element from the apparatus of M. posthernsteini
(PI. 37, fig. 1) represents the most important discovery in the new material and constitutes the
only record of this species from Britain. Used in the biozonation of the Tethyan Trias of southern
Europe, its occurrence in the British late Triassic, where ammonoids are unknown and other
biostratigraphic material is limited, is an important new factor in the correlation of these two
sequences. The assemblage zone characterized by this fossil forms the last conodont biozone in
the geological record (Kozur 1980), corresponding to the higher part of the Rhahdoceras suessi
and greater part of the Choristoceras marshi ammonoid zones of the Tethyan Trias. M. posthernsteini
is widely distributed in, and confined to, Rhaetian sediments, and has been recovered from Tethyan
sequences in Austria (Mosher 1968; Mostler et al. 1978; Krystyn 1980), Czechoslovakia (Kozur
and Mock 1974), and Poland (Kovacs and Kozur 1980), as well as Hungary and Yugoslavia
(Kozur, pers. comm.). Elsewhere it has been found in Rhaetian sequences in Japan (Nagao and
Matsuda 1982; Isozaki and Matsuda 1983) and Papua New Guinea (Skwarko et al. 1976).

Collections from the other localities

Normanton Hills. 158 identifiable conodont elements have been recovered from 12 kg of the
Langport Member calcilutite, making this by far the most productive horizon investigated. 86-7%
of specimens are Sc elements (PI. 37, figs. 5 and 6). Variations in the curvature of the aboral margin
and disposition of costae on denticles may indicate some positional differentiation in these elements,
but overall they comprise a broadly consistent morphological group. Some similarity is apparent
to illustrated Sc elements from southern European Tethyan sequences, e.g. Hindeodella uniforma
Mosher from the Hallstatter Kalk at Steinbergkogel, Austria (Mosher 1968, pi. 1 14, fig. 14), and
Neohindeodella dropla (Spasov and Ganev) from the early Carnian of Hungary (Kozur and Mostler

SWIFT: LATE TRIASSIC CONODONTS FROM BRITAIN

331

1972, pi. 15, fig. 14). The latter name has also been used for coeval forms from Japan (Koike 1981,
pi. 1, fig. 16) and Malaysia (Koike 1982, pi. 8, fig. 32).

The next most numerous form (7-6%) is an M element showing distinctive recurved and twisted
denticles around the cusp (PI. 37, fig. 2). No completely comparable specimens have been recorded
elsewhere, but there are similarities to some illustrated elements attributed to Prioniodina
(Cypridodella) muelleri (Tatge 1956), which has been used to accommodate a range of M
morphotypes from various Triassic sequences (e.g. see Kozur and Mostler 1972, pi. 11; figs. 1-15,
17-21).

The remaining five element types are in each case represented by one or two specimens only,
and whilst some are incomplete, there are morphological similarities which suggest that they may
constitute a single multi-element species. These similarities are particularly apparent in the
denticulation; all elements have denticles which are crimped, sharp-edged, laterally compressed,
and widely spaced. The Pa element (PI. 37, fig. 9) may be related to Misikella , but cannot be
compared to any existing species. The Pb element (PI. 37, figs. 10 and 1 1) possesses a conservative
morphology of a generalized pattern which recurs often in the conodont record, yet has no parallel
amongst described Triassic forms. Similar conclusions apply also to the Sa (PI. 37, fig. 12) and the
two incomplete elements (PI. 37, figs. 13 and 14).

Lavernock Point. A 3 kg sample of the topmost micrite bed of the Langport Member yielded six
identifiable conodont elements from a total of nine isolated. Sc (PI. 37, fig. 7) and M (PI. 37,
fig. 4) elements, exactly comparable to those recovered from Normanton Hills, constitute the
recognizable forms.

Blue Hill. A 3 kg sample yielded six conodont elements, five being identifiable. As at Lavernock
Point, only Sc (PI. 37, fig. 8) and M (PI. 37, fig. 3) elements are identified. These are identical to
those found at Normanton Hills.

CONCLUSIONS

The conodont elements described here are the first from the British Triassic and extend the range
of conodonts in Britain into the youngest beds of that system. A correlation of the Pr e-planorbis
Beds of Nottinghamshire with part of the upper R. suessi and C. marshi zones of the late Triassic
of the Tethyan Realm of southern Europe, and also with Rhaetian sequences in Japan and Papua
New Guinea, is demonstrated by the shared occurrence of M. posthernsteini. The other elements
recovered have, as yet, little correlative value, although some are comparable with late Triassic
forms elsewhere. Low numbers of element types and shared characteristics indicate the presence
of three or less conodont species.

Acknowledgements. The encouragement and advice offered by Dr R. J. Aldridge are appreciated, and I also
record by thanks to him for critically reading the manuscript. I am indebted to Dr G. Warrington for
suggestions which considerably improved the final draft. Mrs J. M. Wilkinson prepared the text-figs, and
table.

REFERENCES

aldridge, r. j., briggs, d. e. g., Clarkson, e. n. k. and smith, m. p. 1986. The affinities of conodonts— new
evidence from the Carboniferous of Edinburgh, Scotland. Lethaia , 19, 279 291.

— smith, m. p., norby, r. d. and briggs, d. e. G. 1987. The architecture and function of Carboniferous
polygnathacean conodont apparatuses. In aldridge, r. j. (ed.). Palaeobiology of conodonts , 63-75. Ellis
Horwood, Chichester.

briggs, d. e. G., Clarkson, e. n. k. and aldridge, r. j. 1983. The conodont animal. Lethaia , 16, 114.
browne, m. 1895. Preliminary notice of an exposure of Rhaetic Beds near East Leake, Nottinghamshire.
Rep. Br. Ass. advmt Sci. (for 1895), 688 690.

fisher, m. j. 1972. Rhaeto-Liassic palynomorphs from the Barnstone railway cutting, Nottinghamshire.
Mercian Geol. 4, 101 106.

332

PALAEONTOLOGY, VOLUME 32

fox-strangways, c. 1905. The geology of the country between Derby, Burton-on-Trent, Ashby-de-la-Zouch
and Loughborough. Mem. geol. Surv. UK, 83 pp.
hallam, a. 1960. The White Lias of the Devon Coast. Proc. Geol. Ass. 71, 47-60.

isozaki, y. and matsuda, t. 1983. Middle and late Triassic conodonts from bedded chert sequences in the
Mino-Tamba belt, southwest Japan. Part 2: Misikella and Parvigondolella. J. Inst. Polytech. Osaka Cy
Univ. 26, 65-86.

ivimey-cook, h. c. 1974. The Permian and Triassic deposits of Wales. In owen, t. r. (ed. ). The Upper
Palaeozoic and Post-Palaeozoic rocks of Wales , 295-321. University of Wales Press, Cardiff.

-and elliott, r. e. 1969. Boreholes in the Lias and Keuper of south Nottinghamshire. Bull. geol. Surv.
Gt Br. 29, 139-151.

Johnson, m. r. w. 1950. The fauna of the Rhaetic Beds in south Nottinghamshire. Geol. Mag. 87, 116-120.
rent, p. e. 1937. The Lower Lias of south Nottinghamshire. Proc. Geol. Ass. 48, 163-174.

— 1953. The Rhaetic Beds of the north-east Midlands. Proc. Yorks, geol. Soc. 29, 117-139.

— 1968. The Rhaetic Beds. In sylvester-bradley, p. c. and ford, t. d. (eds.). The Geology of the East
Midlands , 174 187. Leicester University Press, Leicester.

1970. Problems of the Rhaetic in the East Midlands. Mercian Geol. 3, 361-373.
koike, t. 1981. Biostratigraphy of Triassic conodonts in Japan. Sci. Rep. Yokohama Univ. Sec. II, 28, 25-
42.

— 1982. Triassic condont biostratigraphy in Kedah, west Malaysia. Geol. Palaeont. S.E. Asia, 13, 9-51.
kovacs, s. and kozur, h. 1980. Stratigraphische Reichweite der wichtigsten Conodonten (ohne Zahnreihen-

conodonten) der Mittel- und Obertrias. Geol. Palaont. Mitt. Innsbruck, 10 (2), 47-78.
kozur, H. 1980. Revision der Conodontenzonierung der Mittel- und Obertrias des tethyalen Faunenreichs.
Ibid. 10 (3/4), 79-172.

— and mock, r. 1974. Misikella posthernsteini n. sp., die jiingste Conodontenart der tethyalen Trias. Cas.
Miner. Geol. 19, 245-250.

— and mostler, h. 1972. Die Conodonten der Trias und ihr stratigraphischer Wert. I. Die ‘Zahnreihen-
Conodonten' der Mittel- und Obertrias. Abh. geol. Bundesanst., Wien, 28, 1-53.

krystyn, l. 1980. Triassic conodont localities of the Salzkammergut Region (Northern Calcareous Alps).
In schonlaub, H. p. (ed.). Guidebook Abstracts, Second European Conodont Symposium (Ecos II). Ibid.
35,61-98.

lamplugh, G. w., gibson, w., Sherlock, r. l. and wright, w. b. 1908. The Geology of the Country between
Newark and Nottingham. Mem. geol. Surv. UK, 126 pp.

— wedd, c. b., Sherlock, r. l. and smith, b. 1909. The Geology of the Melton Mowbray District
and South-East Nottinghamshire. Ibid. 118 pp.

mosher, l. c. 1968. Triassic conodonts from western North America and Europe and their correlation. J.
Paleont. 42, 895-946.

mostler, H., scheuring, b. w. and urlichs, m. 1978. Zur Mega-, Mikrofauna and Mikroflora der Kossener
Schichten (alpine Obertrias) vom Weissloferbach in Tirol unter besonderer Beriicksichtigung der in der
suessi- und marshi- Zone auftretenden Conodonten. Schreihe. Erdwiss. Komm. Osterr. Akad. Wiss. 4, 141
174.

nagao, h. and matsuda, t. 1982. ‘Rhaetian problem’ in terms of conodont biostratigraphy— A case study
in bedded chert sequence at Toganoo, in northwest of Kyoto, Southwest Japan. In Proceedings of the
First Japanese Radiolarian Symposium. Spec. Vol. News Osaka Micro. 5, 469-478.
richardson, l. 1905. The Rhaetic and contiguous deposits of Glamorganshire. Quart. Jl geol. Soc. Lond.
61, 385 424.

skwarko, k. s., nicoll, r. s. and Campbell, k. s. w. 1976. The late Triassic molluscs, conodonts and
brachiopods of the Kuta Formation, Papua New Guinea. B.M.R. J. Aust. Geol. Geophys. 1, 219-230.
sweet, w. c. 1981. Morphology and composition of elements, macromorphology of elements and apparatuses.
In robison, r. a. (ed.). Treatise on invertebrate paleontology, part W, supplement 2, Conodonta , W5--W20.
Geological Society of America and University of Kansas Press, Lawrence, Kansas.
swift, a. 1986. The conodont Merrillina divergens (Bender and Stoppel) from the Upper Permian of England.
In harwood, g. m. and smith, d. b. (eds.). The English Zechstein and Related Topics. Spec. Publ. geol.
Soc. Lond. 22, 55-62.

— and aldridge, r. j. 1982. Conodonts from Upper Permian strata of Nottinghamshire and North
Yorkshire. Palaeontology , 25, 845-856.

1985. Conodonts of the Permian System from Great Britain. In higgins, a. c. and Austin, r. l.
(eds.). A stratigraphical index of Conodonts, 228-236. Ellis Horwood Ltd., Chichester.

SWIFT: LATE TRIASSIC CONODONTS FROM BRITAIN

333

sykes, J. h. 1971. A new Dalatiid fish from the Rhaetic Bone Bed at Barnstone, Nottinghamshire. Mercian
Geol. 4, 13-22.

- 1979. Lepidotes sp. : Rhaetian fish teeth from Barnstone, Nottinghamshire. Ibid. 7, 85 91.

— cargill, j. s. and fryer, h. g. 1970. The stratigraphy and palaeontology of the Rhaetic Beds (Rhaetian:
Upper Triassic) of Barnstone, Nottinghamshire. Ibid. 3, 233-264.
tatge, u. 1956. Conodonten aus dem germanischen Muschelkalk. Palaont. Z. 30, 106 147.
trueman, a. e. 1915. The Fauna of the Hydraulic Limestones in South Notts. Geol. Mag. 52, 150 152.

1918. The Lias of South Lincolnshire. Ibid. 55, 64-73, 101 111.

WARRINGTON, G., AUDLEY-CHARLES, M. G., ELLIOTT, R. E., EVANS, W. B., IVIMEY-COOK, H. C., KENT, P. E.,

robinson, p. l., shotton, f. w. and taylor, F. m. t. 1980. A correlation of Triassic rocks in the British
Isles. Spec. Rep. geol. Soc. Loud. 13, 78 pp.

waters, R. a. and Lawrence, d. j. d. 1987. Geology of the South Wales Coalfield, Part III, the country
around Cardiff. Mem. geol. Surv. UK, 1 14 pp.

wiedmann, j., fabricius, f., krystyn, l., reitner, j. and urlichs, m. 1979. Uber Umfang und Stellung des
Rhaet. News/. Stratigr. 8, 133-152.

Typescript received 17 February 1988
Revised typescript received II July 1988

A. SWIFT

Department of Geology
University of Nottingham
University Park
Nottingham NG7 2RD, UK

HETEROCHRONY IN A FOSSIL REPTILE:
JUVENILES OF THE RHYNCHOSAUR SCAPHONYX
FISCHERI FROM THE LATE TRIASSIC OF BRAZIL

by MICHAEL J. BENTON and RUTH KIRKPATRICK

Abstract. A juvenile (?one year post-hatching) specimen of Scaphonyx fischeri, an advanced rhynchosaur
from the Late Triassic Santa Maria Formation (Carnian) of Brazil, is described. This is the youngest
rhynchosaur ever reported, and it was probably 0-38 m long overall, compared to adult body lengths of 1-3-
1-6 m. It is compared with an ontogenetic series of skulls of this species, ranging in length from 55 to 250
mm. An allometric analysis suggests that the posterior portion of the skull roof widens relatively more rapidly
than overall size increase (positive allometry), while the length of the posterior region of the skull shows
negative allometry. These relative shape changes during ontogeny are compared with ancestral rhynchosaurs
whose ontogenies appeared to stop earlier than that of Scaphonyx. This peramorphic shift in ontogeny is
tentatively identified as an example of hypermorphosis.

The recent debates about the relationships between development and evolution have led to a
number of studies of heterochrony, differences in the timing of developmental processes. The fossil
record can contribute a great deal of data to this debate when temporal sequences of ontogenies
of closely related taxa are discovered. So far, most such studies have concerned microfossils and
invertebrates because of the abundance of well-dated specimens.

In the present study, a juvenile (?near-hatchling size) fossil reptile, Scaphonyx fischeri, is described.
Its anatomy is compared with older juveniles and adults of the same species and allometric
relationships are established. The ontogeny of S. fischeri is then compared with earlier, presumably
ancestral, taxa, and a heterochronic explanation is proposed for the major evolutionary changes
in adult morphology.

THE RHYNCHOSAURS

Rhynchosaurs were an abundant and widespread group in the middle and early late Triassic (240-
225 million years ago). Typical forms were 1-2 m in body length, with a heavy body and a massive
skull. The front of the head was characterized by a ‘beak’ formed from the horn-covered
premaxillae. There were large eyes, and the back of the skull was very wide in order to accommodate
the jaw adductor muscles. The dentition consisted of a heavy maxillary tooth-plate (upper jaw)
with numerous rows of teeth, and a V-shaped crest on the lower jaw which fitted perfectly into
the V-groove.

The diet of rhynchosaurs has been variously interpreted as either molluscs or plants. The plant
interpretation seems more likely (Benton 1983a, h , 1984) because the teeth are pointed and not at
all ‘crushers’, the bite is precision-shear and not grinding, the body is massive (to accommodate a
large gut?), and the rhynchosaurs are nearly always abundant in their faunas (generally about 50%
of all animals collected).

Rhynchosaurs are known from all parts of the world, and their anatomy is remarkably uniform.
Indeed, the late Triassic forms ( Scaphonyx from Brazil and Argentina, Hyperodapedon from
Britain and India, and other less well-known forms) are virtually indistinguishable except on minor
points.

These late Triassic forms are distinguished from a middle Triassic group by their dental

IPalaeontology, Vol. 32, Part 2, 1989, pp. 335-353.| © The Palaeontological Association

336

PALAEONTOLOGY, VOLUME 32

anatomy and other characters. The first rhynchosaurs were rather different small forms from the
early Triassic of South Africa (Benton 19836).

MATERIALS AND METHODS

Occurrence and specimens

Scaphonyx fischeri was named in 1907 by Woodward who had received some fragments of bone from the
Triassic beds of the Santa Maria Formation of southern Brazil. He interpreted the fossils as those of
phytosaurs or dicynodonts, although rhynchosaurs had been described from other parts of the world in the
nineteenth century. In 1926, Huene assigned further fragmentary material which he had received to Scaphonyx
and also erected the genera Cephalonia, Cephalastron, Cephalastronius , and Scaphonychimus, together forming
six new species, all of which he believed belonged to a specialized archosaur group, which he called the
Pelycosimia. By the time this study was published, Huene had received further specimens which showed that
his ‘pelycosimians’ were all rhynchosaurs. Huene (1929) retained all of his new taxa of Brazilian rhynchosaurs,
but collections of better material which he made on an expedition to Brazil in 1928 1929 suggested that there
were in fact only two valid species, Scaphonyx fischeri and Cephalonia lotziana (Huene 1942). Cephalonia was
apparently more lightly built than Scaphonyx. Later, Sill (1970) showed that the differences in the relative
proportions of the bones in the two forms was diagenetic rather than genetic. Chemical changes during
preservation of the bones had caused some (‘ Scaphonyx') to expand, which is shown by the ‘exploded’
appearance of the bone tissue in cross-section. Cephalonia is then identical to Scaphonyx , and only one genus
of rhynchosaur is known so far from Brazil. A second Brazilian species, S. sulcognathus , which differs from
S. fischeri in proportions and dental features, has been described (Azevedo and Schultz 1988). The present
specimen is almost certainly S. fischeri since the new species is largely restricted to the Caturrita Formation
which overlies the Santa Maria.

S. fischeri is the most abundant element in the Santa Maria fauna, representing nearly 70% of all individual
specimens found. Well over 150 partial or complete skeletons are known (Benton 1983a). Other Santa Maria
animals include the large herbivorous dicynodonts Dinodontosaurus and Stahleckeria, the small cynodonts
Traversodon , Chiniquodon, and Belesodon, and the large meat-eating thecodontians Prestosuchus and
Rauisuchus , as well as ten other less common genera.

The S. fischeri remains have been found in upper portions of the Santa Maria Formation, in Part ‘C’ of
the Upper Santa Maria Member (Bortoluzzi and Barberena 1967). The localities are mainly close together,
about 3 km south-east of the town of Santa Maria in Rio Grande do Sul province in southern Brazil.
Bortoluzzi and Barberena (1967, pp. 175-177) record the abundance of articulated rhynchosaur skeletons at
several horizons over a 15 m thickness of sediments in Sanga Grande (their locality 4). More recently, S.
fischeri has been found elsewhere in Brazil (C. Schultz, pers. comm., 1988). The upper part of the Santa
Maria Formation is dated as Carnian, possibly middle Carnian (Bonaparte 1982; Olsen and Sues 1986).

In the present study, specimens of S. fischeri were examined in the following institutions: American Museum
of Natural History (AMNH); Bayerische Staatssammlung fur Palaontologie und historische Geologie,
Miinchen (BSP); Institut und Museum fur Geologie und Palaontologie der Universitat, Tubingen (GPIT);
Museum of Comparative Zoology, Harvard (MCZ).

Juvenile and adult material of S. fischeri has already been described (Huene 1929, 1942; Sill 1970, 1971;
Barberena 1971). The ontogenetic series of this species can now be extended back to a juvenile stage that
was probably just post-hatching, by a hitherto unrecorded specimen, MCZ 1664. This specimen is briefly
described and illustrated here before the allometric and heterochronic study.

MCZ 1664 (text-fig. 1) is a skull and partial skeleton. The skull is nearly complete, except for the premaxillae
and most of the snout. As a result of crushing, the braincase has been extruded from the right posterior
margin of the skull and the palate and occiput cannot be viewed. The skull has been preserved with the first
three vertebrae of the vertebral column. Other postcranial remains include cervical vertebrae 4 8, three
dorsals, one sacral, two caudals, most of the forclimbs and pectoral girdles, as well as fragments of the
pelvic region and hindlimb. The three anterior cervical vertebrae have been twisted 90° clockwise and,
when viewed from above, they are in left lateral view. This fact, and the pattern of preservation of the
other bones of the skeleton suggests that the animal was fossilized lying on its right side, with its limbs
extended (text-fig. 2).

The quality of preservation is poor, with the rock matrix encrusted on the bones in a nodular manner.
This makes it impossible to comment on the surface features and textures of the bones. Such poor preservation
may be due to the fossilization process. Sill (1970) demonstrated that the voluminous nature of the bones of

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

337

text-fig. 1. Scaphonyx fischeri (MCZ 1664), juvenile specimen in dorsal view. The skull is at the top left,
with the cervical vertebral column behind. The left shoulder girdle and forelimb (and hand) runs down in
the centre and left of the picture. The pelvic remains are shown at the top right, with parts of the left hindlimb

below. Half natural size.

S. fischeri was due to the chemico-mineralogical reactions during fossilization. Petrographic analysis and
radiographic diffraction showed that such reactions resulted in the entire replacement of the bone with
calcium carbonate, the total destruction of the cellular structure and the expansion of the microstructure
which subsequently refilled with microcrystalline mud.

Preparation of MCZ 1664

There were two stages in the preparation of the material— chemical and mechanical. The initial chemical
preparation involved soaking of the specimen in 6% acetic acid, and subsequent neutralization in water. This
method removed the fine film of surface rock particles, but in no way dislodged the large nodules of rock
matrix. A small hand drill was used for the mechanical preparation. This was successful in removing
prominent nodules of rock matrix, but much of the encrustation could not be removed because of its chemical
and physical similarity to the preserved bone. The exposed bone surfaces were painted with butvar in propan-
2-o 1 (isopropyl alcohol, c. 1 : 10) for protection.

A Home trie study

This study was limited to measurements of the skull. Thirteen skulls of S. fischeri were measured, ranging in
length from 55 to 270 mm. Nine measurements were made on dorsal views of the skull (Table 1; text-fig. 3).
Where possible, symmetrical variables were measured on both sides of the skull and averaged. It should be

338

PALAEONTOLOGY, VOLUME 32

I i

10cm

text-fig. 2. Restoration of the skeleton of Scaphonyx fischeri (MCZ 1664) lying in the position in which it

was fossilized.

table 1. Measurement of skulls of Scaphonyx fischeri (in mm) used in
the allometric study. BARB A D are the four skulls figured by
Barberena (1971). Characters 1-9 are: (1) total mid-line length of skull;
(2) posterior width of skull between lateral margins of quadrates (in
dorsal view); (3) maximum length of orbit; (4) greatest width of frontal;
(5) mid-line length of frontal; (6) greatest width of parietal (excluding
the posterial lateral processes); (7) mid-line length of parietal; (8)
maximum length of upper temporal fenestra; (9) maximum width of
upper temporal fenestra.

Specimen

1

2

3

4

5

6

7

8

9

MCZ 1664

55

60

14

15

15

16

19

18

14

BARB A

90

120

34

30

34

18

32

24

23

BSP 18/4

150

190

45

26

53

22

54

65

60

BSP 19/4

155

190

45

30

48

30

53

50

57

BARB B

164

170

40

40

44

40

58

46

50

GPIT

170

200

55

34

38

26

52

50

55

MCZ 3640

180

180

67

43

72

30

43

49

70

BARB C

200

230

66

60

60

40

58

52

68

BARB D

205

270

60

90

70

60

60

42

72

MCZ 1529

205

210

64

42

71

29

60

78

86

AMNH 7805

205

216

53

60

80

42

55

50

89

AMNH 7799

210

290

61

55

73

40

63

52

68

MCZ 1636

250

300

80

57

73

38

72

100

106

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

339

text-fig. 3. The measurements made on skulls of
Scaphonyx fischeri for the allometric study. Measure-
ments 1-9 are described, and values are listed in
Table 1.

2

1

noted that some of the specimens measured have been distorted during fossilization (but none as badly as
MCZ 1664), so that the measurements of skull length and width, and of skull openings (characters 1 3, 8,
9) may be subject to an error of ±5%.

The measurements were fitted in turn to the logarithmic expansion of the equation of simple allometry,
log Y + log b + a log X, where a is the allometric coefficient, b is a constant, X is a standard comparative
measure (here, total skull length), and Y is the measurement of the part or organ under consideration. The
value of the allometric coefficient, a , gives a measure of the rate of relative growth of the part, and in the
logarithmic form, a is simply the slope of the straight line; a can be greater than 1 (positive allometry), equal
to 1 (isometry), or less than 1 (negative allometry). When compared with phylogenetic changes, the values
of a in an ontogenetic series can give information on the kind of heterochrony, if any, that is involved. The
relative shape changes with growth were also examined by preparing transformed coordinate diagrams on
five skulls of very different sizes.

DESCRIPTION OF MCZ 1644

Class REPTILIA
Subclass DIAPSIDA
Order rhynchosauria
Family rhynchosauridae Huxley, 1877
Scaphonyx fischeri Woodward, 1907

Skull

General appearance. The skull approximates to the shape of a triangle in dorsal view (text-figs. 4a, and 5a).
The height of the skull is difficult to appreciate since it has been subjected to dorso-posterior crushing during
fossilization, but a careful reconstruction (text-fig. 5b) indicates that it was relatively deep. An accurate
determination of the length of the skull cannot be made as the premaxillae, nasals, prefrontals, and lacrimals
are all absent, but the reconstruction (text-fig. 5a) suggests that the skull length is of the order of 55 mm and
approximately equal to the maximum width. The palatal and occipital portions of the skull cannot be viewed
clearly because of the distortion in preservation. Crushing has also resulted in the protrusion of the braincase
at the right posterior margin of the skull (text-fig. 4a). The paroccipital process is clearly visible, which
enables a confident estimation of the width of the skull. All preserved features of the skull confirm that this
is a typical Scaphonyx fischeri specimen, but with juvenile features.

Dermal skull roof. The paired premaxillae are absent. They are restored as beak-like elements as in other
rhynchosaurs (text-fig. 5). Both maxillae are preserved (text-figs. 4 a-d, and 5b), and they bear small markings
which could be blood vessel openings (lateral alveolar foramina; laf, text-fig. 4c, cl). The nasals are absent in
this specimen, and the exact limits of the frontals are also not clear (?f, text-fig. 4a) but their approximate
position and size is illustrated in the skull reconstruction (text-fig. 5a), on the basis of other rhynchosaurs.

340

PALAEONTOLOGY, VOLUME 32

text-fig. 4. Skull of juvenile Sc aphony x fischeri (MCZ 1664) in dorsal ( a ), ventral ( b ), right lateral (c), and
left lateral (d) views. Abbreviations are listed in the Appendix.

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

341

2cm

text-fig. 5. Restoration of the skull of the juvenile Scaphonyx fischeri, based on MCZ 1664, with restored

areas shown by dashed lines.

The parietals are fused to form an inverted T-shaped element in the posterior third of the skull roof between
the upper temporal fenestrae (text-figs. 4 a and 5a). There is the typical crest in the mid-line and a dorsal
ridge on the posterior crossbar. Longitudinal striations are visible on the surface of the parietals which may
reflect the surface texture of the bone.

The lacrimal and prefrontal are not preserved in this specimen, but their approximate positions and sizes
are indicated in the reconstruction (text-fig. 5a, b ), on the basis of other rhynchosaurs. The postfrontal is a
typical three-pronged bone in the posteromedial border of the orbit (text-figs. 4 a and 5a, b). The exact limits
of the left postorbital are obscured by a nodule of rock matrix (text-fig. 4a), but the right postorbital is clearly
visible, although anteriorly displaced. The jugal is a deep four-pronged bone forming the middle part of the
side of the skull (text-figs. 4b, c and 5b) as in adult Scaphonyx. Only the right jugal is preserved in this
specimen and it lacks the posterior quadratojugal process.

Only a partial right quadratojugal is preserved, and it is posteriorly displaced (text-fig. 4c). It would
have been an L-shaped bone that formed the posterolateral angle of the skull (text-fig. 5b). The squamosal
is a large bone which forms much of the posterior margin of the skull (text-figs. 4 a-cl and 5a, b). Both
squamosals are preserved, although they are displaced inwardly towards the parietal, thus reducing the
apparent overall width of the skull in this region. The right postorbital has been displaced from its
point of overlap on the squamosal and the groove on the squamosal in which the postorbital sits is clearly
visible (text-fig. 4c).

Palate. The palatal surface of the skull is obscured by the mandibles. However, in the ventral view of the
skull (text-fig. 46), there are two overlapping long elements of bone protruding from the posterior margin of
the skull, which could be the quadrate wings of the pterygoids.

Quadrate and epipterygoid. The quadrate is a strong columnar element of bone which lies vertically in a
groove on the posterior side of the squamosal and quadratojugal (text-figs. 4a, c, d and 5b). The head of the
right quadrate is visible, as well as the probable distal end. A displaced left quadrate is also indicated (?q,
text-fig. 4c/, d). The epipterygoid cannot be identified in this specimen.

Braincase. In the dorsal view of the skull (text-fig. 4a), elements of the braincase can be seen protruding from
the right posterior margin of the skull. The only identifiable element is the paroccipital process, whose distal
end is oval in cross-section. The shaft of the paroccipital process diverges as it approaches the braincase, but
the midline elements of the braincase are not discernible.

342

PALAEONTOLOGY, VOLUME 32

Lower Jaw

Only the right lower jaw is well preserved in the specimen. It is a deep boat-shaped element, which curves
to a pointed process at its anterior end. When the jaws were closed the premaxillae would have curved down
between the pointed processes of the upwardly curving lower jaws like a pair of tongs (text-fig. 5b).

The dentary constitutes the anterior two-thirds of the mandible (text-figs. 4c, d and 5b). It is not possible
to view the occlusal surface and the teeth of the dentary. In lateral view, near the anterior tip of the dentary,
there are small indentations which could be blood vessel openings (mental foramina; mnf, text-fig. 4c). The
splenials are visible in ventral (text-fig. 4b) and lateral (text-fig. 4c, d) views, but the coronoid , being a medial
element, cannot be seen. The angular forms the base of the mandible in its posterior half, and the surangular
the lateral margin (text-fig. 4c). Only the right surangular is preserved in this specimen, and it can be seen
to contact the dentary in front. The prearticular, a medial element, cannot be viewed in this specimen, and
only the right articular has been preserved (text-fig. 4c).

Axial skeleton

Vertebral column. Fourteen vertebrae can be identified; three cervicals in association with the skull (text-fig.
4a, d), a closely fitting block with a further seven cervical and dorsal vertebrae in association with the shoulder
girdle (text-fig. 6), a sacral vertebra and rib in a block with the left femur and part of the ilium (text-fig. 9a),
and two anterior caudals in association with the right pubis (text-fig. 9e, f). A possible mid to posterior
dorsal is preserved in a separate block (text-fig. 6 f).

The most anterior preserved vertebra is probably the axis, seen only as an ill-defined outline (text-fig. 4a).
This is followed by a poorly delined third cervical and a small part of the fourth. The seven anterior presacrals
(text-fig. 6) presumably consist of cervicals 4-8 (on the assumption that Scaphonyx had eight cervical
vertebrae, as in other rhynchosaurs) and dorsals 9-10. These eight vertebrae are apparently very similar. A
possible posterior dorsal vertebra is represented by a centrum and a partial rib in an isolated small block
(text-fig. 6/).

A single centrum and rib from the right side (text-fig. 9a) represent one of the two sacral vertebrae , and
they are preserved in context on the medial face of the ilium. Two probable caudal centra are preserved with
the right pubis (text-fig. 9e, /).

Ribs. The outlines of five cervical ribs can be traced on the block of seven vertebrae (cr, text-fig. 6a). A
number of fragments of possible dorsal ribs are preserved in isolation, and a proximal end of a rib is
associated with the posterior dorsal vertebra (text-fig. 6 if). No chevron bones or gastralia are preserved in
this specimen.

text-fig. 6. Vertebrae of juvenile Scaphonyx fischeri (MCZ 1664). a, b, cervical vertebrae 4-10, with associated
cervical ribs, in left lateral (a) and dorsal (b) views, c, d , e , cervical vertebrae from the middle of the neck,
restored in left lateral (c), dorsal (d), and anterior ( e ) views./, Centrum of a dorsal vertebra and a dorsal rib
in partial ventral view, embedded in matrix (diagonal shading).

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

343

text-fig. 7. Shoulder girdle and forelimb of juvenile Scaphonyx fischeri (MCZ 1664). a, b, disarticulated
shoulder girdle and forelimb elements (both scapulae, left coracoid, interclavicle, both humeri), in two views,
c, right scapula in lateral view, d , partial left scapula in lateral view, e , interclavicle in ventral view. /, g,
restoration of the shoulder girdle, in left lateral (f) and posterior ( g ) views. The clavicle is unknown.

PALAEONTOLOGY, VOLUME 32

344

text-fig. 8. Forelimb elements of juvenile Scaphonyx fischeri (MCZ 1664). a-d , left humerus, in proximal
(a), dorsal (6), ventral (c), and distal (d) views, e, left hand in dorsal view, with a reconstruction of digit IV
on the right. Matrix is shown with diagonal shading.

Appendicular skeleton

Shoulder girdle. The shoulder girdle is represented by the left and right scapulae, the left coracoid, and the
interclavicle. These elements have been preserved in a block which also includes the left and right humeri
(text-fig. la, b ). There is no trace of the clavicles.

The reconstruction of the left scapula (text-fig. If. , g) has been based on the blade of the left scapula (text-
fig. Id) and the articular end of the right scapula (text-fig. 7c). The coracoid is a saucer-shaped element
approximately half the height of the scapula (text-fig. 7/ , g). The interclavicle is a three-pronged structure
(text-fig. 76, e) with a long dagger-like plate that extends horizontally between and behind the coracoids
(text-fig. If). The lateral prongs would each have met one of the clavicles , but these are missing.

Forelimb. Elements of the forelimb which are preserved are the two humeri and a possible left hand. The
radius and ulna are missing, although they may be represented by some isolated fragments.

The humerus (text-figs, la, b and 8 a-d) is a sturdy bone with a constricted shaft and expanded ends as in
all rhynchosaurs. The poorly preserved left manus (text-fig. 8c) shows a fan-shaped spread of the fingers. The
carpal region is absent and the outline of one digit only can be traced. The pointed tip of the claw of this
digit is visible and an estimation of the number of phalanges suggests that it is digit IV.

Pelvic girdle. The only elements of the pelvic girdle preserved in this specimen are a right pubis and a partial
ilium. Isolated elements of bone may represent fragments of the ischia and the other ilium or pubis.

Only the dorsal blade of the right ilium is visible. It is thin and bowed and slightly thickened at the edges
(text-fig. 9a). The right pubis is a quadrangular element which is seen in ventral view (text-fig. 9c). The
anterior margin is rounded and runs from the mid-line into a short uncinate process directed laterally.

Hindlimb. Elements of the hindlimbs which are present include the left femur, the distal end of the right
femur, a tibia, and a fragment of fibula.

The femur (text-fig. 9 a-d, g-i) has a constricted shaft and expanded ends, the distal end being narrower
than the proximal, as is typical in rhynchosaurs. What may be the proximal end of the right femur embedded
in rock matrix is indicated in text-fig. 9b. The tibia is extremely broad at the proximal end and narrower
distally (text-fig. 9 j, k). From the position of the cnemial crest (text-fig. 9k, l) and the direction in which the
tibia arches, this is most probably a left tibia. Only a fragment of the shaft of the left fibula has been preserved
attached to the left tibia (text-fig. 9 j, k). Three elements of bone attached to the distal end of the tibia may
be the three proximal tarsals, the centrale, the astragulus and the calcaneum (text-fig. 9 j, k). No other elements
of the foot have been preserved.

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

345

text-fig. 9. Pelvis and hindlimb elements of juvenile Scaphonyx fischeri (MCZ 1664). a, b , hindlimb and
pelvic elements (ilium, sacral vertebra and rib, left femur), in ventral (a) and dorsal ( b ) views, c, d , left femur
in proximal (c) and distal ( d) views, e , f pelvic region, showing the right pubis and two caudal vertebral
centra in ventral (pubis) and lateral (vertebrae) views ( e ), and in ventral (vertebrae) view (/). g i, distal end
of the fragmentary right femur, in ventral (g), dorsal (/?), and distal (/) views, j-m, left tibia and fragmentary
fibula, in anterior (J), posterior ( k ), proximal (/), and distal ( m ) views. Some probable tarsal elements are

indicated adhering to the distal end of the tibia.

RECONSTRUCTION AND AGE

The reconstruction of the side view of the juvenile Scaphonyx (text-fig. 10) is based on the available
parts of the specimen, with other portions (dorsal vertebral column and ribs, tail, foot) modified
from adult Scaphonyx (Huene 1929, 1942; Barberena 1971). Total body length is about 380 mm
(skull length, 55 mm; presacral length, c. 180 mm; sacral length, c. 10 mm; tail length, c. 135 mm).
The posture and relationships of the limbs and girdles to the axial skeleton are based on other
more fully preserved (adult) rhynchosaurs (Chatterjee 1974; Benton 1983/)).

346

PALAEONTOLOGY, VOLUME 32

Adult S. fischeri have skull lengths typically of 170-220 mm, and total body lengths of 1-3-
2 0 m (Huene 1942, p. 294). Huene (1942, pp. 294, 295, 313) quoted skull lengths ranging up to
400 or 500 mm (body lengths ?2-5— 3 m), but these large estimates were based on fragmentary
material, and it is unlikely that such huge sizes were attained. The present specimen of S. fischeri
is clearly very young. Is it a hatchling or a juvenile of a year or two old?

Eggs of Scaphonyx have not been reported, but the maximum diameter of a typical egg can be
estimated from the internal width of the pelvis which would have accommodated the birth canal.
This maximum figure ranges from about 60 mm in a 1-3 m long individual to an estimated 70 mm
in a T6 m long specimen. These figures represent either the maximum distance between the medial
faces of the ilia, or the maximum distance between the ventral face of the sacral vertebrae and the
dorsal face of the ischia, whichever is the smaller.

text-fig. 10. Restoration of the juvenile Scaphonyx fischeri based on MCZ 1664 {left), with an adult specimen

(right), drawn to the same scale.

The largest eggs that typical Scaphonyx adults could lay would have been smaller than 60-
70 mm in diameter, which corresponds to lengths of 60-140 mm, depending upon whether the egg
was spherical or ovoid, both of which are known in different reptile groups. A full-term embryo
may be as much as twice the length of its egg, assuming that it is curled up, with the head tucked
down on the chest, and the tail wrapped up in front and meeting the head. Hatchling Scaphonyx
could then have ranged in length from 120-280 mm in total body length. The present specimen,
at 400 mm, could then have been as much as 6 months to a year old.

Another approach to the calculation of hatchling size in reptiles is to use the formula derived
by Currie and Carroll (1984, p. 76) from a study of 120 species of living lizards and crocodilians.
They found that the ratio of adult to hatchling length was described by the formula:

2-7 + 3-3 x (adult length in m)_1

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

347

(misprinted as ‘adult length in mm’ in their paper). The adult length of S. hscheri is typically 1.3-
2.0 m (see above), which gives ratios of 4-3 to 5-2 (i.e. hatchlings are 19-23% of adult body length).
On the assumption then that Scaphonyx had the general physiology and growth rates of modern
reptiles, its hatchling size is predicted as 247-460 mm, depending on which of the ‘adult’ body
sizes is correct. The present specimen falls in the upper half of that range, which might suggest
that it is either a hatchling or aged up to 6 months, on the assumption that rhynchosaurs doubled
their body length in the first year of life as do most modern reptiles (Currie and Carroll 1984).

JUVENILE CHARACTERS

The present specimen, although rather incomplete, shows a number of juvenile characters. For
example, the skull seems to be relatively large in relation to the rest of the body, but not by a
great amount. The ratio cannot be calculated with respect to total body length, since that
measurement is not known independently, but the ratio of skull length : estimated trunk length is
29%, compared to 20-26% in adults (Huene 1942, p. 294). The ratio of skull length : humerus
length in the present specimen ( 1 -28) is within the range of values for adult late Triassic rhynchosaurs
(1 -25- 1-73 for Scaphonyx; 1-24-1 -38 for Hyper odapedon). The articular ends of the long bones are
relatively heavy (the same phenomenon that gives rise to the outsize knees and paws of puppies)
in comparison to undistorted adult specimens of Scaphonyx.

The apparently rather loose association of the skull elements also suggests youth. In adult
Scaphonyx the cranial sutures rarely separate, but most of the skull elements in the present specimen
have shifted apart. Evidently, juvenile rhynchosaurs had rather weakly sutured cranial bones. It
is not possible to comment in detail on the osteological state of the articular ends of the limb
bones because of their slight diagenetic alteration. The roughness of the articular facets suggests,
however, that growth was not complete, and that cartilaginous epiphyses would still have been in
place. Of course, since rhynchosaurs are archosauromorphs (see Benton 19836), rather than
lepidosaurs as once thought, one would not expect to find bony epiphyses.

ALLOMETRIC STUDIES

Regression analysis. The regression analysis indicated that ontogenetic growth of the skull of
Scaphonyx was isometric for most characters (slope close to 1-00; Table 2; text-fig. 1 1). However,
the correlation coefficients (r) are low in most cases, being greater than 0-95 only for skull width,
orbit length, parietal length, and upper temporal fenestra width (characters 2, 3, 7, 9), and between

table 2. Regressions of log skull length against eight other
variables in thirteen skulls of Scaphonyx fischeri. Values for the
slope of the regression lines, the standard deviation, and corre-
lation coefficient are given. One asterisk (*) indicates r>0-90,
while two (**) indicates r>0-95. Utf = upper temporal fenestra.

Variable

Slope

Standard

deviation

Log y
intercept

r

(2) Skull width

100

0 16

0-073

0-966**

(3) Orbit length

1 04

0-18

-0-603

0-954**

(4) Frontal width

0-95

0 16

-0-481

0-838

(5) Frontal length

1 06

019

-0-628

0-933*

(6) Parietal width

0-74

0-13

-0-132

0-810

(7) Parietal length

0-82

0-14

-0-119

0-958**

(8) Utf length

0-95

0-17

-0-411

0-874

(9) Utf width

1 34

0-24

-1-202

0-979**

LOG U.T.F WIDTH LOG PARIETAL WIDTH LOG FRONTAL WIDTH LOG SKULL WIDTH

348

PALAEONTOLOGY, VOLUME 32

text-fig. 1 1. Regression lines of the eight skull measurements (see Table 1) plotted against skull length, all
transformed to logarithms (base 10). Details of the regressions are given in Table 2.

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

349

090 and 0-95 for frontal length (character 5). Thus, linear regressions may not be the best
representations of the points for characters 4, 6, and 8 (frontal width, parietal width, upper
temporal fenestra length). The relatively small sample size (n = 13), however, makes it difficult to
assess the significance of the allometric equations.

The slopes of the regression lines range from 0-74 + 013 (parietal width) to 1-34 + 0-24 (upper
temporal fenestra width). Of the characters that yielded statistically significant slopes (r > 0-90),
three can be interpreted as isometric (slope = 1 -00) : skull width, orbit length, frontal length; one
as showing negative allometry (slope < I -00) : parietal length; and one as showing positive allometry
(slope > 1 -00) : upper temporal fenestra width. The three characters with non-significant slopes
(frontal width, parietal width, upper temporal fenestra length) all show negative allometry.

Transformed coordinates study. The addition of the juvenile specimen described here to the
specimens analysed by Barberena (1971) confirms the conclusions from the allometric equations
(text-fig. 12). The posterior part of the skull roof expands laterally when compared with the anterior
part. The frontal width also appears to increase rapidly in this sequence of skulls, although that
is not confirmed from the larger sample of specimens. In early stages of ontogeny the snout region
of the skull appears to lengthen markedly, while the temporal region (parietal length) diminishes
with respect to overall skull length.

DISCUSSION

The regression analysis indicated that, during ontogeny, the skull of Scaphonyx showed constant
growth in five of the eight measured variables (linear regression lines). Positive allometric growth
occurred in the width of the upper temporal fenestrae, but not in the overall posterior skull width,
according to the present study. This was a surprise, since one would expect both variables to be
broadly linked. However, the maximum skull width character was measured between the lateral
edges of the quadrate condyles which project laterally some way beyond the skull roof proper. It
can be said then that growth in overall skull width is essentially isometric, but the dorsal skull
roof broadens relatively faster than the increase in mid-line skull length. Further, the quality of
preservation of the specimens varies. C. Schultz (pers. comm.) notes that Barberena’s skull D (and
his A and B to a lesser extent) are ‘exploded’ diagenetically which aflects these width measurements.

Negative allometric growth occurred in the parietal length (also in upper temporal fenestra
length, but slope not significant) which indicates that the posterior region of the skull became
shorter relative to the anterior region during growth. The non-significant allometric growth of the
width of the frontal and the width of the parietal indicates that these measurements increased
relatively more slowly than the skull length. However, both characters are possibly sexually
dimorphic in rhynchosaurs (Benton 19836, pp. 612-613) which may have affected the regression
analysis.

The isometric growth of orbit length and frontal length suggests that the anterior part of the
skull grows at the same relative rate as the overall mid-line skull length, while overall maximum
skull width also keeps pace with the skull length. These general results were borne out by the
transformed grid study.

In evolutionary terms the ontogeny of Scaphonyx appears to show heterochronic change when
it is compared with ontogenies of potential ancestors. Regrettably, ontogenetic studies spanning a
broad range of sizes of animals, have not been possible for other rhynchosaur taxa owing to the
rarity of reasonably complete undistorted skulls. However, some comparisons may be made.

The Late Triassic rhynchosaurs, such as Scaphonyx , form a well-defined clade within Rhyncho-
sauria, which is characterized by the following cranial synapomorphies when compared with Early
Triassic taxa such as Mesosuchus , and Middle Triassic taxa, such as Rhynchosaurus and
Stenaulorhynchus ( Benton 19836, 1984, 1987):

1 . Skull is broader than long;

2. Squamosal has broad strap-like ventral process;

350

PALAEONTOLOGY, VOLUME 32

text-fig. 12. An ontogenetic series of five skulls of Scaphonyx fischeri trom Santa Maria, Brazil. Skull a is
MCZ 1664 (based on text-fig. 5a). Skulls b-e are traced from skulls figured by Barberena (1971). Modifica-
tions in relative shape are shown by transformed coordinates (assumed to be quadratic in skull a).

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

351

3. Loss of supratemporal;

4. Lower jaw is very deep;

5. Single groove on maxilla;

6. No teeth on lingual side of maxilla;

7. Occipital condyle in line with quadrates.

Three of these characters (1, 2, 4) are major proportional changes that could be heterochronic in
nature. The positive allometric growth of the width of the dorsal skull roof, reflected in the width
of the upper temporal fenestrae, may represent peramorphosis. The negative allometric growth of
parietal length reflects the fact that Scaphonyx fischeri has a remarkably short temporal region of
the skull when compared to all other rhynchosaurs (length of upper temporal fenestra is one-
quarter to one-third of the total mid-line skull length, compared to values of one-third or more).
This is not a synapomorphy of the Late Triassic rhynchosaur clade, but it could be an autapomorphy
of S. fischeri, and it appears to be peramorphic (text-fig. 13).

1.2

text-fig. 13. The postulated peramorphocline (extended ontogeny) seen in the evolution of the rhynchosaur
skull. Adult skulls of three Early and Middle Triassic rhynchosaurs are shown (based on Benton 19836, in
prep.), and an ontogenetic series of three of the skulls of Scaphonyx fischeri (from text-tig. 12). The skulls
are positioned vertically according to their occurrence in time (stratigraphic column on the left), and
horizontally according to the ratio of posterior skull roof width : mid-line skull roof length (mean values
shown at the tip of the snout of each skull). The posterior skull roof width is measured between the lateral
edges of the intertemporal bars in dorsal view. Abbreviations: A, adult; J, juvenile.

352

PALAEONTOLOGY, VOLUME 32

The particular peramorphic process(es) involved are hard to determine, since the ontogenetic
pattern of ancestral rhynchosaurs has not been described, and the timing of the onset of ontogeny
is unknown. However, since Scaphonyx , and the Late Triassic rhynchosaurs in general, are larger
as adults than the more primitive forms, it is likely that this is a case of hypermorphosis (‘juvenile
development same as ancestor, onset of sexual maturity delayed; adult larger than ancestral adult’:
McNamara 1986, p. 11).

The three best-known rhynchosaur taxa that predate Scaphonyx are Mesosuchus hrowni from
the Early Triassic Cynognathus Zone of South Africa, Stenaulorhynchus stockleyi from the Manda
Formation (Middle Triassic, Anisian?) of Tanzania, and Rhynchosaurus articeps from the Tarporley
Siltstone Formation of England (Middle Triassic, Fadinian?). Very few specimens are known of
these taxa, and the majority are assumed to be adults. The most abundantly represented is R.
articeps, with up to seventeen individual animals, but only five skulls. These range in mid-line
length from 60 mm (estimated) to 82 mm (Benton, in prep.), so cannot offer a great deal of
ontogenetic information. Nevertheless, the ratio of dorsal skull roof width : mid-line skull length
increases from 0-56 to 0-67 from the smallest to the largest specimen, very tentative evidence that
this feature shows positive allometric growth in Rhynchosaurus as well as in Scaphonyx.

Acknowledgements. We thank Ms Elizabeth Purdy for typing the manuscript, and Ms Libby Mulqueeny for
text-figs. It) 13. Cesar Schultz (Rio Grande do Sul, Brazil) very kindly supplied information on Brazilian
rhynchosaurs.

REFERENCES

azevedo, s. a. K. and schultz, c. l. 1988. Scaphonyx sulcognathus sp. nov., um novo rincossarideo neotriassico
do Rio Grande do Sul, Brasil. Anais X Congresso Brasil. Paleont. 1987, 99-113.
barberena, m. c. 1971. Algunas consideraijoes o desenvolvimento de Rincossaurios. An. Acad. bras. Cienc.
43 (suppl.), 403-409.

benton, m. L 1983fl. Dinosaur success in the Triassic: a noncompetitive ecological model. Q. Rev. Biol. 58,
29-55.

— 19836. The Triassic reptile Hyperodapedon from Elgin: functional morphology and relationships. Phil.
Trans. R. Soc. B302, 605-717.

— 1984. Tooth form, growth, and function in Triassic rhynchosaurs (Reptilia, Diapsida). Palaeontology,
27, 737-776.

1987. The phylogeny of rhynchosaurs (Reptilia; Diapsida; Triassic), and two new species. In currie,
p. j. and roster, E. (eds.). Fourth symposium on Mesozoic terrestrial ecosystems, Short Papers. Tyrrell Mus.
Palaeont. Occ. Pap. 3, 12-17.

bonaparte, J. F. 1982. Faunal replacement in the Triassic of South America. J. vertebr. Paleontol. 2, 362-
371.

bortoluzzi, c. a. and barberena, m. c. 1967. The Santa Maria beds in Rio Grande do Sul (Brazil). In
bigarella, J. I., becker, R. D. and pinto, I. D. (eds.). Problems in Brazilian Gondwana Geology, 169-195.
Curitiba, Parana, Brazil.

chatterjee, s. k. 1974. A rhynchosaur from the Upper Triassic Maleri Formation of India. Phil. Trans. R.
Soc. B267, 209 261.

currie, p. J. and Carroll, r. l. 1984. Ontogenetic changes in the eosuchian reptile Tliadeosaurus. J. vertebr.
Paleont. 4, 68-84.

huene, f. von 1926. Gondwana-Reptilien in Sudamerika. Palaeontol. hungarica , 2, 1-108.

— 1929. Ueber Rhynchosaurier und andere Reptilien aus den Gondwana-ablagerungen Sudamerikas. Geol.
pdlaont. Abh. (N.S.), 17, 1 62.

1942. Die fossilen Reptilien cles sudamerikanischen Gondwanalandes. C. H., Beck, Miinchcn.
mcnamara, k. j. 1986. A guide to the nomenclature of heterochrony. J. Paleont. 60, 4 13.
olsen, p. e. and sues, h.-d. 1986. Correlation of continental Late Triassic and Early Jurassic sediments, and
patterns of the Triassic- Jurassic transition. In padian, k. (ed.). The beginning of the age of dinosaurs, 321-
351. Cambridge University Press, Cambridge.

BENTON AND KIRKPATRICK: TRIASSIC RHYNCHOSAUR

353

sill, w. d. 1970. Scaphonyx sanjuanensis, nuevo rincosaurio (Rcptilia) de la Formacion Ischigualasto, Triasico
de San Juan, Argentina. Ameghiniana , 7, 341-354.

1971. Functional morphology of the rhynchosaur skull. Forma Functio , 4, 303 318.
woodward, a. s. 1907. On some fossil bones from the state of Rio Grande do Sul. Revta Mus. paid. 7, 46
57.

Typescript received 23 April 1988
Revised typescript received 21 June 1988

MICHAEL J. BENTON
RUTH KIRKPATRICK

Department of Geology
The Queen’s University
Belfast BT7 INN, UK

APPENDIX

Key to abbreviations used in the figures

SKULL

a

angular

pm

premaxilla

ar

articular

po

postorbital

d

dentary

popr

paroccipital process

f

frontal

prf

prefrontal

j

jugal

Pt

pterygoid

1

lacrimal

ptf

postfrontal

laf

lateral alveolar foramen

q

quadrate

ltf

lower temporal fossa

qj

quadratojugal

m

maxilla

sa

surangular

mnf

mental foramen

sp

splenial

n

nasal

sq

squamosal

o

orbit

utf

upper temporal fossa

P

parietal

POSTCRANI AL

SKELETON

ast

astragalus

fe

femur

calc

calcaneum

fi

fibula

cap

capitellum

h

humerus

ce

centrum

icl

interclavicle

cent

centrale

if

intertrochanteric fossa

cl

clavicle

il

ilium

clw

claw

pub

pubis

cor

coracoid

s

scapula

cr

cervical rib

sr

sacral rib

cv

cervical vertebra

t

tibia

dr

dorsal rib

tr

trochlea

LATE CRETACEOUS AMMONITES FROM THE
WADI QEN A AREA IN THE EGYPTIAN
EASTERN DESERT

by p. luger and m. groschke

Abstract. Ammonites from several transgressive phases of the late Cenomanian to the late Campanian in
the Wadi Qena area (Eastern Desert, Egypt) are described. Taxa included represent mainly Tethyan (south-
west European, North African) and rarer Nigerian and Madagascan species. The recognized species belong
to the genera Neolobites , Pseudocalycoceras (early late Cenomanian); Metengonoceras (late Cenomanian);
Pseudaspidoceras (late late Cenomanian); Nigericeras , Vascoceras, Thomasites (late Cenomanian to early
Turonian); Mammites , Fagesia (early Turonian); Coilopoceras (late Turonian); Metatissotia , Subtissotia
(middle Coniacian); Canadoceras , Manambolites (late middle Campanian); Baculites (middle and late
Campanian); Libycoceras, Nostoceras , and Solenoceras (late Campanian). A correlation of the late Cenomanian
to early Turonian ammonite successions of North Africa, the Middle East, and south-west Europe is
attempted.

Little effort has been spent on the study of ammonites from the late Cretaceous of central Egypt
after a first phase of investigations in the early years of this century (e.g. Eck 1914; Douville 1928).
The present study benefits from the large quantity of new material collected by members of the
Special Research Project 69 'Geoscientific Problems in Arid Areas’ of the German Research
Foundation during field trips in 1985 and 1986 in the central and southern Wadi Qena. Following
detailed investigation of these ammonites, the present paper aims to present the palaeontological
evidence for the biostratigraphical attribution of the lithological units in this region (Klitzsch 1986;
Hendriks et al. 1987; Klitzsch et al. 1989), which is a key area for understanding the geological
history of the Eastern Desert of Egypt.

Wadi Qena, situated in the central part of the Eastern Desert, is a N.-S. directed depression of
approximately 200 km in length, bordered by the Eastern Desert Basement High to the east and
an escarpment of Cenomanian to Eocene sediments in the west (text-fig. 1). Due to its cratonal
position, its depositional history is characterized by transgressive/regressive cycles, which generally
coincide with major eustatic sea-level changes (Luger and Schrank 1 987). Therefore, the Cenomanian
to Campanian stratigraphic column comprises sediments of continental to shallow marine facies
in which the occurrence of ammonites indicates the relative highstand of each transgression from
the late Cenomanian to the late Campanian.

The lithostratigraphy of the Wadi Qena area has been described in detail by Klitzsch (1986) and
Hendriks et al. (1987). A summarized section of the lithological column in the central Wadi Qena
is given in text-fig. 2. A combined section of the Campanian at Gebel Qreiya (text-fig. 1 ) is shown
in text-fig. 4.

STRATIGRAPHY

Cenomanian to Turonian

Late Cenomanian-early Turonian. in the Wadi Qena area the first transgression of the southern
Tethyan sea is documented by the marine sediments ol the Gaiala Formation, overlying fluvial
sandstones ol the Wadi Qena Formation above the crystalline basement (see text-fig. 2; Klitzsch
1986; Hendriks et al. 1987). The lowermost ammonite horizon in the shallow marine deposits of

IPalaeontology, Vol. 32, Part 2, 1989, pp. 355-407, pis. 38-49.|

© The Palaeontological Association

356

PALAEONTOLOGY, VOLUME 32

text-fig. 1. Location of sections 1-9 and Locality A in the Wadi Qena area (see also Hendriks et al. 1987;

Hendriks and Luger 1987).

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

357

c

z

<

Z UJ

b

< 1-

CL <

s -1

>N

<

.c

o

03

QC

o-

< / )

z

<

Z>

z

o

1—

z

c

<

•s

CO

e-

£ «

' 03

j= =>

CO

03

LET

1-

$

03

I

<

Z -1

O -i

(0

< Q

zQ

O s

o

L.r.s.

in

CO

l—

z3

<

<D

z

o

E

O

c CO

cc

C

z >-

=> J

h- — 1

cc

<

Lf .3

LU

c

•3 CM

03 •

E 13

z

o

LL .

<

_co

Z LU

< 1-

CD 7]

o — 1

z

LU

Wadi

o

Qena

F.

* V

^7

conglomeratic sandstone
sand- and siltstone
clayey sand - and siltstone
silty to sandy claystone
claystone
marl

marly limestone
limestone

coquinoid limestone
conglomeratic limestone

sandy limestone

calcareous sand- and
siltstone

coquinoid sandstone
channel

large-scale cross-bedding
small-scale cross-lamination

/ Metatissotia fourneli (BAYLE).M. cf.
I M. sp. Subtissotia afrlcana (PERON)

erosional unconformity

¥ glauconite
© oolites
® Fe-oolite
jk > plant remains
; bioturbations
Ps: paleosoil

Pf: planktonic foraminifera
Bf: benthonic calcareous foraminifera
Af: benthonic agglutinated foraminifera
Am: ammonites
C: corals
E: echinoids
G: gastropods
P: pelecypods

V: vertebrate remains
N: nautiloids

ewaldi CVON BUCHI

®P

Am
♦ PE
♦Am

Coilopoceras requienianum (D'ORBIGNY)
Coilopoceras requienianum (D'ORBIGNY)

ijs f Pseudaspldoceras sp., Vascoceras durandi THOMAS and PERON,

Am D„r- { V.cauvlnl CHUDEAU, V.rumeaui (COLLIGNON),

ctcauvini CHUDEAU, Thomasites cf. subtenue (REYMENT)

—Am PGEC Metengonoceras cf. scutum HYATT
P

. ( Neolobites vlbrayeanus (D'ORBIGNY), Pseudocaiycoceras cf. haug/(PERV.),

NAm pjo y$toguiithes mermeti (COQUAND)

NAm p-*

V* rNeolobites vibrayeanus (D'ORBIGNY), Neolobites sp.
\Angulithes mermeti (COQUAND)

text-fig. 2. Generalized section of late Cretaceous sediments in the central Wadi Qena (sections 7-9
combined, modified after Hendriks et al. (1987, fig. 5)). Lf 13 = Lithofacies 1-3; L.r.s. = Lower regressive
sequence; L.t.s. = Lower transgressive sequence; U.r.s. = Upper regressive sequence; U.t.s. = Upper trans-
gressive sequence.

358

PALAEONTOLOGY, VOLUME 32

the lower Galala Formation contains Neolobites vibrayeanus (d’Orbigny), Neolobites sp. and the
nautiloid Angulithes mermeti (Coquand). Separated by about 2-5 in of clay-/silt-/limestone
intercalations it is overlain by a sandy limestone with N. vibrayeanus and Pseudocalycoceras cf.
haugi (Pervinquiere), together with A. mermeti.

N. vibrayeanus is widely known from the early late Cenomanian prior to the occurrence of
Metoicoceras geslinianum (d’Orbigny) and Vascoceras gamai Chofifat of the Middle East, North
Africa, and Western Europe (see text-fig. 3). According to Berthou (1984, p. 52) and Berthou et
al. (1985, p. 56), N. vibrayeanus is restricted to the naviculare Zone in Portugal. This is also in
agreement with the occurrence of P. cf. haugi , which has been reported from around the boundary
of the crassum and naviculare Zones (Thomel 1972).

The upper part of the Galala Formation is made up of light yellow, massive marls and marly
limestones, containing hermatypic corals at the base. The only ammonite found at the base of
these marls is Metengonoceras cf. acutum Hyatt (section 7, see text-fig. 2). M. acutum is known
from the late Cenomanian pondi to gracile Zones of the North American Western Interior Basin
prior to the occurrence of V. cauvini (see Cobban 1987), which is in agreement with the stratigraphic
position of the present M. cf. acutum.

Further up, irregularly distributed ammonites of latest Cenomanian to earliest Turonian age are
very common. Due to the massive nature of the sediment, the irregular distribution of ammonites
and the severe erosion, most of the ammonites from these layers could not be collected from well-
defined horizons, but from talus surfaces. Therefore, as mixing of different faunal assemblages
may have occurred, it is impossible to give a well-defined ammonite zonation around the
Cenomanian/Turonian boundary in this paper.

Among the present material, V. gamed Choffat was found only at localities where the lower parts
of the massive marls of the upper Galala Formation are exposed (sections 5 and 6, see text-fig. 1).
It was not observed among the assemblage of globose vascoceratids mentioned below. In Portugal,
V. gamed occurs in the geslinianum Zone, although the "V. gamai group’ ranges into the early
Turonian (Berthou 1984, p. 52; Berthou et al. 1985, p. 70). A similar situation is reported from
Algeria, where a Neolobites/ Pseudocalycoceras assemblage (Zone II) occurs prior to V. gamai
(Zone III), which itself is followed by a Wrightoceras/ Bauchioceras assemblage (Zone IV), the
latter being replaced by the vascoceratid assemblage of Zone V (Amard et al. 1981, table 5), see
below. Therefore, we assume the same stratigraphic position for the Egyptian specimens of V.
gamai , i.e. at least partially equivalent to the geslinianum Zone (see text-fig. 3). The ‘ Kanabiceras
Zone’ of Freund and Raab (1969) was shown to be equivalent to the standard geslinianum Zone
by Lewy et al. (1984, p. 72) based on the common occurrence of the zonal index species.

Only one assemblage comprising V. durandi Thomas and Peron, V. cauvini Chudeau, V. cf.
cauvini , V. rumeaui (Collignon), Thomasites cf. subtenue (Reyment) and Pseudaspidoceras sp. is
from a narrow stretch within the higher part of the upper Galala Formation of section 7 (see text-
fig. 2). V. cauvini and V. rumeaui are reported from the cauvini Zone of Israel (Freund and Raab
1969). There V. cauvini has also been shown to co-occur beyond with Metoicoceras geslinianum
and Kanabiceras sp. (‘ K . Zone’ of Freund and Raab 1969) by Lewy et al. (1984). According to
Freund and Raab (1969), V. durandi (Thomas and Peron), V. harttiforme Chofifat, V. cf. amieirense
Choffat, and V. cf. aelonense Choffat, which were all taken into the synonymy of V. durandi by
Berthou et al. ( 1 985), appear for the first time in the overlying pioti Zone; species of Pseudaspidoceras
are reported from both cauvini and pioti Zones by Freund and Raab (1969). In Algeria, V. cauvini ,
V. rumeaui, and Pseudaspidoceras are reported from the Zone V of Amard et al. (1981), which is
at least in part an equivalent of the cauvini Zone. In Portugal, V. durandi is known from the late
Cenomanian juddii Zone to the early Turonian coloradoense Zone, having its acme in the
coloradoense Zone (Berthou et al. 1985). In Tunisia, V. durandi ‘. . . appears above the correlative
of the juddii Zone in association with fragmentary P. cf. flexuosum (Kennedy et al. 1987, p. 68).
Therefore, as V. cauvini and V. rumeaui co-occur in the cauvini Zone, the above mentioned
ammonite assemblage probably comprises the cauvini Zone plus equivalents of the pioti Zone sensu
Freund and Raab (1969).

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

359

Portugal

N* Spain

Middle East

Algeria

Egypt

1)

2)

3)

4)

present study

c

Wrightoceras munierl

c

(M)

Ingrldella malladae

Choffaticeras luciae

3

L

(Mammites nodosoides)

trisellatum

Hoplitoides

Mammites sp.

Choffaticeras quaasi

(VI)

Fagesia cf. superstes

■3

Paramammltes saenzi

<0

HI

K?

Choffaticeras securiforme

Vascoceras pioti

Vascoceras durandi

J

Fallotites subconciliatum

Vascoceras cauvini

Vascoceras cauvini, V. rumeaui

Vascoceras cauvini, V. rumeaui

c

1

(V)

.2

H

G

Wrightoceras, Bauchioceras

E

(IV)

O

F

Vascoceras gamai

Kanabiceras sp.

Vascoceras gamai

Vascoceras gamai

0>

E

Metoicoceras geslinianum

(III)

0

D

Metoicoceras muelleri

Neolobites, Pseudocalycoceras

Metengonoceras cf. acutum

(ID

C

Neolobites vibrayeanus ♦

Neolobites vibrayeanus

Neolobites vibrayeanus

Neolobites vibrayeanus

®

Calycoceras (Lotzeites) lotzei

(1)

is

Eucalycoceras spathl

text-fig. 3. Inter-regional correlation of ammonite successions around the Cenomanian Turonian boundary.
(I) after Choffat (1898) and Berthou (1984); (2) after Wiedmann 1978; (3) after Freund and Raab (1969) and
Lewy et al. (1984); (4) after Amard et at. (1981). Cenomanian-Turonian boundary and Turonian correlation
after Kennedy et al. (1987). Late Cenomanian correlation original.

The cauvini Zone is apparently not well defined and its use varies in different parts of the world.
In Israel it is understood to represent the uppermost Cenomanian by Lewy et al. (1984), who
correlated it with the standard juddii Zone because of the common occurrence of Pseudaspidoceras
of type pseudonodosoides. These authors also described V. cauvini from the preceding ‘A. Zone’,
which they correlated with the geslinianum Zone. However, from the drawings given by Lewy et
al. (1984, figs. 2 and 3, table 1) it is not clear whether in their opinion the cauvini Zone is restricted
to the Cenomanian or extends into the basal Turonian. According to Kennedy (1985, table 6)
and Kennedy et al. (1987, p. 68), who adopted Lewy’s view, the cauvini Zone is restricted to
the late Cenomanian. In the Western Interior Basin of the US a cauvini Zone is distinguished
between the late Cenomanian gracile and juddii Zones (Cobban 1984). Thus, although the exact
relation between the southern Tethyan cauvini Zone and the cauvini and juddii Zones of the
Western Interior Basin are in our view not yet fully understood, the authors follow the attribution
of the entire cauvini Zone of the Middle East to the late Cenomanian as expressed by Kennedy
et al. (1987, text-fig. 13).

Among the material collected from the weathering scree of the upper Galala Formation in
section 5 (see text-fig. 1), Fagesia cf. superstes (Kossmat) and Mammites sp. are represented. The
genera Mammites and Fagesia sensu Wright and Kennedy (1981, pp. 67, 87) are known from the
early Turonian. Hook and Cobban (1981, fig. 3) reported Fagesia sp. together with Neocardioceras
juddii (Barrois and de Guerne) from the Colorado Formation of New Mexico. However, Hook
and Cobban (1981) did not give an illustration of the specimen from this horizon, but instead
figured specimens questionably assigned to Fagesia (Hook and Cobban 1981, pi. 2, figs. 1-2, 5)
from overlying horizons without N. juddii. In a later study Cobban and Hook (1983, p. 16) discussed
the stratigraphical occurrence of Fagesia and noted: ‘ Fagesia is widely distributed in rocks of early

360

PALAEONTOLOGY, VOLUME 32

and middle Turonian age’ (latest Cenomanian — juddii Zone). In the Middle East, F. cf. superstes
is recorded from the early Turonian quaasi Zone (Zone 5, see Freund and Raab 1969, p. 35).
Therefore, we attribute the present specimens of F. cf. superstes and Mammites sp. from unspecified
horizons of the upper Galala Formation to the early Turonian at a higher position than the
assemblage of globose vascoceratids mentioned above.

In the present material no other undoubted early Turonian ammonites, like species of
Choffaticeras ( sensu Freund and Raab 1969, p. 50) which have previously been described from
Egypt by Eck (1914) and Douville (1928), were observed, probably due to paleogeographical
reasons. Thus it is impossible to propose a corresponding biozone of previous authors for the beds
with Fagesia and Mammites.

Late Turonian. The marine early Turonian sediments are overlain by fluvial sandstones with an
erosional contact (lower Umm Omeiyed Formation, see text-fig. 2). A new transgression is
documented by the sediments of the upper Umm Omeiyed Formation, which consist of deposits
of different shallow marine subfacies (Hendriks et al. 1987). Here Coilopoceras requienianum
(d’Orbigny) was found in two well-defined horizons of glauconitic, calcareous sandstones in section
7, which are separated by about 8 m thick intercalations of claystones, marls, and marly limestones,
containing pelecypods and echinoids (see text-fig. 2).

The type material of C. requienianum from the Uchaux Massif was attributed to the late Turonian
neptuni Zone by Kennedy and Wright (1984, p. 285). The species is also recorded from the
deverianum and neptuni Zones of France by Devalque et al. (1982) who include the deverianum
Zone in the late Turonian. According to Kennedy (1984c/, p. 151) the deverianum Zone is an
equivalent of the upper part of the woolgari Zone, which is referred to the middle Turonian sensu
anglico. Collignoniceras woolgari (Mantell) has recently been shown to co-occur successively with
Romaniceras deverianum (d’Orbigny) and Subprionocyclus neptuni (Geinitz) in the eastern Paris
Basin by Kennedy et al. (1986). The C. requienianum Zone of Fewy (1975) hence very likely is an
equivalent of the deverianum and neptuni Zones, here regarded as late Turonian (see also Kennedy
1985, table 11).

Coniacian to Maastrichtian

Coniacian-? early Campanian. The marine late Turonian is overlain by a regressive sandstone
sequence (lower regressive sequence of Hawashya Formation, see text-fig. 2). The succeeding
ammonite assemblage in the middle Wadi Qena consists of Metatissotia fourneli (Bayle), M. cf.
ewaldi (von Buch), Metatissotia sp., and Subtissotia africana (Peron) from a single limestone
horizon intercalated in thick massive oyster beds of the overlying lower transgressive sequence of
the Hawashya Formation.

In Europe, M. ewaldi is known from the middle Coniacian tridorsatum Zone (Kennedy 19846,
p. 128). The species was also reported from the Ca 5 Zone of the Middle East by Fewy (1975)
and Lewy and Raab (1978), together with M. fourneli, M. cf. steinmanni Lisson, Tissotia sp.,
Protexanites sp., and Reesideoceras sp. The CA 5 Zone in the upper Zihor Formation was
tentatively assigned to the late Coniacian by Fewy and Raab (1978, p. 4), although Fewy (1975,
p. 31) attributed the ‘lower chalk’ of the overlying Menuha Formation, barren of ammonites, to
the late Coniacian P. emscheris Zone (= upper tridorsatum + margae + serratomarginatus Zones of
Kennedy 19846, p. 5), since the next ammonites overlying the ‘lower chalk’ were placed by him
within the texanus Zone (early Santonian). This early view of Fewy was confirmed by Reiss et al.
(1985) and Almogi-Fabin et al. (1986, p. 852) who attributed the ‘lower chalk’ of the Menuha
Formation to the lower part of the late Coniacian/early Santonian planktonic foraminiferal Zone
of Dicarinella concavata. Therefore, we propose to correlate the present tissotiid assemblage with
the tridorsatum Zone, i.e. the middle Coniacian sensu Kennedy (19846).

In the central Wadi Qena, the marine Coniacian is overlain by a regressive sandstone sequence
(upper regressive sequence of Hawashya Formation), which is succeeded by marine sediments with
a poor molluscan fauna of pelecypods and gastropods (upper transgressive sequence of Hawashya

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES 361

Formation, see text-fig. 2). Due to the lack of guide fossils, the age of this marine ingression
remains uncertain, but it may be Santonian-?early Campanian since the base of the next datable
unit (Rakhiyat Formation) is attributed to the late middle Campanian (see text-fig. 4).

Qreiya, southern end of Wadi Qena (sections 2 and 3 combined, modified after Hendriks and

Luger 1987, fig. 2).

Middle Camp anion- Maastrichtian. At the Gebel Qreiya at the southern end of Wadi Qena,
Campanian ammonites occur in two phosphatic sequences separated by thick, almost unfossiliferous
claystones (Rakhiyat Formation, see text-fig. 4). The lower assemblage consists of Baculites cf.
ovatus Say, Canadoceras cottreaui Collignon and Manambolites piveteaui Hourcq. Until now, M.
piveteaui and C. cottreaui were known exclusively from the ‘Zone of Delawarella sub delaw arensis
and Australiella australis ’ of Madagascar, which is attributed to the ‘late middle’ Campanian
by Collignon (e.g. Collignon 1977, table 1). Since knowledge about the stratigraphic distribu-
tion of M. piveteaui and C. cottreaui seems to be very limited and no better known ammonites
(nor any other index fossils) were found co-occurring in the Gebel Qreiya section, we hesitate to

362 PALAEONTOLOGY, VOLUME 32

propose a correlation of the Manambolites layer with the European standard Zones of Kennedy
(1986).

The fauna of the overlying claystones exclusively consists of rare casts of pelecypods and
arenaceous foraminifera, possibly of a mixohaline facies (Hendriks and Luger 1987). Shallow open-
marine conditions again are documented at this locality by the overlying intercalations of
phosphoritic, calcareous, partially silicified conglomerates, and marls (upper part of unit 3 of
Rakhiyat Formation, see text-fig. 4). Whereas the marls contain low-diversity calcareous foramin-
iferal assemblages (mainly buliminids, rare planktonics), as well as pelecypods and vertebrate
remains (fish teeth), ammonites have been recovered from two of the conglomeratic horizons
(Hendriks and Luger 1987). Libycoceras sp. ex gr. L. ismaeli (Zittel) was found in the lower one
and an assemblage of heteromorphs ( Nostoceras ( Nostoceras ) sp., N. ( Planostoceras ) sp., Solenoceras
humei (Douville), Baculites subanceps Haughton) in the upper one, about one metre above. This
upper ammonite assemblage clearly corresponds with that of Lewy (1967, 1969) from the upper
Mishash Formation in Israel and that of Barthel and Herrmann-Degen (1981) from the basal
Dakhla Shale Member in the Dakhla Basin. The biostratigraphic position of these assemblages
falls within the late Campanian polyplocum Zone (Reiss et al. 1985).

Generally, the late Campanian deposits in the Wadi Qena and area to the south are overlain
with an erosional unconformity by conglomeratic, phosphoritic marls containing besides pelecypods
( Gryphaea vesicularis Lamarck, pectinids) and vertebrate remains a rich foraminiferal fauna of late
early Maastrichtian age (upper part of the planktonic foraminiferal Zone of Globotruncana
falsostuarti , base of Hamama Marl Member of Dakhla Formation, see Hendriks and Luger 1987).
The Maastrichtian in the central and southern Wadi Qena and the Gebel Rakhiyat area is
characterized by sermpelagic open shelf sediments rich in planktonic and benthic foraminifera, in
which no ammonites have yet been found.

SYSTEMATIC PALAEONTOLOGY

Repository of material. All specimens are stored in the collection of the Special Research Project 69 (SFB),
Technical University of Berlin.

Measurement of dimensions. D, diameter (mm); Wh, Wb , whorl height and breadth as fractions of D\ Ud ,
umbilical diameter as fraction of D.

Suture terminology. Suture terminology after Wedekind (1916) (see Kullmann and Wiedmann 1970). I,
internal lobe; U, umbilical lobe; L, lateral lobe; E, external lobe.

Order ammonoidea Zittel, 1884
Suborder ammonitina Hyatt, 1889
Superfamily desmoceratacea Zittel, 1895
Family pachydiscidae Spath, 1922
Genus canadocf.ras Spath, 1922

Type species. Ammonites newberryanus Meek, 1876, p. 47, by original designation of Spath (1922).

Canadoceras cottreaui Collignon, 1938

Plate 38, figs. 2 and 3

1938 Canadoceras cottreaui Collignon, p. 63, pi. 3, fig. 2.

1955 Canadoceras cottreaui Collignon; Collignon, p. 47.

1970 Canadoceras cottreaui Collignon; Collignon, p. 24, fig. 2301.

Material. Four fragments of body-chambers (SFB C423 C426). Each a half of a whorl. Last suture line just
faintly visible on all specimens. Dimensions could not be determined.

Description. Shell moderately involute. Whorl section highly oval with flat flanks, venter rounded, shallow.

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

363

Umbilical shoulder rounded, umbilicus moderately deep. Ornamentation in general relatively faint, appearing
somewhat irregular. Single ribs radiate to weakly prorsiradiate, turning to the front in the ventrolateral
region, crossing the venter with a strong sweep in direction of growth. In the umbilical area, ribs are first
slightly sharpened, then broaden and smoothen on the flanks, thicken slightly in the ventrolateral area and
are only vaguely recognizable on the venter. Single intercalatory ribs are present, partly reaching as far as
the umbilical region.

Discussion. The present material, although slightly larger, resembles the specimens described by
Collignon (1938, pi. 3, fig. 2) as regards size of umbilicus and ornamentation. In his original
description Collignon (1938, p. 63) mentioned the existence of four specimens of C. cottreaui and
gave the measurements of three of them, while marking the second measured specimen as ‘(type)’.
The measurements of his only figured specimen (Collignon 1938, pi. 3, fig. 2) are identical to those
given on p. 63 for his ‘no. I', and we therefore assume that the holotype of Collignon’s own
designation is not figured.

Occurrence. Rakhiyat Formation: Unit I, section 2. Late middle Campanian.

C. cottreaui is recorded from the late middle Campanian (‘Zone of D. subdelawarensis and A. australis ’)
of Madagascar by Collignon (1970). The holotype of the Mitraiky section (Madagascar) was originally
assigned to the late Campanian by Collignon (1938).

Superfamily hoplitaceae Douville, 1890
Family engonoceratidae Hyatt, 1900
Genus metengonoceras Hyatt, 1903

Type species. Metengonoceras inscription Hyatt, 1903, p. 180, pi. 25, figs. 5-9; pi. 26, figs. 1 4, by subsequent
designation of Diener (1925).

Metengonoceras cf. acutum Hyatt, 1903
Plate 38, fig. 1; text-fig. 6d

cf. 1903 Metengonoceras acutum Hyatt, p. 184, pi. 26, fig. 8; pi. 27, figs. 1 and 2.

cf. 1981 Metengonoceras acutum Hyatt; Kennedy et al ., text-fig. 5a.

cf. 1987 Metengonoceras acutum Hyatt; Cobban, p. 63, pi. 1, figs. I and 2, 7; pi. 2, figs. 4 8; pi. 3 (see

here for complete synonymy).

Material. Nine poorly preserved fragments (SFB C279 C287). One of them with partially preserved body-
chamber. Dimensions could not be determined.

Discussion. The present fragments of internal moulds, which apart from very faint, short fold-like
ribs near the umbilicus lack any trace of ornamentation, very closely resemble M. acutum in their
straight suture trace (text-fig. 6d) and the sharpened venter on body-chamber. However, due to
the poor preservation the present specimens cannot be assigned unequivocally to M. acutum.
M. acutum differs from the similar species M. dumbli (Cragin, 1893) in the sharpened instead
of rounded venter on the body-chamber (Cobban 1987) and the straight instead of curved suture
line (Kennedy et al. 1981).

Occurrence. Galala Formation: Locality A (SFB C279). Middle Galala Formation: Section 5 (SFB C281
C287) and section 7 (SFB C280). Late Cenomanian.

The holotype of M. acutum was assigned to the later Cenomanian (gracile Zone) by Kennedy et at. (1981).
The species is known from the late Cenomanian pondi to gracile Zones of the Western Interior of the United
States (Cobban 1987).

Genus neolqbites Fischer, 1 882

Type species. Ammonites vibrayeanus d’Orbigny, 1841, p. 322. pi. 96, figs. 1-3, by original designation of
Fischer (1882).

364

PALAEONTOLOGY, VOLUME 32

text-fig. 5. Neolobites vibrayeanus (d’Orbigny). SFB C265, section 7, lower Neolobites assemblage,
lower Galala Formation, early late Cenomanian, x 1 .

EXPLANATION OF PLATE 38

Fig. 1 . Metengonoceras cf. acutum Hyatt. SFB C279, fragmentary phragmocone, locality A, Galala Formation,
late Cenomanian.

Figs. 2 and 3. Canadoceras cottreaui Collignon. 2, SFB C424; 3, SFB C423. Fragmentary body-chambers,
both from section 2, basal phosphate horizon of Rakhiyat Formation, late middle Campanian.

Figs. 4 6. Neolobites sp. 4, SFB C252; 5 and 6, SFB C253. Both specimens from section 7, lower Galala
Formation, lower Neolobites- assemblage, late Cenomanian.

All figures x 1.

PLATE 38

LUGER and GROSCHKE, Metengonoceras, Canadoceras, Neolobites

366

PALAEONTOLOGY, VOLUME 32

Neolobites vibrayeanus (d’Orbigny, 1841)

Plate 39, fig. 3; text-fig. 5

1841 Ammonites vibrayeanus d’Orbigny, p. 322, pi. 96, figs. 1-3.

1981 Neolobites vibrayeanus (d'Orbigny); Kennedy and Juignet, p. 23, figs. 3 a-c, 4 a, b , 5, 6 a (see
here for further synonymy and refigured holotype).

1985 Neolobites cf. vibrayeanus (d’Orbigny); Dominik, pi. 14, figs. 4 and 8.

Material. Six internal moulds (SFB C265-C270). The shell is partially preserved in one specimen. The body-
chamber, up to half a whorl, is partially preserved in three specimens, the largest of which shows a diameter
of 146 mm.

Description. The extremely involute, highly compressed specimens, in which the umbilicus reaches only up
to 5% of the diameter, show weak rib-like folds on the inner flanks, which arise from the umbilicus. No
ornamentation is visible on the outer flanks, possibly due to the poor preservation. In the smallest specimen
(SFB C270) the lateral surface is smooth at a whorl height of 22 mm, later on ribs arise and strengthen to
the outer whorl. The venter is narrow and flat in the outer, subsulcate in the inner whorls.

Discussion. The present specimens very closely resemble the holotype as refigured by Kennedy and
Juignet (1981, fig. 3a-c); however, their ornamentation is slightly weaker and the umbilicus a little
narrower.

Occurrence. Lower Galala Formation: Lower (SFB C265) and upper (SFB C266-C270) Neolobites-assemb\age,
section 7. Late Cenomanian.

According to Kennedy and Juignet (1981), N. vibrayeanus is known from the earlier late Cenomanian of
western Europe (France, Spain, Portugal), northern Africa and the Middle East (Morocco, Algeria, Tunisia,
Egypt, Israel, Lebanon, Arabia), as well as South America (Peru, Bolivia).

Neolobites sp.

Plate 38, figs. 4-6; text-fig. 13d, e

Material. Thirteen specimens (C252-C264). Mainly internal moulds of phragmocones, with partial remains
of shell; body-chamber of about half a whorl preserved in two specimens.

D

Wh

Wb

U

SFB C253

38

0-56

0-31

012

SFB C264

86

0-52

0-23

012

SFB C252

115

0-59

0-25

012

Description. Shell compressed with high whorl sides and small umbilicus; coiling becomes excentric on body-
chamber. Venter subsulcate to flat, becoming rounded on body-chamber. Cross-section shown in text-fig.
13d, e. Ornamentation of phragmocone consists of almost straight ribs on the inner flanks arising from the
umbilicus at early growth stages. Ribs are broadening and weakening on the outer flanks. Elongated
ventrolateral tubercles are developed on the phragmocone. On body-chamber ornamentation only consists
of broad folds on the inner flanks.

Discussion. Neolobites sp. differs from N. vibrayeanus (d’Orbigny, 1841) as figured by Kennedy
and Juignet (1981, fig. 3a-c, holotype) and the present specimens by its larger umbilicus, its
excentric coiling of the body-chamber, the elongated ventrolateral tubercles on the phragmocone,
and the rounded venter of the body-chamber. Neolobites sp. is, as regards mode of coiling and
ribbing, very close to the holotype of N. choffati Hyatt, 1903 (pi. 25, figs. 1-4 = N. vibrayeanus
(d’Orbigny) in Choffat 1898, pi. 5, figs. 3, 8-9). However, neither from Hyatt’s nor Choffat’s
descriptions and illustrations can it be recognized whether N. choffati ( sensu Hyatt 1903) bears ventro-
lateral tubercles or not. Neolobites sp. is also very similar to N. peroni Hyatt, 1903, a form
with ventrolateral tubercles which, according to Collignon (1965), shows a narrower umbilicus.

The present specimens of Neolobites are from two well-defined horizons; the thirteen specimens
here referred to as Neolobites sp. are exclusively from the lower unit and show hardly any variation

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

367

text-fig. 6. External sutures of: a, e, Coilopoceras requienianum (d’Orbigny). a, SFB C329; E, SFB C331.
Both x 2. b, Thomasites compressus (Barber). SFB C315, x 2. c, Vascoceras gamai Choffat. SFB C309, x 3.
d, Metengonoceras cf. acutum Hyatt. SFB C279, x 2. f, V. cf. cauvini Chudeau. SFB C306, x2. G, h, V.
cauvini Chudeau. G, SFB C295, x 2; h, SFB C296, x 1. I, Pseudcispidoceras sp. SFB C316, x 2.

368

PALAEONTOLOGY, VOLUME 32

as regards size of umbilicus and ornamentation. Only one specimen of N. vibrayeanus was found
together with them. All the specimens of Neolobites from the upper horizon show the flat, non-
tuberculate venter up to large growth stages of true N. vibrayeanus (see Kennedy and Juignet 1981,
fig. 3 a-c, holotype). The differences between the present N. vibrayeanus and Neolobites sp. could
be related to dimorphism as they co-occur in the lower horizon; this seems, however, unlikely since
Neolobites sp. was not observed in the upper one.

N. choffati and N. peroni have been considered to be junior synonyms of N. vibrayeanus by
Kennedy and Juignet (1981) without a detailed discussion. Collignon (1965, p. 169) discussed the
species of N. vibrayeanus , N. peroni , and N.fourtaui Pervinquiere, 1907, but the detailed stratigraphic
relationships of the figured specimens remained uncertain. This is also the case for Avnimelech
and Shoresh (1962, p. 530) who doubted the validity of N. peroni and N. choffati. However, from
the present material it cannot be decided whether the high variability attributed to N. vibrayeanus
by Kennedy and Juignet (1981) is in accordance with reality or hides biostratigraphical or other
parameters which could justify a separation of distinct species.

Occurrence. Lower Galala Formation: Lower Aco/o^fies-assemblage, section 5 (SFB C263-C264) and section
7 (SFB C252-C262). Late Cenomanian.

Superfamily acanthocerataceae de Grossouvre, 1894
Family acanthoceratidae de Grossouvre, 1894
Subfamily acanthoceratinae de Grossouvre, 1894
Genus pseudocalycoceras Thomel, 1969

Type species. Ammonites harpax Stoliczka, 1864, p. 72 (pars), pi. 39, fig. 1 only, by original designation of
Thomel (1969).

Pseudocalycoceras cf. haugi (Pervinquiere, 1907)

Plate 39, figs. 1 and 2

cf. 1907 Acanthoceras haugi Pervinquiere, p. 270, pi. 14, fig. la, b.

cf. 1972 Pseudocalycoceras ( Haugiceras ) haugi (Pervinquiere); Thomel, p. 97, pi. 31, figs. 7 and 8.

Material. Four poorly preserved fragmentary internal moulds (SFB C275-C278). Dimensions could not be
determined.

Description. Shell moderately evolute. Umbilical shoulder rounded. Whorl section subsquarish, a little broader
than high. Venter flat. Ornamentation coarse; one umbilical and two ventrolateral rows of nodes developed,
which are connected by radial primary ribs. Lateral intercalatory ribs present, also bearing ventrolateral
nodes. In the outer ventrolateral row nodes are claviform. Whorls do not embrace the inner ventrolateral
row of nodes. Suture line not visible.

Discussion. The present specimens are strongly eroded, especially in the ventral regions. However,
the mode of coiling and the ornamentation is very similar to that of P. haugi, but due to the poor
preservation our specimens are only compared with this species. The specimen figured by Thomel
(1972, pi. 31, figs. 7 and 8) from the late Cenomanian of France shows a smaller umbilicus than
our material and the holotype (Pervinquiere 1907, pi. 14, fig. 1 a, b).

explanation of plate 39

Figs. 1 and 2. Pseudocalycoceras cf. haugi (Pervinquiere). 1, SFB C276; 2, SFB C275. Both specimens from
section 7, lower Galala Formation, upper Neolobites- assemblage, early late Cenomanian.

Fig. 3. Neolobites vibrayeanus d’Orbigny. SFB C266, section 7, lower Galala Formation, upper Neolobites-
assemblage, early late Cenomanian.

All figures x 1 .

PLATE 39

LUGER and GROSCHKE, Pseudocalycoceras, Neolobites

370

PALAEONTOLOGY, VOLUME 32

P. haugi , originally attributed to Acanthoceras , was assigned to Pseudocalycoceras by Thomel
(1972) and chosen as subgenotype of Haugiceras Thomel, 1972. Wright and Kennedy (1981,
p. 36) discussed the genus Pseudocalycoceras and rejected the subgenus Haugiceras by noting:
The type species as figured by Pervinquiere is, however, so similar to typical Pseudocalycoceras
that separation is in our opinion unnecessary.’

P. haugi was also reported, but not figured, by Avnimelech and Shoresh (1962, p. 532) as
'Calycoceras' haugi from the Cenomanian of Israel, prior to the occurrence of Neolobites (‘Calcaire
a Acanthoceras'). These authors, apparently tentatively, regarded ‘ A .' palestinense Blanckenhorn,
1905 as figured by Taubenhaus (1920, pi. 2, fig. 3) as synonymous with P. haugi (a view which
was adopted by Thomel 1972). The specimen figured by Taubenhaus (1920) is, however, much
more densely ribbed and the ribs are finer than in Pervinquiere’s holotype, so that we suggest
keeping the two forms separate. Since Avnimelech and Shoresh (1962) did not give an illustration
of their ‘C.’ haugi , this report remains questionable.

Occurrence. Lower Galala Formation: Upper Neolobites- assemblage, section 7. Late Cenomanian.

P. haugi is known from the Cenomanian of Tunisia and the late Cenomanian of France (Thomel 1972).

Subfamily mammitinae Hyatt, 1900
Genus mammites Laube and Bruder, 1887

Type species. Ammonites nodosoides Schluter, 1871, p. 19, pi. 8, figs. 1 4, by monotypy (see Wright and
Kennedy 1981, P- 75).

Mammites spp.

Plate 40, fig. 4

Material. Three fragments (SFB C326-C328). Dimensions could not be determined.

Discussion. Three poorly preserved, variably large fragments of internal moulds of body-chambers
of the genus Mammites are present in the investigated material. Apparently these represent two
different species. In the first form (PI. 40, fig. 4), represented by two dififerent-sized but large
body-chambers, the whorl section is rectangular, little wider than high and the venter is flat. In the
other one the cross-section is higher than wide and the venter is rounded (not figured). Both
forms show the typical ornamentation with large, obliquely projecting ventrolateral nodes,
which are connected with smaller, elongated umbilical bulges by broad, ill-defined ribs. The venter
apparently is smooth. Suture line only faintly visible in one specimen.

Due to the poor preservation the present specimens cannot be determined at the specific level.

Occurrence. Upper Galala Formation: Section 5. Early Turonian.

Genus pseud aspidoceras Hyatt, 1903

Type species. Ammonites footeanus Stoliczka, 1864, p. 101, pi. 52, figs. 1, G-cand2, 2a, by original designation
of Hyatt (1903).

EXPLANATION of plate 40

Figs. I and 2. Fagesia cf. superstes (Kossmat). SFB C323, section 5, upper Galala Formation, early Turonian.
Figs. 3, 6, 8 9. Vascoceras cauvini Chudeau. SFB C480. 3 and 6, outer whorl and 8 and 9, inner whorl of
the same specimen, section 5, upper Galala Formation, late Cenomanian.

Fig. 4. Mammites sp. SFB C326, section 5, upper Galala Formation, early Turonian.

Figs. 5 and 7. V. gamai Choffat. 5, SFB C309; 7, SFB C310. Both specimens from section 5, upper Galala
Formation, late Cenomanian.

All figures x 1 .

PLATE 40

372

PALAEONTOLOGY, VOLUME 32

Pseudaspidoceras sp.

Text-fig. 61

Material. Six fragments of internal moulds (SFB C316-C321). Five of phragmocones and one body-chamber
of half a whorl. Dimensions could not be determined.

Description. Shell very evolute. Whorl section subsquarish with a slightly convex venter in the smaller and a
flat venter in the larger specimens. Large umbilicus with angularly rounded shoulders and steep vertical walls.
Ornamentation consists of elongated ventrolateral nodes and smaller umbilical tubercles. Space in between
them sometimes covered by shallow elevations. Outer ventrolateral nodes not clearly observed due to the
poor preservation. Suture line shown in text-fig. 61.

Discussion. The present specimens cannot safely be determined at the specific level because of their
poor preservation. Flowever, they are similar to P. cf. P. pseudonodosoides (Choffat, 1898) in
Freund and Raab (1969, pi. 1, figs. 10 and 1 1) from the cauvini Zone of the Negev.

Occurrence. Upper Galala Formation: Section 7. Latest Cenomanian.

Family vascoceratidae Douville, 1912
Subfamily vascoceratinae Douville, 1912
Genus fagesia Pervinquiere, 1907

Type species. Olcostephanus superstes Kossmat, 1897, p. 26 (133), pi. 6(17), fig. 1 a-c, by original designation
of Pervinquiere (1907).

Fagesia cf. superstes (Kossmat, 1897)

Plate 40, figs. 1 and 2

cf. 1897 Olcostephanus superstes Kossmat, p. 26 (133), pi. 6 (17), fig. 1 a-c.
cf. 1907 Fagesia superstes Kossmat; Pervinquiere, p. 322, pi. 20, figs. 1 -4A; fig. 122.
cf. 1969 Fagesia cf. F. superstes (Kossmat); Freund and Raab, p. 35, text-fig. If.
cf. 1983 Fagesia superstes (Kossmat); Cobban and Hook, p. 16, pi. 3, figs. 1-2; pi. 13, figs. 6-11;
fig. 12 (see here for complete synonymy).

Material. Three fragments of internal moulds (SFB C323-C325) comprising two phragmocones and one
body-chamber, each about half a whorl. Dimensions could not be determined.

Description. Shell evolute, cadicone. Moderately large umbilicus with angularly rounded shoulders and high
vertical walls. Whorl section broadly rounded, very low, more than twice as wide as high. Ornamentation
consists of large umbilical tubercles, each giving rise to two diverging, coarse ribs which cross the venter
uninterruptedly. Suture line hardly visible.

Discussion. The present specimens very closely resemble F. superstes as figured by Kossmat (1897,
pi. 6, fig. 1) and Pervinquiere (1907, pi. 20, figs. 1-4). However, in the present material the ribs
are coarser and more widely spaced than in typical F. superstes.

Occurrence. Upper Galala Formation: Section 5. Early Turonian.

The species is known from the early Turonian of India, North Africa, and the Middle East (Freund and
Raab 1969) as well as Mexico (Cobban and Hook 1983).

Genus nigericeras Schneegans, 1943

Type species. Nigericeras gignouxi Schneegans, 1943, p. 1 19, pi. 5, figs. 10- 15, by subsequent designation of
Reyment (1955).

Nigericerasl tinrhertense Collignon and Roman, 1981

Text-fig. 7

1981 Nigericeras tinrhertense Collignon and Roman in Amard et al., p. 55, pi. 8, fig. 2.

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

373

text-fig. 7. Nigericeras tinrhertense Collignon and Roman. SFB C312, section 5, upper Galala
Formation, late Cenomanian or earliest Turonian, x 1.

Material. One internal mould (SFB C312). Body-chamber preserved up to three-quarters of last whorl,
phragmocone slightly distorted. Dimensions could not be determined.

Description. Shell evolute, faintly compressed. Medium-sized umbilicus with rounded shoulders and vertical
walls. Whorl section almost subsquarish, only faintly higher than wide. Venter rounded. Phragmocone and

374

PALAEONTOLOGY, VOLUME 32

inner part of body-chamber ornamented by widely spaced coarse ribs crossing the venter, which are almost
straight on the inner and prorsiradiate on the outer flanks. Outer part of body-chamber smooth, as far as
can be observed. Suture line poorly preserved.

Discussion. Although slightly less depressed and larger the present specimen very closely resembles
N. tinrhertense (see Amard et al. 1981). The inner whorls of this species, which was based on a
single specimen, are unknown. Since the present specimen also does not permit further observation
the generic assignment to Nigericeras remains questionable.

Occurrence. Upper Galala Formation: Section 5. Latest Cenomanian or earliest Turonian.

N. tinrhertense was originally described from Zone V of the Tinrhert, Algeria, which was assigned to the
early Turonian by Amard et al. (1981). The fauna of Zone V of these authors very closely resembles that of
the cauvini Zone of Freund and Raab (1969), which is now included in the late Cenomanian by Kennedy et
al. (1987).

Genus vascoceras Choffat, 1898

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

Remarks. Kennedy et al. (1987) emended the definition of Vascoceras and included Paravascoceras
Furon, 1935, Pachyvascoceras Furon, 1935, Paracanthoceras Furon, 1935, Broggiiceras Benavides-
Caceres, 1956, Discovascoceras Collignon, 1957, and Provascoceras Cooper, 1979 as synonyms.
This view is followed herein.

Vascoceras cauvini Chudeau, 1909

Plate 40, figs. 3, 6, 8-9; Plate 41, figs. 1-4; Plate 42, fig. 1; text-figs. 6g, h and 8c.

1909 Vascoceras cauvini Chudeau, p. 67, pi. 1, figs. I a and 2 a\ pi. 2, figs. 3 and 5; pi. 3, figs. 1 b,
2b, 4.

71915 Acanthoceras mantelliil) Sowerby; Greco, p. 207, pi. 18, figs. 1 and 2.

71915 Vascoceras durandi Thomas and Peron; Greco, p. 268, pi. 9, fig. 9.

1921 Thomasites cauvini (Chudeau); Chudeau, p. 463, fig. I.

1933 Vascoceras cauvini Chudeau; Furon, p. 268, pi. 9, fig. 17.

1935 Vascoceras ( Paravascoceras ) cauvini Chudeau; Furon, p. 60, pi. 5, fig. 1.

1943 Paravascoceras cauvini (Chudeau); Schneegans, p. 128, pi. 4, fig. 2; fig. 9 a-f.

1956 Broggiiceras humboldti Benavides-Caceres, p. 470, pi. 56, figs. 3-6.

71957 Paravascoceras aff. cauvini (Chudeau); Barber, p. 37, pi. 14, figs. 2 and 3; pi. 32, figs. 8 and 9.
1969 Paravascoceras cauvini (Chudeau); Freund and Raab, p. 20, pi. 3, figs. 1-3; text-fig. 5 a-b.

1975 Paravascoceras cauvini (Chudeau); Schobel, p. 119, pi. 4, figs. 1-3; pi. 5, figs. 1-4.

71978 Vascoceras ( Paravascoceras ) cf. cauvini Chudeau; Cooper, p. 130, text-figs. 6c h, 35-37.

1981 Paravascoceras cauvini (Chudeau); Amard et al., p. 51, pi. 3, fig. 9.

Material. Twelve specimens (SFB C294-C304, C480). All are internal moulds, four of them with partially
preserved body-chambers of up to three-quarters of a whorl.

D

Wh

Wb

U

SFB C294

91

0-39

0-42

0-28

SFB C295

88

0-43

—

0-26

SFB C296

89

0-41

—

0-24

SFB C297

100

0-43

—

0-27

Description. The whorl section (text-fig. 8c) varies from faintly depressed to compressed with slightly inclined
flanks. Venter well rounded. Umbilicus moderately wide with rounded shoulders and vertical to oblique
walls. Most specimens without ornamentation, as far as preservation permits observation. Only one specimen
shows faint, weak marginal ribs crossing the venter. Suture line poorly preserved in three specimens (see text-
fig. 6g-h).

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES 375

Discussion. The present specimens closely resemble V. cauvini Chudeau, 1909 (pis. 1-3) in size,
shape, and suture line. They differ from Chudeau’s type material in their very weak to absent
ornamentation. According to Schobel (1975), V. cauvini displays variation in the degree of
compression and ornamentation. Therefore, the present specimens mostly represent the unorna-
mented variety in the sense of Schobel (1975). The latter author suggested in his study V. rumeaui
Collignon, 1957 to be a synonym of V. cauvini. In the present material these two forms can clearly
be separated since V. rumeaui shows a considerably broader cross-section and a narrower umbilicus.
According to Berthou et al. (1985, p. 72), V. cauvini is possibly a junior synonym of V. barcoicense
Choffat, 1898.

Occurrence. Upper Galala Formation: Section 5 (SFB C294-C302, C480), section 6 (SFB C304), and section
7 (SFB C303). Late Cenomanian.

V. cauvini is widely known from central and northern Africa, the Middle East, and Peru. The cauvini Zone
of Freund and Raab (1969), originally assigned to the early Turanian, is now placed in the late Cenomanian
by Kennedy et al. (1987).

376 PALAEONTOLOGY, VOLUME 32

Vascoceras cf. cauvini Chudeau, 1909

Plate 42, fig. 2; Plate 43, fig. 3;

i text-figs. 6f and 8b

Material. Two internal moulds of phragmocones
up to half a whorl.

(SFB C305, C306),

one with body-chamber preserved of

Dimensions.

D

Wh

Wb

U

SFB C305 79

0-37

—

0-27

SFB C306 98

0-50

—

0 21

Description. Shell moderately evolute. Whorl section strongly depressed on phragmocone, less depressed on
body-chamber. Moderately large umbilicus with rounded shoulder and high vertical walls. Flanks and venter
well rounded (text-fig. 8b). No ornamentation observed. Initial whorls unknown. Suture line with characteristic
broad V-shaped lateral lobe (text-fig. 6f).

Discussion. In the investigated material V. cf. cauvini differs from the specimens attributed to V.
cauvini and the type material as figured by Chudeau (1909) in its more depressed whorl section at
comparable growth stages. The suture line of V. cf. cauvini falls within the range of variation
attributed to V. cauvini by Chudeau (1921, fig. 1), Schneegans (1943, figs. 9 a-f and 1 1), and Freund
and Raab (1969, fig. 5a). Schobel (1975) attributed a high variability of the whorl section to V.
cauvini , varying from compressed to depressed. According to Berthou et al. (1985, pp. 72, 75), V.
cauvini is more compressed than V. durandi (Thomas and Peron, 1890). Therefore, as the whorl
section of the specimens here referred to as V. cf. cauvini is as much depressed as in the more
evolute species V. durandi , we hesitate to attribute these extremely depressed morphotypes (e.g.
specimen SFB C305) to the generally rather compressed species V. cauvini.

Occurrence. Upper Galala Formation: Section 7. Latest Cenomanian.

Vascoceras durandi (Thomas and Peron, 1890)

Plate 43, figs. I and 2; text-fig. 8a

1890 Pachydiscus durandi Thomas and Peron; Peron, p. 27, pi. 18, figs. 5 8.

1898 Vascoceras amieirensis Choffat, p. 61, pi. 12, figs. 1 and 2; pi. 13, figs. 1 and 2; pi. 21, figs. 17-
21.

1928 Vascoceras sp.; Douville, p. 15, pi. 1, fig. 6.

71957 Discovascoceras cf. amieirense Choffat; Collignon, p. 124.

1969 Vascoceras cf. V. amieirense Choffat; Freund and Raab, p. 32, text-fig. 6 k, /.

1985 Vascoceras durandi (Peron); Berthou et al., p. 72, pi. 4, figs. 4-9; pi. 6, figs. 1-6 (see here for
complete synonymy).

Material. Two internal moulds of phragmocones (SFB C292-C293).

D

Wh

Wb

U

SFB C292

102

0-40

—

0-30

SFB C293

69

0-35

—

0-32

Description. Shell moderately evolute, strongly depressed. Wide umbilicus with angularly rounded shoulders
and steep vertical walls. Whorl section rounded (text-fig. 8a), at first much wider than high, later becoming

EXPLANATION OF PLATE 41

Figs. I 4. Vascoceras cauvini Chudeau. I and 2, SFB C302, fragmentary body-chamber; 3 and 4, SFB C298.

Both specimens from Section 5, upper Galala Formation, late Cenomanian.

Figs. 5 and 6. V. rumeaui (Collignon). SFB C288, section 7, upper Galala Formation, late Cenomanian.

All figures x 1 .

PLATE 41

'**&■'*,

LUGER and GROSCEIKE, Vascoceras

378

PALAEONTOLOGY, VOLUME 32

less depressed. No ornamentation observed even at growth stages of 40 mm diameter (smaller specimen).
Suture line only rudimentarily preserved with a large first lateral saddle and two successively smaller accessory
saddles, the inner one being situated immediately at the umbilical shoulder.

Discussion. The present specimens are very close to V. amieirensis Choffat, 1898 (p. 61, pi. 12,
fig. 1 a-b) as well as to the specimen figured as Vascoceras sp. by Douville (1928, pi. 1, fig. 6).

In a revision of the Portuguese species of Vascoceras , Berthou et al. (1985, p. 72) regarded the
species V. douvillei Choffat, 1898 and V. amieirensis as junior synonyms of V. durandi. The authors
interpreted the presence or absence of umbilical tubercles in the earlier growth stages of V. durandi
as being related to geographical variation. Therefore, as far as can be observed, the present
specimens may represent a smooth variety of V. durandi.

V. durandi is very similar to V. cauvini Chudeau, 1909, from which it differs in its larger umbilicus
and more depressed whorl section at comparable growth stages. For detailed discussion of V.
durandi see Berthou et al. (1985, p. 72).

Occurrence. Upper Galala Formation: Section 6 (SFB C293) and section 7 (SFB C292).

The species is known from the late Cenomanian to early Turonian of Choffat’s beds ‘G’ to ‘L’ in Portugal
(Berthou et al. 1985). In Israel it is reported only from the early Turonian pioti and quaasi Zones (Freund
and Raab 1969). V. durandi is also known from Spain, Algeria, Tunisia, Japan and questionably Angola,
Mexico, and Brasil (Berthou et al. 1985).

Vascoceras gained Choffat, 1898

Plate 40, figs. 5 and 7; text-fig. 6c

1898 Vascoceras gamai Choffat, p. 54, pi. 7, figs. 1-4; pi. 8, fig. 1; pi. 10, fig. 2; pi. 21, figs. 1-4.
1898 Vascoceras gamai var. subtriangularis Choffat, p. 55, pi. 7, fig. 5; pi. 21, fig. 5.

1898 Vascoceras adonensis Choffat, p. 59, pi. 9, fig. 3; pi. 21, fig. 12.

1898 Vascoceras mundae Choffat, p. 56, pi. 8, figs. 2-4; pi. 10, fig. 1; pi. 21, figs. 6-8, 10.

1898 Ammonites (Vascoceras?) grossouvrei Choffat, p. 68, pi. 9, figs. 1 and 2; pi. 22, figs. 37 and 38.
1898 Vascoceras cf. barcoisensis Choffat, pi. 16, fig. 11; pi. 22, fig. 36 (non Vascoceras barcoisensis
Choffat, p. 67, pi. 17, fig. 1; pi. 22, fig. 35).

1928 Vascoceras gamai Choffat; Douville, p. 13, fig. 3; pi. 1, fig. 4.

1975 Vascoceras gamai Choffat; Berthou et al ., p. 81.

1981 Vascoceras gamai Choffat; Wright and Kennedy, p. 86.

71981 Vascoceras gamai Choffat; Amard et al ., p. 102.

1985 Vascoceras gamai Choffat; Berthou et al., p. 66, pi. 2, figs. 1-12; pi. 3, figs. 1-3, 5-7, 10, 13-14.

Material. Five internal moulds of phragmocones (SFB C307-C311), all in rather poor preservation.
Dimensions could not be determined.

Description. Shell moderately evolute, slightly compressed. Moderately large umbilicus with rounded shoulders
and vertical walls. Whorl section rounded, with rounded venter and converging flanks. Ornamentation
consists of five to six strong umbilical nodes from which one or two diverging, slightly sinous prorsiradiate
ribs arise which strengthen while crossing the venter. Minor ribs, arising on the flanks, are intercalated in
irregular distances. Suture line see text-fig. 6c.

Discussion. In a revision of Choffat’s (1898) vascoceratids, based on morphometrical studies,
Berthou et al. (1975) regarded V. gamai var. subtriangularis , V. adonensis, V. grossouvrei, and V.

EXPLANATION OF PLATE 42

Fig. 1. Vascoceras cauvini Chudeau. SFB C294, specimen with partly preserved body-chamber, section 5,
upper Galala Formation, late Cenomanian.

Fig. 2. V. cf. cauvini Chudeau. SFB C305, section 7, upper Galala Formation, late Cenomanian.

Figs. 3 and 4. V. rumeaui (Collignon). SFB C289, section 7, upper Galala Formation, late Cenomanian.

All figures x 1.

PLATE 42

LUGER and GROSCHKE, Vascoceras

380

PALAEONTOLOGY, VOLUME 32

mundae as synonyms of V. gamai (see also Berthou et al. 1985). This concept is followed herein
and adopted in synonymy. The specimen figured by Douville (1928, fig. 3; pi. 1, fig. 4) is almost
identical to our forms. In the present material V. gamai is represented only by juvenile specimens.

Occurrence. Upper Galala Formation: Section 5 (SFB C307, C309, C310) and section 6 (SFB C308, C311).
Late Cenomanian.

In its Portuguese type area V. gamai is known from Choffat’s beds ‘E’ to ‘L’ (late Cenomanian), but is
most abundant in bed ‘F’ ( geslinianum Zone, late Cenomanian, see Berthou et al. 1985, p. 70). Furthermore,
it is known from the late Cenomanian to early Turonian of North Africa (Douville 1928; Amard et al. 1981;
Collignon 1957, 1965) and according to Berthou et al. (1985) from Spain, Mexico, and Brasil.

Vascoceras rumeaui (Collignon, 1957)

Plate 41, figs. 5 and 6; Plate 42, figs. 3 and 4; text-fig. 8d
1957 Paravascoceras rumeaui Collignon, p. 10, pi. 1, fig. 2.

1969 Paravascoceras rumeaui Collignon; Freund and Raab, p. 21, pi. 3, figs. 4 and 5; text-fig. 5c, cl.

Material. Four specimens (SFB C288-C291). Two of them almost complete internal moulds, in which the
body-chamber is preserved up to a length of half a whorl.

Dimensions.

SFB C288
SFB C289

D Wh

NO 0-52

111 0-45

Wb U
0-50 0-22

0-20

Description. Shell moderately evolute, weakly depressed. Deep umbilicus fairly wide with rounded shoulders
and vertical walls. Whorl section rounded and wider than high on phragmocone and inner part of body-
chamber; on outer part of body-chamber suboval with faintly flattened outer flanks and almost as wide as
high (text-fig. 8d). Ornamentation consists exclusively of strong fold-like prorsiradiate ribs on body-chamber,
which are strongest on outer flanks and venter and weaken on the inner flanks. Suture line only poorly
preserved.

Discussion. V. rumeaui apparently varies in the degree of inflation. Whereas the present specimens
closely resemble the holotype in the height/width ratio (see Collignon 1957, p. 10, pi. 1, fig. 2),
Freund and Raab (1969, pi. 3, figs. 4 and 5; text-fig. 5c) figured and described a more depressed
form. V. rumeaui differs from V. chevalieri Furon, 1935, which has a similar cross-section and
coiling-mode, in the lack of ornamentation on the phragmocone and the inner part of the body-
chamber. V. rumeaui was regarded as a junior synonym of V. cauvini Chudeau, 1909 by Schobel
(1975). We cannot follow Schobel’s view since in the present material V. rumeaui differs from V.
cauvini in its considerably more inflated cross-section and the narrower umbilicus. For comparison
with other species see Freund and Raab (1969, p. 23).

Occurrence. Upper Galala Formation: Section 5 (SFB C290, C291) and section 7 (SFB C288, C289). Latest
Cenomanian or (?)earliest Turonian.

V. rumeaui was originally described from Libya, where it was assigned to the early Turonian (Collignon
1957). It is reported from the cauvini Zone of the southern Negev (Freund and Raab 1969), which is now
assigned to the latest Cenomanian by Kennedy et al. (1987).

EXPLANATION OF PLATE 43

Figs. 1 and 2. Vascoceras durandi (Thomas and Peron). SFB C292, section 7, upper Galala Formation, basal
Turonian or ?latest Cenomanian.

Fig. 3. V. cf. cauvini Chudeau. SFB C306, section 7, upper Galala Formation, late Cenomanian.

Fig. 4. Thomasites compressus (Barber). SFB C314, section 5, upper Galala Formation, late Cenomanian or
earliest Turonian.

All figures x 1.

PLATE 43

LUGER and GROSCHKE, Vascoceras, Thomasites

382

PALAEONTOLOGY, VOLUME 32

Subfamily pseudotissotiinae Hyatt, 1903
Genus thomasites Pervinquiere, 1907

Type species. Pachy discus rollandi Peron, 1890, p. 25, pi. 17, figs. 1-3, by original designation of Pervinquiere
(1907).

Remarks. Wright and Kennedy (1981, p. 98) discussed the genera Gombeoceras Reyment, 1954,
Koulabiceras Atabekjan, 1966, and Ferganites Stankievich and Pojarkova, 1969 and regarded them
as synonyms of Thomasites Pervinquiere, 1907.

Thomasites compressus (Barber, 1957)

Plate 43, fig. 4; text-figs. 6b and 8e

1957 Gombeoceras gongilense compressum Barber, p. 41, pi. 19, figs. 2 and 5; pi. 33, figs. 15 and 16.
1978 Gombeoceras compressum Barber; Offodile and Reyment, p. 59, fig. 33.

Material. Two internal moulds (SFB C314-C315), one of them with parts of the body-chamber.

Dimensions.

D Wh Wb U
SFBC314 77 0 55 0-39 0 12

Description. Shell involute, compressed. Umbilicus small with angularly rounded shoulders and vertical walls.
Venter narrow and tabulate on inner to subrounded on outer whorl (text-fig. 8e). Flanks fiat on inner to
slightly convex on outer whorl. Ornamentation only faintly visible, consisting exclusively of minor ventrolateral
tubercles. Suture line see text-fig. 6b.

Discussion. The present specimens very closely resemble the type material as figured by Barber
(1957, pi. 19, figs. 2 and 5; pi. 33, figs. 15 and 16). They differ slightly from it in their more convex
flanks and the weaker ornamentation. The lack of feeble ribs in the present material may be due
to wind erosion. The species was originally assigned to Gombeoceras , which is now placed under
synonymy with Thomasites (see Wright and Kennedy 1981). The assignment of the species to
Thomasites remains somewhat doubtful because it is unknown whether it bears umbilical tubercles
on the early whorls.

Occurrence. Upper Galala Formation: Section 5. Latest Cenomanian or earliest Turonian.

The species is known from Nigeria (Barber 1957; Offodile and Reyment 1978) where species of ‘ Gom-
beoceras’ Reyment, 1954 (now Thomasites ) are restricted to the costatum Zone (Barber 1957). The cosiatum
Zone is placed in the early Turonian by Popoff et al. (1986) and in the late Cenomanian by Kennedy et al.
(1987).

Thomasites cf. subtenue (Reyment, 1954)
cf. 1954 Gombeoceras subtenue Reyment, p. 261, pi. 4, fig. 4; text-fig. 3/

cf. 1957 Gombeoceras gongilense subtenue Reyment; Barber, p. 43, pi. 19, fig. 4; pi. 20, fig. 3; pi. 24,
fig. 3; pi. 33, figs. 13 14, 17.

Material. One poorly preserved internal mould with body-chamber of little more than half a whorl (SFB
C313). Dimensions could not be determined.

Description. Shell involute, compressed. Small umbilicus with angularly rounded shoulders and steep walls.
Cross-section highly arched with rounded venter and almost flattened flanks. Ornamentation consists of weak
ribs rising on the outer flanks, crossing the venter. Ribs thicken ventrolaterally to form elongated short
nodes. Ornamentation on outer part of body-chamber not observable. Suture line only poorly preserved.

Discussion. The present specimen closely resembles ‘ Gombeoceras ’ subtenue Reyment, 1954. It is,
however, slightly more compressed than Reyment’s holotype. The ornamentation of the juvenile
stage is not observable due to the poor preservation. T. subtenue differs from T. compressus (Barber,
1957) in its rounded venter and stronger ornamentation.

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

383

Occurrence. Upper Galala Formation: Section 7. Latest Cenomanian.

T. subtenue was originally described from Nigeria (Reyment 1954; Barber 1957); see also T. compressus.

Family tissotiidae Hyatt, 1900
Subfamily tissotiinae Hyatt, 1900
Genus metatissotia Hyatt, 1903

Type species. Ammonites fourneli Bayle, 1849, p. 360, pi. 17, figs. I 5, by subsequent designation of Roman
(1938).

Metatissotia fourneli (Bayle, 1849)

Plate 45, figs. 2 and 3; text-fig. 10f-g
1849 Ammonites fourneli Bayle, p. 360, pi. 17, figs. 1-5.

1897 Tissotia fourneli Bayle emend. Thomas et Peron; Peron, p. 59, pi. 10, figs. I 8; pi. 17, figs. 9
and 10.

1897 Tissotia grossouvrei Peron, p. 70, pi. 16, figs. 1 and 2; pi. 18, fig. 17.

1903 Metatissotia fourneli (Bayle); Hyatt, p. 45.

1903 Paratissotia grossouvrei (Peron); Hyatt, p. 50.

1907 Tissotia fourneli Bayle; Pervinquiere, p. 372, pi. 26, fig. 5.

71915 Tissotia fourneli Bayle; Greco, p. 225, pi. 21, fig. 6.

1956 Tissotia fourneli (Bayle); Benavides-Caceres, p. 480, pi. 62, figs. 3 and 4.

Material. Ten internal moulds (SFB C345-C354). Five specimens with partly preserved body-chamber of up
to half a whorl. Diameter of largest specimen is 162 mm.

D

Wh

Wb

u

SFB C351

76

0-47

—

015

SFB C354

80

0-55

—

014

Description. Shell involute, rather strongly depressed (text-fig. 10f, g). Pronounced sharp keel which
disappears slowly on body-chamber. Umbilical shoulder rounded. Ornamentation in general not strongly
developed. Consists at first of radial ribs, beginning on umbilical node-like undulations and ending at
ventrolateral, sometimes clavi-like, nodes. Single intercalatory ribs are present, also ending in ventrolateral
nodes. Later ribs disappear, finally on body-chamber only in the umbilical region vague coarse radial folds
are developed. Suture line only poorly preserved.

Discussion. M. fourneli apparently varies in the degree of depression of the shell. Whereas Peron
(1897, pi. 10, figs. 1-8; pi. 17, figs. 9 and 10) figured relatively compressed specimens, Pervinquiere
(1907, p. 372, pi. 26, fig. 5) reported a more depressed form. Also the measurements of additional,
unfigured specimens given by Pervinquiere (1907, p. 372) were derived from depressed forms. The
measurements of the present specimens closely resemble those given by Pervinquiere (1907). The
whorl sections of the studied specimens also resemble Tissotia grossouvrei Peron, 1897 (pi. 16,
figs. 1 and 2). T. grossouvrei differs from T. fourneli only in its slightly weaker ornamentation,
so that, as already mentioned by Pervinquiere (1907, p. 374), a differentiation of the two forms
cannot be justified. M. fourneli differs from the likewise depressed Subtissotia africana (Peron,
1897) in the presence of umbilical sculptural elements. M. nodosa Hyatt, 1903 differs from M.
fourneli in the presence of fiexuous ribs. M. ewaldi (von Buch, 1848) sensu Kennedy (19846) is
much more compressed, has a fastigiate venter, and generally shows a weaker ornamentation than
M. fourneli.

Occurrence. Hawashya Formation: Lower transgressive sequence, section 7. Middle or late Coniacian.

M. fourneli is known from Algeria and Tunisia, poorly dated as Coniacian or ‘lower Senonian’ (Peron
1897; Pervinquiere 1907), from the Coniacian of Peru (Benavides-Caceres 1956), and is reported from the
late Coniacian of the Middle East (Lewy and Raab 1978).

384

PALAEONTOLOGY, VOLUME 32

Metatissotia cf. ewaldi (v. Buch, 1848)

Plate 44, figs. 1 and 2; text-fig. 10e
cf. 1848 Ammonites ewaldi von Buch, p. 221, pi. 1, fig. 4.

cf. 19846 Metatissotia ewaldi (von Buch); Kennedy, p. 127, pi. 28, figs. 4 and 5; pi. 29, figs. 9 11; pi. 30,
figs. 1 -2, 5-6, 8-9, 12; pi. 32, figs. 1 3; text-fig. 40B, E (see here for further synonymy).

Material. One adult specimen with partly preserved body-chamber (SFB C355).

Dimensions.

D Wh Wb U
SFB C355 131 0-53 - 0 06

Description. Very involute compressed shell. Fhnbilical shoulder rounded with a steep wall. Venter nearly
fastigiate, slowly becoming rounded on body-chamber. On phragmocone ornamentation consists of faint
marginal ribs which end in ventrolateral clavi. No ornamentation visible on body-chamber. Suture line only
partly preserved (see text-fig. 1 0e).

Discussion. The present specimen closely resembles the one figured by Kennedy (19846, pi. 29, figs.
9-11). The author notes (Kennedy 19846, p. 127): The line of the clavi is marked by a distinct
facet so that the very narrow angle of the projected flanks is replaced by a less acute, fastigiate
venter.’ This ’facet’, which is clearly visible almost on all specimens figured by him, cannot be
observed on the investigated specimen. Therefore, the present specimen cannot be assigned
unequivocally to the species M. ewaldi (von Buch).

Occurrence : Hawashya Formation: Lower transgressive sequence, section 7. Middle or late Coniacian.

M. ewaldi is known from the middle Coniacian of France, northern Spain, and Austria (Kennedy 19846).
The species is also reported from the late Coniacian of Sinai and Israel (Lewy and Raab 1978).

Metatissotia sp.

Plate 45, figs. 4 and 5; text-fig. 10h
Material. One internal mould of phragmocone (SFB C356).

Dimensions.

D Wh Wb U
SFBC356 81 0-54 0-68

Description. Shell inflated with narrow umbilicus. Whorl section broadly rounded, venter occupied by strong,
clearly set-off keel (see text-fig. 1 Oh). Ornamentation consists of few umbilical nodes and more numerous
node-like elevated ventrolateral clavi. The relation between the number of umbilical nodes and clavi is not
clear due to the poor preservation of the specimen. Each umbilical node is connected with one of the clavi
by a short straight fold-like rib. Short intercalatory ribs arising from the clavi are present. Suture line not
preserved.

Discussion. The present specimen cannot be assigned to a certain species of Metatissotia due to
the strong inflation of phragmocone. It differs from the likewise inflated S. africana (Peron, 1897)
in the presence of umbilical nodes.

Occurrence: Hawashya Formation: Lower transgressive sequence, section 7. Middle or late Coniacian.

EXPLANATION OF PLATE 44

Figs. 1 and 2. Metatissotia cf. ewaldi (von Buch). SFB C355.

Figs. 3 and 4. Subtissotia africana (Peron). SFB C357.

Both species from section 7, lower transgressive sequence of Hawashya Formation, middle or late Coniacian,

x 1.

PLATE 44

mmmi

LUGER and GROSCHKE, Metatissotia, Subtissotia

386 PALAEONTOLOGY, VOLUME 32

Genus subtissotia Hyatt, 1903

Type species. Tissotia tissoti var. inflata Peron, 1897, p. 68, pi. 12, tig. 6, by original designation of Hyatt
(1903).

Subtissotia africana (Peron, 1897)

Plate 44, figs. 3 and 4; Plate 45, fig. 1; text-fig. 9

1897 Tissotia ewaldi de Buch var. africana Peron, p. 63, pi. 11, figs. 5 and 6; pi. 17, fig. 12 (non
pi. 11, figs. 1 -4; pi. 17, fig. 1 1).

1903 Subtissotia africana (Peron); Hyatt, p. 44.

Lectotype. Peron (1897, pi. 11, figs. 5 and 6, lectotype designated here, see below).

Material. Four internal moulds (SFB C357-C360). Body-chamber preserved up to half a whorl in two
specimens.

text-fig. 9. Subtissotia afri-
cana (Peron). SFB C360, sec-
tion 7, lower transgressive
sequence of Hawashya Forma-
tion, middle or late Coniacian,
x 1.

EXPLANATION OF PLATE 45

Fig. 1. Subtissotia africana (Peron). SFB C358.

Figs. 2 and 3. Metatissotia fourneli (Bayle). SFB C354.

Figs. 4 and 5. Metatissotia sp. SFB C356.

All from section 7, lower transgressive sequence of Hawashya Formation, middle or late Coniacian, x 1.

PLATE 45

LUGER and GROSCE1KE, Subtissotia , Metatissotia

388

PALAEONTOLOGY, VOLUME 32

D

Wh

Wb

U

SFB C357

88

0-52

—

010

SFB C360

115

0-64

—

007

Description. Shell involute. Umbilical shoulder rounded. Whorl section variable, moderately depressed. Flanks
flat to rounded. Phragmocone with sharp, pronounced keel, body-chamber rounded in ventral aspect, without
keel. Ornamentation in general weak, at first with ventrolateral clavi situated close to the keel, from which
short, vague broad radial folds arise, terminating in the upper third of the height of whorl. Later, only vague
broad radial folds are to be recognized. The body-chamber is smooth, as far as can be observed due to the
state of preservation. In general, no sculptural elements are visible on the flanks or the umbilical region.
Suture line only poorly preserved.

Discussion. The three specimens figured by Peron (1897, pi. 11, figs. 1-6) differ considerably in
shape. Therefore, their attribution to one and the same species may be doubted. A small, strongly
depressed form (Peron 1897, pi. 11, figs. 5 and 6) and two more compressed forms (Peron 1897,
pi. II, figs. 1-4) face one another. A holotype was not designated by Peron (1897). Hyatt (1903),
although he differentiated Tissotia on subgeneric level and redescribed the species erected by Peron
(1897), did not define a lectotype. The present material closely resembles the specimen figured by
Peron (1897, pi. 11, figs. 5 and 6) as regards ornamentation and shape. Therefore, we would like
to suggest this specimen as the lectotype of S. africana (Peron, 1897) and exclude the other figured
specimen (Peron 1897, pi. 11, figs. 1-4) from this species.

5. africana differs from the likewise depressed M. fourneli (Bayle, 1849) in its lack of any
umbilical sculptural elements. It can be distinguished from S. inflata (Peron, 1897) and S. intermedia
(Peron, 1897) by the absence of the keel-like elongated ventral ridges and the less inflated
phragmocone. M. ewaldi (von Buch, 1848) is more compressed than S. africana.

Occurrence. Hawashya Formation: Lower transgressive sequence, section 7. Middle or late Coniacian.

The species was originally described from Tunisia where it was tentatively assigned to the ‘lower Senonian’
by Hyatt (1903).

Family coilopoceratidae Hyatt, 1903
Genus coilopoceras Hyatt, 1903

Type species. Coilopoceras colled Hyatt, 1903, p. 91, pi. 10, tigs. 5-21; pi. 11, fig. 1, by original designation.

Coilopoceras requienianum (d’Orbigny, 1841)

Plate 46, figs. 1-3; text-figs. 6a, e, 11, 12, 13a-c

1841 Ammonites requienianum d’Orbigny, p. 315, pi. 93, figs. 1-4.

1903 Coilopoceras requienianum (d’Orbigny); Hyatt, p. 99.

71975 Coilopoceras sinaiense Lewy, p. 42, pi. 1, fig. 3; fig. 1 5i l.

71975 Coilopoceras muldcostatum Lewy, p. 42, pi. 1, figs. 1 and 2; fig. 15a-d, h.

1984 Coilopoceras requienianum (d’Orbigny); Kennedy and Wright, p. 282, pis. 35-36; text-figs. 1-5
(see here for further synonymy).

Material. Sixteen internal moulds (SFB C329-C344). Five of them with parts of body-chamber preserved up
to a length of three-quarters of a whorl.

EXPLANATION OF PLATE 46

Figs. 1-3. Coilopoceras requienianum (d’Orbigny). SFB C331, smooth specimen. 2, view on side affected by
wind erosion, umbilicus enlarged by weathering. 3, view on opposite side, where sediment was removed
by manual preparation, umbilicus in natural size. Section 7, upper transgressive sequence of Umm Omeiyed
Formation, late Turonian, x I.

PLATE 46

LUGER and GROSCHKE, Coilopoceras

390

PALAEONTOLOGY, VOLUME 32

text-fig. 10. a-c, external sutures and whorl section of Manambolites piveteaui Hourcq. a, SFB C396, x 2-7;
b, SFB C404, x 2-7; c, SFB C390, x 0-9. d, whorl section of Libycoceras sp. ex gr. L. ismaeli (Zittel), SFB
C432, xO-9. e, suture of Metatissotia cf. ewaldi (Buch), SFB C355, x L9. f, g, whorl sections of M.fourneli
(Bayle). f, SFB C350, x 0-9; G, SFB C353, keel eroded on outer whorl, x 0-9. h, Metatissotia sp. SFB C356,

x 0-9.

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

391

Dimensions.

SFB

SFB

SFB

D

Wh

Wb

U

C329

113

0-55

0-38

005

C330

165

0-56

0-32

008

C33 1

99

0-58

0-32

005

Description. In the investigated specimens two morphotypes, a smooth and a ribbed form of the species,
occur. In the smooth form the shell is involute, oxycone. Whorl section lanceolate with a sharpened venter,
maximum width close to umbilicus (text-fig. 13c). Umbilicus small, with rounded shoulders and almost
vertical walls. The ribbed form is less compressed and the maximum width tends to be situated at the middle
of the flanks (text-fig. I 3a, b). Venter sharpened on phragmocone, becoming blunt on body-chamber.
Ornamentation consists of almost straight, low ribs on inner whorls of phragmocone, which grade into low
radial folds on the outer whorls of the phragmocone and early part of body-chamber. Single inlercalatory
ribs present. Ribs and folds are best developed on the middle flanks; ventral region is smooth. Suture line
observed only in a few specimens (see text-fig. 6a, e).

text-fig. 1 1 . Coilopoceras
requienianum (d'Orbigny).
SFB C329, ribbed speci-
men, section 7, upper trans-
gressive sequence of Umm
Omeiyed Formation, late
Turonian, xl.

Discussion. Among the better preserved specimens of the investigated material the largest specimen
of the smooth morphotype attain a maximum diameter of 96 mm. The ribbed specimens attain
larger diameters (up to 220 mm).

The studied specimens agree well with the description and illustrations of Kennedy and Wright
(1984, p. 283). In their redescription of the type-material of d’Orbigny (1841) these authors
suggested the existence of sexual dimorphism in C. requienianum which is expressed as the existence
of both smooth oxycones and ribbed, less compressed forms. This phenomenon is also reported
from other species of Coilopoceras (Cobban and Hook 1980, p. 1 1).

392

PALAEONTOLOGY, VOLUME 32

The specimens figured by Lewy (1975) as C. sinaiense sp. nov. and C. multicostatum sp. nov.
from the requienianum Zone of the Eastern Desert and Sinai show a close similarity with C.
requienianum. Because of the poor state of preservation of the figured specimens, these species can
only questionably be taken into synonymy with C. requienianum.

SfiSSt

Iff:

ittft

..;.,U7'YY ;

,f ' , , r

- ■

' « ■■ ,

, T ' \ 1 '

text-fig. 12. Coilopoceras requienianum (d’Orbigny). SFB C330, ribbed specimen, section 7, upper transgress-
ive sequence of Umm Omeiyed Formation, late Turonian, x 1.

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

393

Occurrence. Umm Omeiyed Formation: Upper transgressive sequence of Umm Omeiyed Formation, section
7. Late Turonian. The present specimens of C. requienianum are mainly from two horizons of glauconitic
calcareous sandstones, separated by about 8 m thick intercalations of clays, marls, and marly limestones in
which only a few specimens have been recovered. Indeterminable fragments of Coilopoceras have also been
found in the same stratigraphic position of section 4.

The species is known from the late Turonian of France and Germany (Kennedy and Wright 1984) and
probably the Sinai (Lewy 1975).

Family sphenodiscidae Hyatt, 1900
Genus manambolites Hourcq, 1949

Type species. Manambolites piveteaui Hourcq, 1949, p. Ill, pi. 3, fig. 1; figs. 20-22, by monotypy.

Manambolites piveteaui Hourcq, 1949
Plate 47, figs. 1-5; Plate 48, figs. 1 and 2; text-fig. 10a-c

C331. d, e, Neolobites sp. d, SFB C263;
e, SFB C252. All xO-67.

394

PALAEONTOLOGY, VOLUME 32

1949 Manambolites piveteaui Hourcq, p. 1 1 1, pi. 3, fig. 1; figs. 20-22.

71953 Coahuilites ( Mzezzemceras ) pervinquieri Basse, p. 866, pi. 27, fig. 2.

1959 Praelibycoceras sp. 1; Faris and Hassan, pi. 2, fig. 1.

1959 Praelibycoceras sp. 2; Faris and Hassan, pi. 2, fig. 2.

71959 Libycoceras ismaeli var. safagensis Faris and Hassan, p. 194, pi. 2, figs. 4 and 8.

1970 Manambolites piveteaui Hourcq; Collignon, p. 70, fig. 2341.

Material. Thirty-seven specimens (SFB C385-C421). Body-chamber of up to half a whorl preserved in ten
specimens. Numerous fragments.

D

Wh

Wb

u

SFB C385

41

0-53

0 31

—

SFB C389

133

0-55

0-28

—

SFB C393

71

0-61

0-28

—

SFB C399

74

0-62

0-31

—

SFB C404

46

0-59

0-27

—

SFB C406

123

0-57

0-28

—

SFB C407

91

0-58

0-26

—

Description. Shell involute, oxycone, umbilicus extremely narrow (text-fig. 10c). Venter on phragmocone with
very acute, sharp keel on body-chamber, broadening, becoming oval with smooth keel, fastigiate throughout.
Ornamentation smooth, of varying intensity on different specimens. In early stages with short smooth, sickle-
shaped ventrolateral riblets, which do not reach the venter. Later ventrolateral part of riblets is strongly bent
forward to form ventrolateral clavi, which persist, but weaken, on body chamber. At the same growth stage
a row of lateral tubercles of increasing intensity arise; lateral tubercles become smooth or vanish on body-
chamber of large specimens. Fine sinuous growth striae are visible throughout on well-preserved specimens.
Suture-line see text-fig. 10a-b.

Discussion. Among the present specimens assigned to this species, two morphologic varieties can
be recognized. The smaller variety, form A (see PI. 47, fig. 5; PI. 48, fig. 1), is characterized by a
broadening of the venter and a relatively strong ornamentation at an early stage. In specimens of
form A the body-chamber is preserved and occupies at least one-half of the final whorl. These
specimens do not exceed 91 mm in diameter; most of them have diameters between 70 and 80 mm.
The second variety, form B (see PI. 47, figs. 1-4), shows a weak ornamentation and a sharp keel
up to a very late stage and the lateral tubercles are only faintly developed or absent. Form B is
the large variety with diameters up to at least 123 mm (incomplete specimen). As both varieties
are from one horizon and locality, show identical suture lines and mode of development, sexual
dimorphism is assumed with the existence of a stronger ornamented microconch and a smoothly
ornamented macroconch.

M. piveteaui differs from M. dandensis Howarth, 1965 from the late Campanian of Angola,
which shows a very similar ornamentation, in details of the suture line (see Howarth, 1965, p.
397). Coahuilites ( Mzezzemceras ) pervinquieri Basse, 1953 may represent a slightly stronger
ornamented variety of M. piveteaui , as already mentioned by Howarth (1965, p. 397). The genus
Praelibycoceras Douville, 1912 from Tunisia, which is often cited in the Egyptian literature, is
considered to be a synonym of Eulophoceras Hyatt, 1903 (Arkell et ah 1957). Nevertheless, it is
very likely that most of the Egyptian forms attributed to Praelibycoceras belong to M. piveteaui.

EXPLANATION OF PLATE 47

Figs. 1-5. Manambolites piveteaui Hourcq. 1 and 2, SFB C389, almost smooth adult retaining keel until a
late stage (form B). 3 and 4. SFB C396, almost smooth phragmocone with sharp keel (form B). 5, SFB
C410, specimen showing relatively strong ornamentation, venter on body-chamber rounded (form A).

All specimens from section 2, basal phosphate horizon of Rakhiyat Formation, late middle Campanian, x 1.

PLATE 47

LUGER and GROSCHKE, Manambolites

396 PALAEONTOLOGY, VOLUME 32

□ CoBhullltes (Mzezzemceras) pervlnqulerl BASSE 1953 ☆ Manembolltes piveteaui HOURCQ 1949

• Manembolltes dandensls HOWARTH 1965 A Manembolltes plveteaul HOURCQ present material, sharp keel

▲ Manembolltes plveteaul HOURCQ present material, keel broadening early

text-fig. 14. Growth parameters of Manambolites piveteaui Hourcq and related forms.

Occurrence. Rakhiyat Formation: Unit 1, section 2. Late middle Campanian.

According to Collignon (1970) the species is known from the late middle Campanian of Madagascar (‘Zone
of Delawarella subdelawarensis and Australiella australis').

Genus libycoceras Hyatt, 1900

Type species. Sphenodiscus ismaelis Ziltel, 1884, p. 451, fig. 631, by original designation of Hyatt (1900).

Libycoceras sp. ex gr. Libycoceras ismaeli (Zittel, 1884)

Plate 48, figs. 3-6; text-fig. 10d

EXPLANATION OF PLATE 48

Figs. I and 2. Manambolites piveteaui Hourcq. SFB C390, specimen showing relatively strong ornamentation,
venter on body-chamber rounded (form A), section 2, basal phosphate horizon of Rakhiyat Formation,
late middle Campanian.

Figs. 3-6. Libycoceras sp. ex gr. Libycoceras ismaeli (Zittel). 3 and 6, SFB C435; 4 and 5, SFB C432. Both
specimens from section 3 (Gebel Qreiya), upper Rakhiyat Formation, late Campanian.

All figures x 1 .

PLATE 48

V ■ k
.... ■ >. ”
\ A

mmM

|f||p

LUGER and GROSCHKE, Manambolites, Libycoceras

398 PALAEONTOLOGY, VOLUME 32

Material. Five fragmentary internal moulds of body-chambers (SFB C432-C436). Dimensions could not be
determined.

Description. Shell involute, compressed. Umbilical region not preserved. Ornamentation consists of an inner
row of tubercles and an outer row of nodes and broad undulations. The inner row is formed by small pointed
tubercles, situated at about the middle of the flanks. The outer, ventrolateral row consists of strong clavi,
which are less numerous than the inner tubercles. The area in between both rows is occupied by shallow
undulations, which arise from the clavi and vanish before reaching the tubercles. The area between lateral
tubercles and umbilicus is smooth. Whorl section high, subhexagonal, greatest width at about mid-height,
where the row of lateral tubercles is situated (see text-fig. 10d). Venter acutely rounded with a faint keel;
nearly tabulate between the clavi. The tabulate venter is more pronounced on the outer part of the body-
chamber. Suture line not well preserved.

Discussion. The present specimens closely resemble L. ismaeli (Zittel, 1884), from which they differ
in showing an almost tabulate venter between the ventrolateral nodes. In true L. ismaeli the venter
is keeled throughout the phragmocone in specimens which are larger than the present material, as
figured by Quaas (1902, pi. 30, fig. 1).

Occurrence. Rakhiyat Formation: Unit 3, about 1 m below the assemblage of heteromorphic ammonites,
section 3. Late Campanian.

Suborder ancyloceratina Wiedmann, 1966
Superfamily turrilitacea Gill, 1871
Family baculitidae Gill, 1871
Genus baculites Lamarck, 1799

Type species. Baculites vertebralis Lamarck, 1801, p. 103, by subsequent designation of Meek (1876).

Baculites cf. ovatus Say, 1820
Plate 49, fig. 2; text-fig. 1 5i o
cf. 1820 Baculites ovata Say, p. 41.

cf. 1962 Baculites ovatus Say; Reeside, p. 113, pi. 68, figs. I 4.

cf. 1974 Baculites ovatus Say; Cobban, p. 3, pi. 1, figs. I 32; pi. 2, figs. 1 14; pi. 3, figs. 1-6, 9-11;
fig. 4 (see here for further synonymy).

Material. Seventeen fragments (SFB C427). Dimensions could not be determined.

Discussion. The present material consists of small, poorly preserved body-chamber fragments of
different growth stages. Suture line preserved in three specimens (see text-fig. 15o). The largest
fragment has a length of 75 mm, a height of 22 mm, and a width of 17-5 mm at its broader and
of 19 mm and 14-5 mm at its smaller end, respectively. The investigated specimens show an almost
perfect oval cross-section (see text-fig. 15i-n) and their surface is smooth to very faintly ribbed.
Thus they closely resemble B. ovatus , but due to their poor preservation, they can only be compared
with this species.

EXPLANATION OF PLATE 49

Fig. 1. Baculites subanceps Haughton. SFB C470.

Fig. 2. B. cf. ovatus Say. SFB C427.

Figs. 3 and 4. Solenoceras humei (Douville). 3, SFB C467; 4, SFB C468.

Figs. 5, 9-10. Nostoceras (Nostoceras) sp. 5, SFB C452; 9 and 10, SFB C437; both specimens are body-
chambers.

Figs. 6-8. N. ( Planostoceras ) sp. 6, SFB C460, fragment of phragmocone, view on the base of spire. 7, SFB
C463, plastic cast of phragmocone, view on the top of spire. 8, SFB C457; body-chamber.

All specimens from sections I 3 (Gebel Qreiya), Rakhiyat Formation, Campanian, x I .

PLATE 49

' : ^

LUGER and GROSCEIKE, Baculites , Solenoceras , Nostoceras

400

PALAEONTOLOGY, VOLUME 32

Occurrence. Rakhiyat Formation: Unit 1, section 2. Late middle Campanian.

The species is known from the late Campanian to early Maastrichtian of the United States (Cobban 1974).

Baculites subanceps Haughton, 1925
Plate 49, fig. 1; text-fig. 15a-h
1925 Baculites subanceps Haughton, p. 278, pi. 14, figs. 6 8.

1965 Baculites subanceps Haughton; Howarth, p. 368, pi. 5, fig. 3; pi. 6, figs. 6 and 7; pi. 7, fig. 1;
figs. 4, 13-15 (see here for further synonymy).

Material. Thirty-six fragments of body-chambers (SFB C470). Dimensions could not be determined.

Discussion. The present material consists of mainly poorly preserved body-chambers of different
growth stages, the largest of them has a length of 80 mm, width of 24 mm, and height of 28 mm.
The better preserved specimens show the characteristic ornamentation and cross-section (see text-
fig. 15a-h) as described by Howarth (1965). Probably due to the poor preservation, most of the
present specimens do not show the crenulated ribs crossing the slightly sharpened, unkeeled venter.

B. anceps pacificus Matsumoto and Obata, 1963 was treated as a subspecies of B. subanceps by
Howarth (1965), although he noted (Howarth 1965, p. 370): ‘Comparison of the holotype of
pacificus (Matsumoto 1959a: pi. 34, fig. 3) with the lectotype of subanceps (pi. 6, fig. 6) shows that
pacificus has between two and three times as many arcuate ribs as subanceps.' Howarth (1965,
p. 405) attributed the specimens of B. subanceps from Carimba (Angola) to the latest Campanian
(polyp/ocum Zone). B. anceps pacificus was reported as a valid species by Ward (1978) from the
late middle to early upper Campanian (pacificutn and vancouverense Zones) of the Vancouver
Island region (Canada). Due to the differences in ornamentation, as already expressed by Howarth
(1965), and the differing stratigraphic occurrence we cannot follow Howarth’s view and suggest
keeping the two forms separate.

text-fig. 15. a H, whorl sections of Baculites subanceps Haughton. Based on eight specimens, SFB
C470« /;, x 1 . i-o, whorl sections and suture of B. cf. ovatus Say. Whorl sections based on six specimens,

C427 a-fi x 1. Suture, SFB C47(W , x3.

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

401

Occurrence. Rakhiyat Formation: Unit 3, section 3, late Campanian.

The species is known from the late Campanian of Angola (Howarth 1965).

Family nostoceratidae Hyatt, 1894
Genus nostqceras Hyatt, 1894
Subgenus nostoceras Hyatt, 1 894

Type species. Nostoceras stantoni Hyatt, 1894, p. 569, by original designation.

Nostoceras ( Nostoceras ) spp.

Plate 49, figs. 5, 9-10

Material. Sixteen sinistrally and dextrally coiled fragments (SFB C437-C452). Dimensions could not be
determined.

Discussion. The present fragments of hook-shaped body-chambers differ only in size. The whorl
height increases very slightly from the initial part of body-chamber to the aperture (20-5 mm to
21-5 mm in the best preserved specimen). The ornamentation consists of strong ribs, each bearing
widely spaced ventrolateral tubercles. On the initial part of the body-chamber ribs tend to bifurcate,
whereas on the apex bifurcation is only occasionally observed. The ribs are prorsiradiate on the
spiral limb and radiate on the apertural limb. Ribs weaken on dorsal part of the body-chamber.

Since the investigated specimens show distinct size differences, height and breadth of whorl at
the apex of body-chamber were measured in thirteen specimens. The apex has been chosen because
the aperture is preserved only in two smaller pieces. Generally, the height/width ratio varies
between 0-92 and 0-97. However, regarding absolute values of whorl height, apparently two groups
can be distinguished: one with w7?-values of 18 to 24 mm (five specimens), the other showing
w/z-values of 29 to 34 mm (seven specimens). One exceptionally large body chamber has a wh
of 41 mm (PI. 49, figs. 9 and 10).

Kennedy (1986, p. 95) reported dimorphism in N. ( Bostrychoceras ) polyplocum (Roemer, 1841),
being expressed in distinct size differences of the body-chamber. This could also be the case in our
material, but since the preservation of our specimens does not permit determination on the specific
level, no final conclusions can be drawn.

The specimen figured on Plate 49, figs. 9 and 10 (largest specimen) very closely resembles that
of Douville (1928, pi. 6, fig. 17) from Gebel Abu Had, a locality close to that of the present
material, which he attributed to B. polyplocum. The medium-sized body-chambers of the present
material resemble those of Antunes and Sornay’s (1970, pi. 2) specimens, figured as ‘TV. aff.
helicinum (Shumard) Stephenson’, but since the phragmocone of our material is unknown, it is
impossible to decide whether they are conspecific.

Occurrence. Rakhiyat Formation: Unit 3 in section 1 (SFB C437-C447) and section 3 (SFB C448-C452).
Late Campanian.

Subgenus planostoceras Lewy, 1967

Type species. Planostoceras rehavami Lewy, 1967, p. 168, pi. 4, figs. I 4, by original designation.

Nostoceras (Planostoceras) sp.

Plate 49, figs. 6-8

Material. Fifteen fragments (SFB C454-C466). Six fragments of hook-shaped body-chambers, six fragments
of phragmocone, the most complete of which is a cast of one and a half whorls, and three small fragments
of the intermediate part between the phragmocone and the body-chamber. The suture line is only rudimentarily
visible on one fragment. Dimensions could not be determined.

Description. The most complete part of phragmocone is characterized by a very low torticone spire, in which
the whorls are only slightly touching. The ornamentation of phragmocone consists of fine dense ribs which

402

PALAEONTOLOGY, VOLUME 32

bear a pair of ventrolateral tubercles at irregular distances. Ribs weaken on the dorsal side. The fragments
of body-chamber, which are V-shaped with a rounded base, differ slightly in size and density of ribs. The
regularly spaced ribs are simple, first prorsiradiate, on the initial part of the body-chamber with a weak
tendency to bifurcate, later radiate, and each bears ventrolateral tubercles. The aperture is slightly constricted
on the flanks. No tendency for a change of the direction of coiling was observed in the most complete
specimen.

Discussion. The present material very closely resembles the description and illustrations of N. ( P .)
rehavami Lewy, 1967 (pi. 4, figs. 1-4) from the late Campanian of the Negev with regard to the
mode of coiling and ornamentation, but reaches only two-thirds of the size of Lewy’s specimens.
Since no complete specimen is present among the investigated material, due to the smaller size of
our fragments and the differences of the phragmocone, the relation between the present material
and N. (P.) rehavami is not clear.

Occurrence. Rakhiyat Formation: Unit 3, section 3. Late Campanian.

Genus solenoceras Conrad, 1860

Type species. Hamites annulifer Morton, 1842, p. 213, by original designation of Conrad (1860).

Solenoceras humei (Douville, 1928)

Plate 49, figs. 3 and 4

1928 Ptychoceras humei Douville, p. 37, pi. 6, figs. 9 and 10.

1929 Ptychoceras sp.; Picard, p. 436, pi. 9, fig. 2.

1967 Solenoceras humei densicostata Lewy, p. 170, pi. 3, fig. 4.

? 1 969 Solenoceras cf. S. multicostatum Stephenson; Lewy, p. 125, pi. 3, fig. 6.

71969 Solenoceras cf. S. reesidei Stephenson; Lewy, p. 126, pi. 3, fig. 7.

1969 Solenoceras cf. S. texanum (Shumard); Lewy, p. 127, pi. 3, fig. 8.

Material. Two almost complete specimens (SFB C430, C468), twenty-three double-shafted fragments, and
more than fifty single shaft fragments (SFB C467, C469).

Dimensions. In the most complete specimen (SFB C430) the young shaft is 62 mm long, preserved up to the
outer part of the initial stage, and its adult shaft is 43 mm long. In a second, less complete specimen (SFB
C468, see PI. 49, fig. 4) the preserved part of the young shaft is 52 mm and the adult one 46 mm long.

Description. See Lewy (1967, p. 170).

Discussion. The present material, collected near the type-locality of S. humei , consists mainly of
internal moulds in different stages of preservation. The suture line is not preserved. Within the
association of Solenoceras the specimens attributed to S. humei show a considerable variation in
size and ornamentation; height and width of the cross-section of the adult shaft are almost equal.
Almost regardless of the size the density of the ribs varies between four and eight per centimetre
at the same growth stage. Although a tendency is observed for extremely small specimens always
to show dense ribbing, we assume that the subspecies S. humei densicostata Lewy, 1967 is a large,
densely ribbed form within the variability of S. humei.

Occurrence. Rakhiyat Formation: Unit 3, section 1 (SFB C467-C468, C430) and section 3 (SFB C469). Late
Campanian.

The species is known from the late Campanian of the Middle East (Lewy 1967).

CONCLUSIONS

The study of ammonites from the Wadi Qena area contributes to knowledge on the late
Cretaceous transgressive/regressive history of south-eastern Egypt. Since a recent compilation of
palaeontological, micropalaeontological, and palynological data from the Mesozoic to Palaeogene

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

403

strata of southern Egypt regarding the transgressive/regressive cycles has been published by Luger
and Schrank (1987), only a brief discussion on the late Cretaceous shall be given here.

In the Wadi Qena area the lowermost sedimentary unit (Wadi Qena Formation, ?Albian/early
Cenomanian) is characterized by continental deposits. Here the first transgression lasted from the
late Cenomanian to the early Turonian (Galala Formation, sea-level highstand probably during
the latest Cenomanian). A regressive period thereafter is characterized by continental deposits
(lower Umm Omeiyed Formation). During the late Turonian another minor transgression took
place (upper Umm Omeiyed Formation). Subsequently, in the early Coniacian, a return to
continental sedimentation is observed (lower regressive sequence of Hawashya Formation). In the
mid- to late Coniacian a new transgression invaded eastern Egypt (lower transgressive sequence
of Hawashya Formation) and reached further south than the preceding ones (Fuger and Schrank
1987). The Santonian to early Campanian here is characterized by a regressive/transgressive/
regressive development (upper Hawashya Formation); the age of the marine transgression remains
uncertain since as yet no guide fossils have been found in the southern Wadi Qena. During the
late middle Campanian a new transgression took place (uppermost Hawashya Formation and
basal Rakhiyat Formation). Interrupted by a minor regressive phase (unit 3 of Rakhiyat Formation,
see Hendriks and Fuger 1987), a sea-level rise is again observed in the late Campanian (unit 4 of
Rakhiyat Formation, see Hendriks and Luger 1987). A short, but prominent, regressive period
during the latest Campanian/earliest Maastrichtian is indicated by a hiatus between the Rakhiyat
and the Dakhla Formations in the Eastern Desert (Hendriks and Luger 1987). During the late
early Maastrichtian (base of Dakhla Formation in the Eastern Desert) a major transgression
started in Egypt which, interrupted only by minor regressive tendencies, lasted until the early
Eocene (see Luger 1985; Luger and Schrank 1987).

The transgressive/regressive cycles of southern Egypt show remarkable correspondence with
those of other cratonic areas as expressed in the eustatic curves of Haq et al. (1987, fig. 3). The
late Cenomanian/earliest Turonian transgression in Egypt clearly corresponds to cycle UZA 2.5
TR/HS, the late Turonian to cycle UZA 2.7 HS, and the mid or late Coniacian to cycle UZA 3.1.
The late middle Campanian transgression probably corresponds to cycle UZA 4.2 (referred to as
early late Campanian in Haq et a /., ibid.); the late Campanian to Maastrichtian development
corresponds to eustatic cycles UZA 4.3-UZA 4.5. The eustatic cycles UZA 2.6 (middle Turonian)
and UZA 3.2-UZA 4.1 (Santonian to early Campanian) are not recognizable in the sedimentary
succession of southern Egypt (see also Klitzsch 1986; Hendriks et al. 1987). These deviations from
the eustatic cycles may indicate that the transgressive pulses were only of minor importance and
did not reach the investigated area. However, it seems more likely that they are related to major
regional tectonic movements, e.g. extensive faulting and formation of Graben systems in the
Aswan- Abu Simbel area (Hendriks 1987, p. 150) or folding of the Syrian Arc System (Schandelmeier
1988, p. 105). In contrast the times of cycle coincidence represent phases of relative tectonic stability
in south-eastern Egypt.

Acknowledgements. The present study was carried out under the Special Research Project 69 ‘Geoscientific
Problems in Arid Areas’ of the German Research Foundation. The authors are greatly indebted to Professor
Dr E. Klitzsch, who initiated this study, and to Mr W. Herrmann-Degen for the opportunity to use material
collected by him. Many thanks are also due to Professor Dr H. Kallenbach, Professor Dr J. H. Schroder,
Dr Habil F. Hendriks, Dr H. Schandelmeier, Mr J. Bowitz, Mr C. Schreier, Mrs C. Werner (all TU-Berlin),
and Mr M. M. A. El-Helaifi (Egyptian General Petroleum Company, ARE) for their valuable support in the
field-work and discussions. Mrs E. Susin and Mrs B. Dunker (drafting), Mrs G. Banks (correcting our
English), and Mr B. Kleeberg (photography) assisted in preparing the manuscript.

REFERENCES

almogi-labin, a., reiss, z. and caron, m. 1986. Senonian Globotruncanidae from Israel. Eclogae geol. Helv.
79 (3), 849-895, 1 1 pis.

404

PALAEONTOLOGY, VOLUME 32

amard, b., collignon, m. and roman, j. 1981 . Etude stratigraphique et paleontologique du Cretace superieur
et Paleocene du Tinhert-W et Tademait-E (Sahara algerien). Docum. Lab. Geol. Lyon , (hs) 6, 15 173, pis.
1-17.

antunes, m. telles and sornay, j. 1969- 1970. Contributions a la connaissance du Cretace superieur de Barra
do Dande, Angola. Rev. Fac. Cienc. Univ. Lisboa , (2, C) 16 (1), 65-104, pis. 1-10.
arkell, w. j., kummel, b. and wright, c. w. 1957. Mesozoic Ammonoidea. In moore, r. c. (ed. ). Treatise
on invertebrate paleontology. Part L. Mollusca 4. Cephalopoda , L80-L465. Geological Society of America
and University of Kansas Press.

avnimelech, m. a. and shoresh, r. 1962. Les Cephalopodes cenomaniens des environs de Jerusalem. Bull.
Soc. geol. Fr. (7) 4, 528-535, pi. 15.

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

barthel, k. w. and herrmann-degen, w. 1981 . Late Cretaceous and Early Tertiary Stratigraphy in the Great
Sand Sea and its SE Margins (Farafra and Dakhla Oasis), SW Desert, Egypt. Mitt. Bayer. Staatsslg.
Palaont. hist. Geol. 21, 141-182.

Basse, e. 1953. Sur la presence du genre Coahuilites Bose 1927 dans le Sud tunesien et du genre Eulophoceras
Hyatt 1903 dans le Proche-Orient. Bull. Soc. geol. Fr. (6) 3, 865-871, pi. 27.
bayle, e. 1849. Sur quelques fossiles de la Province de Constantine. In fournel, h. (ed.). Richesse
minerale d'Algerie accompagnee d' Eclair cissements historicities et geographiques sur cette partie de PAfrique
Septentrionale. 1, 359-379, pis. 17-18. Imp. Nationale, Paris.
benavides-caceres, v. E. 1956. Cretaceous system in northern Peru. Bull. Am. Mus. nat. Hist. 108 (4), 357-
493, pis. 31-66.

berthou, p. y. 1984. Albian-Turonian stage boundaries and subdivisions in the western Portuguese Basin,
with special emphasis on the Cenomanian- Turonian boundary in the ammonite facies and rudisl facies.
Bull. geol. Soc. Denmark, 33, 41-55.

brower, j. c. and lauverjat, j 1975. Morphometric study of Choffat’s vascoceratids from Portugal.
Bull. geol. Instn. Univ. Uppsala , (ns) 6, 73-83.

— chancellor, G. R. and lauverjat, j. 1985. Revision of the Cenomanian Turonian Ammonite Vascoceras
Choffat, 1898, from Portugal. Comunic. Serv. geol. Port. 75 (1), 55-79, 6 pis.

buch, L. von. 1848. Uber Ceratiten, besonders von denen, die in Kreidebildungen sich finden. Ber. Akad.
Wiss. Berlin for 1847, 214-223, I pi.

choffat, p. 1886 1898. Recueil d’etudes paleontologiques sur la faune cretacique du Portugal. 1. Especes
nouvelles ou peu connues. 2. Les Ammonees du Bellasien, des Couches a Neolobites Vibrayeanus, du
Turonien et du Senomen. Trav. geol. Portugal, 1, 1-40, pi. 1 18 (1886); 2, 41-86, pi. 19-40 (3-22) (1898).
chudeau, r. 1909. Ammonites du Damergou (Sahara meridional). Bull. Soc. geol. Fr. (4) 9, 67-71, 3 pis.

1921. Ammonites turoniennes du Soudan. Bull. Mus. Hist. nat. Paris, 6 (27), 463-470, 1 pi.
cobban, w. a. 1974. Ammonites from the Navesinsk Formation at Atlantic Highlands, New Jersey. Prof.
Pap. US geol. Surv. 845, 21 pp., 11 pis.

— 1984. Mid-Cretaceous ammonite zones, Western Interior, United States. Bull. geol. Soc. Denmark, 33,
71-89.

— 1987. The Upper Cretaceous (Cenomanian) Ammonites Metengonoceras dumbli (Cragin) and M. acutum
Hyatt. Bull. US geol. Surv. 1690, C1-C7, 3 pis.

— and hook, s. c. 1980. The Upper Cretaceous (Turonian) Ammonite Family Coilopoceratidae Hyatt in
the Western Interior of the United States. Prof. Pap. US geol. Surv. 1192, 28 pp., 21 pis.

— 1983. Mid-Cretaceous (Turonian) ammonite fauna from Fence Lake area of west-central New
Mexico. Mem. Inst. Min. Technol. New Mex. 41, i-iv + 5-49, 14 pis.
collignon, m. 1938. Ammonites campaniennes et maestrichtiennes de l’Ouest et du Sud de Madagascar.
Ann. geol. Madagascar, 9, 55-115, pis. 1-9.

— 1955. Ammonites neocretacees du Menabe (Madagascar). 2. Les Pachydiscidae. Ibid. 21, 9-98, 28 pis.
1957. Cephalopodes neocretacees du Tinhert (Fezzan). Ann. Paleont. (Invert.), 43, 1 15-136, pis 16-18.

— 1965. Nouvelles ammonites neocretacees sahariennes. Ibid. 51 (2), 165-202, pis. A-H.

1970. Atlas des fossiles char act eristiques de Madagascar. XVI. Campanien moyen et superieur, iv + 83 pp.,
pis. 607-639. Serv. geol. Tananarive.

— 1977. Essai de comparaison des faunes d’ammonites au Cretace superieur (Turonien a Maestrichtien)
au Japon et a Madagascar. Spec. Pap. palaeont. Soc. Japan, 21, 213-222.

conrad, T. a. 1860. Description of new species of Cretaceous and Eocene fossils of Mississippi and Alabama.
J. Acad. nat. Sci. Philadelphia, (2) 4, 275-297 .

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

405

cooper, m. r. 1978. Uppermost Ccnomanian-basal Turonian ammonites from Salinas, Angola. Ann. S. Afr.
Mus. 75, 52-152.

devalque, CH., amedro, F., philip, J. and robaszynski, F. 1982. Etal des correlations litho- et biostrati-
graphiques dans le Turonien superieur des massifs d’Uchaux et de la Ceze. Les zones d’ammonites et
de rudistes. Mem. Mus. nail. Hist. nat. (Petris), (ns C) 49, 57-69.
diener, c. 1925. Ammonoidea neocretacea. Fossilium Catalogus. 1. Animalia. 29, 244 pp. Junk, Berlin.
dominik, w. 1985. Stratigraphie und Sedimentologie (Geoehemie, Schwermineralanalyse) der Oberkreide von
Bahariya und ibre Korrelation zum Dakhla-Becken (Western Desert, Agypten). Berliner geowiss. Abh. (A)
62, 173 pp., 22 pis.

douville, h. 1912. Evolution et classification des Pulchellides. Bull. Soc. geol. Fr. (4) 11, 285-320.

— 1928. Les Ammonites de la Craie superieur en Egypte et au Sinai. Mem. Acad. Sci. Inst. France, 60,
1-41, pis. 1 7.

eck, o. 1914. Die Cephalopoden der Schweinfurthschen Sammlung aus der oberen Kreide Agyptens. Zeitsclir.
dtsch. geol. Ges. 66, 179-216, pis. 9-19.

faris, m. i. and hassan, m. y. 1959. Report on the stratigraphy and fauna of the Upper Cretaceous-Paleocene
rocks of the Um-el-Huetat, Safaga area. Ain Shams Sci. Bull. 4, 191 207, 2 pis.
fischer, p. 1880-1887. Manuel de Conchy liologie et de Paleontologie Conchyliologieque, xxiv+1369 pp., 23
pis. Masson, Paris.

freund, r. and raab, m. 1969. Lower Turonian ammonites from Israel. Spec. Pap. Palaeont. 4, v + 83 pp.,
10 pis.

furon, r. 1933. Faunes et extension du Cretace au sud de l’Ahaggar (Cenomanien, Turonien, Senonien).
Bull. Soc. geol. Fr. (5) 3, 259-280, 1 pi.

— 1935. Le Cretace et le Tertiaire du Sahara soudanais (Soudan, Niger, Tchad). Arch. Mus. natl. Hist,
nat. (Paris), (6) 13, 96 pp., 7 pis.

greco, b. 1915. Fauna cretacea dell’Egitto racolta da Figari Bey. Parte 1. Cephalopoda. Palaeontographia
ital. 21, 189-231, pis. 17-22.

grossouvre, a. de. 1894. Recherches sur la craie superieure. 2. Paleontologie, Les ammonites de la craie
superieure. Mem. Serv. Carte geol. Fr. ii + 264 pp., 39 pis. (misdated 1893).
haq, b. u., hardenbol, j. and vail, p. r. 1987. Chronology of fluctuating sea levels since the Triassic. Science,
235, 1156 1167.

haughton, s. h. 1925. Notes on some Cretaceous fossils from Angola (Cephalopoda and Echinoidea). Ann.
S. Afr. Mus. 22, 263-288, pis. 12-15.

hendriks, F. 1987. Die Entwicklungsgeschichte des SE-Agyptischen Sedimentationsraumes in der Kreide und
im Alttertidr: Eine Beckenstudie, 167 pp. Habil. thesis (unpublished), TU-Berlin.

— and luger, p. 1987. The Rakhiyat Formation of the Gebel Qreiya area: Evidence of Middle Campanian
to Early Maastrichtian synsedimentary tectonism. Berliner geowiss. Abh. (A) 75 (1), 83-96, 1 pi.

— bowitz, j. and kallenbach, h. 1987. Evolution of the depositional environments of SE-Egypt
during the Cretaceous to Lower Tertiary. Ibid. 49-82.

hook, s. c. and cobban, w. a. 1981. Late Greenhorn (Mid Cretaceous) discontinuity surfaces, southwest
New Mexico. Circ. Bur. Min. Technol. New Mex. 180, 5-21, 3 pis.
hourcq, v. 1949. Paleontologie de Madagascar. XXVIII. Sur quelques ammonites du Senonien. Ann. Paleont.
(Invert.), 35, 89 117, pis. 1113.

howarth, m. k. 1965. Cretaceous ammonites and nautiloids from Angola. Bull. Br. Mus. nat. Hist. (Geol.),
10 (10), 337-412, 13 pis.

hyatt, a. 1894. Phylogeny of an acquired characteristic. Proc. am. Phil. Soc. 32, 349 647, pis. 114.

— 1900. Cephalopoda. In zittel, k. a. von. 1896-1900. Textbook of Palaeontology. 1. Transl. and ed.
Eastman, c. r., 502-604. McMillan, London and New York.

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

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

— 19846. Systematic palaeontology and stratigraphic distribution of the ammonite faunas of the French
Coniacian. Spec. Pap. Palaeont. 31, 160 pp., 33 pis.

— 1985. Integrated macrobiostratigraphy of the Albian to basal Santonian, 91-108. In reyment, r. a. and
bengtson, p. (Compilers). Mid-Cretaceous Events: report on results obtained 1974 1983 by IGCP Project
58. Pubis. Palaeont. Instil. Univ. Uppsala, Spec. Vol. 5, 132 pp.

406 PALAEONTOLOGY, VOLUME 32

Kennedy, w. j. 1986. Campanian and Maastrichtian ammonites from northern Aquitaine, France. Spec. Pap.
Palaeont. 36, 145 pp., 23 pis.

— amedro, F. and collete, c. 1986. Late Cenomanian and Turonian ammonites from Ardennes, Aube
and Yonne, eastern Paris Basin (France). N. Jb. Geol. Palaont. Abh. 172 (2), 193-217.

— and juignet, p. 1981. Upper Cenomanian Ammonites from the Environs of Saumur, and the Provenance
of the Types of Ammonites vibrayeanus and Ammonites geslinianus. Cret. Res. 2, 19-49.

— and Hancock, j. M. 1981. Upper Cenomanian ammonites from Anjou and the Vendee, western
France. Palaeontology, 24 (1), 25-84, pis. 3 17.

— and wright, c. w. 1984. The Cretaceous ammonite Ammonites requienianus d’Orbigny, 1841. Ibid. 27
(2), 281-293, pis. 35-37.

-and Hancock, j. m. 1987. Basal Turonian ammonites from west Texas. Ibid. 30 (1), 27-74, 10 pis.
klitzsch, e. 1986. Plate tectonics and cratonal geology in Northeast Africa (Egypt, Sudan). Geol. Rundsch.
75 (3), 755-768.

— groschke, m. and herrmann-degen, w. 1988. Wadi Qena: Paleozoic and Pre-Campanian Cretaceous
Strata. In said, r. and squires, c. h. The Geology of Egypt. Balkema, Rotterdam.

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

kullmann, j. and wiedmann, j. 1970. Significance of sutures in phylogeny of Ammonoidea. Paleont. Contr.
Univ. Kansas , 47, 32 pp.

lamarck, j. p. b. a. de m. de. 1801. Systeme de Animaux sans vertebres , vii + 432 pp. Deterville, Paris.
lewy, z. 1967. Some Late Campanian nostoceratid ammonites from southern Israel. Israel J. Earth Sci. 16,
165- 173, 4 pis.

— 1969. Late Campanian heteromorph ammonites from southern Israel. Ibid. 18, 109-135, 4 pis.

— 1975. The geological history of southern Israel and Sinai during the Coniacian. Ibid. 24, 19-43, 1 pi.

— Kennedy, w. j. and chancellor, g. r. 1984. Co-occurrence of Metoicoceras geslinianwn (d'Orbigny)
and Vascoceras cauvini Chudeau (Cretaceous Ammonoidea) in the southern Negev (Israel) and its
stratigraphic implications. Newsl. Stratigr. 13, 67-76.

— and raab, m. 1978. Mid-Cretaceous stratigraphy of the Middle East. Ann. Mas. Hist. nat. Nice , 4, xxxii,
1-20, 2 pis.

luger, p. 1985. Stratigraphie der marinen Oberkreide und des Alttertiars im sudwestlichen Obernil-
Becken (SW-Agypten) unter besonderer Berucksichtigung der Mikropalaontologie, Palokologie und
Palaogeographie. Berliner geowiss. Abh. (A) 63, 1-151, pis. 1-24.

— and schrank, e. 1987. Mesozoic to Paleogene transgressions in middle and southern Egypt— Summary
of paleontological evidence. In matheis, g. and schandelmeier, h. (eds.). Current research in African Earth
Sciences, 199-202. Balkema, Rotterdam.

matsumoto, t. 1959. Upper Cretaceous ammonites from California. Part 1. Mem. Fac. Sci. Kyushu Univ.
(D)8, 91 171, pis. 30-45.

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

In hayden, F. v. Rep. US geol. geogr. Surv. Territ. 9, lxiv + 629 pp., 45 pis.
morton, s. G. 1842. Description of some new species of organic remains of the Cretaceous Group of the
United States; with a tabular view of the fossils hitherto discovered in this formation. J. Acad. nat. Sci.
Philadelphia, 8, 207-227.

offodile, m. e. and reyment, r. a. 1978. Stratigraphy of the Keana-Awe area of the middle Benue region of
Nigeria. Bull. geol. Instn. Univ. Uppsala, (ns) 7, 37-66.
orbigny, a d’. 1840 1842. Paleontologie franqaise. Terrains cretaces. 1. Cephalopodes , 1 120 (1840); 121-430
(1841); 431 622 (1 842), 1 5 1 pis. Paris.'

peron, a. 1890 1893. Description des mollusques fossiles des terrains Cretaces de la region sud de Hauts-
Plateaux de la Tunisie recueillis cn 1885 et 1886 par M. Philippe Thomas. Explor. sci. Tunisie, xii + 405
pp., 35 pis.; xii + 1 103 (1890); 104-327 (1891); 328-405 (1893). Masson, Paris.

— 1896-1897. Les ammonites du Cretace superieur de l’Algerie. I. + 11. Mem. Soc. Geol. Fr. (Paleont.), 17,
88 pp., 18 pis.; 6, I 24, pis. 14-19 (1-6) (1896); 7, 25-88, pis. I 12 (7-18) (1897).

pervinquiere, l. 1907. Etudes de paleontologie tunisienne. I. Cephalopodes des terrains secondaires. Carte
geol. Tunisie, v + 438 pp., 27 pis. De Rudeval, Paris.

picard, l. 1929. On Upper Cretaceous (chiefly Maestrichtian) Ammonoidea from Palestine. Ann. Mag. nat.
Hist. 10 (3), 433-456, pis. 9 10.

popoff, m., wiedmann, j. and de klasz, i. 1986. The Upper Cretaceous Gongila and Pindinga Formations,

LUGER AND GROSCHKE: EGYPTIAN CRETACEOUS AMMONITES

407

northern Nigeria: Subdivisions, age, stratigraphic correlations and paleogeographic implications. Eclogue
geol. Helv. 79 (2), 343-363.

quaas, a. 1902. Beitrag zur Kenntnis der Fauna der obersten Kreidebildungen in der libyschen Wiiste
(Overwegischichten und Blatterthone). Palaeontographica, 30 (2), 153-334, pis. 20 33.
reeside, j. b. 1962. Cretaceous ammonites of New Jersey. Bull. Bur. Geol. Topogr. New Jers. (Paleont.), 61,
113- 137, pis. 68-75.

REISS, Z., ALMOGI-LABIN, A., HONIGSTEIN, A., LEWY, Z., LIPSON-BENITAH, S., MOSHKOVITZ, S. and ZAKS, Y. 1985.

Late Cretaceous multiple stratigraphic framework of Israel. Israel J. Earth Sci. 34, 147 166.
reyment, R. A. 1954. Some new Upper Cretaceous ammonites from Nigeria. Colon, geol. Min. Resow. Div.
4, 248-270, 5 pis.

— 1955. The Cretaceous Ammonoidea of southern Nigeria and the southern Cameroons. Bull. geol. Surv.
Nigeria , 25, 99 pp., 25 pis.

roman, f. 1938. Les ammonites jurassiques et cretacees. Essai cle genera, 554 pp., 53 pis. Masson, Paris.
say, t. 1820. Observations on some species of zoophytes, shells, etc., principally fossils. Amer. J. Sci. (1) 2,
34-45.

SCHANDELMEIER, H. 1988. Die Intraplattendeformation Nordost Afrikas — Prakambrische Anlage und Phanerozo-
ische Reaktivierung, 130 pp. Habil. thesis (unpublished), TU-Berlin.
schluter, c. 1871 1876. Cephalopoden der oberen deutschen Kreide. Palaeontographica, 21, 1 24, pis. 1-8
(1871); 21, 25-120, pis. 9-35 (1872); 24, 1-144 (121-264), pis. 35-55 (1876).
schneegans, D. 1943. Invertebres du cretace superieur du Damergou (Territoire du Niger). Bull. Div. Mines
Afr. occid.fr. 7, 87 -150, 8 pis.

schobel, j. 1975. Ammoniten der Familie Vascoceratidae aus dem Unterturon des Damergou-Gebietes,
Republique du Niger. Pub/, palaeont. Instn. Univ. Uppsala, 3, 116 pp., 6 pis.
spath, l. F. 1922. On the Senonian ammonite fauna of Pondoland. Trans, r. Soc. S. Afr. 10, 113 147, pis.
5-9.

stoliczka, f. 1863 1866. The fossil Cephalopoda of the Cretaceous rocks of southern India. Mem. geol.
Surv. India. Palaeont ologia Ind. 3 (1), 41-56, pis. 26-31 (1863); (2-5), 57-106, pis. 32-54 (1864); (6 9), 107
154, pis. 55-80 (1865); (10 13), 155-216, pis. 81 94 (1866).
taubenhaus, h. 1920. Die Ammoneen der Kreideformation Palastinas und Syriens. Z. dt. Palastina-Vereins,
43, 58 pp., 9 pis.

thomel, g. 1969. Sur quelques ammonites turoniennes et senoniennes nouvelles ou peu connues. Ann. Paleont.
(Invert.), 55, 109 140, 7 pis.

— 1972. Les Acanthoceratidae cenomaniens des chaines subalpines meridionales. Mem. Soc. geol. Fr. (ns)
116, 201 pp., 88 pis.

ward, p. d. 1978. Baculitids from the Santonian Maestrichtian Nanaimo Group, British Columbia, Canada
and Washington State, USA. J. Paleont. 52 (5), 1143-1154, 2 pis.
wedekind, R. 1916. Uber Lobus, Sulurallobus und Inzision. Cbl. Miner. Geol. Paldont. for 1916, 185 195.
wiedmann, j. 1978. Die Ammoniten der NW-deutschen, Regensburger und Ostalpinen Oberkreide im
Vergleich mit den Oberkreidefaunen des westlichen Mediterangebietes. In Aspekte der Kreide Europas.
IUGS, Series A, 6, 335-350, Stuttgart.

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

zittel, K. a. von. 1884. Handbuch der Paldontologie. 1 Abt. Palaeozoologie. 2, 893 pp. Munchen and Leipzig.

PETER LUGER

Technische Universitat Berlin, SFB 69
AckerstraBe 71-76
1000 Berlin 65

MANFRED GROSCHKE

Institut fur Geologie und Palaontologie
Technische Universitat Berlin
Ernst-Reuter-Platz 1
1000 Berlin 12

Typescript received 12 August 1987
Revised typescript received 4 August 1 988

LEPTOPTERYGIUS TENUI ROSTRIS AND OTHER
LONG-SNOUTED ICHTHYOSAURS FROM THE
ENGLISH LOWER LIAS

by C. MCGOWAN

Abstract. One of the commonest ichthyosaurs from the English Lower Lias is the long-snouted species
Leptopterygius tenuirostris , known principally from Street, Somerset. Because of the vagaries of preservation
there are few complete skeletons, and the problem is exacerbated by the occurrence of composite specimens.
The authenticity of a quarter of the specimens studied here is in doubt, and hence caution is needed when
working on material from Somerset. The occurrence of a tail bend in L. tenuirostris is confirmed by the
presence of wedge-shaped centra in the caudal region of several skeletons. The vertebral column was probably
not steeply downturned, and may have been essentially straight in life.

Eurhinosaurus , unusual for its abbreviated mandible, may be closely related to L. tenuirostris and is
therefore of interest here. The suggestion that it occurs in the Upper Lias of England is confirmed. The
contention that Eurhinosaurus lacked a tail bend is questioned because a wedge-shaped centrum has been
identified in one specimen.

Two trivial names besides tenuirostris have been used for long-snouted ichthyosaurs: latifrons and
longirostris. The former is a taxon dubium , while the latter should be used only in combination with
Eurhinosaurus.

Ichthyosaurs occur throughout most of the Mesozoic, but they are best known from the Lower
Jurassic, where large numbers of complete or near-complete skeletons have been found, sometimes
in remarkably good states of preservation. Especially prolific have been the Lower Lias (Hettangian,
Sinemurian, and Lower Pliensbachian) outcrops of south-west England, and the Upper Lias
(Toarcian) deposits of southern Germany. Upper Lias ichthyosaurs occur in England, notably in
the Whitby area of Yorkshire, and in the vicinity of Ilminster, Somerset, but neither locality has
been very productive, and the Whitby material is generally not well preserved. The temporal
separation between the Upper and Lower Lias is about 15 million years (Harland et al. 1982) and,
although the two faunas have similar diversities of forms, they are taxonomically distinct. Both
faunas, for example, have a short-snouted form; Ichthyosaurus breviceps in the Lower Lias and
Stenopterygius hauffianus in the Upper Lias. While the present paper is primarily concerned with
Lower Lias ichthyosaurs, some taxonomic problems require reference to Upper Lias material.

The commonest English species, accounting for about half of the determinate skeletons, is
I. communis , a moderately sized ichthyosaur reaching a maximum total length (measured from the
tip of the snout to the tip of the tail) of about 2-5 m (McGowan 1974/)). Less common in terms
of complete skeletons, but abundantly represented by isolated humeri, partial fins and rostral
segments, is Leptopterygius tenuirostris , characterized by its relatively long slender rostrum.
L. tenuirostris is somewhat larger than /. communis , reaching lengths in excess of 2-5 m. While it
has been found at several Lower Lias localities, it is best known from Street and the surrounding
areas of Somerset. Because of the vagaries of preservation, L. tenuirostris is not so well known as
I. communis , and one uncertain point is whether there was a tail bend. The tail bend, a prominent
feature of post-Triassic ichthyosaurs, marks the position of the caudal peduncle, where the vertebral
column is downturned to support the hypocaudal lobe of the tail (McGowan 1 974c/). The most
completely preserved skeletons (BGS 51236 and BMNH R498— see McGowan 19746, figs. 1 1 and
12a) appear to lack a tail bend, raising the question of whether an asymmetrical caudal fin was

(Palaeontology, Vol. 32, Part 3, 1989, pp. 409-427.|

© The Palaeontological Association

410

PALAEONTOLOGY, VOLUME 32

present in this species. Absence of a reversed heterocercal tail, however, may not be unique among
ichthyosaurs because it has recently been proposed that Eurhinosaurus, an Upper Lias genus
bearing a superficial resemblance to the modern swordfish, lacked this feature (Riess 1986). This
is of particular interest here because of the possibility that L. tenuirostris lies close to the ancestry
of Eurhinosaurus (McGowan 1986).

While most long-snouted ichthyosaurs from the Lower Lias have been referred to L. tenuirostris,
two other names, /. longirostris and I. latifrons , have also been used, sometimes synonymously
and frequently causing much confusion. An additional problem is caused by incomplete preservation,
especially of the narrow tip of the snout and mandible, and also by changes that have been made
to specimens during preparation. Some of these modifications are easily recognized, but others
have been so skilfully executed that their detection is difficult, even when it is possible to dismantle
the entire skeleton (McGowan et al., in prep.). It is, therefore, necessary to be especially circumspect
when dealing with material from Somerset localities.

In a previous description of L. tenuirostris (McGowan 1974/?) attention was drawn to the
problem of assigning the species to an appropriate genus. The decision was taken to refer it to
Ichthyosaurus but, in the light of new information presented here, this is no longer appropriate.
The species is accordingly referred to Leptopterygius, a genus erected by Huene (1922) for
L. tenuirostris, and several other species. This usage is consistent with that of Appleby (1979), but
I do not use his ordinal designations, which are based upon the recognition of latipinnate and
longipinnate ichthyosaurs. These latter terms are based on fin structure, and they have been widely
used for classification. I once considered that there were also correlated cranial characters which
could be used to distinguish between the two groups (McGowan 1972), but 1 later questioned the
validity of the dichotomy (McGowan 1976), and concluded that there were no unequivocal
distinctions between latipinnate and longipinnate ichthyosaurs (McGowan 1979, pp. 125-126).
Huene (1922) did not give a diagnosis for Leptopterygius and Appleby (1979, p. 943), considering
it to be a monotypic genus, gave the same diagnosis as for the species L. tenuirostris. Although a
redefinition of Leptopterygius is clearly needed, this requires a review of several other species and
therefore lies beyond the scope of the present work.

There are three primary objectives of this paper: to clarify the taxonomy of I. longirostris and
I. latifrons', to assess the available long-snouted specimens, assigning them to their appropriate
taxa and assessing their authenticity where this is in doubt; and, to use these additional data to
revise the previous description of L. tenuirostris (McGowan 1974/?). Secondary objectives are to
examine the question of whether Eurhinosaurus occurs in the English Upper Lias, and to make
some preliminary remarks on the tail of this genus. The reason for including Eurhinosaurus here
is partly because of the taxonomic confusion which has existed between this long-snouted form
and L. tenuirostris, and also because of the possible phylogenetic relationship between them
(McGowan 1986).

MATERIALS AND METHODS

Of the specimens examined from England, twenty-seven are from the Lower Lias (primarily from Street,
Somerset) and two from the Upper Lias of Whitby. Reference is also made to three specimens of Eurhinosaurus
from the Upper Lias of Germany. Abbreviations used are: BATGM, Bath Geological Museum (the
ichthyosaur material, which is part of the Moore Collection, has been on loan to the National Museum of
Wales, Cardiff, for the last several years); BGS, British Geological Survey, Keyworth, Nottinghamshire (the
ichthyosaur BGS 51236 is currently in the Geological Museum, London); BMB, Admiral Blake Museum,
Bridgwater, Somerset; BMNH, British Museum (Natural History), London; DLR, Dinosaurland, Lyme
Regis, Dorset; LSL Forschungsinstitut Senckenberg, Frankfurt (Natur-Museum, Senckenberg), Germany;
GTS, Alfred Gillett Trust, Street, Somerset (this collection is located in the archives of C and J Clark Ltd.);
LEICS, Leicestershire Museums, Art Galleries and Records Service, Leicester; OUM, Geological Collections,
University Museum, Oxford; SCM, Somerset County Museum, Taunton; SMNS, Staalliches Museum fiir
Naturkunde, Stuttgart, Germany; WM, Wells Museum, Somerset. Specimens lacking catalogue numbers are
referred to by MS numbers, given in quotation marks.

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

41 1

Body lengths were measured along the vertebral column using a steel tape and recorded to the nearest
millimetre. The same tape was used for all measurements in excess of 550 mm. Small dial calipers and large
vernier calipers were used for measurements less than 550 mm, recorded to the nearest 0- 1 and to the nearest
1 mm, respectively. Details of the characters measured and the ratios derived from them are given elsewhere
(McGowan 1974a; 1976) and are summarized in Table 1.

table 1 . Characters recorded and ratios derived.

Vertebral count to pelvis
Vertebral count to tail bend
Total vertebral count
Presacral length
Preflexural length

AXIAL SKELETON

Number of vertebrae from atlas to level of pelvis.

Number of vertebrae from atlas to level of tail bend.

Number of vertebrae from atlas to end of column.

Length from atlas to level of pelvis, measured along vertebral column.
Length from atlas to level of tail bend, measured along vertebral column.

Number of primary digits
Total digital count
Elements in longest digit
Humerus length
Humerus width
Humerus shaft
Femoral length
Femoral width
Pelvic condition
Coracoid length
Coracoid width
Coracoid notching

FINS AND GIRDLES

Number of digits arising from carpus.

Sum of primary and secondary (arising from outside carpus) digits
Number of elements in longest digit, counting from epipodials.
Maximum length, measured between horizontals perpendicular to shaft.
Maximum distal width, measured between verticals parallel to shaft.
Minimum width of shaft.

Maximum length, measured between horizontals perpendicular to shaft.
Maximum distal width, measured between verticals parallel to shaft.
Pelvis tripartite (pubis and ischium unfused), or bipartite (fused).
Maximum anteroposterior length.

Maximum lateromedial width.

Anterior and/or posterior margins emarginated.

Skull length
Jaw length
Orbital diameter
Snout length
Premaxillary length
Prenarial length
Sclerotic diameter
Overbite
Orbital ratio
Snout ratio
Premaxillary ratio
Prenarial ratio
Sclerotic ratio

SKULL

Distance between tip of snout and posterior edge of quadrate.

Distance between tip of dentary and posterior edge of angular.

Internal diameter of orbit, measured along its longitudinal axis.
Distance between tip of snout and anterior (internal) margin of orbit.
Distance between tip of snout and anterior tip of maxilla.

Distance between tip of snout and anterior margin of external naris.
Internal diameter of sclerotic ring measured along its longitudinal axis.
Distance between tip of snout and tip of jaw.

Orbital diameter divided by jaw length.

Snout length divided by jaw length.

Premaxillary length divided by jaw length.

Prenarial length divided by jaw length.

Sclerotic diameter divided by orbital diameter.

In addition, several measurements have been found useful for comparing the relative slenderness of skulls
and jaws. These measurements are obviously sensitive to compression distortion and, since there is no way
of assessing how this might vary from one specimen to another, the measurements are used only for
comparative purposes and do not contribute to the diagnosis of L. tenuirostris.

1. Snout depth at the tip of the maxilla (abbreviated S-M) — the minimum depth of the snout, measured
at right angles to the longitudinal axis of the skull, at the level of the anterior tip of the maxilla.

2. Snout depth at the naris (S-N) — the minimal depth of the snout, measured at right angles to the
longitudinal axis of the skull, at the level of the anterior end of the external naris.

3. Snout depth at the mid-point of the snout (S-S2) — the minimal depth of the snout, measured at right
angles to the longitudinal axis of the skull, at a point one-half of the snout length back from its tip.

412

PALAEONTOLOGY, VOLUME 32

4. Snout depth at the mid-point of the jaw (S-J2)— the minimal depth of the snout, measured at right
angles to the longitudinal axis of the skull, at a point one-half of the jaw length back from its tip.

5. Jaw depth at the tip of the maxilla (J-M)— the minimal depth of the jaw, measured at right angles to
the longitudinal axis of the jaw, at the level of the anterior tip of the maxilla.

6. Jaw depth at the naris (J-N)- the minimal depth of the jaw, measured at right angles to the longitudinal
axis of the jaw, at the level of the anterior end of the external naris.

7. Jaw depth at the mid-point of the snout (J-S2) — the minimal depth of the jaw, measured at right angles
to the longitudinal axis of the jaw, at a point one-half of the snout length back from its tip.

8. Jaw depth at the mid-point of the jaw ( J-J2) — the minimal depth of the jaw, measured at right angles
to the longitudinal axis of the jaw, at a point one-half of the jaw length back from its tip.

The tail bend is usually an obvious feature in skeletons with complete or near-complete vertebral columns
and its identification is therefore usually a simple matter. However, care has to be taken to ensure that a
given tail bend is natural and not an artefact of preparation. This is because a tail bend can be manufactured,
unwittingly or otherwise, simply by inclining a block containing the terminal portion of the vertebral column
to the rest, and I suspect that several ichthyosaur skeletons have been so modified. The position of the tail
bend in specimens lacking an obvious flexion can be estimated by detecting the changes in diameter of the
centra in its vicinity (McGowan 1974a, pp. 4 6). However, the only way to establish unequivocally the
position of the tail bend is to identify the three or so wedge-shaped centra that form its apex (McGowan
1974a, fig. 3b). Since the distinctive shape of these apical vertebrae can only be seen when they are exposed
in lateral view, which is seldom the case, confirmation is usually not possible.

TAXONOMIC STATUS OF I. LONGIROSTRIS AND I. LATIFRONS
I. longirostris

Although this species is usually attributed to Owen 1881 the first description, albeit brief, was
given by Mantell (1851, p. 385) based upon a badly crushed skeleton from Whitby, Yorkshire,
said to be remarkable for its exceedingly slender and elongated snout. Lydekker (1889, p. 91)
identified Mantell's specimen as BMNH 14566, noting that it was from the Upper Lias, and that
it had been figured by Owen (1881, pi. 32, fig. 8). BMNH 14566 is therefore the holotype of
I. longirostris Mantell 1851.

Jager (1856), dissatisfied with Mantell’s brief description, gave one of his own which included
some additional material, namely an almost complete but distorted skull from the Upper Lias of
Germany (SMNS ’438’). He noted that the mandible of this specimen seemed foreshortened, but
that this was apparent rather than real and that the mandible did extend to the tip of the snout.
After examination of the material, I concluded that the mandible really was shortened, and that
this specimen should therefore be referred to the genus Eurhinosaurus (McGowan 1979, p. 131).
Indeed, most authors, including Huene (1922) and Kuhn (1934), give the authority of the species
E. longirostris as Jager (1856). However, the authority for the name longirostris is Mantell 1851,
erected upon BMNH 14566, and it is now necessary to determine whether this material, like Jager’s
(1856) additional material, is referable to Eurhinosaurus.

Huene (1922, p. 39) noted that the skull of BMNH 14566 was exposed from the dorsal aspect
and the mandible was not visible, but he surmised that it was shortened because he believed that
the material appeared to belong to Eurhinosaurus. I have now had the opportunity of examining
this rather incomplete specimen, and agree with Huene’s conclusions (see p. 416). Thus BMNH
14566 appears to be referable to Eurhinosaurus, and is therefore regarded as the holotype of
E. longirostris (Mantell 1851).

Owen’s (1881) description of I. longirostris was founded upon material from the Lower Lias of
Barrow-on-Soar, primarily upon BMNH 36182 which may be regarded as his holotype. But the
name I. longirostris was already occupied by Mantell’s (1851) species, which belongs to a different
genus ( Eurhinosaurus ). Owen somewhat confused the issue further by including a figure of the
skull of BMNH 14566 in his description of I. longirostris , but, since he did not discuss this material,
and since his species was primarily erected upon BMNH 36182, the name I. longirostris Owen
1881 may be regarded as a junior primary homonym and accordingly rejected.

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

413

I. latifrons

Owen (1881, p. 119) correctly attributed this species to Konig (1825, pi. 19, fig. 250) who had
figured a partial skull and incomplete vertebral column from the Lower Lias of Barrow-on-Soar.
This specimen was subsequently identified by Lydekker (1889, p. 90) as BMNH R 1 1 22. Owen
(1881, ph 27, fig. 1) gave a detailed figure of the skull, and (p. 1 19) referred a second specimen to
the species, ‘a skeleton, lacking both ends, but including the trunk, with chief part of the skull . . .
the total length being 4 feet 10 inches’. He gave the locality as Lyme Regis (and also for Konig’s
material), but Lydekker (1889, pp. 90-91), who identified this second specimen as BMNH 38709,
gave the locality as Barrow-on-Soar. I do not consider either specimen to be adequate for the
erection of a species, and therefore reject the name I. latifrons Konig 1825 as a taxon dubium, sensu
Smith (1970).

SYSTEMATIC PALAEONTOLOGY

Identifying specimens that have been modified during preparation proved to be a major problem
during this study. Almost a quarter of the specimens show evidence of having been modified,
and some of these are obvious composites. Others are less readily detected, and it is largely due
to this uncertainty that the descriptive account of L. tenuirostris is tempered with a degree of
caution.

Seven of the twenty-nine specimens studied will not be treated further here as they have almost
certainly been modified during preparation (BMB C2, BATGM M3560, BATGM M3558, BATGM
M3568, BATGM M3575, OUM 10319, BATGM M3564). Four more specimens are too
incompletely preserved to be identified (BMNH R 1 1 23, BMNH 38709, BATGM M3573, GTS
L/AG/Arch/7), and a fifth (BMNH R 1 1 20), which may also be largely indeterminate, was
inaccessible because of building construction. A large and fairly complete skeleton (DLR ’OOF),
which is probably not referable to L. tenuirostris , may represent a new species. Treatment of this
specimen, however, will be postponed until some comparable material, recently acquired by the
City of Bristol Museum and Art Gallery, has been studied.

Twelve of the remaining specimens are referred to L. tenuirostris , mostly without qualification
(OUM J 10305, BMNH R489, BGS 51236, SCM 8372, WM 527, GTS L/AG/Arch/18, BATGM
M3552, BATGM M3556, BATGM M3565, BATGM M3566, LEICS OS.90.1953, DLR ’002’).
Two more specimens (BMNH 2009 and BMNH 36182) probably represent variant individuals of
the species. Two specimens, BMNH 14566 and BMNH 36876, are referred to E. longirostris ,
confirming the occurrence of this genus in the Upper Lias of England.

L. tenuirostris (Text-figs. 1 and 2)

The picture that emerges of L. tenuirostris is that of a long-snouted, long-bodied ichthyosaur with a tail bend
which is probably not steeply downturned and which may in life have been essentially straight. The vertebral
counts to the pelvis and to the tail bend are in the region of 45 and 85 respectively. The forefin has four
major digits, each with relatively few phalanges, and the number of elements in the longest digit is only about
1 5. The phalanges are large, discoidal, and probably well spaced distally. The humerus has a constricted
shaft, broadly expanded distally, with a facet on its leading edge. The radius has a deep notch on its leading
edge and frequently encloses a small foramen along its contact edge with the ulna. Fusion sometimes occurs
between the radius and ulna and between the radius and the humerus (Table 2). The forefin is so distinctive
that it is possible to identify isolated fins, even partial ones. However, since similar forefin features are also
found among Upper Lias ichthyosaurs, such identifications can only be made if the material is known to be
from the Lower Lias.

The pelvic girdle is tripartite, with distinct and separate ilium, ischium, and pubis. Fusion sometimes occurs
between the pubis and the ischium, but this is only partial and does not give rise to the essentially single
ischio-pubic element as found in the Upper Lias genus Stenopterygius (McGowan 1979). The coracoid is
rounded and, while an anterior notch, often small and discrete, always appears to be present, there is usually
not a posterior one. The coracoid seen in BGS 51236 has an unusual rectilinear shape and might not be
natural.

414

PALAEONTOLOGY, VOLUME 32

A

B

C

text-fig. 1 . Leptopterygius tenuirostris. A, OUM J10305, x 0-23; b, BGS 51236, x 0-22; c, SCM 8372, x 0-22.

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

415

The architecture of the skull is dominated by the extremely long slender snout and the equally slender
mandible. The orbit, which often approaches a perfect circle, dominates the post-rostral segment of the skull
but, relative to the length of the skull, it is fairly small. The orbital ratio is therefore relatively low, usually
lower than that of the commonest Lower Lias species, I. communis. The external naris, rather than being a
simple opening, is more often a bilobed structure which is sometimes quite complex. In some instances it
appears to be drawn out anteriorly into a narrow slit, but this might be a result of preservation. The teeth
are slender rather than conical and there is a tendency towards tooth reduction, both in their size and
number, with increasing maturity. A similar situation occurs in the common Upper Lias species S. quadriscissus
(Huene 1922, p. 40; McGowan 1979, pp. 102-104). A variation seen in some specimens is for the tip of the
snout to extend beyond that of the mandible, giving the skull an overbite. There is a possible overbite of
9 mm in OUM J10305, a definite one of 19 mm in BGS 51236, and an overbite of between 60 and 70 mm
in BMNH 2009, which is about 15% of the snout length.

A B 0

text-fig. 2. Leptopterygius tenuirostris. a, OUM J10305, xO-32; b, SCM 8372, xO-32; c, BMNH R 1 1 27, an
isolated partial fin, x 0-62. For ease of comparison, photographs b and c have been laterally inverted; all

three appear to be left fins in dorsal view.

Emended diagnosis. Vertebral count to tail bend (which may be indistinct) > 79; vertebral count
to pelvis at least 44 but probably not exceeding 50; orbital ratio < 0-25 and may be < 0-20; snout
ratio > 0-70; premaxillary ratio > 0-48; prenarial ratio > 0-56; sclerotic ratio 0-34; teeth
predominantly slender and may be relatively small; forefin probably with four digits; humerus with
constricted shaft, widely expanded distally with a facet on leading edge; radius notched; occlusal
edges of radius and ulna usually enclosing a small foramen; radius and ulna, sometimes also

416 PALAEONTOLOGY, VOLUME 32

table 2. Forefin features of Leptopterygius tenuirostris.

Specimen

Humerus wide
distally

Humerus with
leading edge
facet

Foramen between
radius and ulna

Radius

notched

Phalanges

rounded

Fusion between
radius and ulna

OUM .110305

Yes

Yes

Yes

Yes

Yes

Yes

BMNH R489

Yes

Yes

Yes

Yes

Yes

No

BGS 51236

Yes

Yes

Yes

Yes

Yes

No

SCM 8372

Yes

Yes

Yes

Yes

Yes

No

WM 527

Yes

Yes

No

Yes

Yes

Yes

GTS L/AG/Arch/18

Yes

Yes

Yes

Yes

Yes

No

BATGM M3552

Yes

Yes

Yes

Yes

Yes

No

BATGM M3556

Yes

Yes

Yes

Yes

Yes

No

BATGM M3565

Yes

Yes

No

Yes

Yes

No

BATGM M3566

Yes

Yes

No

Yes

Yes

No

LEICS OS. 90. 1953

Yes

Yes

Yes

Yes

Yes

No

DLR ‘002’

Yes

Indeterminate

Yes

Yes

Yes

No

BMNH 36182

Yes

Yes

Yes

Yes

Yes

Yes

humerus, may be partially fused; phalanges discoidal, relatively large, probably well spaced distally;
femur with slender shaft, expanded distally; tibia notched, probably also tibiale, notches probably
broad; pelvic girdle essentially tripartite, though pubis and ischium may be partially fused; coracoid
probably rounded and probably always with an anterior notch.

Geological range. All specimens here referred to I. tenuirostris are from the Lower Lias (and uppermost
Triassic; see below), primarily from Street in Somerset but material has also been collected from other
localities including Barrow-on-Soar, Leicestershire; Lyme Regis, Dorset; Pinhay Bay, Devon; and Stogursey,
Somerset. A pair of partial forefins (BMNH 41253) referable to I. tenuirostris were collected from Tewkesbury,
Gloucestershire.

Most of the reptilian remains from Street were collected from the Pr e-Planorbis Beds (Arkell 1933), and
while there has been some discussion on whether this horizon should be placed at the base of the Jurassic
or at the top of the Triassic, the latter has been recommended (Harland et al. 1982). At the other end of the
range the youngest material is represented by material from Lyme Regis, and this probably does not extend
beyond the earlier part of the Sinemurian. The geological range of L. tenuirostris is therefore from the
Rhaetian to the Early Sinemurian.

Description of individual specimens. See Appendix.

Eurhinosaurus longirostris

Huene’s (1922, pp. 39-40) suggestion that E. longirostris occurred in England was based on the evidence of
two poorly preserved specimens from the Upper Lias of Whitby, Yorkshire (BMNH 14566 and BMNH
36876), which he tentatively assigned to the species. The skull of the first specimen, described as being badly
damaged, was said to be exposed from the dorsal aspect, with no mandible visible. The second specimen was
described as being a badly damaged skull, again without evidence of a lower jaw.

BMNH 14566 has skull and snout lengths of approximately 860 and 680 mm, and an orbital diameter of
approximately 100 mm. If it were assumed that the specimen was not a eurhinosaur and that the mandible
was about as long as the skull, the snout and orbital ratios would be approximately 0-79 and 012. This
snout ratio is consistent with L. tenuirostris , despite the fact that the species is not known to extend into the
Upper Lias, but the orbital ratio is considerably smaller than that of any other Jurassic ichthyosaur. An
orbital diameter of only 100 mm, however, would be consistent with a eurhinosaur skull of 860 mm length.
BMNH R5465, for example, the smallest eurhinosaur with comparable data, has a skull length of 1035 mm
and an orbital diameter of approximately 125 mm, which gives a value of 012 for the orbital ratio, the same
as in BMNH 14566. Further preparation or radiography is required for confirmation, but the evidence
suggests that BMNH 14566 is referable to E. longirostris.

The second specimen, BMNH 36876 (Lydekker 1889, p. 91), is a rather poorly preserved, partially three-
dimensional skull, exposed from the left side. The snout, which is in several sections, is long and slender and
projects at an angle from the main block. Judged from the narrowness at its terminal end, the snout is

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

417

probably almost complete, and has a preserved length of 1025 mm. The orbit is large, almost circular, and
contains a fairly well-preserved sclerotic ring. According to Huene (1922, p. 39) there was no lower jaw, but
the mandible has been preserved and, although incomplete, its broken tip is only 18 mm deep, indicating
that little is missing. The preserved length of the jaw is 600 mm, which is less than half that of the skull
(approximately 1280 mm). There can be no doubt of the eurhinosaurian identity of the material, and
comparison with BMNH R3938, a similarly sized skull of E. longirostris from the Upper Lias of Germany,
shows a close correspondence in their measurements (Table 3).

table 3. Comparison of BMNH 36876 with an identified specimen of Eurhinosaurus

longirostris.

Specimen

Locality

Skull

length

Jaw

length

Orbital

diameter

Sclerotic

diameter

BMNH 36876

Whitby,

Yorks.

1280*

600*

174

70

BMNH R3938

Probably

Holzmaden

1312

648

179

71

* Measurement approximate, with no allowances for missing parts.

A second specimen, also numbered BMNH 36876, is also from the Upper Lias of Whitby. Tentatively
referred to L. tenuirostris by Lydekker (1889, p. 88), it comprises a partial snout and mandible, exposed from
the ventral aspect. The snout and mandibular sections are 850 mm and 600 mm long, giving an overbite of
250 mm. This estimate is based on the assumption that the snout and mandible have retained their natural
relationship with one another, but this might not be so. While it is unlikely that the mandible has shifted
backwards during preservation, it might have shifted forwards. This is because the mandible could be shifted
back from its present position and still retain a close correspondence between its width and that of the snout.
There might, therefore, have been a more extensive overbite than the present 250 mm. The material is too
incomplete to make a definitive identification, but it is more likely referable to Eurhinosaurus than to any
other taxon.

The structure of the tail. According to Riess (1986, p. 103): ‘ Eurhinosaurus . . . did not have a tail-bend. The
vertebrae at the end of the tail do decrease in size but there is no sudden decrease of vertebrae in the tail
nor a triangular vertebra which would indicate a tail-bend. Furthermore the pictures of unprepared
skeletons . . . and sketches of finds which were made available to me by R. Wild . . . all speak one plain
language: none of them show even an indication of a tail-bend.’ (Translation by E. Wolf.)

Riess illustrated his point by reference to the photographs of FSF 4155 taken before and after preparation
(Riess 1986, pi. 1, fig. 2; Hauff and Hauff 1981, pis. 40 41), in which there appears to be no evidence for a
natural tail bend. It is certainly true that it has been a common practice during the preparation of Holzmaden
ichthyosaurs to remove almost all of the matrix from around segments of the skeleton, such as the tail, and
then to drop these into recessed limestone reliefs. The tail bends depicted in such restorations are therefore
entirely unfounded, but this does not mean that a tail bend was not originally present, nor that all specimens
have been so modified, and each case must be judged on its own merits. Without extensive preparation, the
authenticity of an undocumented specimen will always be in doubt, but the identification of wedge-shaped
vertebrae in the vicinity of the tail bend is persuasive evidence that a tail bend was present. Given the
infrequency with which vertebrae are exposed from the lateral aspect it is not surprising that Riess (1986)
should report the absence of wedge-shaped centra in Eurhinosaurus. However, I have found a wedge-shaped
centrum in the specimen he illustrated (FSF 4155). This vertebra (no. 91) has a diameter of 30 mm and is
15-5 mm wide dorsally and 12 mm wide ventrally. Furthermore, it occurs at a point where there is a marked
decline in the rate of reduction of vertebral diameters. It is difficult to deny that a tail bend was present in
this specimen, but the angle of the tail bend as depicted in the prepared skeleton may bear little relationship
to reality. A careful re-examination of all Eurhinosaurus material is obviously needed, but in the meantime
it may be noted that this genus probably had a tail bench

418

PALAEONTOLOGY, VOLUME 32

DISCUSSION

A fairly wide range of variation is seen among the individuals referred to L. tenuirostris, which
raises the question of whether they really do all belong to the same species. How likely is it, for
example, that one member of a species has a complex external naris while another has a simple
opening, or that one individual has a snout which is considerably longer than that of another?
Some measure of the amount of individual variation to be expected within an ichthyosaurian
species is obviously required, but there are no criteria for recognizing biological species within the
fossil record (except in those rare instances provided by maternal ichthyosaurs— see McGowan
1979). In the absence of direct means of assessing individual variation, E. longirostris would appear
to serve as a suitable yardstick. This is because, being so highly specialized and distinctly different
from all other ichthyosaurs, it is likely to represent a single species, as is its modern analogue, the
swordfish ( Xiphias gladius). It must be remembered, though, that there is great variety in the range
of individual variation among living animals, even among closely related ones. E. longirostris can,
therefore, only give an indication of the degree of individual variation that might be expected
within an ichthyosaurian species.

A wide range of variation has been found in Eurhinosaurus , both in continuous and discontinuous
characters, and some may be attributable to sexual dimorphism (McGowan 1979). Some individuals,
for example, have a total digital count of five, others four; the counts to the pelvis and to the tail
bend range between 45 and 49 and between 91 and 95; some individuals have a complex bilobed
naris while others have a simple opening, and the snout ratio varies between 1-42 and 1-93. The
swordfish, similarly, has a wide range of variation in its snout and mandibular proportions
(McGowan 1988), and some differences in body proportions are possibly attributable to sexual
dimorphism (Alvarado Bremer 1988). The variability seen among specimens here referred to
L. tenuirostris , therefore, probably does represent individual variation rather than the unwitting
lumping together of individuals belonging to separate biological species.

The tendency in L. tenuirostris for the tip of the rostrum to extend beyond the mandible lends
support to its possible ancestral relationship to Excalibosaurus costini , a species characterized by
an extensive overbite (McGowan 1986). The overbite in E. costini amounts to about 35% of the
snout length, compared with a maximum of about 15% in BMNH 2009, here described as a variant
individual of L. tenuirostris (see Appendix). But the most extreme rostral development is seen in
Eurhinosaurus longirostris, where the overbite approaches 60% of the snout length. The possibility
that E. longirostris may have been derived from Excalibosaurus costini is discussed elsewhere
(McGowan 1986; in press).

The extension of the geographical range of Eurhinosaurus into the Whitby locality of England,
suggested by Huene (1922), is now established. This is not surprising in view of the similarity in
age of the Whitby and Holzmaden localities and of their close proximity (less than 1000 km). Nor
is this unprecedented— the predominently German species Stenopterygius hauffianus, for example,
also occurs in the Upper Lias of Ilminster, Somerset. Wide geographical ranges appear to be the
rule rather than the exception for ichthyosaurian species (McGowan 1978).

Acknowledgements. I am sincerely grateful to a large number of people who have given access to material in
their care, or who have helped in other ways. With apologies for any names inadvertently missed I thank
the following: Annabel Ainsworth, Cathy Arden, Chris Arden, Michael Bassett, Joan Burke, Alan Charig,
Chris Collins, John Cooke, David Costain, Ronald Croucher, Jill Crowther, Peter Crowther, Mark Davis,
Paul Ensom, John Fowles, Chris Hawks, David Hill, Steven Howe, Andrew Kendal, Michael Lambert,
David Langdon, Cindy Langham, Peter Langham, Robert Langham, Andrew Leitch, Rosemary McDougal,
John Martin, Angela Milner, Philip Palmer, C. W. Pamplin, Dennis Parson, Phillip Powell, Dianne Smith,
David Sole, Michael Taylor, Jeffrey Thomason, Alan Timms, and Erica Wolff.

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

419

REFERENCES

alvarado bremer, j. r. 1988. Quantitative comparisons of allometric growth and shape in the swordfish
(Xiphias gladius). M.Sc. thesis (unpublished). University of Toronto.
appleby, R. m. 1979. The affinities of Liassic and later ichthyosaurs. Palaeontology , 22, 92U946.
arkell, w. J. 1933. The Jurassic System in Great Britain , 681 pp. Clarendon Press, Oxford.

HARLAND, W. B., COX, A. V., LLEWELLYN, P. G., PICKTON, C. A. G., SMITH, A. G. and WALTERS, R. 1982. A
geological time scale , 1 3 1 pp. Cambridge University Press, Cambridge.
hauff, b. and hauff, r. b. 1981. Das Holzmadenbuch , 136 pp. Hauff, Holzmaden.

hawkins, T. 1834. Memoirs of Ichthyosauri and Plesiosauri , extinct monsters of the ancient earth , 58 pp. Relfe
and Fletcher, London.

HUENE, F. VON. 1922. Die Ichthyosaurier des Lias und ihre Zusammenhange, 1 14 pp. Gebruder Borntraeger,
Berlin.

jager, G. F. 1856. fiber eine neue Species von Ichthyosauren (Ichthyosaurus longirostris Owen et Jager). Nova
Acta Acad. Leop.-Carol. 25, 937-967.

Johnson, r. 1979. The osteology of the pectoral complex of Stenopterygius Jaekel ( Reptilia: Ichthyosauria).

Neues Jb. Geol. Palaont. Abh. 159, 41 -86.
konig, c. D. E. 1825. leones fossilium sectiles , 4 pp. Centuria Prima, London.

kuhn, o. 1934. Ichthyosauria. In quenstedt, w. (ed.). Fossilium Catalogus , 63, 3-75. W. Junk, Berlin.
lydekker, R. 1889. Catalogue of the fossil Reptilia and Amphibia in the British Museum ( Natural History ),
part 2, 307 pp. British Museum (Natural History), London.
mcgowan, c. 1972. The distinction between latipinnate and longipinnate ichthyosaurs. Life Sci. Occ. Pap.
R. Out. Mus. 20, I 8.

— 1974a. A revision of the longipinnate ichthyosaurs of the Lower Jurassic of England, with descriptions
of two new species (Reptilia: Ichthyosauria). Life Sci. Contr. R. Out. Mus. 97, 1 -37.

— 19746. A revision of the latipinnate ichthyosaurs of the Lower Jurassic of England (Reptilia:
Ichthyosauria). Ibid. 100, 1-30.

1976. The description and phenetic relationships of a new ichthyosaur genus from the Upper Jurassic

of England. Can. Jl Earth Sci. 13, 668-683.

— 1978. Further evidence for the wide geographical distribution of ichthyosaur taxa (Reptilia: Ichthyo-
sauria). J. Paleont. 52, 1155 1162.

— 1979. A revision of the Lower Jurassic ichthyosaurs of Germany with descriptions of two new species.
Palaeontographica, A166, 93-135.

— 1986. A putative ancestor for the swordfish-like ichthyosaur Eurhinosaurus. Nature , Land. 322, 454 456.

— 1988. Differential development of the rostrum and mandible of the swordfish (Xiphias gladius) during
ontogeny and its possible functional significance. Can. Jl Zool. 66, 496-503.

mantell, G. A. 1851. Petrifications and their teachings ; or , a hand-book to the gallery of organic remains of
the British Museum , 496 pp. H. G. Bohn, London.

OWEN, R. 1881. A monograph of the fossil Reptilia of the Liassic formations. Part 3. Ichthyopterygia , 83-134.
Palaeontographical Society, London.

pamplin, c. m. 1987. ‘The Leaping Ichthyosaur.’ Fossil Forum , Palaeo-Enterprises Publications, 1 (3), centre
page illustration.

riess, J. 1986. Fortbewegunsweise, Schwimmbiophysik, und Phylogenie der Ichthyosaurier. Palaeontographica ,
A 192, 93-155.

smith, h. m. 1970. Nomina and Taxa Dubia. Syst. Zool. 19, 94.

Typescript received 20 April 1988
Revised typescript received 20 August 1988

c. MCGOWAN

Department of Vertebrate Palaeontology
Royal Ontario Museum
100 Queen’s Park
Toronto, Canada M5S 2C6
and

Department of Zoology
University of Toronto

420

PALAEONTOLOGY, VOLUME 32

APPENDIX

Specimens referred to L. tenuirostris

1. OUM J10305 (text-figs. 1a and 2a). This almost complete and rather well-preserved skeleton, from Street,
comprises a number of blocks set into a plaster relief (McGowan 19746, fig. 12 6). The matrix lacks chisel
marks (see below) and there are no grounds to question the authenticity of the specimen. Because some of
the anterior vertebrae are overlain by other bones, difficulties were encountered in making vertebral counts,
but the error is not likely to be greater than + 1. The vertebral column is not sharply downturned, but the
presence of a tail bend is confirmed by the presence of wedged-shaped centra at levels 85 and 87 (the feature
is obscured in vertebra 86). Thus the centrum of vertebra 85, which is 21 mm high, has dorsal and ventral
widths of 8 and 7 mm; measurements for vertebra 87 are 19, 8, and 6 mm, respectively. Because of incomplete
preservation, and some displacement of the pelvic girdle, it is not possible to determine the vertebral count
to the pelvis with precision; the count is between 45 and 47, and the median value of 46 will be recorded.
The pelvic condition is indeterminate.

The tip of the snout has a small (9 mm) extension which may be a displaced tooth, or a bony process from
the premaxilla. The former assumption has been made, but if this should prove to be incorrect, 9 mm would
have to be added to all relevant measurements. This would elevate the snout, premaxillary, and prenarial
ratios from 0.72, 0.52, and 0.60 to 0.74, 0.53, and 0.62, respectively. The external naris has a somewhat
bilobed appearance, with an expanded posterior portion. The teeth are slender, especially towards the tip
of the rostrum, and are relatively small. One of the larger teeth, for example (at the tip of the maxilla), is
12-5 mm long and 4 mm wide.

There are two partial forefins, the anterior one overlying much of the other and comprising four digits
with large discoidal phalanges. The humerus has a relatively narrow shaft, much widened distally, with the
leading edge expanded proximodistally forming a facet. This leading edge facet, a distinguishing feature of
the species (see Table 2), is also seen in many Upper Lias species (McGowan 1979, pis. 2, 4, 5) and may
have served for the transmission of the radial artery and nerve (Johnson 1979, p. 68). The radius has a notch
on its leading edge, and the occlusal edges of the radius and ulna enclose a small foramen. An unusual
feature of the anteriormost forefin is that the radiale has a small emargination on its distal margin. The
otherwise normal appearance of this oblong element discounts the possibility that it has simply been rotated
through 90°. The coracoid is largely indeterminate but appears to lack a posterior notch. The hindfin appears
to have three digits, the femur is slender shafted and widely expanded distally. The anterior margins of the
tibia and tibiale are broadly emarginated.

2. BMNH R498. This specimen, from Street, Somerset, comprises several blocks set in plaster, colour-
matched to the matrix (McGowan 19746, fig. 12a). The matrix bears chisel marks, but the pattern is not like
that seen in BGS 51236 or in BMB C2 and, although there are patches of plaster, some with chisel patterns,
most of the matrix appears to be original. The tail bend is not an obvious feature because of the dorsal
exposure of the skeleton. However, there is evidence of a change in the diameters of the centra at the level
of the 85th vertebra, and this probably marks the position of the tail bend. This is confirmed by the presence
of wedge-shaped centra; vertebra 84 is slightly wedge-shaped, 85 is markedly so (vertical height of centrum
approximately 23 mm, dorsal and ventral widths approximately 7 0 and 4-2 mm), and vertebra 86 is the most
strongly wedge-shaped (measurements 20, 9, and 4-5 mm, respectively). The vertebral count to the pelvic
girdle is 45. The pelvic condition is indeterminate.

Because the skull, which has been dorsoventrally compressed, lies partially embedded in matrix, few reliable
measurements can be made. The external naris is not well preserved but there appears to be an expanded
posterior portion, as in OUM J 10305. The teeth are slender.

The forefin, seemingly complete on the right but obviously incomplete on the left side, has a total digital
count of four, with 15 elements in the longest digit. The individual phalanges, which are discoidal, are well
spaced distally and this appears to be natural rather than manufactured. The humerus has a relatively narrow
shaft, much expanded distally, with a well-developed leading-edge facet. The leading edge of the radius is
deeply notched. There is no evidence of a foramen between the contact edges of the radius and ulna for the
right forefin, but there is evidence of one for the left side. The coracoid is indeterminate. The femur is broadly
expanded distally and has a narrow shaft. The tibia and tibiale have broad emarginations on their anterior
edges, there are only three digits, and the distal phalanges are discoidal.

3. BGS 51236 (text-fig. 1b). The initial impression is one of a complete skeleton lying in a single block of
matrix that has been worked with a chisel (McGowan 19746, fig. 11). However, some judicious probing with

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

421

a mounted needle reveals that the specimen comprises several separate blocks surrounded by an artificial
matrix, the whole having been veneered with what appears to be a grey pigmented plaster bearing chisel
marks. In places this veneer has cracked, and the original matrix, which has a more yellow colour, can clearly
be seen beneath. Many cracks run across the specimen, especially through the post-sacral vertebral column,
and some of these are wide and infilled with the grey plaster. It is quite likely that the specimen has been
tampered with but this could only be established by an extensive and invasive investigation. The following
description, which emends the previous account (McGowan 19746), is therefore given with the reservation
that the authenticity of the material is in question.

The vertebral column lacks any obvious tail bend, the tail having been thrown into coils in this region.
However, there is evidence of a constriction at level 84 indicating a tail bend, as previously reported, and
this now appears to be confirmed by identifying three vertebrae with slightly wedge-shaped centra; the centra
of vertebrae 85 and 86 are 21 and 19 mm high, with dorsal and ventral widths of 9 mm and 7 mm and 7
and 6 mm, respectively. The centrum of vertebra 87, which is 18 mm high, is narrower dorsally (7 mm) than
ventrally (8 mm) and this would cancel out the downturn effect of the previous centrum. Whether any of the
centra anterior to vertebra 84, or posterior to 87, are wedge shaped is not known. In any event it seems
unlikely that there was an effective tail bend, which would explain why there is no obvious tail bend in this
specimen. The count to the pelvic girdle is 47. The pelvic girdle is tripartite, the pubis and ischium being
quite separate.

Close inspection of the skull reveals that the extreme tip of the snout is missing, but it is already so narrow
at this point that it is unlikely that very much has been lost, and this is estimated to be about 10 mm. Making
allowances for the missing tip changes the previously given cranial ratios; the snout ratio increases from 0-72
to 0-74, the premaxillary ratio from 0-52 to 0-54, and the prenarial ratio increases from 0-60 to 0-61. The
mandible, which appears to be complete, stops 9 mm short of the broken tip of the snout, and when
allowances are made for the missing tip of the snout this overbite is increased to 19 mm. The orbital ratio
has been modified from 0-23 to 0-24. The external naris is bilobed, with a narrow tongue of bone separating
a narrow lower portion from the rest. The teeth are very slender; one from the anterior tip of the maxilla is
13 mm long and about 4 mm wide.

Note was taken in the previous account that the forefin measurements were unreliable because the fins
had been reconstructed distally. Nevertheless, there is good agreement between left and right sides in the
measurements of the humerus, radius and ulna, suggesting that measurements of these elements are reliable.
The proximal articular surface of the left radiale looks as if it has been modified to articulate with the radius.
The coracoids are rectilinear rather than being typically rounded, raising doubts about their authenticity.

4. SCM 8372 (text-figs, lc and 2b). This specimen, which lies on its right side, consists of two main blocks
set into a plaster relief, but this has not been made to blend with the matrix and the relationship between
the two main blocks, and between their component parts, is good. This is, therefore, considered to be one
of the most reliable of the specimens referred to L. tenuirostris.

There is a tail bend, which appears to be natural, and vertebra 82, which is at the beginning of the tail
bend, is wedge-shaped. The height, dorsal, and ventral widths of this centrum are approximately 19, 9-3, and
6-5 mm, respectively. The centrum of vertebra 84 also appears to be wedge-shaped, but its dorsal and ventral
widths are about the same. The apex of the tail bend therefore appears to lie between vertebrae 82 and 84
and the count will be recorded as 83. The vertebral count to the pelvic girdle is 45. The pelvis is tripartite,
with no evidence of fusion between the pubis and ischium.

The skull is essentially complete but has been badly crushed, making it difficult to interpret some of its
features. There is a particular problem at the tip of the rostrum and it is not clear whether the tips of the
right mandible and right side of the snout are being seen, or whether the mandible stops short and what is
being seen are the tips of the left and right halves of the snout. The assumption is made that the mandible
extends to the tip of the snout. A sub-terminal portion of the rostrum— a segment about 140 mm long— was
stolen from the specimen while on display, and has been restored in plaster. Teeth are not plentiful, though
several can be seen, and these are small and very slender. A tooth from close to the tip of the snout, for
example, is 9 mm long and about 1 mm wide. The external naris appears to be bilobed, but further preparation
is needed to determine its shape. The naris appears to be continued anteriorly as a narrow slit, reminiscent
of the condition in Excalibosaurus costini (McGowan, 1986) but this could be an effect of preservation. This
crack-like extension is separated from the external naris proper by a constriction, and the latter has been
taken as the anterior boundary of the naris in all measurements. If the anterior extension is later determined
to be an integral part of the naris, it will be necessary to decrease the prenarial ratio from the present value
of 0 61 to 0-59.

422

PALAEONTOLOGY, VOLUME 32

The forefin, which has only 10 elements in the longest digit and is probably not complete, has three or
perhaps four primary digits and a total digital count of four, possibly five. The humerus has a narrow shaft,
is broadly expanded distally, and has a prominent leading edge facet. The anterior edge of the radius is
broadly notched and a prominent foramen is enclosed between the contact edges of the radius and ulna. The
coracoid is rounded, not angular as in BGS 51236, with a fairly broad anterior notch. The posterolateral
margin appears to be scalloped but this has probably been caused by its being crushed against the underlying
ribs.

5. WM 527 . This specimen, from Mandeville, near Street, is preserved from the dorsal aspect and faces
towards the left. It would appear to lie on a single slab of matrix, but this is much cracked and infilled and,
without an extensive investigation, it cannot be determined whether all the parts belong together. However,
with the possible exception of a crack immediately posterior to the pelvic girdle, there is nothing to arouse
suspicion and the assumption is made that the specimen is authentic. Although the entire length of the skull
is preserved, the skull roof has been lost so that few measurements can be taken. Furthermore, although the
posterior limit of the mandible can be determined, its anterior tip cannot. The assumption is made that the
mandible extended to the tip of the snout because this is the usual situation. The vertebral count to the pelvic
girdle is approximately 45. The pelvis is fairly well preserved and clearly shows that the pubis and ischium
are not fused.

The most complete of the two forefins is well preserved but is incomplete distally. The humerus has a
narrow shaft, is much widened distally, and has a prominent leading edge facet. The radius and radiale are
notched but there is no foramen between the radius and ulna. The radius is partially fused with the humerus
and with the ulna. The coracoid is largely indeterminate, being overlain by other elements, but it appears to
be rounded, not rectilinear as in BGS 51236.

6. GTS L/AG/Arch/18. This near-complete skeleton from Street lies on its right side, and comprises two
major blocks, set in plaster. Some of the matrix bears chisel marks, but these appear to be genuine. The
vertebral column has a distinct tail bend, but an oblique crack across the matrix at this level has been infilled
with plaster, painted to match the colour of the matrix. This raises the possibility that the specimen has been
tampered with but it is most unlikely that the bend has been manufactured from an originally straight tail
because this would have required inserting a wedge of matrix, which has not been done. The possibility
cannot be dismissed that the segment posterior to the crack has been added from a second specimen, though
the two broken edges, which are only separated by a gap of about 3 mm, appear to correspond with one
another fairly well. For the present it will be assumed that the vertebral column is complete. The vertebral
counts are difficult to make because of displacements in the thoracic region, and could be underestimated by
between two and three; the counts to the pelvis and to the tail bend are 44 and 86. The centrum of vertebra
86, which is approximately 20 mm high, is slightly wedge-shaped, being 9 mm wide dorsally and 7-5 mm
ventrally. The pubis and ischium are unfused.

An unusual feature of this specimen is that most of the rostral portion of the skull is missing, as are the
anterior portions of both mandibles. Indeed, all that remains of the mandibles are the surangulars, and while
this preservation is unusual it is not without precedent because there are two isolated surangulars in the
collections of the British Museum (BMNH 2122X and 2129X). No measurements are possible for the skull.

The forefins are typical of L. tenuirostris. The humerus has a narrow shaft, widely expanded distally, with
a leading edge facet. The radius is deeply notched and encloses a prominent foramen with the ulna. There
are four digits, the phalanges are discoidal and there are only 1 1 elements in the longest digit of the most
complete (right) fin, though there are probably a few elements missing terminally. Neither the coracoids, nor
the hindfins, are well preserved.

7. BATGM M3552. This skeleton, which lies on its right side, comprises a skull —complete except for its
tip — a fairly complete and well-preserved forefin and coracoid, and scattered vertebrae and ribs. The blocks
bearing the bones are set in plaster but the relationships between them appears to be good. No vertebral
counts are possible.

The tip of the snout and mandible have been partially and inexpertly restored in plaster and it is estimated
that approximately 20 mm has been lost. This has been taken into account in the cranial measure-
ments, which are therefore partly estimated. The snout is long and slender, typical of the species. Teeth are
numerous, slender, and are about 1 cm long. The orbit is well rounded; the shape of the external naris is
indeterminate.

The humerus is fairly broad distally, and has a leading edge facet. The radius and radiale are both notched,
and a prominent foramen is enclosed between the radius and ulna. The phalanges are discoidal, but the fin

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

423

is too incomplete to determine either the number of elements in the longest digit, or the number of digits —
three digits are preserved. The coracoid is rounded, with a broad anterior, and a posterior emargination.

8. BATGM M3556. This partial skeleton from Street appears to be exemplary of the species but, because of
its generally poor preservation, few measurements can be taken. The skeleton lies on its left side, but the left
forefin has been flipped up above the vertebral column, where it rests on its ventral surface. The block bearing
the skeleton is set in a plaster relief and has numerous cracks running through it, some deep and wide. Little
remains of the postsacral skeleton; the vertebral count to the pelvis is 46, possibly 45. The pelvic girdle is
well preserved and there is partial fusion between the pubis and ischium.

Most of the prenarial segment of the skull is missing, and that which remains is incomplete and displaced.
The few scattered teeth that can be seen are relatively small, and while some are very slender others are
conical.

The forefin has four digits, the phalanges are discoidal, the radius is notched and there is a foramen
enclosed between the radius and ulna. The humerus is broadly expanded distally and has a well-developed
leading edge facet. The coracoid is longer than it is wide, tends to be rounded rather than rectilinear, and
has a single, round, anterior notch that is relatively small. Little remains of the hindfin; the femur is fairly
slender and widens distally, as in other referred specimens. The tibia has a broad notch that occupies much
of its leading edge.

9. BATGM M3565. This partial skeleton, which lies on its left side, comprises several blocks set in plaster.
The skull has several cracks and repairs, and while the contact edges of the broken elements appear to match
satisfactorily, it is not known how good these matches are. The accuracy of the cranial measurements are
therefore uncertain. No vertebral counts could be made; the pelvic girdle is indeterminate.

The snout and mandible are both long and slender, typical of the species, but the orbit appears rather
large. The orbital ratio is correspondingly high (0-24) but is within the limits previously diagnosed for
L. temiirostris. The external naris, which also appears to be relatively large, has something of a posterior
expansion, but its shape is partially obscured by a displaced bone. Teeth occur throughout; they are slender
and relatively small and one of the largest ones is 1 1 mm long and 4 mm wide.

The forefin is robust. The humerus is broad, widely expanded distally, and has a prominent leading edge
facet. There appear to be four digits and the phalanges are discoidal. The radius and radiale are both notched,
but there is no foramen enclosed between the radius and ulna. The coracoid is longer than broad, with a
relatively long and straight intercoracoid facet and a single anterior emargination.

10. BATGM M3566. This large and incomplete skeleton lies with the skull, which lacks jaws, exposed from
the ventral aspect. The specimen essentially occupies a single block set in plaster. No vertebral counts can
be made and there is no determinate pelvis.

The skull is approximately 690 mm long and is the second largest referred specimen treated here. The
snout is long and slender but few measurements can be taken because of its orientation. Few teeth are
preserved and these are relatively small and are conical rather than slender. One of the largest teeth is 9-6 mm
long and 4 mm wide.

The humerus has a constricted shaft and is widened distally, typical of L. tenuirostris , but it has a relatively
small leading edge facet. The radius is notched but there is no foramen enclosed with the ulna — the remainder
of the fin is indeterminate. The coracoid is considerably longer than it is wide (length 125 mm, width 75 mm)
with a discrete anterior notch; the posterior edge appears to be broadly emarginated. The well-preserved
scapula is robust.

1 1 . LEICS OS.90.1953. The snout of this rather poorly preserved skeleton, from Barrow-on-Soar, Leicester-
shire, is incomplete, and it is not possible to estimate how much has been lost. Conceivably the snout could
have been extremely long, as in BMNH 36182 from the same locality, and the specimen might therefore be
atypical of the species (see p. 425 below). Few measurements can be taken from the skull, or from the rest
of the skeleton, but the material is considered important enough to be included here, especially since it is
only one of three specimens from this locality. The skeleton is exposed from the right side and comprises
several blocks that have been set into a plaster relief. The vertebral column is depicted as being straight but
the caudal vertebrae, which have been largely freed from matrix, have been arranged in a line without any
obvious keying-up of the individual segments. Therefore, if a tail bend had been present when the specimen
was found, this would have been lost during preparation. A disturbance in the regular order of the vertebrae
occurs at about level 85 which may mark the position of a tail bend, but wedge-shaped centra cannot be
seen. The vertebral count to the pelvic girdle is 48.

424

PALAEONTOLOGY, VOLUME 32

The external naris requires some preparation, and although it appears to have a simple petaloid shape
there is indication of a slit-like extension of its anterior boundary. Poor preparation obscures most of the
details of the teeth, but they are numerous and appear to be slender.

The incomplete forefin has a humerus with a relatively narrow shaft, much widened distally and with a
leading edge facet. The radius bears a prominent notch on its leading edge and a small foramen is enclosed
between the contact edges of the radius and ulna. The slender femur has a wide distal expansion.

text-fig. 3. DLR ‘002’, probably representing a mature individual of L. tenuirostris , x0-06. Drawing by
C. M. Pamplin, who kindly gave permission for its inclusion here.

12. DLR ‘ 002 ’ (text-fig. 3). This rather important skeleton is tentatively identified as being a mature individual
of L. tenuirostris. It was found in 1979 by Robert Langham in Pinhay Bay, just west of Lyme Regis, in the
adjoining county of Devon. The horizon is given as that of Arietites bucklandi , which places the material at
the beginning of the Sinemurian. The skeleton, which is almost complete, lies with its left side embedded in
a thin sheet of matrix which has been trimmed to the approximate outline of the skeleton. Its arched posture
has given rise to the epithet, ‘the leaping ichthyosaur’ (Pamplin 1987). The material has undergone considerable
compression; the skull, for example, is only about 13 mm thick at the level of the external nares. This has
undoubtedly distorted the specimen, exaggerating the depth of the skull and mandible and making the
humerus appear much broader than it was in life. The tip of the skull broke off during collection, an estimated
loss of approximately 10 mm (Robert Langham, pers. comm.) having been allowed for in all relevant
measurements. The single hindfin, which is incomplete, was placed in its present position during preparation
and is therefore not in its natural position.

There is no obvious tail bend but there is a disturbance in the orderly arrangement of the vertebrae at
about level 83, and this is accompanied by a marked decrease in the diameters of the centra. Close inspection
reveals that vertebra 85 is wedge-shaped; the centrum has a height of 31 mm and is 14 mm wide dorsally
and 1 1 mm wide some two-thirds of the way down (the ventral width, which is less than 1 1 mm, cannot be
measured without further preparation). The pelvic girdle has not been preserved and since the hindfin is not
in its natural position it cannot be used to determine the level of the pelvis. However, given that ichthyosaurian
ribs become reduced in length at the level of the pelvic girdle, it is possible to estimate the position of the
pelvis by detecting this change. The rib associated with vertebra 46 is 70 mm long while those of vertebrae
47 and 48 are both 53 mm long. This indicates a vertebral count to the pelvis of approximately 47.

The skull has a long slender snout and the snout ratio of 0-75 falls within the diagnosed limits for
L. tenuirostris. The orbit is difficult to measure because of uncertainties in its posterior margin, but an
estimate of its diameter gives an orbital ratio of 0-18. While this is lower than that of specimens that

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

425

have been referred, without reservation, to L. tenuirostris, it is comparable to that of one tentatively referred
specimen (BMNH 36182— orbital ratios 016). Because of poor preservation the anterior tip of the maxilla
is difficult to discern, but an estimate of its position gives a value of 0-53 for the premaxillary ratio, which
is within diagnosed limits for L. tenuirostris. The same holds true for the prenarial ratio of 0-58. The external
naris is not immediately apparent because of the effects of crushing, which lias caused the elements forming
its borders to be flattened against the bones of the other side of the skull. The naris is remarkable for its
large size and for its complex bilobed shape. Anteriorly it is continued as a narrow slit, as in SCM 8372, but
whether this is truly part of the naris or merely a result of crushing cannot be determined. The teeth are far
more sparse than they are in most other specimens (e.g. OUM 10305, BGS 51236, and LEICS OS.90.1953)
and are relatively minute compared to the size of the skull; one of the largest teeth is only 13 mm long and
3 mm wide.

The best preserved forefin appears to be complete, but has been restored in plaster distal to the first row
of phalanges (Peter Langham, pers. comm.). It is, therefore, not possible to determine the total number of
elements in the longest digit, and the total number of digits may have exceeded the three that have been
preserved. The humerus appears to be rather broad, but this is almost certainly attributable to crushing and
the shaft is obviously narrow, typical of L. tenuirostris. It is not possible to determine whether there is a
leading edge facet because this region is partially overlain by the displaced left mandible. The radius is
notched, and a foramen is enclosed between the radius and ulna. Both coracoids are fairly well exposed and
are rounded, almost discoidal in shape, with a small notch on the anterior margin.

Referring DLR '002’ to L. tenuirostris extends the upper size limit of the species considerably, but it is
fairly certain that it is correctly assigned because its other features are typical of the species, namely: relatively
long snout, complex external naris, vertebral count to the tail bend of approximately 85, distally expanded
humerus, notched radius, foramen enclosed between the radius and ulna.

Atypical specimens probably representing variant individuals of L. tenuirostris

1. BMNH 2009 (text-fig. 4a). This well-preserved skull, which is exposed from the left side, was figured by
Hawkins (1834, pi. 13). Initially the skull appears to be a perfectly ordinary example of /. tenuirostris , but
closer inspection reveals two departures from the norm, namely an extensive overbite and an unusually
slender mandible. Before these features can be discussed it is necessary to consider the state of preservation.

The tip of the snout appears to lie at the same level as that of the mandible, but this is not so because the
two halves of the mandible have moved relative to one another and what appears to be the tip of the left
mandible, lying in line with the tip of the left side of the snout, is in fact the right half of the mandible. The
tip of the left half of the mandible lies about 18 mm posterior to the tip of the snout, giving an overbite of
the same amount. But the overbite is in fact greater than this, for two reasons. First, the very tip of the snout
is missing— between 10 and 20 mm. Secondly, the left mandible has been displaced forwards, the posterior
end of the left mandible having come to rest about 20 mm anterior to the posterior margin of the orbit. To
restore the left mandible to its natural position would require moving it back about 30 mm. Allowing for
this, and for the missing tip of the snout, gives an overbite of between some 60 and 70 mm, which is about
15% of the length of the snout. Having an elongated snout and an abbreviated jaw gives the relatively high
value of 0-89 for the snout ratio.

Remarkable too is the relative slenderness of the jaw. This is revealed by comparing BMNH 2009 with
OUM J 10305, a typical specimen of L. tenuirostris of similar skull length (Table 4). Although their snouts
are similar in depth, the jaw is only about half as deep in BMNH 2009, and the snout is longer and the jaw
shorter. In other characters, including the possession of slender teeth and a relatively long external naris (its
outline is partly obscured by a displaced tooth), BMNH 2009 is typical of L. tenuirostris. I conclude that the
specimen probably represents a variant individual, and the data will not be used to contribute to the diagnosis
of the species.

2. BMNH 36182 (text-fig. 4b). Owen’s 'least incomplete’ exemplar of /. longirostris (1881, pp. 124-126,
pi. 32, fig. 7), is a large, almost complete skeleton which lies with its right side exposed. The material is quite
well preserved, but its preparation could be much improved by further work. The bone and matrix are a
distinctive dark brown, which appears to be characteristic of its Barrow-on-Soar locality, and the matrix is
very hard. Several deep cracks run through the specimen, and there is a large gap in the vertebral column
at the level of the pelvic girdle which has been filled in with a cement. However, the specimen has probably
not been tampered with. The most striking feature is the remarkably long and slender snout, unsurpassed in
any of the material hitherto referred to L. tenuirostris. There arc no comparably sized skulls of L. tenuirostris
with adequate data for comparison, but in BGS 51236, which is less than two-thirds as long, the depths of

426

PALAEONTOLOGY, VOLUME 32

text-fig. 4. Specimens probably representing variant individuals of L. tenuirostris. a, BMNH 2009, x 0-26.

B, BMNH 36182, xO-19.

table 4. Comparison of BMNH 2009 and OUM J10305.

Specimen

Skull

length

Jaw

length

Orbital

diameter

Snout

length

BMNH 2009

539 0*

482-0

100-0

429-0*

OUM J 10305

523-0

550-0

1 14-0

398-0

SNOUT DEPTH

S-S2

S-J2

S-M

S-N

BMNH 2009

14 6

15-3

20-5

32-0

OUM J 10305

14 0

20-0

20-0

33-0

JAW DEPTH

J-S2

J-J2

J-M

J-N

BMNH 2009

110

8-5

10-0

14-6

OUM J 10305

16 0

21-0

22-0

25-0

Allowance made for missing tip of snout.

McGOWAN: LONG-SNOUTED ICHTHYOSAURS

427

table 5. Comparison of BGS 51236 and BMNH 36182.

SKULL LENGTH SNOUT DEPTH

Specimen

S-S2

S-J2

S-M

S-N

BGS 51236

570-0*

21-5

26-3

28-0

40-8

BMNH 36182

760.0*

14-0

1 8-0

25-0

36-0

JAW DEPTH

J-S2

J-J2

J-M

J-N

BGS 51236

17-2

18-0

19 0

26-0

BMNH 36182

1 0 0

12-0

14-0

16-0

* Allowance made for missing tip of snout.

the snout and jaw all exceed those of BMNH 36182 (Table 5). The snout appears to be relatively long, but
the snout ratio (0-78) is the same as that of BMNH R498. Although the orbit appears to be prominent (Owen
1881, p. 125 commented that it was relatively larger than in L. tenuirostris) it is relatively smaller (orbital
ratio 0T6) than in any referred specimens of L. tenuirostris. Few teeth can be seen, and then only their tips,
and it is not clear whether this is due to poor preservation, poor preparation, or to tooth reduction. The
external naris is largely obscured by overlying bone, but it appears to have the shape of a curved ellipse.

There is no tail bend and it is not possible to discern any wedge-shaped centra. However, there is a
disturbance in the orderly arrangement of the centra between vertebrae 80 and 84, accompanied by a marked
decrease in their diameters, and this might indicate the position of a tail bend. The vertebral count to the
pelvis is probably 45 (Owen counted 48). The pelvic girdle is partly indeterminate, but the pubis and ischium
are certainly not fused proximally.

The best preserved forefin is incomplete and the humerus is partially fused with the radius which makes
it difficult to determine its shape. The humerus appears to be expanded distally, there are four digits, and
the phalanges appear to be discoidal. There is a well-developed foramen between the radius and ulna in one
fin (the least complete fin), with some indication of one in the other, and the radius is deeply notched. In all
of these features the material is typical of L. tenuirostris. The coracoid is indeterminate.

The unusually slender rostrum no doubt influenced Owen’s (1881) decision to refer this specimen to a
separate species from L. tenuirostris. However, aside from this feature, and the relatively small orbit, the
material is consistent with what is known of L. tenuirostris , and it is concluded that it probably represents
a variant individual of the species. Further preparation might help clarify the situation, but for the present
the material is tentatively referred to L. tenuirostris , though its data were not used to contribute to the
diagnosis of the species.

SIZE-SELECTIVE TRANSPORT OF SHELLS BY BIRDS
AND ITS PALAEOECOLOGICAL IMPLICATIONS

by GERHARD C. CADEE

Abstract. Size-selective transport of shells is demonstrated in the Dutch Wadden Sea for the Herring Gull
( Larus argentatus ) and the Oystercatcher (Haematopus ostralegus ): only larger shells were transported. Size-
selective transport of shells by predators is one of the taphonomic processes altering the size-frequency
distribution of shells in the death assemblage: it results in mortality not recorded in the death assemblage
forming where the animal lives. Such transport will occur particularly in intertidal areas. It hampers the use
of size-frequency distribution for studies of population dynamics in fossil assemblages.

Size-frequency distribution in a fossil assemblage is dependent on the interplay of growth rate,
mortality rate, taphonomic processes, and usually also time averaging (Cummins et al. 1986). It
is therefore like an equation that cannot be solved because there are too many unknown variables.
Nevertheless, palaeoecologists have tried to use size-frequency distributions to extract data on
transport or population dynamics from fossil assemblages ever since the early stimulating papers
by Boucot (1953) and Kurten (1953) (see Hallam 1972; Cadee 1968, 1982; Cummins et al. 1986
for other references).

The formation of a death assemblage is the initial step in the creation of a fossil assemblage.
During this step the skeletal components of preservable organisms are subjected to such taphonomic
processes as dissolution, breakage, bioerosion, abrasion, transport, and time averaging (shell
condensation), all of which may alter the size-frequency distribution (Cummins et al. 1986).
The study of these ‘biostratinomic’ processes has a long history (see Schafer 1962; Muller 1976,
1979).

The importance of predation in producing shell fragments has been stressed by, among others,
Schafer (1962), Cadee (1968), and Trewin and Welsh (1976). In this paper I will concentrate on
transport of shells by predators which represents a process of mortality not recorded in the death
assemblage forming in the localities where the animals live. After 20 years of problems with
punctured bicycle tyres on my way to the laboratory due to shell-smashing activities of gulls, I
learned to see this activity as a mechanism by which shells are transported from the intertidal area
to land. Do these gulls select shells of a certain size? Such a selection was reported by Zwarts and
Drent (1981) for the Hooded Crow which uses the same smashing method. Moreover, a nearby
roost of Oystercatchers offered the opportunity to study transport of Mvtilus edulis to the roost
on land and possible size selection by this species. Such a transport was first reported by Leopold
et al. (1985) and found to be non size-selective.

Transport of shells from sea to land by birds foraging in the intertidal area has been reported
repeatedly. Most data pertain to gulls (Sunkel 1925; Schwarz 1932; Teichert and Serventy 1947;
Remane 1951; Goethe 1958; Schafer 1962; Barash et al. 1975; Siegfried 1977; Kent 1981) and
the palaeoecological importance has been stressed: marine shells transported to a terrestrial
habitat may pose problems in the correct interpretation of the terrestrial palaeoenvironment.
Transport of shells from the intertidal area, however, will also alter the death assemblage
of shells left behind, particularly if such a transport is quantitatively important. If such transport
is size selective, it will influence the size-frequency distribution of the death assemblage, presenting
problems in the use of size-frequency distribution in palaeoecology (e.g. for population
dynamics).

[Palaeontology, Vol. 32, Part 2, 1989, pp. 429-437. |

© The Palaeontological Association

430

PALAEONTOLOGY, VOLUME 32

text-fig. 1. Location of tidal flat (stippled), shell-smashing locality of Herring Gulls and Oystercatcher roost

studied.

OBSERVATIONS

A large sample was collected in February 1988 of fresh Mytilus shells recently dropped by Herring
Gulls ( Lams argentatus ) on the dike bordering a tidal flat area of the Dutch Wadden Sea along
the south-east coast of Texel, near polder Ceres (text-fig. 1 ). Most shells still contained some
adherent soft tissue. Shell length was measured to the nearest millimetre and the resulting

CADEE: SIZE-SELECTIVE SHELL TRANSPORT

431

SHELLS SMASHED BY HERRING GULLS
DIKE CERES TEXEL
20-2-1988
N = 722

LIVING ADULT MYTILUS (>35MM)
TIDAL FLAT CERES (TEXEL)
20/22-2-1988
NM89

LIVING MYTILUS
TIDAL FLAT CERES (TEXEL)
20-2-1988
N = 34 1

MYTILUS FROM OYSTERCATCHER ROOST
TEXEL NEAR CERES
24-2-1988
N = I 4 I

text-fig. 2. Size-frequency distributions of Mytilus edulis shells, a, dropped by Herring Gulls; b , living
population; c. living adults only; d , from Oystercatcher roost. Size classes 5 mm, all measurements to the

nearest mm.

size-frequency distribution is given in text-fig. 2a. Gulls collected these mussels from mussel-beds
on the tidal flat bordering the dike. The living mussel population was sampled there at a number
of localities within this feeding area for size-frequency distribution measurement (text-fig. 2b). A
comparison of the two size-frequency distributions reveals a marked difference: the size-frequency
distribution of the living population is bimodal, that of the dropped shells shows a normal
distribution. The smaller mussels, all year class 1987, were well represented in the living population
but absent in the assemblage of shells transported by gulls. Average size and standard deviation
were 25-9+ 19-4 and 58-7 + 6-9 mm for the living and the dropped population, respectively. This
clearly indicates a size selection by gulls. Only the larger mussels, of older year classes, were
transported.

To measure the size-frequency distribution of this older living population more exactly, a larger
sample was collected of only older Mytilus (text-fig. 2c). Size-frequency distributions of older living
Mytilus and shells transported to land and smashed are very similar (average and standard

432

PALAEONTOLOGY, VOLUME 32

SHELLS SMASHED BY HERRING GULLS
DIKE CERES TEXEL
FEBRUARY/MARCH 1988
N = 3 3 1

text-fig. 3. Size-frequency distribution
of Ensis directus shells smashed by Her-
ring Gulls.

deviation respectively 56-2 + 5-8 mm and 58-7 + 6-9 mm). Although the difference of 2-5 mm is
statistically significant (t-test), this most probably does not indicate size selection by gulls but
spatial variation within the mussel population and the difficulty in collecting exactly the population
which the gulls were feeding on.

Smashing activity of gulls occurs the year round but shows a seasonal variation with a peak in
winter. The condition index (ratio meat weight : shell volume) of Mytilus is lower in this period
than in summer (Zandee et al. 1980) because phytoplankton concentration is lowest in winter
(Cadee and Hegeman 1979). For the same amount of food a gull has to collect in February/March
twice as many Mytilus as in August/September. So it is not food quality which makes the gulls
select Mytilus particularly in winter, but probably a lower availability of other food types.

Between the smashed Mytilus shells on the dike a number of American razor clams (Ensis
directus) were found, also with adhering soft tissue, which apparently were smashed by Flerring
Gulls too. Subsequently, such shell smashing of Ensis by Flerring Gulls was indeed observed. E.
directus is a recent immigrant in the Dutch Wadden Sea (Essink 1985). Size-frequency distribution
of these shells also showed that only large specimens were transported (text-fig. 3). Not all Ensis
shells had been broken when dropped. Apparently some shells opened without breaking. Kent
(1981) reports the same for Argopecten dropped by Herring Gulls. Because of the difficulties in
sampling razor clams adequately (they are fast movers in the sediment), I did not try to collect a
representative sample of the living population for comparison. Instead growth was measured using
the growth rings on the shells to estimate the age of the transported razor clams. Measurements
of annual growth rings are given in text-fig. 4. Length increase is comparable to that found by
Swennen et al. (1985) in the German part of the Wadden Sea. From these growth rate data it is
concluded that Herring Gulls dropped only older Ensis shells with two to four winter rings.

E. directus is a recent addition to the food resources of birds in the Wadden Sea (Essink 1985;
Swennen et al. 1985). E. directus was an important food item for Herring Gulls and Oystercatchers
( Haematopus ostralegus ) foraging at extreme low tide on the tidal flat studied. The Herring Gulls
did not collect all these razor clams and mussels themselves, but stole some from Oystercatchers
(‘kleptoparasitism’). Swennen et al. (1985) observed that some Oystercatchers have specialized in
feeding on E. directus , a new kind of prey in the Wadden Sea where before its introduction no
razor clams occurred. Herring Gulls have also adapted to this new prey which they collect when
swimming above the flat at low tide and in water depths of 10 to 20 cm. The Herring Gulls used
the smashing method for Ensis only sporadically; in most cases they were able to feed on Ensis
without smashing.

CADEE: SIZE-SELECTIVE SHELL TRANSPORT

433

GROWTH CURVE ENSIS DIRECTUS text-fig. 4. Growth curve of Ensis di-

TIDAL FLAT CERES TEXEL rectus from tidal flat near Ceres.

FEBRUARY/MARCH 1988

M. edulis shells collected on the Oystercatcher roost showed a size distribution comparable to
that of the shells dropped by Herring Gulls: the smaller Mytilus of year class 1987 were not
transported, only older larger shells were found (text-fig. 2d). They were somewhat smaller (average
and standard deviation 52-3 + 7-4 mm) than the older shells in the living population. This may be
due to the fact that the Oystercatchers roosting here feed over a larger area of the tidal flat, which
was not adequately sampled. Inspection of other mussel-beds in this tidal flat area indeed indicated
such differences; in some places also the small mussels of the 1987 year class were almost lacking.
Another explanation could be that the Mytilus taken to the roost are the last ones the Oystercatcher
was able to collect before the feeding area became submerged. Mytilus living higher in the intertidal
zone are smaller because they have a shorter feeding period per tide (Dare 1976).

DISCUSSION

A number of methods by which birds transport shells have been described in the literature,
particularly concerning the Herring Gull. Shells, some not used as food by this bird, are transported
to the nest and used as ‘decoration’ (Goethe 1937). Males present quantities of bivalves to the
female (Tinbergen 1953; Goethe 1958), and these bivalves largely remain uneaten and intact near
the nest. Some small shells are brought to the breeding colony as food for chicks, but according
to Spaans (1971), fish form the main food for chicks. Quantitatively, however, transport of shells
for these purposes will be small compared with transport of bivalves as food for (sub)adults.
Smaller bivalves are ingested whole when small (9-23 mm length: Harris 1965; Spaans 1971) and
broken in the muscular stomach. Their remains are dropped as faeces or regurgitated as pellets
(cough balls) on land, on roosts, or in the breeding colony. Remane ( 1951 ) calculated this transport
to be 1450 tons per year for the German coast. Since then the number of Herring Gulls has
increased considerably (Smit and Wolff 1981).

Remane (1951) did not include the smashing of shells by Herring Gulls in his estimates. Larger
shells are first broken by smashing (Oldham 1930; Tinbergen 1953; Goethe 1958; Ingolfsson and
Estrella 1978; Kent 1981; personal observations) and the meat is extracted leaving the shells behind.
Shells for smashing are taken some 4 to 10 m in the air and dropped on the tidal fiat, but sometimes
they are brought to dry land. Most successful are those birds that drop shells on a hard surface

434

PALAEONTOLOGY, VOLUME 32

(rocks, pebbly beach, artificial dike, road), but Herring Gulls do not select hard surfaces according
to Oldham (1930) and Tinbergen (1953); moreover, hard surfaces are not always available (Kent
1981). However, Ingolfsson and Estrella (1978) found a marked preference for hard surfaces. I
found that whereas some gulls seem to select the road or the hard surface of the dike, others
dropped shells on the tidal flat, on grass, or even in water. Shells dropped on land break after
being dropped one or several times. Oldham (1930) and Smit and Wolff (1981) report smashing
of shells also for the Common Gull ( Larus canus). For a number of species of gulls in other parts
of the world the smashing method has been described (Teichert and Serventy 1947; Barash et al.
1975; Siegfried 1977; Kent 1981; Maron 1981), indicating its widespread occurrence among the
larger gull species. The Hooded Crow ( Corvus corone comix ) also displays this smashing method,
but they confine themselves to bombing hard soils (Tinbergen 1953). Whereas only the older
mussels were observed smashed (the smaller can be swallowed whole), without size selection among
these older mussels, Kent (1981) found that his Herring Gulls showed a decided preference for
smashing the largest available prey. Siegfried (1977) reports such a size selection among the larger
shells for Kelp Gulls ( L . dominicanus ), as do Zach (1978) and Zwarts and Drent (1981) for the
crows C. caurinus and C. corone cornix respectively.

Oystercatchers open shells (M. edu/is , Cerastodenna edide, Macoma balthica , E. directus) with
their strong beak and extract the meat without ingesting the shell (Tinbergen and Norton-Grilfiths
1964; Norton-Griffiths 1967; Hulscher 1982; Swennen et al. 1985). Shells are left on the feeding
area and usually no transport is involved except when Oystercatchers try to escape kleptoparasitism
by gulls or other Oystercatchers (Leopold et al. 1985; and personal observations). However,
Leopold et al. (1985) reported long-distance transport of shells by this bird from the intertidal
feeding area to the high-tide roosts. This was confirmed by the finding of Mytilus shells at the
roost on Texel. Leopold et cd. (1985) did not find size-selective transport. This difference with the
results presented herein can be explained by the differences in the size-frequency distribution of
the living populations. The smaller Mytilus were almost absent in the living populations they
studied (mean shell length, SD, and range given as 37-9 + 7 (20-62) and 48-2 + 7-1 (28-71) mm).
The size selection observed in the current study on Texel agrees with size selection of mussels used
for feeding as reported in the literature; Norton-Grilfiths (1967) observed Oystercatchers to select
larger mussels, none smaller than 16 mm being eaten. Zwarts and Drent (1981) found a peak in
the living population of Mytilus at 10 mm length which was not consumed by the Oystercatcher.
However, Zwarts and Drent also observed a size selection for larger shells from the older Mytilus
population. Compared with the ‘modal mussel’ of 44 mm, the relative risk of a mussel of 50, 54,
or 58 mm being consumed by an Oystercatcher was 3-6, 6-7, and 10-5 times as high, respectively.
Mytilus transported to the roost do not show this size selection for larger shells (Leopold et cd.
1985; this study). This could be due to the fact that they were taken from high in the intertidal
zone (see above). Shell transport to the small roost on Texel studied here was far less important
than that observed by Leopold et al. (1985) who estimated an annual transport of 1-3 tonnes of
Mytilus shells on a large Oystercatcher roost.

Another example of a bird that transports shells in the Wadden Sea is the Eiderduck ( Somateria
mollissima). They collect their food by diving and therefore are also able to feed subtidally and on
tidal flats during high tide. In the Dutch Wadden Sea this bird feeds largely on Mytilus and
Cerastodenna. They ingest these molluscs entire and crush the shells in their muscular stomachs.
The crushed fragments, most between T5 and 4 mm (Trewin and Welsh 1976), leave the birds as
faeces. During low tide part of the eider population roosts on tidal flats, and in the breeding period
eiders roost ashore near the nesting site, with only sick birds roosting ashore in winter (Swennen
1976). Most of the defaecation of the shell fragments occurs during roosting, and thus away from
the feeding area, involving a transport of (fragmented) shells from the habitat where the molluscs
lived. Eiderducks do select for size: they avoid eating the largest bivalves probably because larger prey
offers a greater risk of internal injuries (Swennen 1976). The population of Eiderducks in the Dutch
Wadden Sea (63 000 on average the year round) was estimated to consume 32 000 tonnes dry
weight of flesh or 1-2 g.m_2-yr~1. More than 80% of this food consists of the bivalves mentioned

CADEE: SIZE-SELECTIVE SHELL TRANSPORT

435

(Swennen 1976, 1981). As shell weight is about 15-20% of dry weight, this indicates a considerable
and size-selective transport of shells by these birds of 3800-5100 tonnes per year, albeit as fragments.
Compared with the total carbonate production of shells in the Wadden Sea (Beukema 1982) it
amounts to c. 3%.

Fewer quantitative data are available on other birds in the Wadden Sea that feed size-selectively
on molluscs and produce faeces and/or cough balls on land containing shell fragments. Examples
of birds feeding mainly on molluscs are Shellduck ( Tadorna tadorna ), Common Scoter ( Melanitta
nigra). Curlew (Numenius ar quota ), and Knot (Calidris canutus ), the latter taking mainly spat-size
molluscs. A number of other birds in this area feed partly on molluscs (van der Baan et al. 1958;
Swennen 1975; Smit and Wolff 1981).

Size-selective predation on the tidal flats is not confined to birds. During high tide, aquatic
predators forage on the tidal flats (plaice, see Kuipers 1973, 1977; shore crab, see Klein Breteler
1976; shrimps, see van der Veer and Bergman 1987). From Kuipers’ (1977) data it may be concluded
that plaice transport shells (as fragments) from the tidal flat to the channels.

PALAEOECOLOGICAL RELEVANCE

Hallam’s (1972, p. 78) statement that no pronounced size-selective effects of predation have been
clearly demonstrated in the fossil record is probably still correct. However, this illustrates only the
difficulty in demonstrating such effects, not that they do not occur. Data from Recent environments
indicate that size-selective predation is more the rule than the exception (see Daan 1973, Ursin
1973, and Kuipers 1977 for fish; van der Veer and Bergman 1987 for shrimps). Vermeij (1978) and
Reise (1985) give examples of size-selective predation on molluscs. Size-selective transport by
predators is demonstrated in this paper for the Wadden Sea where it might involve at least 5 to
10% of the total shells produced.

Size-selective transport by predators is probably best developed in intertidal areas where
predators can feed during only part of the tidal period. All these predators show tidal migrations
between feeding areas on the tidal flat and resting areas on land or in deeper water. It may also
occur in subtidal waters visited for instance by predators during the night that return to deeper
water during the day. Parrish (1987) gives examples of this ‘guild of daily commuters’ from the
reef environment.

It will be clear that, although difficult to demonstrate in the fossil record, size-selective predation
(including fragmentation and/or transport) may be significant enough to hamper the use of size-
frequency distribution of fossils for studies in population dynamics; particular size groups may
not be fully represented in the fossil assemblage. This supports the conclusion reached by Cummins
et al. (1986) that the value of size-frequency distribution in fossils to assess a species’ population
dynamics is doubtful.

Acknowledgements. The author is very grateful to his colleagues M. F. Leopold, C. Swennen, and J. J. Zijlstra
for discussions, help with literature, and critically reading the manuscript.

REFERENCES

baan, g. van der, blok, a., nijhoff, p. and swennen, c. 1958. Een inleidend onderzoek naar de betrekkingen
tussen wadvogels en bodemfauna, 27 pp. Vogelwerkgroep Ned. Jeugdbond Natuurstudie.
barash, d. p., donovan, p. and myrick, r. 1975. Clam dropping behaviour of the Glaucus-winged Gulls
(Larus glaucescens). Wilson Bull. 87, 60-64.

beukema, j. j. 1982. Calcimass and carbonate production by molluscs on the tidal flats in the Dutch Wadden
Sea: II The edible cockle, Cerastoderma edule. Neth. J. Sea Res. 15, 391 405.
boucot, a. j. 1953. Life and death assemblages among fossils. Am. J. Sci. 251, 25-40.
cadee, G. c. 1968. Molluscan biocoenoses and thanatocoenoses in the Ria de Arosa, Galicia, Spain. Zool.
Verh. Leiden , 95, 1 121.

436 PALAEONTOLOGY, VOLUME 32

cadee, G. c. 1982. Low juvenile mortality in t'ossil brachiopods, some comments. Interne Verst. Nederl. Inst.
Onderz. Zee, 1982 (3), I -29.

— and hegeman, j. 1979. Phytoplankton primary production, chlorophyll and composition in an inlet of
the western Wadden Sea (Marsdiep). Neth. J. Sea Res. 13, 224 241.

cummins, h., powell, e. n., Stanton, r. j. and staff, G. 1986. The size-frequency distribution in paleocology:
effects of taphonomic processes during the formation of molluscan death assemblages in Texas Bays.
Palaeontology 29, 495-518.

daan, n. 1973. A quantitative analysis of the food intake of North Sea cod, Gadus morhua. Neth. J. Sea Res.
6, 479-517.

dare, p. j. 1976. Settlement, growth and production of the mussel Mytilus edulis L. in Morecambe Bay,
England. Fish. Invest. (II) 28, 1-25.

essink, k. 1985. On the occurrence of the American jack-knife clam Ensis directus (Conrad, 1843) (Bivalvia,
Cultellidae) in the Dutch Wadden Sea. Basteria, 49, 73-80.
goethe, f. 1937. Beobachtungen und Untersuchungen zur Biologie der Silbermowe ( Laras a. argentatus
Pontopp.) auf der Vogelinsel Memmertsand. J.f Ornithol. 85, 1-119.

1958. Anhaufungen unversehrter Muscheln durch Silbermowen. Natur u. Volk, 88, 181-187.
hallam, a. 1972. Models involving population dynamics. In schopf, t. j. m. (ed.). Models in Paleobiology ,
62-80. Freeman, Cooper and Cy, San Francisco.

Harris, m. p. 1965. The food of some Larus Gulls. Ibis, 107, 43-53.
hulscher, J. b. 1982. The Oystercatcher as a predator of Macoma. Ardea, 70, 89-152.
ingolfsson, a. and estrella, b. t. 1978. The development of shell-cracking behaviour in Herring Gulls. Auk,
95, 577-579.

kent, b. w. 1981. Prey dropped by Herring Gulls ( Larus argentatus) on soft sediments. Ibid. 98,
350-354.

klein breteler, w. c. m. 1976. Migration of the shore crab, Carcinus maenas, in the Dutch Wadden sea.
Neth. J. Sea Res. 10, 338-353.

kuipers, b. r. 1973. On the tidal migration of young plaice ( Pleuronectes platessa) in the Wadden Sea. Ibid.
6, 376-388.

1977. On the ecology of juvenile plaice on tidal flats in the Wadden sea. Ibid. 11, 56-91.
kurten, b. 1953. On the variation and population dynamics of fossil and recent mammal populations. Acta
Zool. Fennica , 76, 1-122.

Leopold, m. f., MARTEiJN, E. c. l. and swennen, c. 1985. Long-distance transport of prey from the intertidal
zone to high-tide roosts by the Oystercatcher. Ardea, 73, 76-82.
maron, J. L. 1981. Shell dropping behaviour of Western Gulls (Larus Occident alis). Auk , 99, 565-569.
muller, a. h. 1976. Lehrbuch der Palaozoologie Bd. I , Allgemeine Grundlagen, 3rd edn. 423 pp. Gustav
Fischer, Jena.

— 1979. Fossilization (Taphonomy). In moore, r. c., robison, r. a. and teichert, c. (eds.). Treatise on
Invertebrate Paleontology. Part A. Introduction, A2-A78. Geological Society of America and University of
Kansas Press, Boulder, Colorado and Lawrence, Kansas.

norton-griffiths, m. 1967. Some ecological aspects of the feeding behaviour of the Oystercatcher Haematopus
ostralegus on the edible mussel Mytilus edulis. Ibis, 109, 412-424.
oldham, c. 1930. The shell-smashing habit of Gulls. Ibid. 12, 239-244.

parrish, J. d. 1987. Characteristics of fish communities on coral reefs and in potentially interacting shallow
habitats in tropical oceans of the world. Unesco Rep. mar. Sci. 46, 171-218.
reise, K. 1985. Tidal Flat Ecology, Ecological Studies, 54, 1-191 Springer-Verlag, Berlin.
remane, a. 1951. Marine Schillablagerungen im Siiszwasser und aeolischer Hydrobienschill. Kieler Meeres-
forsch. 8, 98-1 10.

Schafer, w. 1962. Aktuo-palaontologie nach Studien in der Nordsee, viii + 666 pp. Verlag Waldemar Kramer,
Frankfurt am Main.

Schwarz, a. 1932. Mowengewolle. Natur u. Museum, 62, 305-310.

Siegfried, w. r. 1977. Mussel-dropping behavior in Kelp Gulls. South African J. Sci. 73, 337-341.
smit, c. J. and wolff, w. j. 1981. Birds of the Wadden Sea , 308 pp. Balkema, Rotterdam.
spaans, a. l. 1971. On the feeding ecology of the Herring Gull Larus argentatus Pont, in the northern parts
of the Netherlands. Ardea, 59, 73-1 18.
sunkel, F. 1925. Mellum 1924. J.f. Ornith. 73, 1 10-127.

swennen, c. 1975. Aspecten van voedselproduktie in Waddenzee en aangrenzende zeegebieden in relatie met
de vogelrijkdom. Vogeljaar, 23, 141-156.

CADEE: SIZE-SELECTIVE SHELL TRANSPORT

437

— 1976. Population structure and food of the Eider Somateria m. mollissima in the Dutch Wadden Sea.
Ardea, 64, 311-371.

— 1981. Eider ( Somateria mollissima L.). In SMIT, c. J. and wolff, w. j. (eds.). Birds of the Wadden Sea ,
78-88. Balkema, Rotterdam.

— Leopold, m. f. and stock, m. 1985. Notes on growth and behaviour of the American razor clam Ensis
directus in the Wadden Sea and the predation on it by birds. Helgolander Meeresunters. 39, 255-261.

teichert, c. and serventy, d. l. 1947. Deposits of shells transported by birds. Am. J. Sci. 245, 322-328.
Tinbergen, n. 1953. The Herring Gull's World xvi + 255 pp. Collins, London.

and norton-griffiths, M. 1964. Oystercatchers and mussels. British Birds , 57, 64 70.

trewin, n. h. and welsh, w. 1976. Formation and composition of a graded estuarine shell bed. Palaeogeogr.,
Palaeoclimatol. , Palaeoecol. 19, 219 230.

ursin, e. 1973. On the prey size preference of cod and dab. Meddr Kommn Damn. Fisk, -og Havunders. (n.s.)
7, 85-98

veer, H. w. van der and bergman, m. j. n. 1987. Predation by crustaceans on newly settled O-group plaice
Pleuronectes platessa in the western Wadden Sea. Neth. J. Sea Res. 35, 203-215.
vermeij, g. J. 1978. Biogeography and Adaptation, xvi + 332 pp. Harvard University Press, Cambridge, Mass.
zach, R. 1978. Selection and dropping of whelks by Northwestern crows. Behaviour , 67, 134 148.
zandee, d. l, kluytmans, J. h., zurburg, w. and pieters, it. 1980. Seasonal variations in biochemical
composition of Mytilus edulis with reference to metabolism and gametogenesis. Neth. J. Sea Res. 14,
1-29.

zwarts, l. and drent, r. h. 1981. Prey depletion and the regulation of predator density: Oystercatchers
( Haematopus ostralegus) feeding on mussels (Mytilus edulis). In jones, n. v. and wolff, w. j. (eds.). Feeding
and Survival Strategies of Estuarine Organisms , 193-216. Plenum Press, New York.

GERHARD C. CADEE

Netherlands Institute for Sea Research
PO Box 59
1 790 AB Den Burg

Typescript received 5 April 1988 Texel

Revised typescript received 26 September 1988 The Netherlands

A PEAFOWL FROM THE PLIOCENE OF
PERPIGNAN, FRANCE

by CECILE MOURER-CHAUVIRE

Abstract. A tarsometatarsus from the Pliocene of Serrat-d’en-Vacquer, Perpignan, attributed to Gallus
bravardi Gervais, actually belongs to the recent genus Pavo , and hence is designated as Pavo bravardi
(Gervais). Fossil Peafowls are also present in other Pliocene and Lower Pleistocene localities in France and
Moldavia. The Perpignan form is very similar to the recent species of Pavo and differs from the African form
Afropavo, suggesting that these two genera diverged from a common ancestor prior to the Pliocene.

The species Gallus bravardi was described by Gervais (1849, 1848-1952) from a fragment of
tarsometatarsus bearing a strong spur, which came from the site of Arde, or Ardes, in the
’Montagne de Perrier’, near Issoire (Puy-de-Dome, France). This specimen, collected by M.
Bravard, was deposited in the collection of the Paris Museum national d'Histoire naturelle, but it
has not been possible to locate it again in this collection.

Deperet (1890) described from the Pliocene of Perpignan, in Roussillon, an upper part of a right
coracoid and an almost complete left tarsometatarsus which he referred to G. bravardi Gervais.
The part of the tarsometatarsus which bears the spur resembles in all its details the specimen
described by Gervais as G. bravardi , its size being just a little smaller. However, the morphological
characters of the proximal and distal parts indicate that this tarsometatarsus does not correspond
to the recent genus Gallus , the Junglefowl, but to the recent genus Pavo, the Peafowl. The species
bravardi is therefore transferred to the genus Pavo. However, the coracoid from Perpignan is
morphologically similar to the genus Gallus and can be provisionally referred to as Gallus sp.

The specimen from the Roussillon Pliocene, illustrated by Lambrecht (1933, p. 875, fig. 193h)
as G. bravardi Gervais, is actually G. aesculapi Gaudry, from Pikermi.

The Perpignan tarsometatarsus, described as complete, was made up of two fragments stuck
together in such a way that the dorsal face of the proximal part was in continuity with the plantar
face of the distal part, as can be seen in Deperet’s illustration. This tarsometatarsus has been
restored by placing its dorsal and plantar faces correctly in line and, by comparison with recent
forms, its shaft has been lengthened by plaster so that its present length is 158 mm. This is an
estimate of its minimal size (its previous length was 143 mm).

The occurrence of a fossil Peafowl in the European Pliocene and Lower Pleistocene is confirmed
by the fact that the specimens from Seneze (Stehlin 1923) and Saint-Vallier (Viret 1954), referred
to as ?G. bravardi, also have the morphological characteristics of Pavo and differ from Gallus.
They can also be attributed to the species P. bravardi (Gervais). Another Peafowl, P. moldavicus ,
recently described by Bochenski and Kurochkin (1987) from the Moldavian Roussillon is here
placed in synonymy with P. bravardi.

SYSTEMATIC PALAEONTOLOGY

Order galliformes (Temminck, 1820)

Family phasianidae Vigors, 1825
Subfamily phasianinae (Vigors, 1825)

Genus pavo Linnaeus, 1758
Pavo bravardi (Gervais, 1849)

Plate 50, figs 1-3, 7, 8

IPalaeontology, Vol. 32, Part 2, 1989, pp. 439-446, pi. 50.]

© The Palaeontological Association

440

PALAEONTOLOGY, VOLUME 32

1844 ‘Gallinace’ Gervais, p. 22.

1 849 Gallus Bravardi Gervais, p. 220.

1848-1852 Gallus Bravardi Gervais; Gervais, t. 1, p. 238; t. 2, explanation of pi. 51; t. 3, pi. 51, fig.
1-1 a.

1859 Gallus Bravardi Gervais; Gervais p. 418, pi. 51, fig. 1-1 a.

1869-1871 Gallus Bravardi Gervais; Milne-Edwards, t. 2, p. 250.
non 1890 Gallus Bravardi Gervais; Deperet, p. 134, pi. 13, fig. 11-1 In (= Gallus sp.).

1 890 Gallus Bravardi Gervais; Deperet, p. 1 38, text-fig. 3 a, b.

non 1892 Gallus Bravardi Gervais; Deperet, p. 691 (= Gallus sp.).

1923 1 Gallus Bravardi Gervais; Stehlin, p. 278.

1933 Gallus Bravardi Gervais; Lanrbrecht, p. 443.
non 1933 Gallus Bravardi Gervais; Lambrecht, p. 875, text-fig. 193h (= Gallus aesculapi Gaudry).

1954 1 Gallus bravardi Gervais; Viret, p. 173.

1964 Gallus bravardi Gervais; Brodkorb, p. 318.

1987 Pavo moldavicus Bochenski and Kurochkin, p. 89, pi. 18, figs 13 and 14.

Holotype. Middle part of the shaft of a left tarsometatarsus with a spur. This specimen was in the collection
of the Museum national d’Histoire naturelle de Paris but, as yet, it has not been possible to find it.

Type stratum and locality. Lower Villanyiunr or Lower Villafranchian, Neogene Mammal Unit 16 (Mein
1975). Arde near Issoire, Puy-de-Dome, Prance.

Additional material. An almost complete left tarsometatarsus from Serrat-d’en-Vacquer, near Perpignan,
Pyrenees-Orientales, Prance. Upper Ruscinium, Neogene Mammal Unit 15. This specimen is in the collection
of the Musee Guimet d’Histoire naturelle de Lyon (Pp 269). Other specimens referable to the same species
are known from the sites of Saint-Vallier, Drome, Prance, Upper Villanyium, Neogene Mammal Unit 17
(Collection Musee Guimet d’Histoire naturelle de Lyon), Seneze near Brioude, Haute-Loire, Prance, Lower
Biharium, Neogene Mammal Unit 18 (Collection Museum of Natural History of Basel), and Lucheshty,
Kagul district, Moldavian SSR, Upper Pliocene, Moldavian Roussillon (Collection of the Palaeontological
Institute of the USSR Academy of Sciences in Moscow).

Original diagnosis (Gervais 1848-1852). ’Portion intermediate du tarse de Gallinace. . . . Ce
fragment est la partie la plus voisine de l’eperon. Celui-ci est long de 0m,021, quoique son sommet
ait ete casse; sa base a 0m,013 de hauteur verticale. Cet eperon est assez comprime; il est creuse
en gouttiere pres de sa base, pour le passage des tendons. A cet endroit la face externe du tarse a
0m,015 et elle diminue brusquement a 0m,05 au dessous de la base de l’eperon par la cessation de
la crete posterieure de l’os, qui n’est que la soudure au canon du metatarsien du pouce. Ce fragment
a plus de rapport avec la partie correspondante du canon du Coq qu’avec la meme partie chez le
Paon ou les autres Gallinaces auxquels je l’ai compare. II indique un oiseau voisin des Coqs, et
dont la taille etait intermediate a celle du Paon et du Coq ordinaire, mais que je ne crois pas de
la meme espece que ce dernier, quoiqu'il lui ressemble plus qu’aux autres oiseaux du meme ordre.’

Remarks. This specimen is illustrated in natural size and there are probably two misprints in the
dimensions given by Gervais in its description. On the figure the spur length, as preserved, is
31 mm instead of 21 mm, and the lateral face of the tarsometatarsus below the level of the bony
spur core is 5 mm deep instead of 50 mm (0 05 m).

Description and comparisons

The anatomical descriptions follow the terminology of Baumel (1979).

Comparison with Aide specimen (text-fig. 1). The spur of the Perpignan specimen looks very similar to the
Arde one. It is long, slightly curved and elliptical in section. It arises on the medial face of an intertendinal
ossified septum running down the plantar face of the tarsometatarsus. This septum disappears a little below
the level of the bony spur core so that the depth of the medial face of the shaft decreases considerably. On
the base of the bony spur core and on its medial side, one can see a tendinal groove which is very similar
to the tendinal groove illustrated in the Arde specimen (Gervais, 1848 1852, pi. 51, fig. In). Although the
shaft of the latter looks a little more robust than the Perpignan one, the dimensions are very close (Table 1).

MOURER-CHAUVIRE: PLIOCENE PEAFOWL

441

text-fig. 1. Pavo bravardi , holotype, left tarsometatarsus, central part of the shaft, with
the spur, from Arde, Puy-de-Dome, France (after Gervais 1848 1852, pi. 51, fig. 1-la),
a, lateral view; b , medial view. Natural size.

Comparison with the recent genera Gallus and Pavo. The general shape of the Perpignan bone is slender, the
shaft in the middle is rather thin while in the large-sized Gallus (which are artificially produced by selection,
the wild Gallus being much smaller) the bone is proportionally shorter, and the shaft more robust. The
Perpignan form has the same general outline as the two recent species of Peafowls, P. muticus , the Green
peafowl, and P. cristatus , the Blue one.

In the proximal part the disposition of the hypotarsal calcaneal ridges is different in Peafowls and
Junglefowls (PI. 50, figs. 6 and 7). In Peafowls the medial calcaneal ridge (crista medialis hypotarsi of Baumel
1979) (PI. 50, fig. 7, a) is prolonged by an ossified intertendinal septum along the medio-plantar corner of
the bone and bears the bony core of the spur. In Junglefowls the medial calcaneal ridge is short and completely
separated from the septum upon which the bony core arises (PI. 50, fig. 5). In Pavo the lateral calcaneal ridge
(crista lateralis hypotarsi of Baumel 1979) projects very little and is more or less included in the hypotarsal
block (PI. 50, fig. 7, b); in Gallus this ridge is well developed, flattened on its plantar face, and enlarged at
its distal part (Ballmann 1969) (PI. 50, figs. 5 and 6, b). The groove between these two ridges is deeper in
Gallus than in Pavo. On the lateral side of the hypotarsus there is a short accessory lateral ridge, more
developed in Gallus (PI. 50, fig. 6, c) than in Pavo (PI. 50, fig. 7, c). On the Perpignan tarsometatarsus the
shape of the calcaneal ridges is similar to Pavo. The medial one was probably prolonged down to the spur
by an ossified intertendinal septum which has been preserved but is no longer in situ.

In the proximal part the shaft narrows considerably below the articular part in Peafowls, while in
Junglefowls it narrows gradually (PI. 50, figs. 1 and 4). In Peafowls the sulcus extensorius is situated nearer

table 1. Measurements of the bony spur core in Pavo bravardi and in recent P. muticus, male, in mm.

Pavo bravardi

P. muticus male

Arde Perpignan

MNHN Paris

Pp 269

1891-1022 1875-50 1880-1397

Length of the bony spur core

31 (1)

31 (I)

300

260

310

Height of the spur core at its base

13

12

110

110

1 15

Depth of the spur core at its base

—

7-3

8-0

6-5

c. 8-0

Depth of the medial face of the tarso-
metatarsus, above the spur core

15

11

Depth of the medial face of the tarso-
metatarsus, below the spur core

5

6

—

—

—

( 1) as preserved.

442

PALAEONTOLOGY, VOLUME 32

the medial side while in Junglefowls it is in the middle of the dorsal face. On the Perpignan tarsometatarsus
the shaft also narrows considerably below the upper articular part, but the bone has been damaged and
restored and it is not possible to see the position of the sulcus extensorius.

In the distal part, at the level of the trochleae, the shaft is widened symmetrically in Pavo (PI. 50, fig. 2)
while in Ga/lus it is wider on the medial side (PI. 50, fig. 5). In Pavo the outline of the lateral side of the bone
curves, while in Gallus it is almost straight. In Peafowls the trochlea metatarsi tertii projects with regard to
the dorsal face of the shaft, while in Junglefowls it is almost on the same plane. The trochlea metatarsi tertii
is also proportionally larger in Peafowls than in Junglefowls (PI. 50, figs. 1-2 and 4-5). The trochlea metatarsi
secundi is less strongly directed backwards in Peafowls than in Junglefowls.

Taking into account all these characteristics, the distal part of the Perpignan tarsometatarsus agrees with
the genus Pavo and differs from the genus Gallus. It also presents, above the incisura intertrochlearis medialis,
a small foramen, which is very conspicuous on the plantar face (PI. 50, fig. 2). The occurrence of this foramen
has been considered as characteristic for the Meleagrinae (Howard 1927; Olson and Farrand 1974) but it
can also be seen in Pavo and in some specimens of Gallus. Steadman (1980) indicates that it may be present
in several other genera of Phasianinae.

The bony spur core of the Perpignan form is long, elliptical in section, and slightly curved. In Peafowls
the spur is often not very developed and variable in shape but several individuals of P. muticus in the
collection of the Paris MNHN bear a spur very similar in shape and size to the Perpignan one (Table 1).
Milne-Edwards (1867 1871) also illustrated a tarsometatarsus of P. muticus with a strong, elongated, and
curved bony spur core. In Turkeys spur length increases with age and is fully developed only in 2-3-year-
old individuals (Steadman 1980); it is probably the same in Peafowls. By this feature the Perpignan form is
also similar to Junglefowls which have long, sharply pointed spurs (PI. 50, figs. 4 and 5).

In its absolute dimensions the Perpignan Peafowl is very close to the male specimens of the Green peafowl,
P. muticus , which are generally larger than the Blue ones, P. cristatus. It is also nearer to the Green peacocks
by the ratios between the different measurements. In recent Peafowls the spur position differs in males and
females, the spur being situated lower in Peacocks than in Peahens. The only difference between the recent
and the fossil forms is that in the fossil the spur is situated in a lower position than in the recent males (Table
2). The mean distance between the middle of the spur core and the distal part of the bone, expressed as a
percentage of its total length, is 401% in the Green peacocks and 33-3% in the Blue peacocks. In P. bravardi
the minimal total length being 158 mm, the percentage of 29-7% for the position of the spur is a maximal
value. In this form the bony spur core is situated immediately above the fossa metatarsi I, while in the recent
Peafowls there is some distance between them.

Comparison with the recent genus Afropavo. Another Peafowl, Afropavo congensis from Africa (Chapin 1936)
seems from osteological, myological, and karyological evidence to be more closely related to the genus Pavo
than to any other gallinaceous bird (Lowe 1939; Hulselmans 1962; de Boer and van Bocxstaele 1981). I
compared the Perpignan Peacock with two specimens of A. congensis males from the American Museum of
Natural History in New York, and one female from the Lyon Collection which was presented to me by the
Royal Zoological Society of Antwerp. The Perpignan tarsometatarsus differs from the African ones by its
much larger size (Table 2). In the two males there is a supplementary ridge between the crista medialis and
the crista lateralis hypotarsi, which does not reach the distal part of the hypotarsus, but this supplementary
ridge does not exist in the female specimen. This supplementary ridge also exists in some specimens of

EXPLANATION OF PLATE 50

Figs. 1-3. Pavo bravardi (Gervais). Musee Guimet d’Histoire naturellc de Lyon, n°. Pp 269. Upper Ruscinium
of Serrat-d’en-Vacquer, near Perpignan (Pyrenees-Orientales), France. Left tarsometatarsus. 1, dorsal view,
xl. 2, plantar view, xl. 3, medial view, xl.

Figs. 4 6. Gallus gallus (Linnaeus). Departement des Sciences de la Terre de Lyon, n°. 456-2. Recent. Left
tarsometatarsus from a wild bird. 4, dorsal view, x 1. 5, plantar view, x 1. 6, proximal view, x 2, a—
crista medialis hypotarsi, b— crista lateralis hypotarsi, c— accessory lateral ridge.

Figs. 7 and 8. P. bravardi (Gervais). Left tarsometatarsus, n°. Pp 269. 7, proximal view, x 1 -5, a, b, c, same
as in fig. 6. 8, distal view, x 1-5.

Fig. 9. Afropavo congensis Chapin. Departement des Sciences de la Terre de Lyon, n°. 1988-1. Recent. Left
tarsometatarsus, proximal view, x 2, a, b, c, same as in fig. 6, d— tubercle on the latero-plantar corner of
the external cotyla.

table 2. Measurements of the tarsometatarsus in Pavo bravardi and in recent Peafowls, in mm.

444

PALAEONTOLOGY, VOLUME 32

aS

Oi

<L> ^
O 22

o a -o
<u ° S

P B r-

'5

CO G

co <l> OJ ~

2

•rr

P

i on vr>
p m

m

VO

d-

o ^

fn

Os

P _

co ^

d"

rn

vb ^
1 rn

ITT

d"

m

d*

oo

©

d- T
i o d-
p d-

ON

m

VO

IT) ^

' <N d
O i/^

©

m

<N

d

fT)

d-

d-

OO ITT
r<r rn

<N

TT

O IT)
d CM
r~> ON

(N

r<T

O c*-

d

-P <D £

-*-* co VO

-d ^

d p;
co d- «/n

s £

ON _
(N ON

©

d-

o

On vo

<N OO

oc

.A

CN

ON —
OO ON
OO

NO

OO rn
ON OO

l O <N

OO vo

UN O

d £
c/3 "O d"

S £

— I ON <d

(N ,—

1 d
P
O
<N

NO

ON

(N

»/N irr

ON d

rP

Q- m
<D
-d

d

s 6

'* 22 <n

2 ?

Cl,

o

ON

"T

r5 T

^ Os <N
d ,— .

©

ON

m

<N T

d

P (N

O

<^1

OO <d

p

<N

d i/t

p o

p 2;
d

I i/T <N

d *5

p 00,

r° C
f-H 0)

ON ^

«3

d

m

i © m

oo ^

2 ^
’•£2 d

ON

2 d

d

£ p
^ ^ m
O

ON —

O

oo

tn j i/n

ON

NO

<N

<N

IT) OO ^
d- 2

o

Cl

Si

Dh

Oo

^ C

K

o

Co

S

Co

S

^ 2

Nj

2

« u

a wj

o

p*.

.50

^ J

73

1 I

e S,

J-H

C3

JJ

73

e

■O (X

<o

Np

s

ai

ai

. congensis ,

female, n = 1 79-0 13-0 12-6 11-3 9-1 44 34 0 43-6

MOURER-CHAUVIRE: PLIOCENE PEAFOWL

445

Pavo (Howard 1927). It is generally present in the Meleagrinae (Steadman 1980). The medial intertrochlear
foramen is absent in the two males but is present in the female.

In A. congensis there is a strongly developed tubercle on the latero-plantar corner of the lateral cotyla (PI.
50, fig. 9, d). This tubercle exists also in Pavo but is smoother. At the distal part the trochlea metatarsi tertii,
which is very wide in Pavo , is proportionally narrower in Afropavo. The ossified intertendinal septum which
extends from the hypotarsus to the fossa metatarsi I and from which the bony spur core arises exists also in
Afropavo, as in Pavo and many other phasianids. The spur core is long, tapered, elliptical in section, and
slightly curved. It is situated proportionally higher in the female than in the male, and at the same level as
in the Asiatic Peafowls, i.e. proportionally higher than in P. bravardi (Table 2).

In conclusion, the Perpignan form is more similar to the Asiatic Peafowls, and particularly to the Green
one, than to the Congo Peafowl.

Comparison with P. -moldavicus Bochenski and Kurochkin. This species was described from an incomplete
lower part of a coracoid, from the locality Lucheshty, in Moldavia. Its age ‘Moldavian Roussillon’ is the
same as the Perpignan fauna. Its size is indicated as being larger than any living species of the genus Pavo
(15% larger than the largest bone of P. muticus). It is not possible to compare P. bravardi and P. moldavicus
directly. The Perpignan specimen, as well as the other specimens from Saint-Vallier and’Seneze, share the
same size as the recent individuals of P. muticus. Moreover, in the collection of the Paris MNHN, one
specimen of P. muticus, male, is larger than the Moldavian Peafowl. The medial part of its facies articularis
sternalis is 10-4 mm thick on the left coracoid, and II 9 mm on the right one, while in P. moldavicus this
dimension is 7-5 mm (Bochenski and Kurochkin 1987).

Given that the Moldavian specimen is the same age as the Perpignan one, and considering the great
variation in size in the recent Peafowls, I think that they probably belong to the same species.

PALAEOBIOGEOGRAPHICAL AND PALAEOECOLOG IC AL IMPLICATIONS

It is necessary to emphasize the occurrence of an element of eastern Asiatic affinities in the
Perpignan avifauna. Southern and eastern Asiatic elements, such as Cathaya, Sciadopilys,
Pterocarya, Parrotiopsis cf. jacquemontiana, and Microtropis fallax, have also been found in the
palynological study of the hill of Serrat-d’en-Vacquer, Sondage FI de la Mutualite agricole de
Perpignan (Cravatte et al. 1984). The genus Pavo is now restricted to Pakistan, India, Nepal,
Bangladesh, Sri Lanka, south-west China, Burma, Thailand, Indochina, the Malay Peninsula, and
Java, but its occurrence in the Pliocene and Lower Pleistocene of France and Moldavian SSR
shows that it was much more widespread in the past. The discovery of a form related to it and
isolated in Africa seems therefore much easier to understand. However, the Perpignan form is
more closely related to the Asiatic form than to the African one, so the separation from an
ancestral Peafowl into the two recent genera must have occurred before the Pliocene, and probably
during the Miocene.

Recent Peafowls live in dense jungles or open forests, near water. Thus the presence of a Peafowl
in the Perpignan fauna confirms that the environment was mainly forested, as has already been
established by palynological data.

REFERENCES

ballmann, p. 1969. Die Vogel aus der altburdigalen Spaltenfiillung von Wintershof (West) bei Eichstatt in
Bayern. Zitteliana, 1, 5-60.

baumel, j. j. 1979. Osteologia. In baumel, j. j. (ed.). Nomina Anatomica Avium. An annotated anatomical
dictionary of birds, 53-121. Academic Press, London and New York.

bochenski, z. and kurochkin, e. n. 1987. New data on Pliocene phasianids (Aves, Phasianidae) of Moldavia
and S. Ukraine. Acta zool. Cracov. 30, 8 1 96.

boer, l. e. m. de and bocxstaele, r. van. 1981. Somatic chromosomes of the Congo Peafowl ( Afropavo
congensis) and their bearing on the species’ affinities. Condor, 83, 204 208.

brodkorb, p. 1964. Catalogue of fossil birds: Part 2 (Anseriformes through Galliformes). Bull. Florida State
Mus. 8, 195-335.

chapin, j. p. 1936. A new Peacock-like bird from the Belgian Congo. Rev. Zool. Bot. Afr. 29, 16.

446

PALAEONTOLOGY, VOLUME 32

cravatte, j., matias, i. and sue, J. p. 1984. Nouvelles recherches biostratigraphiques sur le Pliocene du
Roussillon. Geol. Fr. 1-2, 1984, 149-163.

deperet, c. 1890. Les animaux pliocenes du Roussillon. Mem. Soc. geol. Fr ., Paleont. 3, 1 1 98.

— 1892. Sur la faune d'Oiseaux pliocenes du Roussillon. C.r. hebd. Seanc. Acad. Sci. Paris, 114, 690-692.
gervais, p. 1844. Remarques sur les oiseaux fossiles. These Univ. Paris. Fac. des Sci. 1 40.

— 1848-1852. Zoologie et paleontologie franqaises ( animaux vertebres) ou nouvelles recherches sur les
animaux vivants et fossiles de la France. 3 vols., 271 pp., 80 pis. Arthus Bertrand, Paris.

1849. Oiseaux et reptiles fossiles de France. Acad. Sci. Lettres Montpellier, Mem. Sect. Sci. 1, 220-222.

— 1859. Zoologie et paleontologie franqaises. Nouvelles recherches sur les animaux vertebres dont on trouve
les ossements enfouis dans le sol de la France. 2nd edn., 544 pp. Arthus Bertrand, Paris.

Howard, h. 1927. A review of the fossil bird Parapavo calif ornicus (Miller), from the Pleistocene Asphalt
beds of Rancho La Brea. Univ. Calif. Publ. geol. Sci. 17, 1-62.
hulselmans, j. l. j. 1962. The comparative myology of the pelvic limb of Afropavo congensis Chapin, 1936.
Bull. Soc. R. Zool. Anvers, 26, 25-69.

lambrecht, k. 1933. Handbuch der palaeornithologie, 1024 pp. Borntraeger Verlag, Berlin.
lowe, p. R. 1939. Some preliminary notes on the anatomy and systematic position of Afropavo congensis
Chapin. C.r. 9° Congr. ornitli. int., Rouen 1939, 219-230.
mein, p. 1975. Biozonation du Neogene mediterraneen a partir des mammiferes. IUGS, RCMNS. Report on
Activity of the RCMNS Working Groups, Bratislava, 78-81.
milne-edwards, A. 1867- 1871. Recherches anatomiques et paleont ologiques pour servir a Thistoire des oiseaux
fossiles de la France. 2 vols., 474 + 627 pp. Victor Masson et Fils, Paris.
olson, s. l. and farrand, j. Jr. 1974. Rhegminornis restudied: a tiny Miocene turkey. Wilson Bull. 86, 114

Steadman, d. w. 1980. A review of the osteology and paleontology of turkeys (Aves: Meleagridinae). Contr.
Sci. nat. Hist. Mus. Los Angeles Co. 330, 131 207.

stehlin, h. g. 1923. Die oberpliocene Fauna von Seneze (Haute-Loire). Eclog. geol. Helv. 18, 268-281.
viret, j. 1954. Le loess a bancs durcis de Saint-Vallier (Drome) et sa faune de mammiferes villafranchiens.
Nouv. Arch. Mus. Hist. nat. Lyon, 4, 1-200.

120.

Typescript received 4 July 1988

Revised typescript received 2 September 1988

CECILE MOURER-CHAUVIRE

Centre de Paleontologie stratigraphique et Paleoecologie
de rUniversite Claude Bernard, Lyon-I
Laboratoire associe au CNRS (UA 11)

27-43 boul. du 11 novembre
69622 Villeurbanne Cedex, France

HEMIHOPLITID AMMONOIDS FROM THE
LOWER CRETACEOUS OF SOUTHERN
PATAGONIA

by a. c. riccardi and m. b. aguirre urreta

Abstract. The genus Hemihoplites Spath is locally common in the Hauterivian- Barremian of the Austral
Basin, southern Patagonia, Argentina. The taxonomic status of Hemihoplites is discussed, and the new species
H. varicostatus and H. ploszkiewiczi are described. Extensive intraspecific variation occurs, and sexual
dimorphism is reported for the first time within this genus. This is the first record of the Hemihoplitidae in
the Southern Hemisphere.

The marine Cretaceous of southern Patagonia is poorly fossiliferous. Tithonian-Berriasian strata
with a rich but badly preserved ammonite fauna are overlain by a thick pelitic sequence containing
mostly Favrella R. Douville, 1909 and Hatchericeras Stanton, 1901. Ammonites become diverse
again in the Aptian and Albian (see Riccardi 1988).

The apparent absence of ammonites and the endemism of Favrella and Hatchericeras have been
used as evidence for a Valanginian-Barremian biostratigraphic hiatus and unconformity (Leanza
1963). However, studies carried out in the last two decades have shown that ammonite faunas,
even if rare or poorly preserved, are present also between the Berriasian and Aptian faunas. They
document the Yalanginian, Hauterivian, and Barremian, and the late early-late Hauterivian and
Barremian ages of Favrella and Hatchericeras respectively (see Riccardi 1984a, 6, 1988; Aguirre
Urreta and Klinger 1986; Riccardi et al., 1987). Some ammonites of these levels have already been
described or figured, and others will be described in the near future. This paper deals with a very
distinctive family, the Hemihoplitidae Spath, 1924, which, however, is rather rare even in the
Northern Hemisphere.

The ammonoids studied here were collected by one of us (A.C. R.) in 1972, V. A. Ramos and
M. A. Palma in 1979, G. Marin and M. B. Aguirre Urreta in 1981, and J. V. Ploszkiewicz and
V. A. Ramos in 1982, from three different localities (text-figs. 1 and 2; Table 1). They were found
in outcrops of the Rio Mayer Formation, immediately below, above, and within levels with
Favrella spp., and in association with Hatchericeras spp.

The material contained in Table 1 is important in documenting the range of variation and the
existence of dimorphism in the Hemihoplitidae, as well as their presence in the Southern Hemisphere.
It also gives additional support to the previously reported age of Favrella and Hatchericeras (see
above).

FOSSIL ASSEMBLAGES AND AGE

Within the zonal scheme proposed by Riccardi (1984a, b ) for the Cretaceous of southern Patagonia,
the hemihoplitids studied here occur in the following zones, in ascending sequence:

1. F. americana Assemblage Zone. Hemihoplites ploszkiewiczi sp. nov. has been collected (text-
fig. 2; Table 1) 5 m below levels with F. americana (Favre) and is here included in the same zone.
F. americana is usually associated with Aegocrioceras sp. and Belemnopsis patagoniensis (Favre),
and has also been recorded immediately above Acanthodiscus sp. (Olivero 1982). The age of this
zone is late early to early late Hauterivian (inversum Zone).

IPalaeontology, Vol. 32, Part 2, 1989, pp. 447-462, pis. 51-52. |

© The Palaeontological Association

448

PALAEONTOLOGY, VOLUME 32

Table 1. Occurrences of ammonoids at Rio Belgrano, Chorrillo Rivera, and Veranada de la Vinca; samples

collected in situ at levels indicated in text-fig. 2.

Taxa

Locality

Rio Belgrano

Chorillo Rivera

Veranada
de la Vinca

12 3 4

5

6

1 2

3 4

5

6

1 2 3

Hemihoplites varicostatus

+

+

H. ploszkiewiczi

+

Hatchericeras spp.

+

+

+

+ +

+

+

Cryptocrioceras yrigoyeni

+

Sanmartinoceras africanum

+

Protaconec. patagoniense

+ +

Favrella wilckensi

+

F. americana

+

+

Crioceratitinae indet.

+

+

Ancyloceratitinae indet.

+ +

+

+

2. F. wilckensi Assemblage Zone. H. varicostatus sp. nov. has been found (text-fig. 2; Table 1)
alone, and associated with Protaconeceras pcitagoniense (Favre), or with earliest Hatchericeras (see
below). The occurrence of P. pcitagoniense in the late Flauterivian ( gottschei Zone) of Europe
establishes the age of this zone in Patagonia.

3. H. patagonense Assemblage Zone. Plemihoplites varicostatus sp. nov. occurs (text-fig. 2; Table
1) with Platchericeras sp. This latter genus is widespread in the area, and is associated with
Cryptocrioceras yrigoyeni (Leanza), and Sanmartinoceras africanum insignicostatum Riccardi et al.
The stratigraphic position of this assemblage, above the F. wilckensi Assemblage Zone and below
the Colchidites Assemblage Zone, suggests Lower-Middle Barremian.

SYSTEMATIC PALAEONTOLOGY

Dimensions. Dimensions of specimens are given in mm, in the following order: diameter (D), whorl height
(H), whorl breadth (W), and breadth of umbilicus (U). P and S refer to number of primary and secondary
ribs in half whorl. Figures in parentheses refer to dimensions as a percentage of diameter.

Suture terminology. The conventional suture terminology is followed here: I = Internal lobe, U = Umbilical
lobe, L = Lateral lobe, E = External lobe.

Abbreviations used for collections. CPBA, Catedra de Paleontologia, Facultad de Ciencias Exactas y Naturales,
Universidad de Buenos Aires, Ciudad Universitaria, Pabellon 2, 1428 Buenos Aires, Argentina: MLP,
Division Paleozoologia de Invertebrados, Museo de Ciencias Naturales, Universidad Nacional de La Plata,
Paseo de Bosque s/n., 1900 La Plata, Argentina.

Class cephalopoda Zittel, 1884
Order ammonoidea Zittel, 1884
Suborder ancyloceratina Wiedmann, 1966
Superfamily ancylocerataceae Gill, 1871
Family hemihoplitidae Spath, 1924

Comments. Spath (1924, p. 84) introduced the Family Hemihoplitidae for Hemihoplites , Pseudo thurmannia ,
and Metahoplites without further elaboration. Although the Hemihoplitidae were retained as an independent
subfamily or family by Basse (1952), Wright (1957, 1981), and Luppov et al. (1958), the crioceratitid septal
suture of Hemihoplites prompted Wiedmann (1962, 1966) to include this genus in the Family Ancyloceratidae,

RICCARDI AND AGUIRRE URRETA: ARGENTINIAN CRETACEOUS AMMONOIDS 449

text-fig. I . Index map for locating the stratigraphic sections at Rio Belgrano, Chorrillo Rivera, and Veranada

de La Vinca.

450

PALAEONTOLOGY, VOLUME 32

RIO

BELGRA NO

CHORRILLO

RIVERA

VERANADA
DE LA VINCA
fault

Rio

Belgramo

Fm.

Rio

Mayer

Fm.

o o o o o

V • 2

uzr

-A — 6

:5

7" 7 • /

/ amM

v v v
v v

EL QUEMADO
Complex

- 1

:l /

-1/

Sandstones

=

Shales

V V V

Dacites & tuffs

1-6

Fossil levels

f

200 m

LLq

V V V
V V

/VVAAA

text-fig. 2. Stratigraphical sections at Rio Belgrano, Chorrillo Rivera, and Veranada
de la Vinca (see text-fig. 1) with fossil locality numbers indicated.

Subfamily Crioceratitinae. The material included in Hemihoplites, however, shows no hint of other crioceratitid
features such as crioceratitid coiling, presence of fine ribs intercalated with periodic stronger ribs, and bi-
and trituberculation on the inner whorls with a tendency to fade on the outer. These features are also missing
in Pseudo thurmannia , as represented by its type species, P. angulicostata (d’Orbigny) (see Lapeyre 1974, p.
82, figs. 1 -8). The presence and relative importance of crioceratitid features in Pseudo thurmannia depends on
the species included. Busnardo (1970, pp. 135 136) noted that if all species previously assigned to this genus
were retained, its distinction from Crioceratites Leveille would become almost impossible (see Immel 1978).
A detailed assessment of the generic status of all these species is beyond the scope of this study. Hemihoplites
and Pseudo thurmannia seem to be distinctive enough to be included in a different family. We therefore follow
Wright (1957, 1981) in retaining the Family Hemihoplitidae.

According to Wright (1957, p. 212) the Hemihoplitidae may, in addition to Pseudo thurmannia and
Hemihoplites , include Pascoeites Spath 1933. Pascoeites includes P. budavadensis Spath, the type species, and
P. crassus Spath (1933, p. 827, pi. 126, figs. 5, 7, 12), a species based on three specimens which were not
described. Their stratigraphic position is unclear and no new material has been figured. As pointed out by
Wright (1957, p. 212) therefore, Pascoeites remains a poorly known genffs. Metahoplites has already been
included (Wright 1957, p. 827) in the Holcodiscidae.

Genus hemihoplites Spath, 1924
(= matheronites Renngarten, 1926)

Type species. Ammonites feraudianus d’Orbigny, 1841, by original designation (Spath 1924, p. 84).

Diagnosis. Rather evolute, with rectangular to subquadrate whorl section, flat or slightly curved
flanks and venter. Ribs simple or branching, long and short, changing from dense to well spaced;

RICCARDI AND AGUIRRE URRETA: ARGENTINIAN CRETACEOUS AMMONOIDS 451

straight or slightly flexuous, with subdued to distinct umbilical and/or lower and/or upper
ventrolateral tubercles. Crioceratitid suture with trifid L, and U located on inner flank.

Comments. Hemihoplites was introduced without diagnosis by Spath (1924, p. 84), with A. feraudianus
d’Orbigny (1841, p. 324, pi. 96, figs. 4 and 5) as type species. Renngarten (1926, p. 27) subsequently proposed
Matheronites as a subgenus of Acanthohoplites Sinzow, with A. soulieri Matheron (1878, pi. C-21, fig. la, b)
as type species. In spite of the alleged (Renngarten 1926, p. 97) poor knowledge of H. feraudianus, the figures
of A. soulieri show close resemblance in most features, except for the ribbing which in H . feraudianus is more
flexuous and more often bifurcated on the umbilical margin. This similarity was made plain when Fallot and
Termier (1923, p. 67, pi. 6, fig. 1; text-fig. 29) included in ‘ Parahoplites soulieri Math.’ a specimen that is in
fact closer to the type of H. feraudianus.

The subgenus Matheronites has been subsequently raised to generic status, usually in connection with
records of M. soulieri and related species (see below) from eastern Europe and the Caucasus. This probably
explains the common usage (see Rouchadze 1933; Eristavi 1955; Luppov et al. 1958; Drushchits and
Kudryastev 1960; Breskovski 1966; Dimitrova 1967; Bogdanova 1971) of the name Matheronites, whilst
Hemihoplites has remained a poorly known genus. Nevertheless, Wright (1957) and Wiedmann (1962, 1966)
considered Matheronites to be a junior synonym, and Kakabadze (1981, p. 92) regarded it as a subgenus, of
Hemihoplites. Kakabadze (1981, p. 31), however, placed ‘ A 1 soulieri in the subgenus Hemihoplites, whilst
considering ‘A.' ridzewskyi Karakasch as the type species of the subgenus Matheronites.

The species referred to Matheronites and Hemihoplites by previous authors could on morphological grounds
be placed into two different groups. One of them is characterized by straight or slightly flexuous ribs, simple
or bifurcated, and by umbilical, frequent ventrolateral, and rare lateral tubercles, which develop through
ontogeny. Material with umbilical or with umbilical and ventrolateral tubercles has been described under
'Ml soulieri (Matheron 1878, pi. C-21, fig. la, b, Fallot and Termier 1923, p. 67, pi. 6, fig. la c; Rouchadze
1933, p. 201, pi. 3, fig. 5), ’ M 1 khwamliensis Rouchadze (1933, p. 202, pi. 3, fig. 6; ?Kakabadze 1981, pi. 5,
fig. 2a, b), 'Ml turcmenicus Luppov (1936, p. 122, pi. 1, figs. 1 3), TM.' ridzewskyi (Drushchits and Kudryastev
1960, p. 287, pi. 30, fig. 3a, b, Dimitrova 1967, p. 71, pi. 32, fig. 5), and M. brevicostatus Bogdanova (1971,
p. 337, pi. 6, figs. 1 and 2). Material with three rows of tubercles has been figured under 'Ml ridzewskyi
(Karakasch 1896, p. 108, pi. 4, figs. 9 and 10; Renngarten 1926, p. 29, pi. 2, figs. 9 and 10; Luppov et al.
1958, pi. 46, fig. la, A; Drushchits and Kudryastev 1960, p. 287, pi. 30, fig. 2a, b, Kakabadze 1981, pi. 2, fig.
2a, b), 'Ml astarte (Fallot and Termier 1923, p. 70, pi. 6, figs. 2-5; Wiedmann 1966, pi. 6, fig. 6 a-c), 'M.
soulieri' (Dimitrova 1967, p. 70, pi. 32, fig. 1), 'M. coheni' (Dimitrova 1967, p. 69, pi. 32, fig. 2), H. (M.)
trispinosus (Koenen)’ (Kakabadze 1981, pi. 2, figs. 3a, b- 5), and TH. (M.) brevispinus (Koenen)’ (Kakabadze
1981, pi. 2, fig. 6 a, b).

A different, twofold separation of most of these species has been proposed by Kakabadze (1981, p. 93)
which uses the frequency of intercalatories and the density of ribbing; Matheronites is retained as a subgenus
of Hemihoplites, but with a different type species H. (M.) ridzewskyi (Karakasch)!

Most other species referred to ‘ Matheronites ’ are characterized by crioceratitid coiling, and an ornament
consisting of weak or strong ribs intercalated with prominent periodic ribs, and of umbilical, lateral, and
ventrolateral tubercles which are well developed in the inner whorls and show a tendency to become
ontogenetically weaker, i.e. 'Ml alpinus (d’Orbigny, 1850, p. 100; Cottreau 1937, p. 63, pi. 78, figs. 16 and
17; Dimitrova 1967, p. 68, pi. 34, fig. 3), Ml hammatoptychum (Uhlig 1883, p. 262, pi. 30, figs. 1 and 2 a-c;
H. Douville 1916, p. 1 11, pi. 14, figs. 1-5; Dimitrova 1967, p. 67), 'Ml heberti (Fallot 1884, p. 296, pi. 9, fig.
2 a-c; Thieuloy 1979), 'Ml suessi (Toula 1892, p. 338, pi. 2, fig. I; Dimitrova 1967, p. 67, pi. 32, fig. 3), 'M.'
barremense (Kilian in Kilian and Leenhardt 1895, p. 978; Uhlig 1888, p. 95, pi. 4, fig. 3 a-c; Simionescu 1900,
p. 14, pi. 1, figs. 4 and 5; Collignon 1948, p. 79, pi. 12, fig. 4, 4a; Sarkar 1955, p. 86, text-fig. 13; Nikolov
1964, p. 122, pi. 2, fig. 2a, b, Dimitrova 1967, p. 69, pi. 32, fig. 4), 'Ml anthulai (Eristavi 1955; Anthula 1899,
p. 125, pi. 12, fig. 2 a-c; Thieuloy 1979, p. 307), 'Ml parolinianus (Rodighiero 1919, p. 114, pi. 13, fig. 7;
Dimitrova 1967, p. 70, pi. 33, fig. 3), 'Ml coheni (Sarkar 1955, p. 86, pi. 7, fig. 2), Ml limentinus Thieuloy
(1979, p. 307, pi. 1, figs. 1-4; pi. 2, fig. 5; ?Haug 1889, p. 215, pi. 11, fig. 5; Drushchits and Kudryastev 1960,
p. 292, pi. 34, fig. la, b), and the poorly known 'Ml ukensis Dimitrova (1967, p. 68, pi. 33, fig. 5). As pointed
out by Bogdanova (1971, p. 337), these species have ‘crioceratid’ (ancyloceratid) affinities and should not be
grouped with those mentioned above under the same genus/subgenus. An assessment of their correct generic
status and relationships is beyond the scope of this paper. The same holds true for the type material of 'Ml
brevispinus v. Koenen and 'Ml trispinosus v. Koenen (1902, pp. 363, 366, pi. 35, figs, la-c, 2a, b, 3a, b, 4a,
b, 5a, b, 6a-c, la-c, 8a, b pi. 39, figs, la, b and 2) (see Kemper 1973, p. 51). Material from the western USA
described under ?//. popenoi Murphy (1975, p. 38, pi. 8, figs. 1-5) does not belong in either of these groups.

452

PALAEONTOLOGY, VOLUME 32

The morphological differences between H. feraudianus (d'Orbigny 1841, p. 324, pi. 96, figs. 4 and 5;
Wiedmann 1966, p. 43, pi. 6, fig. 3a, b) and 'M.' soidieri (Matheron 1878, pi. C-21, fig. la, b) are not large
enough to support a generic/subgeneric distinction, as is clearly evident from Kakabadze’s (1981) inclusion
of the latter species in Hemihoplites. The affinity between these two type species cannot be dismissed because
of an artificially enlarged concept of Matheronites (see Dimitrova 1967; Thieuloy 1979), and the designation
of another type species for Matheronites , as proposed by Kakabadze (1981, p. 93), is not permissible under
ICZN rules. Thus, Matheronites is regarded as a junior synonym of Hemihoplites , as previously indicated by
Wright (1957) and Wiedmann (1962, 1966).

The differences between Hemihoplites and Pseudo thurmannia are difficult to determine because inter-
pretations of the latter genus vary. The large number of species and material included in Pseudo thurmannia
imply a large range of variation (see d’Orbigny 1841; Simionescu 1900; Koenen 1902; Rodighiero 1919;
Eristavi 1955; Sarkar 1955; Luppov et al. 1958; Drushchits and Kudryastev 1960; Wiedmann 1962; Dimitrova
1967; Thomel 1964; Breskovski 1966; Busnardo 1970; Immel 1978). When restricted to the type species P.
angu/icostata (d’Orbigny) (see Lapeyre 1974) and related species, it can be seen that Pseudo thurmannia tends
to become uncoiled with growth, to have tubercles on the umbilical margin (from which one or several ribs
and intercalatories are born which are usually restricted to the upper part of the flank), and periodic stronger
ribs on the outer whorl.

Material that could be referred to Hemihoplites has been described from the Barremian and/or Lower
Aptian of the Balearics (Fallot and Termier 1923), Bulgaria (Dimitrova 1967), ?Canada (Jeletzky 1976),
France (d’Orbigny 1841; Matheron 1878; Fallot and Termier 1923; Wiedmann 1966), Italy (Capellini 1881),
?Mexico (Imlay 1938), Spain (Wiedmann 1966), Caucasus and Turkmenia in the USSR (Karakasch 1896;
Renngarten 1926; Rouchadze 1933; Luppov 1936; Eristavi 1955; Luppov et at. 1958; Drushchits and
Kudryastev 1960; Bogdanova 1971; Kakabadze 1981), and Yugoslavia (Petkovic and Miletic 1949).

Hemihoplites varicostatus sp. nov.

Plate 51, figs. 1 9; Plate 52, figs. 1-3; text-figs. 3a-c, 4, 5a-/

71949 Crioceras angulicostatum d’Orb.; Petkovic and Miletic, p. 134, pi. 2, figs. 11 and 12.

1987 Hemihoplitidae indet. Riccardi et al ., p. 109.

Holotype. The incomplete phragmocone of a macroconch (MLP 20647), figured in Plate 51, figs. 1 and 2,
from level 4, Rio Belgrano, Santa Cruz, Upper Hauterivian.

Allotype. The incomplete phragmocone and body-chamber of a microconch (CPBA 14152) figured in text-
fig. 5a, from the same locality as the holotype.

Derivatio nominis. Latin for the wide range of variation in rib density.

Diagttosis. A species of Hemihoplites with rather evolute coiling, subquadrate to subrectangular
whorl section, rounded to shallow inner margin; flexuous ribbing, usually bifurcating at umbilical
margin of inner whorls and simple on outer whorls,, variable density, but decreasing through
ontogeny; inner whorls with tubercle-like swelling present irregularly at furcation points.

Material. The holotype and seven incomplete phragmocones (MLP 20648-20650, 22044, CPBA 1 1096, 14150-
14151) probable macroconchs. The allotype, two incomplete specimens (CPBA 14153 and 14155) probably

EXPLANATION OF PLATE 51

Figs. I 9. Hemihoplites varicostatus sp. nov., Rio Belgrano. 1 and 2, MLP 20647, holotype, incomplete
phragmocone of a macroconch, lateral and ventral views. 3 and 4, CPBA 1 1096, incomplete phragmocone
of a macroconch, lateral and ventral views. 5, MLP 20648, incomplete phragmocone of a macroconch,
lateral view. 6, MLP 20649, incomplete phragmocone, lateral view. 7, CPBA 14151, incomplete phragmo-
cone, lateral view. 8 and 9, MLP 20650, incomplete phragmocone of a macroconch, lateral and apertural
views. All figures x 1.

PLATE 51

RICCARDI and AGUIRRE URRETA, Hemihoplites

454

PALAEONTOLOGY, VOLUME 32

representing microconchs, and one juvenile? incomplete phragmocone (CPBA 14154). One (CPBA 14150)
from level 3, Chorrillo Rivera, and all the others from the same level and locality as the holotype. Collected
by A. C. Riccardi (1972), V. A. Ramos and M. A. Palma (1979), and G. Marin and M. B. Aguirre Urreta
(1981).

Description

Macroconch. The incomplete phragmocones range from 45-2 mm to 77 mm in diameter and are rather evolute
(U/D = 0-35 0-44). The whorl section varies from subquadrate to subrectangular (H/W = 1011-11), with
rounded umbilical shoulder and steep to shallow inner margin. The flanks and venter are flattish to slightly
convex (text-fig. 3 a-c). The ornament consists of rounded, blunt ribs which increase in strength throughout
ontogeny and from the lower to the upper part of the flanks. They cross the umbilical wall with an adoral
projection, and some of them, typically on the inner whorls, become thick and bifurcate on the umbilical
margin; all ribs form a shallow adapical bow on the lower third of the flank, are slightly projected on the
upper part of the flank, and cross straight over the venter. The number of ribs decreases throughout the last
two whorls (PI. 51, figs. 1-4).

a b

text-fig. 3. Cross-sections through the phragmocone and body-chamber (stippled) of Heniihoplites spp. a
c, H. varicostatus sp. nov. a , MLP 20647. holotype; b , MLP 20648; c, CPBA 1 1096. d, e, H. ploszkiewiczi sp.
nov. d, CPBA 14146, holotype; e, CPBA 14149, allotype.

The septal suture (text-fig. 4) is moderately complex and strongly protracted. L is slightly deeper than E,
moderately wide and symmetrically trifid; U is located on the umbilical edge, is also trifid but about half as
deep as L; there follow another well-developed saddle and a small lobe. The internal suture was not
investigated.

Microconch. The adult body-chamber is 20 to 41 mm in diameter, rather evolute (U/D = 0-33-0-35), with
subrectangular whorl section (H/W = I -26- 1 -35), shallow inner margin, flat flanks, and flat to slightly concave
venter. Thick and blunt flexuous ribs, simple or bifurcating on the umbilical margin, increase in thickness
through ontogeny and from the lower to the upper part of the flank, and become more widely spaced on
the last two whorls (text-fig. 5a). The aperture is not preserved.

RICCARDI AND AGUIRRE URRETA: ARGENTINIAN CRETACEOUS AM MONOIDS 455

i

b

text-fig. 4. External septal sutures of Hemihoplites varicostatus sp. nov. a , CPBA 11096
at H = 18-6 mm and W = 15-6 mm; b, MLP 22044 at H = 17-4 mm and W = c. 18-4 mm.

Dimensions.

D

H

W

H/W

U

P

S

Holotype

MLP 20647

phragmocone

72-3

30-7 (0-42)

28-4 (0-39)

1-08

24-5 (0-34)

17

20

phragmocone

63-7

26-8 (0-42)

26 0 (0 41)

1 03

22-1 (0-35)

—

—

MLP 20648

phragmocone

73-0

26-3 (0-36)

25-9 (0-35)

1-02

29-0 (0-40)

22

30

phragmocone

54-2

19 0 (0-35)

18-5 (0-34)

1-03

20-8 (0-38)

—

—

MLP 20649

phragmocone

77-0

—

—

27-6 (0-36)

17

23

MLP 20650

phragmocone

45-2

16 6 (0-37)

14- 1 (0-31)

1-18

12-9 (0-29)

11

17

CPBA 14150

phragmocone

64-2

28 0 (0-44)

24-0 (0-37)

117

20-5 (0-32)

12

19

CPBA 11096

phragmocone

65-2

22-8 (0-35)

20-5 (0-31)

111

22-8 (0-35)

14

16

CPBA 14152

body-chamber

410

14-8 (0-36)

11-0 (0-27)

1 35

14-3 (0-35)

11

14

CPBA 14153

body-chamber

300

8-2 (0-41)

6-5 (0-33)

1 26

6-5 (0-33)

12

16

CPBA 14154

phragmocone

15-2

6 0 (0 39)

5-5 (0-36)

1 09

4-2 (0-28)

—

—

Comments. H. varicostatus includes some specimens (PI. 51, figs. 3 and 4) which are similar to H. feraudianus,
but the intraspecific variation of both species is completely different. Although d’Orbigny (1841, p. 324, pi.
96, figs. 4 and 5) figured only one specimen of H. feraudianus from the Emeric Collection, d’Orbigny’s
Collection included thirty-two specimens. A search for the type series in the Museum National de Histoire

456

PALAEONTOLOGY, VOLUME 32

Naturelle (Paris) by one of the authors (A.C. R.), has revealed that the specimen figured by d’Orbigny
appears to be lost. Nevertheless, the collection includes a plaster cast of a specimen (no. R.923) from the
Barremian of the Maritim Alps, Emeric Collection, labelled as 'Type’, and most of d’Orbigny originals. The
plaster cast of Emeric’s specimen is partially crushed, and agrees in most features with those figured by
d’Orbigny (1841, pi. 96, figs. 4 and 5), including total absence of tubercles. Of twenty-nine specimens examined
in the d’Orbigny collection, most of them (twenty-one) are characterized by the presence of tubercles, usually
on the upper half of the flank and on the ventrolateral shoulder. The remaining specimens are similar to
Emeric’s specimen in the absence of tubercles. Even if the specimens came from different localities (S. Martin,
La Loire, Angles, Barremc) there is a coexistence of both tuberculated and non-tuberculated specimens in
the same locality. Thus, d’Orbigny's type series indicates that the presence of tubercles in H. feraudianus is
quite variable. In this respect it is worth noting that both morphotypes seem to have been included in
‘ Parahoplites soulieri ’ and P. astarte ’ by Fallot and Termier (1923).

Part of the type series and the specimen figured by Wiedmann (1966, pi. 6, fig. 3 a, b ) indicate that the
species is characterized by rather evolute coiling, subrectangular whorl section with flatfish venter, and by
blunt ribs which may bifurcate from incipient umbilical tubercles, are slightly fiexuous and projected on the
flanks, with obsolete tubercles on the ventrolateral shoulder, and cross the venter without interruption. The
same features are present in the specimen referred to "P. soulieri Math.’ by Fallot and Termier (1923, p. 67,
pi. 6, fig. 1 a-c). This specimen, also with slightly fiexuous and projected ribs quite often bifurcating from
umbilical tubercles, differs from A. soulieri Matheron (1878, pi. C-21, fig. I a, b) and the clearly related H.
astarte (Fallot and Termier 1923, p. 70, pi. 6, figs. 2-5). These two specimens have common, simple, and
straight ribs with more clearly developed ventrolateral tubercles. Similar features appear also to be present
in the specimen figured by Rouchadze (1933, pi. 3, figs. 5 and 6) under ‘M. cfr. soulieri and ‘ M . khwamliensis'
Rouchadze. But all features of H. feraudianus are quite clearly represented in the incomplete phragmocone
figured by Luppov (1935, pi. 1, figs. 1-3) under ‘AT turcmenicus'.

However, the inclusion of all these specimens in H. feraudianus is not enough to characterize adequately
the range of variation of this species, especially considering the possible inclusion of morphotypes with up
to three rows of tubercles (see above). Furthermore, the presence of sexual dimorphism, as shown by the
Patagonian material, needs also to be explored. Existence of sexual dimorphism in the European Hemihopliti-
dae is suggested by material figured by d’Orbigny (1841, pi. 42, fig. 3), and later referred to Pseudothurmannia
(see Busnardo 1970, p. 1 35). However, the specimen examined by one of us (A. C. R.) in the Museum National
de Histoire Naturelle, is a poorly preserved impression with aperture poorly defined, and its taxonomic status
is uncertain.

Thus, the material from Patagonia included in H. varicostatus sp. nov. appears to have a range of variation
unlike that of H. feraudianus. Furthermore, in specimens referred in the European and Soviet literature to
either H. feraudianus and related species or even to the genus/subgenus Pseudothurmannia , the subquadrate
whorl section and dense ribbing shown by some of the Patagonian specimens is absent. We therefore conclude
that the Patagonian species is characterized by a different range of variation, in spite of some morphological
overlap with H . feraudianus related to specimens with more subrectangular whorl section and sparser ribbing
without tubercles.

Although specimens with subrectangular whorl section and sparser and stronger ribbing included in H.
varicostatus are very close to European material referred to H. feraudianus , closer similarity was found to a
specimen referred to as ‘ Crioceras angulicostatum by Petkovic and Miletic (1949, pi. 2, figs. 1 1 and 12). This
Yugoslavian species, as well as the specimen figured by these authors on Plate 2, figs. 7 9, do not belong to
Pseudothurmannia as indicated by other authors (see references under genus discussion).

EXPLANATION OF PLATE 52

Figs. 1 3. Hemihoplites varicostatus sp. nov. I, CPBA 14151, incomplete phragmocone, lateral view, Rio
Belgrano. 2 and 3, CPBA 14150, incomplete phragmocone, ventral and lateral views, Chorrillo Rivera.

Figs. 4 9. H. ploszkiewiczi sp. nov., Veranada de la Vinca. 4, CPBA 14147, ?mature phragmocone with nearly
complete body-chamber, lateral view. 5 and 6, CPBA 14149, allotype, complete? microconch, lateral and
ventral views. 7 and 8, CPBA 14148, nearly complete microconch, ventral and lateral views. 9, CPBA
14145, incomplete phragmocone of a macroconch, lateral view.

All figures x 1 .

PLATE 52

5#Sli

RICCARDI and AGUIRRE URRETA, Hemihoplites

458

PALAEONTOLOGY, VOLUME 32

text-fig. 5. a-f, Hemihoplites varicostatus sp. nov., Rio Belgrano. a, CPBA 14152, allotype, mature?
incomplete phragmocone and body-chamber of a microconch, lateral view, b, c, CPBA 14155, incomplete
phragmocone and body-chamber of a ?mature specimen, lateral and ventral views, d , CPBA 14154,
phragmocone of a ?juvenile, lateral view. e,f CPBA 14153, incomplete phragmocone of a ?juvenile, lateral
and ventral views, g, /;, H. ploszkiewiczi sp. nov. CPBA 14146, holotype, incomplete phragmocone of a
macroconch, lateral and ventral views, Veranada de la Vinca. All figures x 1.

Hemihoplites ploszkiewiczi sp. nov.

Plate 52, figs. 4 9; text-figs. 3 d. e , 5 g, /?, 6

Holotype. The macroconchiate incomplete phragmocone (CPBA 14146), figured in text-fig. 5 g, h , from the
basal part (level 1) of the Rio Mayer Formation, Veranada de La Vinca, Santa Cruz, late early Hauterivian.

Allotype. The complete? microconch (CPBA 14149), figured in Plate 52, fig. 4 a, b , from the same level and
locality as the holotype.

Derivatio nominis. After J. V. Ploszkiewicz, who collected the material here described.

Diagnosis. A species of Hemihoplites , with subrectangular and compressed whorl section, and
dense, fine, flexuous ribbing on phragmocone, becoming more widely spaced on body-chamber;
microconchiate body-chamber with faint ventral tubercles and more frequent intercalatory ribs.

Material. The holotype, the allotype, one macroconch with mature? phragmocone and nearly complete body-
chamber (CPBA 14147), one incomplete macroconchiate phragmocone (CPBA 14145), and one nearly
complete microconch (CPBA 14148); all from the same level and locality as the holotype collected by J. V.
Ploszkiewicz and V. A. Ramos (1982).

Description. Coiling is relatively evolute, with impressed dorsum. On the inner whorls, the whorl section is
subquadrate, slightly higher than wide, with flat flanks, rounded venter, and angular umbilical and
ventrolateral margins. With increasing diameter, the whorl section becomes more compressed (H/W = 1 05
114) with flanks flat but slightly converging towards the venter and the umbilical border more rounded. The
maximum width is on the lower third of the flank (text-fig. 3d. e). The umbilicus is relatively large and
shallow, and becomes smaller with increasing diameter (U/D: 0-38-0-32).

On the phragmocone the ribs arise at the umbilical margin, curve forward, then backwards, cross the
flanks with a flexuous curve, and are straight over the venter. Most of the ribs arise singly from the umbilical
edge, some are in pairs, while a few are intercalated higher on the whorl flanks. The macroconchiate
phragmocones have 17-19 primary ribs and 30 35 secondary ribs per half-whorl. One microconchiate
phragmocone shows 17 primary and 21 secondary ribs. They are fine, dense, and evenly separated by
interspaces as wide as the ribs.

RICCARDI AND AGUIRRE URRETA: ARGENTINIAN CRETACEOUS AMMONOIDS 459

text-fig. 6. External septal suture of Hemihoplites ploszkiewiczi sp. nov. CPBA 14147
at H = 21-7 mm and W = c. 19 mm.

On the body-chamber of both macro- and microconchs, ribbing retains the pattern of the phragmocone,
but becomes coarser and more spaced. The microconchiate body-chamber has more intercalatories, arising
from the dorsal third of the flank, and faint tubercles, which border a siphonal depression. There are 14
primary and 23-24 secondary ribs per half-whorl.

The septal suture (text-fig. 6) is moderately complex and strongly protracted. L is deeper than E, moderately
wide and symmetrically trifid; U is located on the umbilical margin and is also trifid; there follows another
saddle and a small lobe. The internal suture was not investigated.

Dimensions.

D

H

W

H/W

u

P

S

Holotype

CPBA 14146

phragmocone

616

23-3 (0-38)

20-4 (0-33)

1 14

22-5 (0-36)

19

35

Allotype

CPBA 14149

body-chamber

50-5

20-2 (0 40)

17-8 (0-35)

1 14

16 0 (0-32)

14

23

phragmocone

34-7

15-3 (0-44)

14-5 (0-42)

1 05

12-0 (0-34)

18

21

CPBA 14148

body-chamber

52-5

218 (0 41)

191 (0-36)

1 14

17-1 (0-32)

14

24

phragmocone

37-5

15-4 (0-41)

14 3 (0-38)

1 07

13-9 (0-37)

—

—

CPBA 14147

phragmocone

85-8

c. 32-6 (0-38)

—

—

c. 28-0 (0-33)

—

—

phragmocone

74-2

29 0 (0-40)

—

—

25-8 (0-35)

—

—

phragmocone

52-2

22-5 (0-44)

—

—

19-9 (0-38)

—

—

CPBA 14145

phragmocone

56-8

22-7 (0-40)

20-5 (0-36)

1 10

20-0 (0-37)

17

30

Comments. H. ploszkiewiczi n. sp. differs from H . feraudianus (d’Orbigny 1841, p. 324, pi. 96, figs. 4 and 5;
Wiedmann 1966, pi. 6, fig. 3a, b) in the denser ribs on the phragmocone at similar diameter, and the seemingly
larger size. Comparison is hindered by the poorly known sexual dimorphism in H. feraudianus. However, H.
ploskiewiczi is similar to some Northern Hemisphere specimens described under other specific names, but
here referred to H. feraudianus (see under genus and below).

The new dimorphic species H. varicostatus (see above), differs in the range of variation of ribbing and
whorl section of the macroconchs, which at similar phragmocone diameter includes specimens with sparser
and stronger ribs and subrectangular whorl section. When the microconchs are compared, H. ploszkiewiczi
is larger and has more numerous and finer ribs.

H. ploszkiewiczi differs from all other species of Hemihoplites (see under genus) in the denser ribbing. Only
the microconch resembles quite closely some Northern Hemisphere representatives of Hemihoplites described
under different species, but here referred to H . feraudianus (see Fallot and Termier 1923, pi. 6, fig. la, b\
Rouchadze 1933, pi. 3, figs. 5 and 6; Luppov 1936, pi. 1, figs. 1-3; Drushchits and Kudryastev 1960, pi. 38,
fig. la, b). These specimens, however, are more involute and no possible macroconch is known that could
be compared to the Patagonian species. H. ploszkiewiczi has some superficial resemblance to Pseudo thurmannia
in coiling and rib density. However, the Patagonian form lacks a distinctive row of umbilical tubercles, the

460

PALAEONTOLOGY, VOLUME 32

intercalatories are born near the umbilical margin, and the ribs are bent on the lower part of the flank and
straight on the upper; it shows no hint of progressive uncoiling and of periodic, stronger ribs throughout
ontogeny (see under generic discussion).

Acknowledgements. Laboratory research was financed by the Consejo Nacional de Investigaciones Cientificas
y Tecnicas de la Repiiblica Argentina (CONICET). Field-work was supported for A. C. R. by the CONICET,
and for M.B.A.U. by the Servicio Geologico Nacional (SGN). J. V. Ploszkiewicz, V. A. Ramos, M. A.
Palma and G. Marin, SGN, Argentina, collected part of the material studied. P. Doyle (British Antarctic
Survey, UK), the late R. Forster (then at the Bayerische Staatssammlung fur Palaontologie und historische
Geologie, FRG), I. Mijailova (Moscow University, USSR), I. Sapunov (Bulgarian Academy of Sciences,
Sofia), and G. Thomel (Museum d’Histoire Naturelle, Nice, France) provided literature that was not available
in Argentina. A. Schindler (La Plata) and T. Sedlecki (Buenos Aires) helped with Russian and Yugoslavian
texts. Part of the text was edited by G. E. G. Westermann, McMaster University, Canada.

REFERENCES

aguirre urreta, m. b. and rlinger, h. c. 1986. Upper Barremian Heteroceratinae (Cephalopoda,
Ammonoidea) from Patagonia and Zululand, with comments on the systematics of the subfamily. Ann. S.
Afr. Mus. 96, 315-358.

anthula, D. J. 1899. Uber die Kreidefossilien des Kaukasus. Beitr. Palaont. Geol. Ost.-Ung. 12, 55-159.
Basse, E. 1952. Classe des Cephalopodes. In piveteau, j. (ed.). Trade de Paleontologie , 2, pp. 461-755. Masson
et Cie, Paris.

bogdanova, t. n. 1971. New Barremian ammonites from western Turkmenia. Paleont. J. 3, 334-344.
breskovski, s. t. 1966. Biostratigraphie du Barremien au sud du village de brestak dans la region de Varna.
Trav. geol. Bulgur ie, Ser. Paleont. 8, 71-121.

busnardo, R. 1970. Les Pseudo thurmannia (Ammonoidea) de FHauterivien superieur de la Montagne de
Lure. Docums Lab. Geol. Fac. Sci. Lyon, 37, 133-145.

capellini, g. 1881. 11 magigno di Porretta e le roccie a Globigerine dell’Appennino bolognese. Memorie R.

Accad. Sci. 1st. CL Sci.fis. Bologna, ser. Y, 2, 175-194.
collignon, m. 1948. Faune Neocomienne des couches a Crioceras de Belohasifaka, Cercle de Sitampiky,
Madagascar. Annls. geol. Serv. Mines Madagascar, 15, 55-83.
cottreau, j. 1937. Types du prodrome de Paleontologie stratigraphique universelle de d’Orbigny. B.

Neocomien superieur ou Urgonien. Annls. Paleont. 26, 17-48.

Dimitrova, N. 1967. Les fossiles de Bulgarie. 4. Cretace inferieur. Cephalopoda (Nautiloidea et Ammonoidea),
424 pp. Academy of Bulgarian Science, Sofia.

douville. h. 1916. Les terrains secondaires dans la Massif du Moghara. Mem. Acad. Sci. Inst. Fr. (2) 54, 1
48.

douville, r. 1909. Sur des Ammonites provenant des environs de Lima et sur une nouvelle coupure de la
famille des Hoplitides: Favrella n. gen. C. r. Soc. geol. Fr. (4) 9, 164-166.
drushchits, v. v. and kudryastev, m. p. 1960. Atlas of the Lower Cretaceous fauna of northern Caucasus
and Crimea. Trudy vses. nauchno-issled. Inst. priv. Gazov. 1-700. [In Russian.]
eristavi, m. s. 1955. Lower Cretaceous fauna of Georgia. Monogr. Inst. Geol. Akad. Nauk Gruz. SSSR , 6,
1-224. [In Russian.]

fallot, e. 1884. Note sur un gisement cretace fossilifere des environs de la gare d’Eze (Alpes-Maritimes).
Bull. Soc. geol. Fr. (3) 12, 289-300.

fallot, p. and termier, h. 1923. Ammonites nouvelles des lies Baleares. Trab. Mus. nac. Cienc. nat. (Geol.),
32, 1-83.

haug, e. 1889. Beitrag zur Kenntniss der Ober-Neocom Ammoniten fauna der Puezalpe bei Corvara
(Subtirol). Beitr. Palaont. Geol. Ost.-Ung. 7, 193-229.

imlay, r. 1938. Ammonites of the Taraises Formation of Northern Mexico. Bull. geol. Soc. Amer. 49, 539-
602.

immel, h. 1978. Die Crioceratiten (Ancyloceratina, Ammonoidea) des Mediterranen und Borealen Hauterive-
Barreme (Unter Kreide). Palaeontographica, A163, I 85.
jeletzky, j. a. 1976. Cretaceous Macrofaunas. Geol. Surv. Canada, Economic Geology Report, IB, 649-662.
kakabadze, m. v. 1981. The ancyloceratids of the south of the USSR and their stratigraphical significance.
Trudy geol. Inst. Tbilisi, 71, 1-196. [In Russian.]

RICCARDI AND AGUIRRE URRETA: ARGENTINIAN CRETACEOUS AMMONOIDS 461

karakasch, N. 1896. Partie Paleontologique. In inostranzeff, a. (ed.). An trovers de la Chaine Principale du
Caucase , 205 pp. St Petersbourg.

kemper, E. 1973. Die Unterkreide im Untergrund der Gehrdener Berge und in der Deister-Mulde. Ber.
Naturhist. Ges. 117, 29-54.

kilian, w. and leenhardt, f. 1895. Sur le Neocomien des Environs de Moustiers Ste. Marie (Basses-Alpes).
Bull. Soc. geol. Fr. (3) 23, 970 981.

koenen, a. von. 1902. Die Ammonitiden des Norddeutschen Neocom (Valanginien, Hauterivien, Barremien
und Aptien). Abh. preuss. geol. Landesanst. nf 24, 1-451.
lapeyre, j. f. 1974. Revision de l’ammonite-index Pseudo thurmannia angulicostata d’Orbigny. Ann. Mus.
Hist. not. Nice , 2, 81 -86.

leanza, a. f. 1963. Patagoniceras gen. nov. (Binneyitidae) y otros ammonites del Cretacico superior de
Chile meridional con notas acerca de su position estratigrafica. Bo In. Acad. nac. Cienc. Cordoba , 43,
203-225.

luppov, n. p. 1936. Ammonites from Barremian deposits of eastern Karabugag area (Northwestern
Turkmenia). Trudy Leningr. Obshch. Yestestvoispyt. 65, 116-124. [In Russian.]

— eristavi, m. s. and drushchits, v. v. 1958. Superfamily Berriasellacea. In orlov, y. a. (ed.). Osnovy
Paleontologii , 6, 96-104. Nedra Press, Moscow. [In Russian.]
matheron, p. 1978. Recherches paleontologiques dans le Midi de la France. 3 partie, pi. C-21. Marseille.
murphy, m. a. 1975. Paleontology and stratigraphy of the lower Chickabally Mudstone (Barremian-Aptian)
in the Ono Quadrangle, Northern California. Univ. Calif. Pubis geol. Sci. 113, 1-52.
nikolov, t. 1964. Barremian ammonites from north-eastern Bulgaria. Trav. geol. Bulgarie , Ser. Paleont. 6,
117-141.

olivero, e. b. 1982. Estratigrafia de la cuenca sur del lago Fontana. Tests de hi Universidad de Buenos Aires ,
147 pp. (unpublished).

orbigny, A. d’. 1840-1842. Paleontologie Franqaise. Terrains Cretaces , I (Cephalopodes), 662 pp., 151 pis.
Masson, Paris.

— 1850. Prodrome de Paleontologie stratigraphique universelle des animaux mollusques et rayonnes . . ., 2,
427 pp. Masson, Paris.

petrovic, k. and miletic, o. 1949. Sur les ammonites barremiens trouves dans les calcaires urgoniens de
Kosutnjak (environs de Belgrade) et leur importance. Ann. geol. Penins. Balkan. 17, 123-139.
renngarten, v. 1926. La Faune des depots cretaces de la region d’Assa-Kambileevka, Caucase du Nord.
Mem. Com. geol. Rus. NS 147, 1 132.

riccardi, A. c. 1984a. Las Asociaciones de Amonitas del Jurasico y Cretacico de la Argentina. 9 Congreso
Geol. Argentino , 4, 559-595.

— 19846. Las zonas de Amonitas del Cretacico de la Patagonia (Argentina y Chile). 3 Congreso
Latinoamericano Paleont. 346-405.

— 1988. The Cretaceous System of Southern South America. Mem. geol. Soc. America , 168, 1-143.

— aguirre urreta, m. b. and Medina, f. A. 1987. Aconeceratidae (Ammonitina) from the Hauterivian
Albian of Southern Patagonia. Palaeontographica, A196, 105 185.

rodighiero, a. 1919. 11 sistema Cretaceo del Veneto occidentale compreso fra f Adige e il Piave, con speciale
riguardo al Neocomiano dei Sette Comuni. Palaeontogr. Ital. 25, 39-125.
rouchadze, j. 1933. Les ammonites aptiennes de la Georgie occidentale. Bull. Inst. geol. Georgie , 1 (3), 165—
273.

sarkar, s. s. 1955. Revision des ammonites deroulees du Cretace inferieur du Sud-est de la France. Mem.
Soc. geol. Fr. ns 72, I 176.

simionescu, I. 1900. Notes sur quelques ammonites du Neocomien frangais. Trab. Lab. Geol. Univ. Grenoble ,

5, I 17.

spath, l. f. 1924. On the ammonites of the Speeton Clay and the subdivisions of the Neocomian. Geol. Mag.
61, 73-89.

— 1933. Revision of the Jurassic cephalopod fauna of Kachh (Cutch). Mem. geol. Surv. India , Palaeont.
Indica , ns 9, 2 (6), 659-945.

Stanton, t. w. 1901. The Marine Cretaceous Invertebrates. Rep. Princeton Univ. Exped. to Patagonia , 4, 1
43.

thieuloy, j. p. 1979. Matheronites limentinus n. sp. (Ammonoidea) espece-type d’un horizon-repere Barremien
Superieur du Vercors meridional (Massif Subalpin Frangais). Geobios , Mem. spec. 3, 305 317.
thomel, g. 1964. Contribution a la connaissance des Cephalopodes Cretaces du Sud-Est de la France. Mem.
Soc. geol. Fr. ns 101, 1 80.

462

PALAEONTOLOGY, VOLUME 32

toula, F. 1892. Geologische Untursuchungen im ostlichen Balkan und in andener Theilen von Bulgarien und
Ostrumelien. Dunkschr. math.-naturwiss. Kl. K. Akad. Wiss. Wien. 57, 323-400.
uhlig, v. 1883. Die Cephalopodenfauna der Wernsdorfer Schichten. Ibid. 46, 127-290.

— 1888. Ueber neocome Fossilien von Gardenazza in Siidtirol nebst einem Anhang fiber das Neocom von
Ischl. Jb. K.-K. geol. Reichsanst., Wien , 37, 69 108.

wiedmann, j. 1962. Unterkreide-Ammoniten von Mallorca. I. Lieferung: Lytoceratina, Aptychi. Abh. math.-
naturwiss. Kl. Akad. Wiss. Mainz , 1962, 1,1 148.

1966. Stammesgeschichte und System der postriadischen Ammonoideen. Ein Uberblick. Neues Jb. Geol.
Paldont. Abh. 127 (1), 13-81.

wright, c. w. 1957. Mesozoic Ammonoidea. In moore, r. c. (ed.). Treatise on invertebrate paleontology.
Part L, Mollusca 4 , L80-L465. Geological Society of America and University of Kansas Press, Kansas.

— 1981. Cretaceous Ammonoidea. In house, m. r. and senior, j. r. (eds.). The Ammonoidea. Systematics
Assoc. Spec. Vol. 18, 157-174.

A. C. RICCARDI

Division Paleozoologia de Invertebrados
Museo de Ciencias Naturales
1900— La Plata, Argentina

Typescript received 15 February 1988
Revised typescript received 4 August 1988

M. B. AGUIRRE URRETA

Departamento de Geologia
Universidad de Buenos Aires
Ciudad Universitaria
1428— Buenos Aires, Argentina

NOTES FOR AUTHORS

The journal Palaeontology is devoted to the publication of papers on all aspects of palaeontology. Review articles are
particularly welcome, and short papers can often be published rapidly. A high standard of illustration is a feature of the
journal. Four parts are published each year and are sent free to all members of the Association. Typescripts should conform
in style to those already published in this journal, and should be sent to Dr 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 1985. 28, pp. 793 800).

Special Papers in Palaeontology is a series of substantial separate works conforming to the style of Palaeontology.

SPECIAL PAPERS IN PALAEONTOLOGY

In addition to publishing Palaeontology the Association also publishes Special Papers in Palaeontology. Members may
subscribe to this by writing to the Membership Treasurer: the subscription rate for 1989 is £37-50 (U.S. $66) for Institutional
Members, and £18-50 (U.S. $32) for Ordinary and Student Members. A single copy of each Special Paper is available to
Ordinary and Student Members only , for their personal use, at a discount of 25 % below the listed prices. Non-members may
obtain these (at cover price) from Basil Blackwell, 108 Cowley Road, Oxford OX4 1JF, England. Older issues are available
to Members and Non-members from the Marketing Manager.

RECENT PALAEONTOLOGICAL ASSOCIATION
PUBLICATIONS

Special Papers in Palaeontology

Numbers 2-26 are still in print and are available (post free) together with those listed below:

27. (for 1981): Late Devonian Acritarchs from the Carnarvon Basin, Western Australia, by G. playford and r. s. dring.
78 pp., 10 text-figs., 19 plates. Price £15 (U.S. $23).

28. (for 1982): The Mammal Fauna of the Early Middle Pleistocene cavern infill site of Westbury-sub-Mendip. Somerset,
by M. j. bishop. 108 pp.. 47 text-figs., 6 plates. Price £25 (U.S. $38).

29. (for 1982): Fossil Cichlid Fish of Africa, by J. a. h. van couvering. 103 pp., 35 text-figs., 10 plates. Price £30 (U.S. $45).

30. (for 1983): Trilobites and other early Arthropods. Edited by d. e. G. briggs and p. d. lane. 276 pp., 64 text-figs., 38
plates. Price £40 (U.S. $60).

31. (for 1984): Systematic palaeontology and stratigraphic distribution of ammonite faunas of the French Coniacian. by w.
j. Kennedy. 160 pp., 42 text-figs., 33 plates. Price £25 (U.S. $38).

32. (for 1984): Autecology of Silurian organisms. Edited by M. G. bassett and J. D. lawson. 295 pp., 75 text-figs., 13 plates.
Price £40 (U.S. $60).

33. (for 1985): Evolutionary Case Histories from the Fossil Record. Edited by j. c. w. cope and p. w. skelton. 202 pp., 80
text-figs., 4 plates. Price £30 (U.S. $45).

34. (for 1985): Review of the upper Silurian and lower Devonian articulate brachiopods of Podolia, by o. i. Nikiforova, t.
l. modzalevskaya and M. G. bassett. 66 pp., 6 text-figs., 16 plates. Price £10 (U.S. $15).

35. (for 1986): Studies in palaeobotany and palynology in honour of N. F. Hughes. Edited by d. j. batten and D. E. G.
BRIGGS. 178 pp., 29 plates. Price £30 (U.S. $50).

36. (for 1986): Campanian and Maastrichtian ammonites from northern Aquitaine, France, by w. j. Kennedy. 145 pp.. 43
text-figs., 23 plates. Price £20 (U.S. $35).

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. $35).

38. (for 1987): Taxonomy, evolution, and biostratigraphy of late Triassic-early Jurassic calcareous nannofossils, by p. r.
bown. 1 18 pp.. 19 text-figs., 15 plates. Price £30 (U.S. $50).

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. $50).

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. $50).

Field Guides to Fossils

These are available from the Marketing Manager. Prices do not include postage. Reduced prices are available to Members.

1. (1983): Fossil Plants of the London Clay, by m. e. collinson. 121 pp., 242 text-figs. Price £7-95 (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 (LfS. $18).

3. (1988): Zechstein Reef fossils and their palaeoecology, by n. hollingworth and t. Pettigrew, iv + 75 pp. Price £4-95

(U.S. $8).

Other Publications

Please contact the Marketing Manager, do not send payment in advance.

1982. Atlas of the Burgess Shale. Edited by s. conway morris. 31 pp., 24 plates. Price £20 (U.S. $30).

1985. Atlas of Invertebrate Macrofossils. Edited by j. w. Murray. Published by Longman in collaboration with the
Palaeontological Association, xiii + 241 pp. Price £13-95. Available in the USA from Halsted Press at U.S. $24-95.

© The Palaeontological Association, 1989

Palaeontology

VOLUME 32 • PART 2

CONTENTS

Palaeotidal characteristics determined by micro-growth patterns
in bivalves

T. OHNO 237

A late Permian freshwater shark from eastern Australia

M. R. leu 265

A new genus of osmundaceous stem from the Upper Triassic of
Tasmania

R. S. HILL, S. M. FORSYTH and F. GREEN 287

Original mineralogy of trilobite exoskeletons

n. v. wilmot and a. e. fallick 297

The mammal-like reptile Rechnisaurus from the Triassic of India

S. BANDYOPADHYAY 305

Iocrinus in the Ordovician of England and Wales

s. k. donovan and A. GALE 313

First records of conodonts from the late Triassic of Britain

A. swift 325

Heterochrony in a fossil reptile: juveniles of the rhynchosaur
Scaphonyx fischeri from the late Triassic of Brazil

M. J. BENTON and R. KIRKPATRICK 335

Late Cretaceous ammonites from the Wadi Qena area in the
Egyptian Eastern Desert

P. LUGER and M. GROSCHKE 355

Leptopterygius tenuirostris and other long-snouted ichthyosaurs
from the English Lower Lias

C. MCGOWAN 409

Size-selective transport of shells by birds and its palaeoecological
implications

G. C. CADEE 429

A peafowl from the Pliocene of Perpignan, France

C. MOURER-CHAUVIRE 439

Hemihoplitid ammonoids from the Lower Cretaceous of southern
Patagonia

A. C. RICCARDI and M. B. AGUIRRE URRETA 447

Printed in Great Britain at the University Press, Cambridge ISSN 0031-0239

S3IHVH0S1

’institution2

z

: to

INSTITUTION

saiavaan

'institution2

z

LIBRARIES SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSH1IINS S3IHWH0n L

m

NOIiniIlSNrNVINOSHims','s3 I ava a n~LI BRAR I eswsmithsonian‘ INSTITUTION
z £2 z t/> 2 ...

"I . v | §= < A

c/> (ftg£ Sp) to yjf/MW in • to /*»> (j)

X vXke.^si o W//'Wi. X g x

> S ^ ^ > N^riiixs^/ 2

LI B RAR I ES^SMITHSONIAN INSTITUTION NOIlfUliSNI NVINOSHilWS^SB S HVB 0 n"

M = to — .... co :

O

NOIinillSNI^NVINOSHlIWS^SaiMVaan^LIBRARIES^SMITHSONIAN^INSTITUTION^NC
r“ > z r- z r~ z

o

LI B RAR I ES ^SMITHSONIAN INSTITUTION^NOJifUUSNI NVIN0SH11WS S3JHV08n LI

tr% *5? U% 2!

<

NOIlfUliSNI NVIN0SHIHNS'W'S3 I U VH 8ll*\lB RAR S ES^SMITHSONIAN^INSTITUTION ^NC

w _ = w „ = _ to

w

uj

z

to

H

w

to

H

UJ

\M7 O "' V55*' ~ 'yjvAi.ty o

Z -1 z J z

LIBRARIES SMITHSONIAN INSTITUTION NOllIUUSNi NVINQSH1ISAIS S1SHVH9I1
^ «“ z r* z

rn

N0llfUllSNI_NVIN0SHllWS_S3 I HVB 8 I1_LI B RAR I ES„SMITHSONIAN“lNSTITUTIONCrt NC

to

to

I 'k >

L! BRAR I E S ^SMITHSONIAN ^ INSTITUTION^ NOIlfUliSNI^ NVINOSHIIWS^S 3 I B VB S II LI

m

NOIlfUliSNI NVINOSHUWS S3 I HVB a II LIBRARIES SMITHSONIAN INSTITUTION NC

ro

33

>

33

m

^ _l 2 _j 2 -J ■*-

ES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IIAIS SBIBVHan LIBRARIES SMITHSONIAN
s” r- z, r- z 8” 2

o
I-

VI

in

m

t 'tft

SNI NVINOSH1IINS S3IHVHail LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVINOSH1IIAIS

— tn zr

E S^SMITHSON IAN ^ INSTITUTION W/NOIinillSNI^NVINOSHllWSC° S3 I aVH 8 II^LI B RAR I ES^SMITHSONIAN
m ~ <n — m ~ to

UJ

»NI JNVIN0SH1IIA!S^S3 I UVh Q II^LI B RAR I ES^SMITHSONIAN^INSTITUTION^NOlinillSNI^NVINOSHlIINS
r° 2 r™ z :?■ r* v

m

E S ^SMITHSONIAN "INSTITUTION^ NOlinillSNI ~NVINOSHllWS</> S3 lavaan "LIBRARIES SMITHSONIAN

,n -3 »»_*/» Z tn ' Zr >■_

05

05

CO

2

05

2

05

;ni__nvinoshii^s saiavaan libraries'smithsonian institution NouniiisNi_NviNosHiiws
2 ^8!^ I I 1 5

ES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IWS S3IHVH3n LIBRARIES SMITHSONS
z *“ 2 F= 2 S” 2

O

3 - ^

—

W

w

SNI NVIN0SH1IIAIS S3 1 H VH S tl LIBRARIES SMITHSONIAN INSTITUTION NOlinillSNI NVIN0SH1IIAIS

in

in

w

ES^ SMITHSONIAN^ I NSTITUTION^ NO I inillSNI^NVINOSHlU^S^S 3 I B V£J 8 11^ LI B RAR I ES^SMITHSONIAN

^ Z ^ w . 05 5 C/5

| S ^ ^ 5 *

SNI^NVINOSHlIINS^SB IBVaan^LIBRARI ES^SMITHSONIAN'-'lNSTiTUTfON^NOlinillSNI^NVINOSHlIlAIS
r-v2 r- 2 r~ z r~ >